[ 3 / biz / cgl / ck / diy / fa / g / ic / jp / lit / sci / tg / vr / vt ] [ index / top / reports / report a bug ] [ 4plebs / archived.moe / rbt ]

Due to resource constraints, /g/ and /tg/ will no longer be archived or available. Other archivers continue to archive these boards.Become a Patron!

/diy/ - Do-It-Yourself

View post   

[ Toggle deleted replies ]
File: 323 KB, 807x1086, WOM.jpg [View same] [iqdb] [saucenao] [google] [report]
2032342 No.2032342 [Reply] [Original] [archived.moe]

>I'm new to electronics. Where to get started?
It is an art/science of applying principles to requirements.
Find problem, learn principles, design and verify solution, build, test, post results, repeat.

>Project ideas:

>Don't ask, roll:

>Archive of Popular Electronics magazines (1954-2003):
>Some guy’s list of electronics resources:
>Microchip Tips and Tricks PDF:
>Li+/LiPo batteries required reading:

>Principles (by increasing skill level):
Mims III, Getting Started in Electronics
Geier, How to Diagnose & Fix Everything Electronic
Kybett & Boysen, All New Electronics Self-Teaching Guide
Scherz & Monk, Practical Electronics for Inventors
Horowitz and Hill, The Art of Electronics

>Design/verification tools:
NI Multisim
iCircuit for Macs
KiCAD (PCB layout software, v5+ recommended)
Logisim Evolution

Mouser, Digi-Key, Arrow, Newark, LCSC (global)
RS Components (Europe)
eBay/AliExpress sellers, for component assortments/sample kits (caveat emptor)
Local independent electronics distributors

>Related YouTube channels:
Ben Eater

>I have junk, what do?
Shitcan it

>> No.2032344


>> No.2032382

What's the going rate for a pristine 200 byte patch of magnetic core memory these days?

>> No.2032385

Actual vintage, BTW, not a reproduction. I think it's from 1968.

>> No.2032397

200B is pretty large, do you know what kind of machine it was from? You'd likely be selling it to a collector, so the price would be rather volatile depending on who wants what.
While there are still a few people who use magnetic core memory in their projects, I suspect they just use it to store single bits, so you'd probably not have a practical market for a large array like that. It's the sort of thing people build a project around, or possibly something people include in a replica of a particular machine.

>> No.2032405
File: 2.66 MB, 1632x1224, mem1.jpg [View same] [iqdb] [saucenao] [google] [report]

No idea what it was designed for and I haven't been able to find any identical modules online.

Apparently here in Australia we had a whole bunch of weird stuff nobody outside the country ever saw much of, and little of it was documented properly or preserved. I'll throw some pictures up, maybe you guys will know what it is.

>> No.2032408
File: 2.63 MB, 1632x1224, mem2.jpg [View same] [iqdb] [saucenao] [google] [report]

Also noticed there's a condensation patch from the mug that was there just before, on the desk, hope that isn't some sort of damage risk for this old stuff. It's back in the little plastic slip it lives in now, patted it down first.

>> No.2032410
File: 2.16 MB, 1632x1224, mem3.jpg [View same] [iqdb] [saucenao] [google] [report]

Close ups of the codes. I think this one says "16 - 7 - 68886" or possibly "16 - 7 - 68 8 86"? The first 8 could also be a 9.

>> No.2032411
File: 2.07 MB, 1632x1224, mem4.jpg [View same] [iqdb] [saucenao] [google] [report]

"150 659 24"?

>> No.2032441
File: 1.08 MB, 638x360, 1564426767991.webm [View same] [iqdb] [saucenao] [google] [report]

If I want to make my own monitor colorimeter should I use a color sensor or a spectrometer sensor?
Pic unrelated.

>> No.2032444

Depends on what you're trying to measure. If you just care about the brightness of the three colour channels, then the requirements aren't very strict. But if you want a spectral analysis, perhaps to calculate CRI, a spectrometer will be a must. A simple 3-colour sensor could be made with an RGB LED, with all three diodes acting as photodiodes. Would definitely require some calibration. A spectrometer on the other hand, will be quite expensive. The ones I've seen are CCDs with diffraction mirrors baked into opposing silicon to form a diffraction pattern, which sounds like a complex process. But if you can find a cheap spectrometer sensor, definitely post it here, because I want one.
And read the datasheet either way.

>> No.2032537 [DELETED] 
File: 22 KB, 600x299, spyder colorimeter.jpg [View same] [iqdb] [saucenao] [google] [report]

>monitor colorimeter

i found a couple at the thrift store for a couple of bucks each, but found them completely useless. i was expecting the software to tell me, ''bring up the red a lil bit, and turn the green down a smidgen." instead, it gives me page after page of colored diagrams, tables, and charts which tell me nothing except what i already know: i dont have any patience for bullshit.

>> No.2032539
File: 22 KB, 600x299, spyder colorimeter.jpg [View same] [iqdb] [saucenao] [google] [report]

>monitor colorimeter

i found a couple at the thrift store for a couple of bucks each, but found them completely useless. i was expecting the software to tell me, ''bring up the red a lil bit, and turn the green down a smidgen." instead, it gives me page after page of colored diagrams, tables, and charts which tell me nothing except what i already know: that i dont have any patience for bullshit.

>> No.2032544

What did you expect, these equipment are intended for pro use.
The average consumer will never have any need for a proper calibrated monitor.

>> No.2032547

AMS has a few.
They also have some fancy color sensor that's a bit hard to get. It's $16 on mouser but some company sells a module with it for $27. Probably the most practical choice for a DIY colorimeter, since the sensor is hard to solder.

>> No.2032577

I'm looking to build a power supply that can operate in constant current or constant voltage modes, automatically if possible.
For example, lets say I want to have a maximum supply of 5V+ at 1A. As long as the current draw is under 1A, the voltage stays at 5V. But if the current pull is too much, it operates in a constant current mode so the voltage will drop to keep the current under 1A at all times.
I can build a simple 5V CV or 1A CC supply with an LM317, but is there a way to build a single unit that will operate both ways, depending on the load? Thanks.

>> No.2032609
File: 43 KB, 691x559, 1586195684494.png [View same] [iqdb] [saucenao] [google] [report]

draw in the voltage source from -3V to ground to make the KVL application more obvious

none, we're all sold as is for scrap here

super hysteresis is better than a latch. imagine one transistor controlled by the comparator output pegging one of the inputs to the appropriate rail, with a momentary NC switch to reset

>bring up the red a little
they're not designed to tell you that. they're designed to tell your OS that. the better ones are designed to also track brightness/color of ambient illumination and coordinate monitor brightness/color in real time

that's exactly how most current limited supplies work already, something like a diode-logic analog AND-gate. pic related is one low-precision implementation, drawn from memory. Q1 is a pnp pass transistor driven by current from Q2 controlled by the opamp. R4/Q3 is a simple current sensor that conducts when Iload * Rsense ≥ 0.6V, arranged here to steal base drive current from Q2 when in the overcurrent condition

>> No.2032610
File: 119 KB, 671x377, It's the electrified truck endorsed by a clown..jpg [View same] [iqdb] [saucenao] [google] [report]

What's the most efficient way to get heat from high voltage AC?

>> No.2032647

>most efficient way to get heat
they're the same picture. 100W of CPU = 100W of nichrome wire, as far as the law of conservation of energy is concerned. getting and keeping that heat where you want it is a mechanical matter
with Ohm's Law you can use tables like https://www.easycalculation.com/engineering/electrical/nichrome-wire-chart.php to choose a gauge and length of nichrome wire. basically, start with a voltage and wattage, figure current I=P/E, then figure required resistance R=E/I, then try out different wire gauges with their particular resistance-per-length values to get the length of that wire required to produce the resistance, then choose the wire and length that best fits within your mechanical constraints
with aliexpress you can order heater assemblies

>> No.2032661

I was just about to specify that I meant something like a space heater or radiant floor heating, and if there was any difference between using low voltage with high amperage and high voltage with lower amperage.

>> No.2032663

If I'm doing my first ardunio without a breadboard is it easier to soldier with or without pins? the prepinned version is the cheapest, and seems like it would be the easiest to soldier directly to the pins, but I know they're supposed to be for breadboarding. my soldiering has improved over the years but I've never done boards before. thoughts?

>> No.2032665

Are you talking about pin headers? Normally you'd use DuPont cables to connect stuff to them in the prototyping phase.

>> No.2032666

amps are what makes the wires glow. resistance makes amp draw higher. also this is probably the least efficient possible way to get heat, i.e. you're not going to get better than a mass produced space heater, there is no way to make heating metal (be it space heater, baseboard, soldiering iron, burnishing tool, foam cutter, or gutter-ice melter) significantly cheaper.

>> No.2032670

Generating heat is one thing, but storing it is another. Research thermal mass.

>> No.2032671
File: 716 KB, 498x208, raw.gif [View same] [iqdb] [saucenao] [google] [report]

That seems like a bit of an oversimplification.

>> No.2032673
File: 195 KB, 1250x568, pins.jpg [View same] [iqdb] [saucenao] [google] [report]

the presoldiered pins version is slightly cheaper. the loose pin version would be used without the pins, directly soldiering wires into the holes.

it's about 6-9 connections, not prototyping anything, nothing fancy being built.

>> No.2032678

and yet for purposes of the discussion it's more that adequate.

>> No.2032679

If you insist.

>> No.2032680

If you know your schematic functions properly then just solder your wires without the pins. If you want to use the board for something else in the future, use the pin headers with jumpers.

>> No.2032697

feel free to point out where it's inadequate based on provided information.

>> No.2032717

Impedance? Reactance?

>> No.2032719

irrelevant to the power supplies, resistance wire sizes, and heat management capabilities likely available to the experimenter

>> No.2032721

now explain how either of those is relevant.

>> No.2032725

The one on the right is higher quality. Look at the difference between components.

>> No.2032736

>if there was any difference between using low voltage with high amperage and high voltage with lower amperage
No. same power

>resistance makes amp draw higher
Lower. Conductance makes it higher
conductance = 1/R

>> No.2032743

I DID say it's a high-voltage AC in my first post, so they are relevant.

>> No.2032766

Everything I'm reading basically just says to use jugs of water painted black to store heat.

>> No.2032771

>super hysteresis is better than a latch
Oh that's a good idea. Is there a good way of making a T-Flip-Flop too?

Also that's an interesting looking dual-mode linear power supply. My designs always require rail-to-rail op-amps due to the current sense resistor being somewhere out of reach, but I haven't thought of using a transistor circuit for it.

A compressor-based heat pump is the most efficient commonly used method. The smaller the difference between the hot and cold side temperatures, the more efficient it runs. Might need a voltage converter or whatever, depending on the type of motor you want to use.
A desiccant-based heat-pump might be more efficient, that's something complicated to look into. They're kinda bulky though, and likely not very good at high temperatures.

>> No.2032784

That would work. Sandbags are good too. Whatever you use should be insulated from the ground.

>> No.2032787

Also, you don't need to paint them black unless they're being heated by infrared light, but insulating them with expanding foam or something similar is best, like a big thermos.

>> No.2032800
File: 46 KB, 523x300, 1610312309566.jpg [View same] [iqdb] [saucenao] [google] [report]

>Is there a good way of making a T-Flip-Flop too?
no good one. you'd need an edge detector, which means capacitors, which gets fiddly. or you'd need a whole master-slave DFF
>Also that's an interesting looking dual-mode linear power supply
it's just a conceptual sketch. I hadn't even had coffee yet when I drew it
>My designs always require rail-to-rail op-amps
otoh my sketch requires a floating input supply because of the load-current-dependent offset introduced by the sense resistor
easier high-side current measurement: use a lower-voltage RRIO op amp (LMV321 or similar are nearly jellybean by now) as a voltage-to-current converter measuring some milliohm sense resistor with respect to V+. power it with picrel, Vdd tied to V+, but Vss tied to a zener and emitter follower

it's hard to beat the specific heat of water

>> No.2032809

>it's hard to beat the specific heat of water
It's less dense so it loses heat faster.

>> No.2032854

>power it with picrel
Why, because RRIO op-amps often have low supply ranges?

>> No.2032870

Whats a good first Bench PSU?
Right now i have a repurposed 5v 1.5A wall wort for powering breadboard prototypes but its getting annoying.
I like to salvage/recover batteries just to see if i can, if it matters.

>> No.2032949
File: 80 KB, 868x702, 1583675946853.png [View same] [iqdb] [saucenao] [google] [report]

the low voltage versions are abundant and usually priced within a factor of 2 of straight non-RRIO LM324 clones. I imagine them being used extensively in mobile devices and IoT crap
I designed a lighting project a while back where the driver IC already used a high-side current sensor, but the input control voltage law was poorly specified, I couldn't drive the driver IC with digital PWM because of flicker, and I wanted to measure and monitor the load voltage and current (and cooling, etc.) with a micro. the venerable picrel circuit was perfect to translate the current reading from 42V world to the low side

if it works, it works. what isn't working? do you want current limiting? (apply it after the fact with a small circuit) better regulation? (get a better wall wart) is it a form factor or connector issue? (build a front panel on your bench with banana jack posts, or pick up one of those breadboard-mounted barrel jack + regulator boards)

>> No.2032953

I considered building one myself but im not entirely sure what it should have.
A wider range of voltages and current limiting are probably the bare minimums.
I suppose I could cannibalize a 24/36v laptop charger and build it up, what magic would I need besides a buck to push down the voltage?

>> No.2032959

What the fuck would you use a WOM for? You really only write on them and cannot read back? Why? What's the fucking purpose?

>> No.2032962

I didn't even have a look at that datasheet. Magic.

>> No.2032965

you'll probably want a current limiter. some buck boards, designed for battery charging, come with adjustable current limiting and overvoltage protection built-in, but are suitable for bench supply use. replace their trim pots with panel pots, and mount the lot in an enclosure with some nice terminals and any meters you like

in woke modernity, someone using company style sheets to produce such a thing would probably be c&. press F to resurrect Bob Widlar's sheep

>> No.2032970

you have a suggestion for a buck board with a limiter?
I can 3D print a case for it easily, ill dig around in my spare charger box and see what volt/amp combos are in there.
Id guess a couple 8-segment displays for power and some banana plug adapters is all else thats needed?

>> No.2032999

you can get little voltmeters like this one that are ready to drop right in https://www.aliexpress.com/item/1005001741247866.html
this one has a current readout too https://www.aliexpress.com/item/32762936294.html
you'll probably also want a barrel jack for the input, and don't forget the front panel pots to replace the trimmers (values depend on model)
or, if you're super lazy and don't want to risk failure https://www.aliexpress.com/item/4001294534288.html

>> No.2033019

Haven’t read the other reply, but just buy a 3-pot LM2596. And buy some full-sized pots, you’ll need both 10k and 100k. Then just put a generic current+voltage panel meter. Total parts cost is well under $10, excluding the case and connectors.

>> No.2033053

Thanks anon, I picked out some parts.
Ill post when its done, assuming I don't burn my house down.

>> No.2033066

>unique SEX process

>> No.2033091 [DELETED] 

Why is voltage constantly across parallel?
For example, if I have a 5V source connected in parallel to the top one having a resistor of 1K and the bottom one having a voltage source of lets say 1V, how come at the end of the parallel, it is 4V?

>> No.2033094

Why is voltage constant across parallel?
For example, if I have a 5V source connected in parallel to the top one having a resistor of 1K and the bottom one having a voltage source of lets say 1V, how come at the end of the parallel, it is 4V?

>> No.2033106
File: 31 KB, 1116x2506, energyBro.png [View same] [iqdb] [saucenao] [google] [report]

a retarded analogy:
how much energy does the ball release if it reaches half the mountain? It will drop from 20 to 10, so it will exert 10 units, even if it drops from 20

>> No.2033160

Uhh, I think it releases 10 units of energy...? Sorry, I basically have like no theory knowledge...

>> No.2033186
File: 9 KB, 380x251, Screenshot 2021-02-16 211548.jpg [View same] [iqdb] [saucenao] [google] [report]

Does anyone know why from the perspective of R equivalent, 6 and 3 are in parallel? I don't get why they're in parallel, but if they are, why not 7 and 6 too?

>> No.2033192

literally just look at it.

>> No.2033195

Does anyone have any resources/advice on graduating from the arduino ecosystem into actually developing with native microcontroller libraries and such in my own IDE?

>> No.2033196

mentally slide 3 left and down then rotate it without disconnecting any nets.
if you can't even see that you're not gonna make it far, good luck.

>> No.2033199

I see it now, thank you, kinda retarded question now that I think about it kek

>> No.2033201

atmel sam using ASF 3 (ASF 4 is hot garbage) or STM32 and stm32cubemx are good ones.
both give full configurable source that lets you make with GCC in any IDE you wish.

>> No.2033205

I'm working with the STM32 all the way. It's just intimidating going from "Press here to upload :))))" to having to do everything myself, I don't even really know where to start. I just know that Arduino is limiting me, and I need the training wheels to come off if I'm to do what I really want to do.

I'll look into what you mentioned. I know the term "HAL" is thrown around a lot too though I don't much know what it is or how to use it, other than it stands for "Hardware Abstraction Layer" and is possibly just an API wearing a different hat

>> No.2033212

HAL is for portability, ie the reason you can use arduino sketches on any compatible platform with a minimal amount of code change for pinouts.
without HAL your code running on micro A setting direct memory mapped registers won't work on micro B because those memory locations aren't registers.
vendors all have their own basic HAL libraries that give basic peripheral usage but beyond that you have to make your own higher level abstractions if you want true portability.
also take a look at RTOS's. learn them early and it makes tackling large projects a whole lot easier.

>> No.2033214

>make with GCC
What the fuck does that even mean
I just want some simple program that takes code I’ve written in some language, had a window full of options to select which MCU I’m programming, what method I’m using, and maybe also choose fuses and stuff, and shoots it fully compiled into my MCU. So long as the functions/libraries/etc. it’s accepting for whatever MCU family isn’t limiting, and it gives me the opportunity to manually manage memory addresses and registers and such, I can’t see why it would be any worse than what you suggested.

>> No.2033217 [DELETED] 

>Why is voltage constant across parallel?

I've spent years asking myself the same thing.

>> No.2033218

take your meds

>> No.2033219

it means you should probably stick with arduino if you want a click n' go programming experience with microcontrollers.

>> No.2033221

Alright that makes a lot more sense. I plan on doing DSP with my STM32s, and I want to be as close to bare metal as possible to have the best performance possible, so learning how to remove as much abstraction as feasible is top priority.
I looked under the hood of some of the libraries I had been using from Arduino and the way they were written it's no fucking wonder my projects ran so slow.

>> No.2033227


>> No.2033230

Yes the HAL is just a high level C library that does the low-level memory mapped register manipulations for you. And the source is available so you can see what's going on.

I don't really have advice about graduating from arduino because I never used its dev tools.
But for what it's worth I didn't have trouble going from PIC/AVR to STM32 ARM. I bought a couple stm32 nucleo dev boards and downloaded their IDE and got a hello-world problem going without any trouble. Then I fiddled around with the port/peripheral/clock configuration tool (cubemx I think) to see what code it generated (and I made lots of backups of my C code as I was doing this in case it got trashed). This gave me good examples of the canonical, manufacturer's way of setting things up. Then I skimmed the HAL library documentation and the data sheet for the particular nucleo board I was using. Also I read the data sheet of the processor when I was working with, and I think I also skimmed thru ARM's documentation for the particular cortex-m core I was working with.
So get used to reading a lot of manufacturer documentation.

I also find it helps to have a specific project goal, something more than just wanting to move away from arduino.

>> No.2033247

I just want something (free) that doesn't force me into some bloaty window with a cluttered toolbar, something outside of my regular programming space. I want to write my code in an external text editor like notepad++ (with appropriate synax addon), then just bang and go. Sure I wouldn't get that line-by-line debug analysis, but I don't think I'd miss it unless I was doing real high-end shit.

The arduino environment itself, while in concept not terrible, has gone too far in order to be the end-all be-all development environment. Though it was probably fatally flawed from the start; even without the bootloader a lot of the default libraries are inefficient.

>> No.2033249

Well he said HF, so I can assume it's somewhere between 3MHz and 30MHz, which is at least somewhat informative.

>> No.2033283

Start with an official IDE then move to your editor of choice

>> No.2033468

mplab X is good.

>> No.2033481

Try arduino-cli. It's arduino without the shitty IDE. I'm pretty sure you can set up a button in notepad++ to run the commands.
>arduino-cli install STM32:stm32
>arduino-cli compile -b STM32:stm32:GenF1
>arduino-cli upload -b STM32:stm32:GenF1
That last one doesn't work for my chinkshit counterfeit STM32s so YMMV.

>> No.2033487

I meant "not microwave," where there are a ton of cool, low loss things you can do with microstrip (with several GHz bandwidth), and not VHF/UHF, meaning an actual 1/4 wave cable would be unreasonably long
Right now I'm looking at 6 feet of coax and I hate it

>> No.2033518

i know exactly what you mean. you want MPLAB's IPE. i write in np++, compile with cli xc8, and upload with that tool.

>> No.2033527

You want MPLAB. But you still need to know what a compiler is you stupid schizo. MCUs and processors do not speak "C" or "python" or "java", they speak in assembly. you need a tool that translates your monkey code into machine code. Microchip has several free tools (they lack the super optimization of the paid ones). For 8-bit mcus you want the one called XC8. (You will have to install it when you start using MPLAB or the program will not work and scream. There are other alternatives such as GCC. Regardless of what you use, you should skim through the documents to see what actually happens when you declare a variable, create a for loop, use structs and what default functions they have.

>> No.2033530

MPLAB IDE is for programming and MPLAB IPE is for cooooding up your chip.

>> No.2033540

>that's exactly how most current limited supplies work already
Where is the load attached in you schematic, the nodes above and below RV1?
I was hoping there was a pre-made IC that would handle the logic from current and voltage sensing that I could build off of, but I wasn't expecting to be so lucky.
Thanks for the explanation, I feel like I'm on the right track, I'm trying to keep the part count to a minimum.

>> No.2033548

Its 2021 you idiots, a GUI IDE is fucking standard. I like writing code/debugging/programming/ect.. from one program that gives me lots of options to rotate code between devices and tools. I don't want to use multiple programs that aren't designed to work together and hunt for problems from a command line.

>> No.2033550

MPLab IDE is for everything and IPE (Integrated Programming Environment) is for programming. If you don't want to use their IDE, the IPE allows you to manage firmware on devices.

>> No.2033554
File: 654 KB, 4738x1873, lm723 power supply.jpg [View same] [iqdb] [saucenao] [google] [report]

>a pre-made IC that would handle the logic from current and voltage sensing

The LM723 is often used for that in an all-analog setup. it's old so there may be newer better chips tho.

>> No.2033557

You didn't join the /diy/ telegram group already? Are you gay or something?

>> No.2033562

yup, also this.
I like programs with user interfaces.
>I want to take traffic off the general
eat shit and die retard.

>> No.2033565

I have misread your ramblings. Mplab and its tools are a GUI IDE you stupid monkey. you still need to choose a compiler you schizoid.

>> No.2033571

>I want to take traffic off the general
Not that anon, the traffic here doesn't look thaaat bad

>> No.2033593

>What the fuck would you use a WOM for? You really only write on them and cannot read back? Why? What's the fucking purpose?
To know it in your bones.

>> No.2033606

I'm not the other schizoid, but you are correct, you need to choose a compiler... if you're writing in C of course. The MPLab installer guides you to their XC series of compilers to download and install, so I didn't specifically add that. MPLAB is a basic starting point, but then you can add on extra features that all work together within the IDE as one big program.

>> No.2033607
File: 475 KB, 4000x3000, IMG_20210216_131539.jpg [View same] [iqdb] [saucenao] [google] [report]

Can someone identify this IC? It must be some kind of buck converter or LED driver. Google isn't very helpful with chinese ICs.

>> No.2033631

Interesting, never see that one before. That is a bit more complex than I was hoping for.
I've seen some schematics of two LM317's, one in CC mode feeding another in CV mode. Apparently this works well, although it burns off a lot of excess power, but I'm not looking for efficiency right now.
I've come up with a design using an INA260 to monitor voltage and current, interfacing it with a microcontroller which in turn can adjust the voltage output from some power supply.
Too many options.
Thanks for the help.

>> No.2033639

Is it in some work/tuck light? Duoxingzhe typically makes those types of things. I can't find any info on it either, but I'd bet money its some constant current led driver.

>> No.2033659
File: 119 KB, 1021x700, Screenshot_20210217_091004.png [View same] [iqdb] [saucenao] [google] [report]

>I was hoping there was a pre-made IC that would handle the logic from current and voltage sensing that I could build off of,
There are a bunch of chips with this topology, I know the MC33063 can be had for like $0.50
Buck, Boost, voltage inverting, etc.

>> No.2033687


>> No.2033715

I have used them and do not recommend them at all. They suck ass greatly. There are better and still cheap and old switching regulators available.
MC33063s are only good for very light loads despite what the datashit tells you.

>> No.2033749

just fold the shit out of it

3x063s are okay all-around switching *controllers*. they're good for learning about switchers (and the importance of proper lead dress) on plastic breadboards, or as one-chip platforms for special applications like Nixie power supplies or LED drivers. being controllers, their internal switch is designed to drive an external switch, not primarily to pass load current. it's good to keep a few of them around just as one would keep a small handful of 555s around

>> No.2033755

Sharing an incredibly satisfying video:


>> No.2033757

I have the weirdest boner right now.

>> No.2033762

i do that with my chink mill all the time for 10x smaller price
literally everything from them is overprices as fuck, they even rip you ass open over copper rivets for vias

>> No.2033768
File: 7 KB, 437x241, Capture.jpg [View same] [iqdb] [saucenao] [google] [report]

Is it safe to throw a capacitor directly on the digital output of a logic gate like this? I want to lowpass the output by using the chip's resistance instead of adding a resistor. It seems to be working.

>> No.2033769

>copper rivets
Can you share a good source for those, anon?

And what model of mill?

>> No.2033776

What chip? If your chip's output resistance is not characterized in the datasheet then it's not a good idea. A capacitor will briefly look like a short circuit and those current spikes can be stressful on parts.

>> No.2033793

In what way do they suck? Not delivering the full 1.5A? Voltage sag?
If it's just ripple, I plan on following them with 317/337's

>> No.2033795

overheat easily, frequency behaviour is not really good (just read the datasheet on how they work), it changes a lot in a non predicable way, too many losses, no short circuit protection (it has over current protection but in a weird way). As I said, it is good for low loads (in relation to datasheet parameters) even with an external switch they deliver meh results
no unless the datasheet says so

>> No.2033799

buy lm2596 with THT packages and go on.
>not delivering full 1.5A
that is also a problem.

>> No.2033801

How does LM2596 compare to XL4005? Pros and cons.

>> No.2033804
File: 66 KB, 975x315, 151591820_792356181490795_231507362008440079_n.jpg [View same] [iqdb] [saucenao] [google] [report]

I think my question doesn't deserve separete thjread so...
can I solder wires from picrel CPU fan to some USB plug and plug into picrel charger? Will it work?

>> No.2033805

USB output is 5V and your fan is 12V. If it runs at all it will be slow and probably stall. Get a boost converter or 5V fan.

>> No.2033807

Nevermind, the charger in your pic says it's rated for 12V 1.5A output so it should work.

>> No.2033808

if you can ensure it outputs 12V, sure

>> No.2033809

that's the point - it should work but idk if it won't broke my charger
I have no idea how charger knows what the output voltage should be

>> No.2033811

LM2596: higher max input voltage, higher min output voltage, lower frequency, higher reference voltage, higher quiescent current, lower switch current, slightly higher drop-out voltages, probably more expensive

the USB plug has to speak whatever quick charge protocol to get the higher voltage. I haven't played with quick charge, maybe the right combination of 1% resistors would do it

>> No.2033817

Yes, USB-PD requires bidirectional communication between the sink and source. I don't think pullups work in this scenario.

>> No.2033840

Is there something like TAOE but for power electronics? I wish it would mention Triacs, Optocouplers and SCR a bit more.

>> No.2033854

Yeah, it's one of those cheapo truck lights. I wanted to see if I could add a dimmer by putting a pot on one of those resistors.

>> No.2033856
File: 66 KB, 689x600, 3457375713.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.2033857

App notes and papers.

>> No.2033859

LEDs are dimmed with PWM rather than a pot.

>> No.2033867
File: 246 KB, 1033x1461, 1612556613689.png [View same] [iqdb] [saucenao] [google] [report]

I did a little googling and, unfortunately, resistors without an active handshake can only ask to increase the current limit of the charger. Qualcomm QC looks a bit simpler than USB-PD and can call to increase the voltage, but requires some handshaking, probably nothing a tiny micro can't help you with

many LED driver chips accept an analog voltage on the dimming input. some are dual-mode and accept either analog or PWM, or even both at the same time

>> No.2033878

>many LED driver chips accept an analog voltage on the dimming input
It can be done that way, but it's less efficient and changes the color of the LED.

>> No.2033887

they might have their own pwm generator internally, as opposed to changing their current reference

>> No.2033899

>I have no idea how charger knows what the output voltage should be
USB power is 5V unless negotiated otherwise by the device.
If you connect something smart like a phone, the charger will start at 5V, the phone will ask what other voltages the charger can do, and request it to output 12V for fast charging or whatever.
Just connecting a dumb device like a fan will guarentee 5V.

>> No.2033945
File: 26 KB, 346x302, 1594254889192.png [View same] [iqdb] [saucenao] [google] [report]

>It can be done that way, but it's less efficient
where? the analog dimming changes the current set point of the driver, which drives more or less current through the LED. luminous efficiency typically decreases as current increases. as for the driver, I don't think any changes in continuous vs. discontinuous mode were implied
>and changes the color of the LED.
plants or B&W cameras wouldn't care about a slightly reduced CRI or a spectral shift. nonetheless it is true, and something to be aware of

>> No.2034021

Don't know if you can really compare them. Right is a Mega32U4, left seems to be a ATTiny84/85 series. The Mega needs more peripheral components.

>> No.2034026

It’s possible that the change in luminous efficacy from the slightly different effective bandgap and whatever any phosphors are doing at lower intensities more than make up for the increase of efficiency. Possible.

Where’s you get that idea? The left one is an ATmega324 (or maybe 164 if it’s a cheaper one). It uses a CH340 of some sort for a USB serial interface (can be used to program it too, so long as it has the bootloader firmware), while the pro micro’s 32U4 has native USB.

>> No.2034029
File: 57 KB, 750x750, 1599526678465.jpg [View same] [iqdb] [saucenao] [google] [report]

yes, the photo on the right is higher quality. have a pocky

the Pro Micro is designed for single-sided stuffing. the bottom of the nano board probably has decoupling caps, crystal load caps, maybe a CH3x0 chip, etc.

>> No.2034113

I don't know how anyone could actively enjoy or want to do VLSI/semiconductor physics/semiconductor design

that kind of person must be grim indeed

>> No.2034116

>Its 2021 you idiots, a GUI IDE is fucking standard.
you think using a GUI in an IDE absolves you from configuring linker scripts or GCC compile parameters.
IDE's are for code debugging, you still have to fucking set it up with whatever micro you're using.

>> No.2034118

visual studio code configured properly (Watch a youtube video) does this. for any micro that supports GCC.

>> No.2034127
File: 31 KB, 979x204, 1613203172443.png [View same] [iqdb] [saucenao] [google] [report]

>I was hoping there was a pre-made IC that would handle the logic from current and voltage sensing
what exactly are you trying to do? You can get a power monitor and read voltage + current over a serial interface

>> No.2034142

I'm trying to design a DCO synthesizer (like the tb303) for fun. I asked last thread about the best way to control the parameters with a digital signal. Originally the plan was to have my microcontroller (or a bunch of daughter mcus) use their PWM to generate a control voltage for the parameters.

All the feedback said to just use a single DAC and a multiplexer. I want to get some clarification.
Does this mean getting a complete. dedicated external DAC? As opposed to just using the PWM output of my stm32? And, is the multiplexer just there so that I can change duty cycles and round-robin cycle through the parameters, pulsing each one with their respective control voltages?

>> No.2034158

Not him, but I think ideally you’d have one that has an analog output. So probably analog internals too. Have it measure current and voltage, and reference the two to some external programming resistors. The output is the max(A,B) of the input signals, intended to be fed to an op-amp that controls the PWM reference, or a linear pass transistor I guess. Because it’s just the close-to-the-rail current sensing and the shitty diode combination of the two references that gets me in the shitter.

>> No.2034231

>a complete. dedicated external DAC
Depends on how good your MCU's DAC is. Most MCUs have no DAC or a shitty one, so it probably means getting an external one. How many bits of resolution do you think you'll need? Do some calculations to see what sort of output voltage resolution is required. If you only need like 8 bits of resolution or less, then PWM might be fine. If you need high resolutions like 12 bits or more, then a DAC is definitely the best way to go, unless your MCU has dedicated hardware for high-resolution high-speed PWM.

>is the multiplexer just there so that I can change duty cycles and round-robin cycle through the parameters, pulsing each one with their respective control voltages?
Have a look at the circuit that was posted, it was basically a large array of parallel sample+hold circuits. The capacitors charge to the analog value when the relevant analog switch is closed, the rest of the time when the switch is open they hold it. The only draining current is out of the capacitor and into the op-amp (high input impedance), and also into the analog switch (should also be high). These input impedances set the minimum frequency at which you refresh these values. The capacitor will also take some amount of time to charge, depending on what kind of circuit is on the output of the DAC. And the analog switch IC will also have some internal impedance, that will slow the capacitor charge time too. This will put a limit on the minimum time spent on each capacitor. But properly designed, the result allows you to use just a single DAC or PWM+filter output (plus some address pins) to create a bunch more analog outputs. At the expense of a lower refresh rate. No need for extra MCUs.

>> No.2034262

you seem to have the general idea right. the STM32F103 doesn't have a true DAC but other STM32 members do. the digital PWM signal must be low-pass filtered to get the analog voltage representing its duty cycle. the choice of filter frequency requires that you balance settling time vs. ripple vs. refresh speed. compared to PWM, a true DAC settles quickly and has no ripple, so you can re-level the entire set of capacitors more often, all else equal
you can work around ripple, sort of, by always switching your mux to the new capacitor at exactly the same time in the PWM cycle. there can be nonlinearity at one or both ends of the range due to carryover from the previous PWM value, if you do not switch at precisely the zero time of the PWM timer

>> No.2034273

>a true DAC settles quickly and has no ripple
You still get glitches on changing (like from 7FFF to 8000), but if he strobes the enable pins of the analog switch ICs with a small delay, he'll be able to avoid that entirely.

>switching your mux to the new capacitor at exactly the same time in the PWM cycle
Sounds kinda shit to me. I'd lean towards multipole filters, be they cascaded increasing-impedance RCs or proper sallen-keys or some combination. The factor between the PWM frequency and the RC corner frequency determines the efficacy in any case. So the more room he has to increase his PWM frequency but keep the multiplexing/refresh rate low, the more viable PWM will be compared to a DAC. So he should do both calculations to see what the minimum required resolution is, and what the resultant ripple would be. That's the sort of thing I'd do in a spreadsheet.

Actually I wonder what a properly tuned notch filter does to PWM. Or maybe an inverse comb filter, to kill the harmonics too.

>> No.2034281

oh shit that makes a lot of sense, the DAC outputs to the sample-and-hold which not only maintains the voltage until the DAC says otherwise, it'll help smooth out the ripple too (I think)

I left it out of my other posts, but I was going to use 2nd order active sallen-key low pass filters to filter the PWM signal.
I'm going to look into standalone DAC ICs now. Any recommendations?
I also still have a spare iCESugar FPGA collecting dust, are they stable and accurate enough to serve as DACs?

>> No.2034291

>2nd order active sallen-key low pass filters to filter the PWM signal
That's an op-amp for each output, which isn't bad, but sallen-keys also have finite input impedances to care about.

>> No.2034319

if he strobes the enables, he's going to get a slightly different voltage into the S&H depending on when in the PWM cycle he does it, even after the PWM post-filter has settled on a stable mean. the lower the post-filter's cutoff frequency, the longer settling time and the slower refresh

you could implement delta-sigma modulation in that iCE at a reasonably high frequency over as many hardware channels as you have pins for. I've seen Verilog/VHDL modules for them. I'd build an I2C/SPI/QSPI/8086 bus interface to update parameters, and include the control for the muxes too, for automatic scanning of up to 256 virtual channels. if you really wanted to get fancy you could do slew rate limitation in hardware, for super-smooth portamento and LFOs and such. those FPGAs have dual-port block RAM, right?

>> No.2034322

>when in the PWM cycle he does it
Strobing was a suggestion to avoid DAC glitches, not for PWMing. While strobing the enable at particular times with PWM could do something towards removing ripple as >>2034262 suggested, but it would make the dc/voltage curve somewhat nonlinear.

>> No.2034407
File: 10 KB, 500x500, pd_trigger_board.jpg [View same] [iqdb] [saucenao] [google] [report]

>maybe the right combination of 1% resistors would do it
nah, that would be too easy for some factory to fuck up trying to cut corners and burn somebody's house down. "pd trigger" appears to be the magic set of keywords to type into your local chinkshit dealership.

>> No.2034492

>you still have to fucking set it up with whatever micro you're using.
No shit you have to set it up.
You can do that within the IDE.
>Stop doing things the easy way! Use the CLI! Use multiple programs!
Stop being such autists and let people enjoy things.

>> No.2034496

man I'm *doing* a signal processing degree and I feel like this is stuff they won't ever teach me in class (granted I'm still doing my base level EE classes but still)
thanks for being patient with me. I feel like a project like this is what I need in order to go from "hobbyist" to "semi-pro". I feel up to the task, in that I think my current knowledge is enough for me to build off what I currently know, to learn what I don't. But that doesn't mean I don't have a shitload of learning ahead of me. Some stuff feels crystal clear, others I have no idea where to start.

I just want to make sure I'm clear before I make any assumptions and bite my ass/incorrigibly sink a project half way through again. Are the "PWM-> sample-and-hold", and "just use a DAC" methods two different approaches I could take? Or is it supposed to be that I feed the DAC into the sample-and hold circuit?
I'm leaning towards the latter, in that all of the units suggested so far feed into one circuit, so that the DAC-set voltage is sustained via the sample-and-hold circuit. Where does the analog switch come into play, though? Or is that just an alternate name for the multiplexer?

I wasn't aware of the finite impedances. What other filtering topologies offer better performance for something like this?

On the one hand, I2C is a lot slower but I wouldn't need to sacrifice pins for SPI chip-select. I'm also not extremely profficient with FPGAs yet, let alone doing home-grown FPGA<->Microcontroller communication, but at least I know it's feasible (and probably preferred considering the gains to be had with delta-sigma, and all the hardware expansion)

>> No.2034497

>what exactly are you trying to do?
On a basic level, I'm making a battery charger.
Yes, a lot of battery charger IC's exist, but I need tight control of current/voltage monitoring because my design is revolving around maximum battery life (over the span of 100's of charges.
Its for multi-cell handling in a larger battery pack for higher voltage.

>> No.2034501

Do any of you guys have a career in electronics? Or is this just a hobby?

>> No.2034504

I'm getting an EE/DSP degree so I'm hoping to make a career out of it. Optimally I would either join an existing firm designing audio electronics or start my own business altogether. The latter route would probably wind up with me bankrupt and my designs all over alebayexpress though

>> No.2034506

>granted I'm still doing my base level EE classes
I've found that the first year of EE classes are a lot of basics like passive components then into semiconductors like diodes and transistors. I didn't even touch microcontrollers until my second year, and it was very basic with small 8-bit mcus. Give it some time, focus on your basics, you'll hit the good stuff soon enough.

>I feel like this is stuff they won't ever teach me in class
I'm going to slap you with some wisdom.
They aren't going to teach you shit in college. College is to teach you a good foundation in your area so you can walk with professionals. A lot of hardcore stuff will be learned outside of college either from your own research or a professional environment.
Hence why the fresh out of college young kids are typically looked down upon when they come strutting in to a new engineering job thinking they own the place with their shiny new degree. They think they can do anything, but end up having to be broken before they realize college doesn't teach you *everything* you need to know.
I earned my EE degree almost 13 years ago, but I've kept researching, studying, experimenting, and applying what I learned from college. Now I've exceeded my professor, and hes been having me teach him some of the new stuff I know.

>> No.2034521

What are you're thoughts on technical certificates?

>> No.2034528
File: 426 KB, 1281x938, 2021-02-18-102830_1281x938_scrot.png [View same] [iqdb] [saucenao] [google] [report]

Another pipe-dream of mine is working a stint at McMurdo Base in Antarctica. My full degree is "Communications and Signal Processing" so I'm maybe they would hire me on to do some sort of satellite stuff.

I'm 23 now and I've been "working" with electronics since I was 18, but only really got serious the past two years or so. I still feel like I don't know shit. But (not to suck myself off or anything) from working with my peers in these classes, I can tell they know nothing of EE and have no passion for it. It kills me but it at least gives me hope that I'll stand out on the applicants list.

These first year classes are torture though. I know enough about basic linear circuits that I feel like I could skip a lot of it. I've taken an "Engineering Analysis with Applications to Embedded Design" class which was "C for Idiots and Engineers", where the only embedded design was XORing the LED register of some mcu in the very last lecture of the class. I learned a lot about C but nothing else (like, for example, how to actually properly program for an embedded system). I'm still 2 semesters off from stuff I actually want to do. I can't even talk to my professors or build a rapport like I really want to, because everything's remote now.

The end result is, even though I think I can tackle a project like this, a lot of the stuff I'm looking into for it seems like its own semester's worth of a rabbit hole that I just don't know about yet. I'm determined but I feel like it'll come crashing down soon enough. I think the process would be a lot smoother and a lot less painful if there was someone knowledgeable to guide me on this, because this project touches on a huge amount of topics and fields (ASP, DSP, Embedded programming, low-noise op-amp design, analog filtering...). I planned on being able to talk about this kind of stuff with my professors, but then covid happened just before I transferred up from community college to start my actual EE degree.

>> No.2034532

>but it at least gives me hope that I'll stand out on the applicants list.
Just make sure to check off the "brown, gay, non-Christian" boxes and you'll get an executive-level position at any corporation you choose. EZPZ

>> No.2034537

>technical certificates
I depends on the cert and who does it. Kinda like the CompTIA A+ cert. At first it was gold. Now its trash.
At the end of the day, its actual knowledge that you can demonstrate and experience with hands-on work that really matter.

>but only really got serious the past two years or so. I still feel like I don't know shit.
You don't know shit. Not being an asshole, but it takes a lot of time and *application* to be able to fit everything together in your head. You can learn what a capacitor does and its textbook definition, but the experience is what will teach you how to use it.

>I know enough about basic linear circuits that I feel like I could skip a lot of it.
You might be able to, but don't get too cocky. There could still be some minor thing you don't know about that is really important.
Not to mention that these classes assume most people don't know very much and they are covering everything to be safe. My first year of college was draining because I did know about 90% of the material already. You gotta do it though, and like I said, there is probably something you don't know which will be really important.

>> No.2034545

I should clarify - it's mainly my current "Linear Circuits 1" course that's killing me. It's just KCL, KVL, Mesh/Node analysis, and DC analysis of inductors and capacitors. It's just my Physics II class all over again.

Other than that, yeah I don't know jack shit and every project I try and do makes me painfully aware of it. I want to get better at the application side of things but I get stuck at roadblocks every step of the way

>> No.2034553

>I want to get better at the application side of things but I get stuck at roadblocks every step of the way
That's a good mindset to have, but you'll have to wait until you learn more to break through some of those roadblocks. Doing your own research will help, but sometimes you have to learn basic concepts of things like DSP before you can really dive into more complex aspects of it.

>> No.2034562

it worked, I'm triggered

above all, a course of study in engineering teaches you how to engineer: how to use engineering data and methods to compose components into a structure that performs a function. the real components and circuits, and their implications, and when and where you can and cannot ignore those implications, are something mostly learned through practice and tearing one's hair out tracing anomalies to their cause
>I feed the DAC into the sample-and hold circuit
this. once you have generated the single control voltage under digital command (DAC or PWM or R/2R ladder even), you then have to route it to a temporary storage location (S&H). two steps to coordinate
>I2C is slower
sure, but is it fast enough? let's go numeric. I estimate you would like to be able to refresh the entire frame of control voltages in about 1/1000s, for really smooth transitions. for that you'd need a DAC that can output 1000*8 samples per second with a reasonable settling time. at 400kbps (2.5µs), that leaves you with 50 bit times (125µs) to send 3 bytes (27 bits including ACKs) over the interface and open the correct sampling gate for long enough for the hold cap to charge. even if the DACs are slow, and you only allow 20% of a sample time to open a 74HC4051 and charge a 1000pf polystyrene cap to 99.9% (7 time constants), it is feasible. a faster DAC or a longer frame time would allow proportionately more samples in a frame

a job at some small software house writing VSTis for the vaporwave world wouldn't be the worst fate

>if there was someone knowledgeable to guide me on this
we have something better at /ohm/: plenty of 3/4-assed folk knowledge AND fully equipped shitposts. you will become an infinitely better engineer with a well-practiced bullshit detector

>> No.2034582

Hey bros, due to a bunch of unfortunate events related to shitty implemented protocols and miscommunication, I have to procure my own lab equipment to use for a month. I won't have time to use aliexpress shipping so are there any suggestions you can give for the equipment you can find on Amazon? As long as they don't cost like 50-100 usd each I should be able to afford it
I need a:
1) multimeter
2) oscilloscope
3) powersupply
4) function generator
5) whatever the fuck you use to measure resistance
Fuck me man.

>> No.2034592

bro you are way too ignorant to be fucking with electronics in any fashion. Are you a student? And your uni is making you procure your own lab equipment?

>> No.2034593

>$100 oscilloscope
RIP in peace b*pis

>> No.2034598

Lol yeah, just starting my lab for the first time for electronics in college. They implemented this cancer ass corona procedure where even if you're corona-free, they won't let you in unless you're on this outdated list they printed a month or something ago.
>And your uni is making you procure your own lab equipment?
Pretty much yeah, either that or I do badly on material that I missed out on-- and then drop. It's not too big of a deal because I'll be using them in my own free time too heh... Shitty college yea.
Damn... as long as it's not $1000 then...

>> No.2034607

maybe you can get this on academic pricing quickly enough https://store.digilentinc.com/analog-discovery-2-100msps-usb-oscilloscope-logic-analyzer-and-variable-power-supply/

>> No.2034613

All resistive heating elements are 100% efficient. A common heating element is nichrome wire. It doesnt matter if you use 20 volts at 1 amp or 1 volt at 20amps, you will still get the same amount of heat, 20 watts.

>> No.2034614

listen, the best thing you could do is hurt yourself trying to do your school work and then you sue the ever loving fuck out of your uni for asking you to do dangerous things.

Anyways, any decent multi-meter will measure resistance, so that takes care of 1, and 5. Do you know what the labs are going to ask you to do? Its hard to recommend a power supply unless we know what you're going to be using it for. It sounds like they want you playing with AC if you need an oscope and a signal generator. But afaik AC lab supplies aren't the most common thing in the world.

Also depending on what the labs want from you, a 2 in 1 oscope and function generator might be good enough and kills two birds with one stone.

If its not too late get a refund and/or take a different class because god damn thats fucking shit and not really worth putting up with IMO They're basically making you pay money out the ass so you can teach yourself, which is entirely fucking beside the point of a god damn university, especially when it comes to shit thats potentially dangerous

>> No.2034617

Velleman LAB-2 maybe.

>> No.2034646

The Technical Certificate is being offered at the community college I am attending. Unfortunately, the program does not teach soldering. They do teach it, it's just not offered in the program... for some reason. I've been having a lot of doubt about this certificate and I was wondering if you could take a look at it and give me your thoughts.

The course description is as follows, "The Electrical/Electronic Fundamentals Certificate Program emphasizes the basic skills needed to begin careers in either the electrical or electronics engineering technology fields. Designed for high school graduates or those entering industry for the first time, the program covers six important areas. These areas include an introduction to electrical/electronic technology; engineering technology techniques; electric circuits; electronic circuits; CAD drawing; and computer systems and applications such as word processing and spreadsheets.

Students who complete this certificate program will be qualified for entry-level positions in industry.

Certificate holders can work as technicians in any area involving electricity and electronics, for example: warehousing and distribution, automation control systems, medical electronics, networks and telephones, power generation and distribution, safety and security, design, production, and maintenance. Five of the six courses (CENT 1310 ,EETC 1313 , EETC 1321 , ENST 1313 and ENTC 1114 ) in the certificate program will transfer to the Electrical Engineering Technology A.A.S. degree (Electrical Design Concentration)."

Here is the link, thanks for your insight anon

>> No.2034648

is this only a semester? (non american here so no idea about what your credits are worth). If yes and you do not have anything else to do then it sounds ok I guess for an introduction. what you showed is like the first semester of a 2-year technical course here in brazil. (you get out of one as a techie and can work as a certified technician)

>> No.2034649

The lab is supposed to 'introduce students to the building, activating and measuring the characteristics of both analog and digital circuits." We'll be going through voltage/current division, superposition, RC circuit (frequency and timed response), and digital clock. We'll also do "waveform measurements using the oscilloscope: P-P voltages, Waveform Freq, DC Offset, Waveform period, simultaneous waveform display" if that helps, I have no idea what it's saying haha. I think we'll only go up to 30V? That's what the triple channel DC power supply goes up to.
Wow, that looks so unique, I didn't even know that this was equipment... I'll give it some more research!
I didn't know there was so many equipment that can do so much at the same time, I'll give this a look too!

>> No.2034655

Two semesters actually. The last three classes will be taken in the fall. BTW, are you >>2034537?
Also, what is brazil actually like? Do all the ladies have really nice rumps?

>> No.2034659

>Also, what is brazil actually like?
Good and bad. cruising through some major historical events that will decide the fate of our fledging democracy. If you are white and middle to high class you are ok I guess. If you are poor or black then it is not good at all. most brazilians just wish bolsonero was killed in his stabbing attack back in 2018, things would have been better..
>nice rumps
yep. my girlfriend is 1/2 polish and 1/2 japanese and has fat ass.
>are you the other anon

>two semesters
then it is a bit scarce I guess. don't you have any trade schools or technical courses available? If you work or do another thing then it is understandable that it is given in 2-semesters.
>does not teach soldering
unironically buy kits from aliexpress and a 30W iron at homedepot and practice until you can do a good solder. It is not hard and it is not like welding were you have several types and a lot of shit to learn.

>> No.2034663

you can even buy unsoldered arduino nanos (without pins) and then they'd be cheaper, you can solder 30-40 something pins and learn how to use a MCU.

>> No.2034670

it's a trainer designed to be good enough for student labs. the experience can be described as a bit like driving a car with a joystick instead of wheels and pedals. you will learn the traffic laws and how the car moves. you won't develop the feel of hands on knobs, but that will vary somewhat by equipment and anyway that's what /b/ is for

>It is not hard and it is not like welding were you have several types and a lot of shit to learn.
I'll bet your gf's phat bum-bum you have never gotten good at SMD

>> No.2034690

Can someone give me a schematic example of a RC circuit in practical use?
Im trying to wrap my head around how/why they are used and searching for examples brings up a shitload of papers with math that tells me nothing practical.

I need to know WHY things are used before I can get to HOW they work.

>> No.2034694

Passive 1st order low/highpass filter
2nd order sallen key (an active filter w/ op amps)
555 timer's timing circuit

>> No.2034708

Good and bad. cruising through some major historical events that will decide the fate of our fledging democracy. If you are white and middle to high class you are ok I guess. If you are poor or black then it is not good at all. most brazilians just wish bolsonero was killed in his stabbing attack back in 2018, things would have been better..
sonuds rough, anon. I hope things works out for you and your country.
>yep. my girlfriend is 1/2 polish and 1/2 japanese and has fat ass.
>then it is a bit scarce I guess. don't you have any trade schools or technical courses available? If you work or do another thing then it is understandable that it is given in 2-semesters.
There is a trade school but it's very expensive for me right now. And I would rather learn carpentry or welding at that school. I'll definitely ask my professor tomorrow some questions. He was in the navy and that's where he learned electronics, he ought to know something.
>unironically buy kits from aliexpress and a 30W iron at homedepot and practice until you can do a good solder. It is not hard and it is not like welding were you have several types and a lot of shit to learn.
I'll look into it, anon. Thanks

>> No.2034727

Pretty much any relaxation oscillator uses RCs for timing, not just the 555.

>> No.2034736

Yupp. Forgot to include schmitt oscillators, those are based

>> No.2034738

BRtard here. There was a US military/marines training course on electronics that helped me when I was struggling with the basics. I can't seem to remember the name. (It was a long ass thing, with 10 volumes or so). If I find it again I'll post it here.
what do you mean
>I'll bet your gf's phat bum-bum you have never gotten good at SMD
You are right. I only use certified organic and whole THT in my bench. SMD use is a feminine trait.
>I'll look into it, anon. Thanks
Big clive has several videos on how to solder, it is basically ASMR of a boomer guiding you in the shop.

>> No.2034741

It was something like this but older. I guess they changed the name. Every book ever introduces you to electronics with the semiconductor physics. Do not get scared, try to understand it but it is better if you understand the circuit laws than semiconductor physics.

The link has several chapters on everything, you probably do not want all of them. Regardless of what you are studying, the circuit laws, ohm law,Thevenin's equivalence theorem are the most important things to learn.

>> No.2034748
File: 38 KB, 988x768, 1593175060335.png [View same] [iqdb] [saucenao] [google] [report]

try looking at them in the frequency domain. you can analyze the RC low-pass filter as a simple voltage divider whose output is a frequency-dependent fraction of the input. the higher the frequency, the smaller the impedance of the C, and the more energy of that frequency gets sent to ground instead of the output
the why of any particular RC low-pass filter is dependent on the whys of the circuit around it. for a sampling application, you need to filter out high-frequency energy that create artifacts in the samples (Nyquist criterion, for those playing along at home). for a communications application, you might want to smooth off high frequencies before sending them over a long cable (= antenna) or, at the receiving end, sink any that might have been coupled onto the wire (= antenna). in a digital audio playback application, you will want to smooth the sharp steps of your output DAC to avoid generating inaudible ultrasonic harmonics that cloud up the high end
inversely, the RC high-pass filter is often deployed specifically to block dc, the lowest frequency energy. one practical example is the extraction of audio signal from a powered electret microphone, pic. the R supplies current to the class-A preamp inside the microphone capsule. the C couples the ac portions of the signal to the output. the value of C must be selected according to the RC filter equation so as not to lose too much audible low-frequency content

some anons link to the NEETS course for obvious reasons

>SMD use is a feminine trait
we're reaching boomer levels that shouldn't even be possible

>> No.2034750

Thanks, anon. I'm pretty worried if my Technical certificate is going to mean anything.

>> No.2034762
File: 45 KB, 988x768, 1604249134201.png [View same] [iqdb] [saucenao] [google] [report]

oops, my bad, the RC filter I wanted to show in the electret mic example is actually composed of the C and Rl, the load resistor on the output that establishes the dc level, that wasn't shown in the original drawing. pic
properly considered, the bias resistor Rb is part of the input impedance and placed in series with C. the resulting transfer function becomes Vout = Vin * (Rl/(Rb+Xc+Rl)), but since Rb is typically small (5k) compared to Rl (50k-100k) and the exact corner frequency is not particularly critical, it can be neglected in practice in this application

>> No.2034772

Im gonna play with this more in falstad.
I think theres just something fundamental im not getting.
I get that its a "filter" but without some kind of example showing with/without filtering I just cant "see" it.

Ill try with the 555 timers, that looks to be the simplest.
Generally the way I learn is looking at an example, building it somehow and then changing values/components to see the differences and figuring it out that way.
Im from a mechanical background, shoehorning abstract concepts into a tangible example is painful to say the least.

I do appreciate the help though.

>> No.2034861

good strategy. even simpler, in the circuits menu, there are RC filter examples that sweep the frequency of a sine wave source into an RC filter. one for LP, one for HP. each example is outfitted with two scopes, one on the input, one on the output. you can watch the amplitude of the output change as the frequency varies. try changing the source to a fixed frequency, if you want to get down to numbers. try a different waveform and turn on the spectrum analysis in the scope to develop a feel for the relations between the frequency domain and the time domain

>> No.2034875

>Are the "PWM-> sample-and-hold", and "just use a DAC" methods two different approaches I could take
Sample+hold is the multiplexing method. A way of turning one high refresh-rate analog output into multiple lower refresh-rate analog outputs.
DAC or PWM+filter are the methods of creating the analog value in the first place. For both, you could either have multiple outputs (gets kinda expensive with DACs) or just one output being fed into a multiplexing circuit. If you do go for PWM, you likely don't need to bother with an analog multiplexing circuit.

>I wasn't aware of the finite impedances
>What other filtering topologies offer better performance for something like this
Well to have a high input impedance, and low output impedance, with just one op-amp, you can only have one pole of a filter. Basically an op-amp integrator/differentiator with an extra gain-limiting resistor across the cap. To have both those impedances and two poles, you'd need two op-amps, basically a buffer before a sallen-key. Could also do three poles with two op-amps, with an integrator/differentiator preceding a sallen-key.

That said, it's not like you require the extremes of op-amp input and output impedances to interface with surrounding circuits or chain filters together. In the RF domain, it's common to chain multiple poles of passive RC (or LC) filters together. Say, for rounding a square wave into a sine. An example would be chaining a 1k/1µF, 10k/100nF, 100k/10nF together for three poles at the same point. If you use it as a feedback network to a non-inverting op-amp, then that's three poles with ideal impedances. Likely won't behave as well as a sallen-key though.

I'm just some physics guy so don't take my word for it, do some simulations or breadboarding.

>> No.2034881

>I get that its a "filter" but without some kind of example showing with/without filtering I just cant "see" it.
go to falstad, plot input, plot output. badabing badaboom

>> No.2034888

>badabing badaboom
you weren't supposed to finish it

>> No.2034891

If you really wanna blow your mind check out some negative resistance circuits and oscillators. That's not a joke, they're real.

>> No.2034892

what did I do wrong here? these LM386 chips do not fit this breadboard at all, I didn't even consider something like this before buying. Total noob obviously but how do i buy the right one?

>> No.2034894
File: 1.59 MB, 3264x2448, IMG_0515.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.2034895

You buy the one that comes in a DIP-8 package. You bought an SOIC (surface mount) part.

>> No.2034896

Solder them on an appropriate breakout board.

>> No.2034898

That's a SMT package, dood. You're looking for a THT package (DIP), and pay attention to pin pitch.

>> No.2034900

see this fella
>SMD use is a feminine trait.

>> No.2034902
File: 2.08 MB, 4032x3024, distortion soldered.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.2034903
File: 60 KB, 1000x750, 1599557768436.jpg [View same] [iqdb] [saucenao] [google] [report]

welcome to the 21st century. git gud at working with these, and whole worlds of interesting components open to you. now you just need some of picrel and some pin headers

chad is secure enough in his sexuality that he doesn't need to care

>> No.2034905

thanks for the quick feedback fren, I'll order that now. luckily this stuff is so cheap that mistakes are pretty painless (other than my $50 replacement soldering iron i bought the other day.. ).

so normaly I'd try shopping around, but Amazon is the king of quick delivery, so how does this look?:

it says DIP-8, looks like it'll work just fine? wish I could just buy 1-2 for a dollar instead of 20 for $5

>> No.2034908

what is this thing? lol

i just want to into breadboarding and soldering so i can into DIY synths and other cool bleepbloppers

>> No.2034910

>how does this look?
That will fit. You'll have to bend the pins inward a bit (normal for new DIP ICs) so be careful.

>> No.2034916

cool thanks bud! the world of electronics is very foreign to me. im just following the projects in Nicholas Collins HandMade Electronic Music and was gonna follow the youtube playlist below for actual synth biulding:

i built the PSU, but was too scared about shocking myself so i went to the other book as it only deals with batteries. the guy at the electronic store told me i could off myself ezpz with the power supply i was building

>> No.2034921

No problem, fren. Knowledge will come to you over time. Don't try to rush it, just keep reading and asking questions. The electronics store guy gave you good advice. You'll know when you're ready for the dangerous stuff. Also, having an isolation transformer and/or variac will make working with mains powered devices much safer.

>> No.2034935

hahahaha how are you having this much trouble with a bargain bin part lmfao

>> No.2034948

He's new. Cut him some slack.

>> No.2034951

nah I get it.
don't use amazon. their userbase is naive americans, the phrase "caveat emptor" is foreign to them, and it shows in the junk they peddle
At least if you get scammed on ebay or aliexpress, you won't be paying american prices

>> No.2034985

>I was wondering if you could take a look at it and give me your thoughts

This is the anon from earlier.
The Certificate looks like it would be a good way to jump start into a work field without dragging through all the extra coursework for an Associates.
Personally, I think it's garbage and really feels like a community college's attempt at a cash grab from those who don't want to go through a 2-year program. Granted, if you had the cert and applied to a job with others that were straight out of high school with nothing, you would stand out more. Then again, that is pretty much the same for other degrees, it really depends on the job you apply for and other candidates that apply.
As far as simply learning electronics, it looks like a very theory-heavy set of classes and much less hands-on, which IMO is a waste of time because technician jobs are 99% hands on.
If you are working towards the cert, I would keep going and put those courses towards an associates, minimum.

>> No.2035001

I'll second the Digilent Analog Discovery 2.
National Instruments owns Digilent, check this out and contact them, you can get a discount and its the same thing:

>30W iron at homedepot
Don't bother with a fire-starter heat stick "soldering iron". Get an adjustable temperature soldering iron, preferably a digital display so you know what kind of heat you are using, It will make you life easier. My most recent iron only cost $150, so you can get a half-ass one for under $100 that is still much better than shitty home depot garbage.

>> No.2035003

it's an adapter board for SOIC chips to DIP footprints, shown with pin headers in place. the idea is you solder the SOIC onto the board, solder the pins into the board, and plug it into your solderless breadboard or a DIP socket or footprint when you want to use it. protip: plug pin headers into the breadboard, set adapter board on top, tack solder into place, or just solder fully if you don't mind maybe melting your breadboard. you can do much the same retaining and aligning function with pin socket strips or even good ol' 2.54mm shorting jumpers.
audio guys are probably going to be fine to stick with THT for another decade, but after that it's anyone's guess

you certainly could. it doesn't hurt to drop a little money on a well-built commercial power supply if you've any doubts of your own capabilities

>> No.2035018

no one in any of the original replies from >2033548 mention using a CLI, only GCC which is the backend of most IDE's
you must be retarded to automatically assume GCC == CLI

>> No.2035032

>so you know what kind of heat you are using
there are only two heats - solder melts, solder doesn't melt.
A cheapass 60W iron will do fine. "375C" for normal pads, "420C" for large parts/ground planes.

>> No.2035039

>Get an adjustable temperature soldering iron, preferably a digital display so you know what kind of heat you are using
Seconding this, there are digital temp-controlled mains irons for only $30 or so.

>there are only two heats - solder melts, solder doesn't melt.
There's more than that. Tip temperature controls how fast solder melts, but also how fast the flux burns off, and how fast you cook your components. A poorly regulated iron will burn off flux too quickly for some work, and possibly cook delicate parts. Unregulated irons will get too cold during heavy soldering and get too hot between contacts with the work.
Your tips must have pretty shitty thermal conductivity or there's a temp offset somewhere in your station, because that sounds too high even for lead-free solder. I run at 320-340 normally, jumping up to 380 for ground planes and XT60s and stuff. Only when working on an aluminium-backed COB PCB did I ever need to get up to 400C.
Cartridge tip master race.

>> No.2035048

it's a $15 iron, and it works.
I have never needed to replace a tip, ever.

>> No.2035049

i just compile with a batch file desu. debugging embedded code inherently takes time, it's not like you can just spam F5 until it does what you want.

>> No.2035056

>I have never needed to replace a tip, ever.
ever burnt a part, delaminated a trace, or burnt off too much flux?

>> No.2035061

only during desoldering, which is hell on its own

>> No.2035063

get a hot air gun at any cost. it's so fucking convenient. idk if the cheap ones are any good though.

>> No.2035070

the ones that are thermostatically controlled are good, like this one https://www.aliexpress.com/item/32920981752.html
some others have only a PWM dimmer and a temperature readout, not actually thermostatically controlled. a rotary temperature control is the tell. avoid

>> No.2035188

>so how does this look?

just read the reviews:
- "Only 2 out of the 20 worked"
- "4 burnt up right away and the other 5 i used are working well after 2 weeks"

>> No.2035194
File: 301 KB, 1744x835, Untitled.png [View same] [iqdb] [saucenao] [google] [report]

one of the largest PC case manufacturers had to a massive recall because one of their cases shipped with a PCB part that involved a screw touching 12V. how do i avoid these kinds of fuckups when designing my own PCBs, im worried as a noob i will design shock hazards / housefire traps


>> No.2035196

I'd guess ERC rules for via clearance

>> No.2035214
File: 501 KB, 2560x1440, 976310_screenshots_20201125220227_1.jpg [View same] [iqdb] [saucenao] [google] [report]

I live in an RV and get power from my neighbor's garage via 2 150ft extension cords daisy chained 2gether. Around 2-3 other RV's are also getting power in this similar way and one guy has a habit of using power hungry equipment every day in the afternoon. When this happens, every outlet in my RV and my neighbor's garage drops around 10-20 volts in addition to whatever else is already being drawn. This isn't a HUGE problem for me except for that when my voltage drops below 80v my computer shuts off and restarts, and God forbid I try to play a game with my PC and my 3080/5600x starts drawing a lot of juice. Any advice as to how I can get more voltage? Could a UPS potentially fix this?

>> No.2035218

>Could a UPS potentially fix this?

depends on the model. some will trip on brownouts, others only on complete blackouts. also consider the battery capacity. you wanna do the math so it will last longer than the brownout.

alternate strategy is to use parallel extension cords, maybe one dedicated just to the energy hog.

>> No.2035220

So long as the power is from the same source I don't see how a dedicated cord would magically add provide more volts. Also wouldn't a UPS have to be supplied a standard voltage range to function properly? pls correct me if I'm wrong.

>> No.2035225

>magically add provide more volts.

the big reason for voltage drop is that there's a loss due to the resistance of a long wire. if you give the energy hog is own dedicated extension, that drop will NOT occur in the other extension.

>wouldn't a UPS have to be supplied a standard voltage

is there a law that requires that? manufacturer's choose whatever design they want, and one common practice is to offer diff models with diff capabilities at diff price points.

>> No.2035271

I want to build a simple clock with a custom strip led display, any suggestion on how to start?
PIC, Arduino clones, Raspberry Pico?

>> No.2035273

An STM32 board (e.g. blue/blackpill) might be good because it has some internal RTC features.
But it doesn't matter that much what you choose.

You can get cheap RTC modules if you want easy mode. Those usually use I2C and will work with any board.

>> No.2035281
File: 34 KB, 520x390, oh no soot.jpg [View same] [iqdb] [saucenao] [google] [report]

>add usb ports to motorcycle
>mixed up + and -

>> No.2035285 [DELETED] 
File: 29 KB, 1052x628, clock chip.gif [View same] [iqdb] [saucenao] [google] [report]

>PIC, Arduino clones, Raspberry Pico?

how about you use a clock chip to make a clock. you may not be able to find one at Digikey (havent looked) but you can get clock radios from the thrift store for a couple of books, and remove the chip. or keep all the clock electronics, and just redo the display.

>> No.2035286
File: 29 KB, 1052x628, clock chip.gif [View same] [iqdb] [saucenao] [google] [report]

>PIC, Arduino clones, Raspberry Pico?

how about you use a clock chip to make a clock. you may not be able to find one at Digikey (havent looked) but you can get clock radios from the thrift store for a couple of bucks, and remove the chip. or keep all the clock electronics, and just redo the display.

>> No.2035292

Thanks, the RTC module should be good.
Forgot to say, I want to make a custom display, where the numbers are composed by various segments of the strip led. I guess I need to write some kind of driver for each bit.

>> No.2035298

has any one ever found something odd with old or to92 bjts? i noticed yesterday that they seem to be pseudo triacs but the data sheet was bjt and these were bought back in the 90s when nitinol wire pistons were on the market. same company was selling lots of parts and all the diodes came individually packed in their own bags but not the bjts. NTE123ap. i found the old box of parts my dad bought that were never used from when i was young

these npns will pass current either way if base has power but the data sheet doesnt indicate this kind of operation AND im using a 3 volt watch battery (no real current) with my homemmade bjt/ fet tester. its got a sucket and thats for collecter emitter/ source drain junction then you touch either the extra positive or negative. its got 2 leds on it and the bjts can do either direction but only when given power to the base

the n channels i bought recently have a body diode so i dont know if this is a issue with to92 or what . i havent seen it with other to92s i salvage from boards so i dont think this is common

>> No.2035301

>I want to make a clock
>just buy one
take your meds

>> No.2035314

>there are only two heats - solder melts, solder doesn't melt.
You're not wrong, but when heating up a pin/lead to apply solder, you don't want it too hot and damage the component. Most components have maximum temperature for soldering joints so you don't damage the part, or de-solder the internal connection on an IC. Not to mention burning up the flux or the physical PCB.

>get a hot air gun at any cost. it's so fucking convenient. idk if the cheap ones are any good though.
This too, if working with any volume of SMD's on the regular, hot air is so nice. My first station was a cheap digital one from Am*zon, I think it was under $100, but it did its job and I really got my moneys worth. The element finally burnt out so I just got a better station after that.

Lots of checking and double-checking. We've been in a massive PC component demand phase for a while now and component manufactures tend to push for new products as fast as possible. This may have been avoided with extra testing. There also aren't any standards for the specific problem, so NZXT got to discover the problem first hand.

>When this happens, every outlet in my RV and my neighbor's garage drops around 10-20 volts
If the voltage is dropping in your neighbors garage where the power originates, the other guy is pulling too much power and overloading the lines. If its only your outlets, after your extension cord, then you need thicker extension cables.
A UPS is mainly for backup in case you lose all power and need to safely shut stuff down, not keep running equipment. Yes it would be a solution as the UPS is basically a battery being charged by the input, and an inverter that uses the batteries to make its own AC voltage. Keep in mind that some crappy quality ones won't be very stable with power fluctuating a lot.

>> No.2035317

Alright lads, I'm angry. Angry at bulbs. I'm trying to investigate what components are available on the market for making my own. I figure if they don't require to be filled with a noble gas anymore, why the fuck are we still buying them when they burn out in 5 years or so? I'm trying to figure out if the yellow led strips found in current bulbs are available on the market anywhere, and if so for how much.
If that's out of the question, what modules do you recommend to get a decent lumens/price performance? The ws2812b's don't seem like they have the raw output and that they'll be very power hungry, so I'm trying to find something more suitable in the meantime. Trawling through digikey atm but if anyone has suggestions that'd be appreciated.

>> No.2035318
File: 4 KB, 250x250, 1613161214684.jpg [View same] [iqdb] [saucenao] [google] [report]

Anybody know how to get my hands on the motherboard schematics of the lenovo ideapad 100-15ibd? It's my first time searching for schematics, I don't know where to look. I've only looked at the FCC webpage, with little luck. Thanks.

>> No.2035319

You can buy bare LED filaments on ebay or Ali.

>The ws2812b
Do you want color changing bulbs?

>> No.2035320
File: 17 KB, 460x460, 1603493878307.jpg [View same] [iqdb] [saucenao] [google] [report]

Forgot to add an image showing the type of led strip in bulbs I'm referring to. No idea what the proper name for these is.
That's the word I was looking for! Cheers anon.

>> No.2035322

>colour changing
I was potentially looking at that to replace the smart bulbs that get sold, but a regular solid colour is what I'm aiming for to begin with.

>> No.2035327

Search for "led bulb e27".

>> No.2035336

You steal this from a Canberra hackerspace per chance? They had some old shit lying round. If I recall it was from one of the machines in the uni's decommissioned particle accelerator, that may be where your memory came from.

>> No.2035339

Really? That could be an interesting lead.

I didn't nick this, though my uncle (no longer with us) could have taken it home as a "work perk" years ago, he used to work with Hawk Measurement back in the early days. Maybe they were involved with that.

>> No.2035342

Basically my aim with this is to reduce the power going through each led filament to increase the lifespan of the bulb in general, being able to choose how many filaments I have on it etc.

>> No.2035346

A kookaburra stole it.
>laughs in gumdrops
Then he dropped it and a koala picked it up.
>Laughs in Vick's Vapo Rub
The koala got syphilis and caught fire. He traded the bubble memory for a bottle of water. The person who gave him the water? Your uncle.
>laughs in fire sale

>> No.2035351

while its not a t flip flop depending on the implementation a sr latch made from 2 bjts ... which you probably want to control with signals and thats 2 more bjts or fets so total 4 and if you want to read 5 transistors which is way less than 8 or 9 or even 22 that most will tell you to use

example im currently trying to loop a 8 bit shift register with a extra bit making it 9 bits. the sr latch goes in between begining and end. the 8th bit activates the side the serial input will read and when the serial input gets moved to the storage register in the 8 bit serial in parallel out shift register that is picked up by a fet to turn off the side to read. i could make a more complicated D latch like im supposed to or jk flip flop but im not. i could try a priority encoder and set output 2 to be read and bit 1 to shift it to output 1 on which negates output 2 and bit 8 to turn off output 1 path but im not. sr latch using only 2 bjts with 2 fets as switches and a 5th to read the side to read is fine

if you need full functionality this wont work so if your circuit needs a clock on the flip flop its not going to work

>> No.2035355

I know this is a kinda dumb question but I'm hoping someone can point me in the right direction:
Looking at setting up a small off-grid solar array for my office, just a small scale test, so far the plan is:
>1KW of used solar panels
>solar charge controller
>2KW/h+ of deep cycle batteries
>500W pure sine wave inverter (total overkill, I only draw 150W peak, but fuckit)

Basically wanna charge the batteries up during the day, while also powering my shit, and then use the batteries to keep everything alive through the night.
Only really have a couple questions:
1: If I'm using used panels, all originally from the same system so relatively well matched, do I just put all the panels in parallel? For grid tied systems series seems to be the consensus, but to charge up a 12V battery system, parallel seems like what the controllers want, yeah?
2: As for the controller, just find something that matches the combined current at Pmax of all the panels, as well as the type of battery I want?

3: Finally: and here's the really dumb question: If I wanna supplement the power I generate from solar with a bicycle generator, that puts out around 100W, what's the best way to do this? I doubt any charge controllers will happily handle solar and bullshit as an input, what about regulating it to the batterie's bulk charge or float voltage and dumping it in? Would that confuse the solar charger?
If it's too bad I could always just disconnect the solar charger and charge by bike, but bleh. I'd rather something that allows both at once, if need be.

(Also fuck 4: what's the go-to ezpz battery chemistry for deep cycle these days?)

Thanks for all the help, appreciate it!

>> No.2035360

for guitar pedals is it a good idea to use a 9V battery and then regulate it to 5v (or another voltage idk) ?? Guitar forums mostly use unregulated supplies but if I use a 5V regulator I will have a good stable supply regardless of the battery charge status right? Dynamic range does not seem to be a problem because the guitar signal is usually smaller than 1V and it is going to the amplifier regardless.

>> No.2035373

The battery already gives you perfectly smooth DC. Does it really matter if the voltage drops slowly with charge?

>> No.2035390


regulators waste power. you'll be paying for that at a premium.

>> No.2035398

Does anyone know how to troubleshoot a refrigerator board?
I made a thread but no help yet :(

>> No.2035403

Just done some reading, looks like I want an MPPT controller, and a hybrid wind/solar controller at that, and I can just feed literally whatever the fuck I want into the wind plug, whenever I want, and worst case if the battery's full it'll just dump it.
Even looks like these things take in 3-phase from the wind, which means I can feed it raw from my bike generator. Ez pz!

>> No.2035410

Unless you get them from Lenovo directly, you have virtually zero chance of getting them.

>do I just put all the panels in parallel?
So long as the input for your solar charge controller is within the output voltage of your panels, yes, parallel.
Series is used to make higher voltage, so the current is lower. If you have ten 100w panels, at 12v, all ten in parallel makes about 12v at 83-ish amps. That is a lot for a charge controller to handle on its input. All ten panels in series is 120v, but only 8.3 amps. Or two sets in parallel of five panels in series makes 60v at 16.6 amps. Its all part of the design for what you need, vs what a controller can handle.
>As for the controller
MPTT type. As you said, you'll need one that handles the voltage and current levels from your solar panels.
>If I wanna supplement the power I generate from solar with a bicycle generator
There are lots of charge controllers out there, you'd be surprised. Try searching for a "Solar and wind battery charge controller".
Never made a system with a bike, but both systems spin a motor, so it should work in theory.
>battery chemistry
If your system is running full time, lithium is best, but they need specific chargers and are expensive up front. They are the best value due to lifespan and capacity, but you gotta pay up front.
Second best is flooded lead acid, but requires maintenance like watering.
From there you could do sealed AGM, which is most common because you hook them up and forget about it until they die. Cost is pretty good too.

>> No.2035424
File: 239 KB, 1364x1276, 81zyzoLkOOL._AC_SL1500_.jpg [View same] [iqdb] [saucenao] [google] [report]

All I can find is controllers that do 800W wind, 600W solar, whereas fuck it, I want a couple kilowatts solar and whateverthefuck wind.
Kinda tempted to go for +1KW solar, 600w wind, plus whatever else I can feed into it.
Some forums are saying that you can hook multiple charge controllers up to one battery and they'd all be happy, looks like I'd be happier too if it means I can get a bunch of good charge controllers rather than one controller shit at everything, but I'm not sure how multiple controllers would work together, perhaps with them all tricking each other into thinking the battery was charged?

>> No.2035439
File: 825 KB, 1048x677, Screen Shot 2021-02-20 at 01.45.43.png [View same] [iqdb] [saucenao] [google] [report]

So it looks like wind charge controllers are dumb as bricks, just a three phase bridge rectifier, a big fuck-off relay which dumps it all into a resistor if the battery voltage gets too high, and a battery sense controller to switch the relay. Doesn't look like it's doing any charge regulation at all.
So the question still stands: Two inputs, one battery.
>There are lots of charge controllers out there, you'd be surprised.
>Try searching for a "Solar and wind battery charge controller".
You're on the money right here, looks like I just need to find the right one.
That said, now that I know what I'm after, it's a lot easier to filter down to the exact unit I want.
>sealed AGM, which is most common because you hook them up and forget about it until they die. Cost is pretty good too.
With a sales pitch like that I'm not surprised most people go for them.
I'm more than familiar with charging lithium, I was just hoping that there's something a little more robust when it comes to deep discharge available now.
Basically I can either buy something which fully discharges overnight, (let's say 1KWh) and is robust enough to handle that, or something with 5x the capacity which only discharges 20% each night, perhaps not as robust but hopefully cheap enough to compensate.
I wouldn't think anything's THAT cheap though, or that it would last long enough to compensate, so just grabbing some AGM's and flogging them 'till they die probably sounds best.

>> No.2035461
File: 126 KB, 411x769, jumper the relay.jpg [View same] [iqdb] [saucenao] [google] [report]


you say you know how to solder, so run a wire across the relay switch. plug it in, and see if compressor/fan start to work. if the relay is bad (and it's one of the things that tends to die soonest) then both will work. dont let them too long, tho, just a quick peek.

BEFORE you do that, tell the thing to cool hard, and then tap the relay (big black box) with a pencil to see if it will operate. if not, tap with something a little heavier, like the back of a screwdriver. if you cant access the top of the relay, hit it from below by tapping the PCB where it lives.

>> No.2035465


oh, and test continuity of the fuse, of course.

>> No.2035467
File: 143 KB, 976x303, Screen Shot 2021-02-20 at 02.11.48.png [View same] [iqdb] [saucenao] [google] [report]

Alright, at prices like this, sketchiness aside ($1.50/Ah seems standard in Aus), two of these fuckers would get me 3KWh, and if I only use 1KWh overnight that's not gonna wear them out too much, so should be fine.
Given how cheap these are I don't blame people for using them.

>> No.2035478

Digi-Key has $5 first class mail shipping. you could know for certain you're getting the right part, and also order some other bits and pieces for your stock to make it more worthwhile

whichever one you feel comfortable with. there are feature sets that make the best use of each. design the custom LED strip display (consider TPIC6B595/TPIC6C595 or MAX7219/MAX7221, either connects to any micro's SPI port) first, since that will determine what kind of micro you'll need to control it all. the timekeeping part is basically solved by any number of chips, you just have to map the time to segment on/off values and deliver them to the LEDs
the STM32 internal RTC is good enough, but you have to do the calendar in software. if you store seconds-since-1970 in the RTC's counter, you can use localtime() and mktime() to convert to and from a human calendar

>these npns will pass current either way if base has power
NTE123 are replacements for jellybean npn small signal transistors. 2N2222, 2N3904, etc.
transistors can act in both directions. usually the gain in the "other" direction is poor, on the order of 1 or 2, due to the sharpness of the edges of the doped regions in the transistor. see the wikipedia article on transistors to understand why in terms of structure

do you really need the hipster fashion, or is brightness enough? it takes little more than basic wiring, basic series/parallel circuits, and basic electrical safety to assemble LED lamps with a heat sink and a mains-powered driver module. look around for COB (chip-on-board) LEDs and you'll find all kinds of modules and assemblies

>is it a good idea to use a 9V battery and then regulate it to 5v
not really. for anything outside of yet another shitty preamp/fuzz box, you want all the headroom you can get. if you want to create voltages, you would do better to create a -9V rail so you can use cleaner opamps than the LM324 design

>> No.2035483

>do you really need the hipster fashion, or is brightness enough?
Brightness is enough. My main issue is just looking for what can give me the most brightness for the least power. I saw these while looking around and they seemed relatively efficient.

>> No.2035485

Cool clock Ahmed

>> No.2035509

plastic screws

>> No.2035513
File: 32 KB, 1753x758, TunnelDiode.png [View same] [iqdb] [saucenao] [google] [report]

I think I have a good idea of whats happening now, configure the RC to a specific frequency and thats what you get out the other side regardless of what trash is added to the input.
Stupid question, whats the difference between high and low pass, from the falstad graphs I can see it does the same thing except the wave form is deep/shallow based on high/low.

Well aint that some shit.

>> No.2035519

I had already checked the fuse and it was good, I think it’s what carries power over to the transformer and steps it down for the rest of the systems.

I will reinstall it and give the black box a few love taps first, and then try the jumper. That should show conclusively it’s the relay, and that the compressor is operational. That relay I can buy for cheap if that’s the problem.

Thanks for the reply! I’ll post an update

>> No.2035547

Where's the C?
high pass = passes high freq
low pass = passes low freq

>> No.2035573
File: 28 KB, 474x487, 1596523663201.jpg [View same] [iqdb] [saucenao] [google] [report]

to optimize for lumens per watt, look at what the dude lmao growers are using. about a year or two ago, the hotness was the Samsung LM301B, a 1/5W white emitter that delivers up to 220lm/W. emitters just over half as efficient are commonly and cheaply available. if you use those, you'll need to think more about assembly and heat sinking, like metal-core pc boards and reflow ovens and so on
here's a first-result-in-the-search alibay seller that sells chip-on-board modules in a variety of configurations, ready for you to add a driver and an enclosure and whatever thermal management you feel like adding https://sumbulbs01.aliexpress.com/store/group/LED-COB-light-source/5796097_517293722.html

>configure the RC to a specific frequency and thats what you get out the other side
did you turn on the spectrum view? did you try different waveforms? did you even look at the amplitudes? you are not done with the RC lab yet and it is too early to draw conclusions
if nothing else, understand that any function f(x) of period 2*pi can be expressed as the sum of a[n]*sin(n*x+b[n]) for integers n≥1, or intuitively, that any periodic wave can be composed by or decomposed into a fundamental and its harmonics, each of its particular amplitude and phase offset. it's not possible to understand ac circuit analysis without it
low-pass filters pass low frequencies from input to output, high-pass filters pass high frequencies, band-pass filters pass a band of frequencies. what does "pass" mean at the output? usually, the majority of the signal, but maybe you want more or less than that. how much of any particular frequency of interest goes from input to output? enter the math

>> No.2035592

>a few love taps first

there's a step i missed in my instructions: before you tap the relay you should wait at least 10 minutes, coz compressors are never supposed to be turned on/off too quickly, so there's a good chance the lil computer will refuse to turn on the relay until 5-10 minutes after power-up.

>> No.2035599

>reduce the power going through each led filament to increase the lifespan of the bulb in general
Already exists, see big clive's video on the dubai bulbs. Deliberately underdriven for the purpose of power saving and longevity. Actually clive does a lot of "hacking" of existing bulbs to decrease their power consumption.

Well my competition isn't a T-ff out of discrete transistors, but rather a 14-pin 7400 or 4000 series logic chip, or something with a 555 maybe. One comparator + two BJTs and some passives is pushing it a bit, but probably still better if I'm using SMDs.
I don't need two inputs on it, just a clock/toggle input. For clock division like in your case, to turn a pushbutton or IR signal into something more permanent.

>> No.2035620

Hawk does so much it's impossible to say which sort of industry it came from, all I know is that plenty of core memory from the canberra synchrotron ended up floating around when it was decommissioned, uni types and whatnot, everyone grabbed a piece.
Sounds about right

>> No.2035621

>I'll second the Digilent Analog Discovery 2.
Woah... Can it do all of the functions it says at the same time? This sounds too good to be true

>> No.2035624

I'll see if I can find anything similar in local computer museums, then. Thanks.

>> No.2035638
File: 235 KB, 1062x1375, 1588355351743.png [View same] [iqdb] [saucenao] [google] [report]

>but rather a 14-pin 7400 or 4000 series logic chip
bet you can't beat picrel

>> No.2035642

>perfectly smooth DC
It isn't as perfect as you'd think. But it's about as good as a classic like a 7805 will give you.

>to optimize for lumens per watt, look at what the dude lmao growers are using
They don't care about CRI though. Some of them still use those awful magenta lights.


>> No.2035670
File: 290 KB, 757x612, china.png [View same] [iqdb] [saucenao] [google] [report]

is a ipad 4 retina display to DisplayPort still worth it? or are there better alternatives for screen size/resolution for the price, at 25eu per display?
Trying to find whether eDP can be directly connected to DP and there's this chinese company which apparantly sells cables doing just that, but it doesnt sit right with me that this is the only source for a ready made solution.

Also, what is objectively good solder, if possible I dont want to go through soldering tiny pin hell with my awful dollarstore solder.

>> No.2035675

>objectively good solder,

kester brand, sn63pb27, 0.8mm thick or less, in 1 pound roll. #44 is their most affordable model at around $40-$50.

>> No.2035688

>They don't care about CRI though.
they don't, but they can and should. most white LEDs have a big trough at the blue edge of green. contrary to popular belief, plants do use some of the green light energy. the LM301B chips were designed for general lighting, and are available with CRI of 70, 80, or 90
>Some of them still use those awful magenta lights.
who knows why. they're just a gimmick or a cheap lark, not the state of the art anymore

get some small-diameter solder, 0.5mm or less. it will really help you avoid getting too much tin on the work and bridging pins. also get a paste flux that works with your chosen alloy, it makes hand microsoldering even possible
also, if you haven't done microsoldering before, you should probably equip and practice enough so that you can solder 0402 or smaller chip components onto a board successfully

>t.reddit boomer

>> No.2035702

kester's fine, and only like $2 more than dubious chinkshit
i'd only recommend 1lb or 0.5lb ($10-$20) for a beginner though

>> No.2035710

Redpill me on no-lead vs lead solders, guys. I hear a lot about the no-lead ones having problems with tin whiskers and the high temperatures causing issues. On the other hand, lead solders have lead in them, though I've also been told that's only a problem if you go breathing the dust from one of those brass solder sponge things.

>> No.2035714

Lead-free is the choice of homosexuals, so use it if you're a gayboi.

>> No.2035720

whiskers are not a problem for hobbyists. Metallic lead is not easily absorbed by the body, the smoke you see is flux (which is not good to breath so work with a window open or something).

>> No.2035726

my boyfriend and I won't go near unleaded solder with a 10 foot pole

>> No.2035729

if you burn the shit out of your solder, lead oxide might be a problem.
when it gets all crunchy-toothpaste like, is that more lead oxide or tin oxide?

>> No.2035736

I'm a Multicore fanboy myself but have no hate for Kester. it's the
>0.8mm thick
that is suboptimal for the microsoldering job that anon seems to want to do

be bi, use both :^)

>> No.2035765

uh.. never happened to me.

>> No.2035775

>started making an order full of SMDs
>chose 0603
>trying to decide which diode package to use
>saw SOD-123 is barely a millimetre long
>freak out and go for SOD-323 or whatever instead
>look at the size of 0603s
>doubly freak out and remake my entire cart with 0805s instead
>costs an extra $5 or so
I am so disappointed in myself. Hope I won’t regret this.

Also I bought a stereo FM radio modulating IC for whatever reason, should be fun to mess about with. FM radio = low-latency moderate-fidelity long-range wireless audio. BT can suck it.

>> No.2035805

0603s are cake to hand solder, assuming your solder and tip aren't too chunky. and sod-323s fit on an 0603 footprint. if you order anything else you should just get 0603, 0805s are too big for dense designs.

>> No.2035811
File: 88 KB, 905x762, 1610204930438.png [View same] [iqdb] [saucenao] [google] [report]

>SOD-123 barely a millimeter long
you read the print wrong, they're the same size as 1206

>> No.2035812

this is what happens when you let interns do all the technical stuff

>> No.2035820

you can probably blame indian outsourcing. i'm sure there's one white person "reviewing" their work, except he can't review it because 6 hours of his day are spent in other meetings.

>> No.2035840

> because 6 hours of his day are spent in other meetings
I was a senior electronic engineer in a big company and where i was doing most technical stuff. I was receiving orders from the senior engineer and that was it.
3 years later i moved up to a senior position. And holy fuck did i hate it there. There was at least a reunion every 2 days with the project manager a representative of the outsourcing firm who have ZERO knowledge in this field : they were all commercials, managers or whatever the fuck they teach in business schools.
Took the 1st PhD offer that flew into my radar and now im looking into becoming a professor or do more post PhD research.
Fuck big companies, fuck outsourcing and fuck jews

>> No.2035841

>except he can't review it because 6 hours of his day are spent in other meetings.
it hurts how true this is in corporate engineering

>> No.2035842

You need to be careful with AGM though, you can't let it run until it dies unless your inverter has a low-voltage cutout feature. It really depends on the battery and how well its made, but AGM shouldn't ever be discharged below 20% of capacity. So a 100Amp AGM battery will give you no more than 80 Amps. Discharging more than that will hurt the battery.
That is also a maximum as some should only be allowed to hit 50%.

>> No.2035846

>Can it do all of the functions it says at the same time?
It uses Digilents WaveForms software and you can open various "tools" in the GUI to run stuff at the same time
Its free to download and check out. You can configure the amount of on-board memory to give to different peripherals for different levels of performance. But yes, you can do a lot at the same time.
You can also program your own custom tools with the Waveforms software, but I think it uses java as the programming language.

>> No.2035860

Leaded solder is fine for hobby or small-scale commercial use. The only reason not to use it is if you need to sell the product in the EU, where leaded solder is illegal for most uses.

>> No.2035861

>eurocucks can't use LDRs because of minute ammounts of cadmium

>> No.2035889

Spanish, I buy 60Sn38Pb2Cu at a local brick and mortar electronics store. Made in France. It's businesses that are not allowed, ie plumbers.

>> No.2035895

I know they’re not that difficult, I’ve done them before. That’s why I feel ashamed at freaking out over their size. But I don’t make terribly high-density designs anyhow, I’ve got a big library of DIPs and intend on using them with SMD passives and transistors for the near future, so my density is already subpar in most instances. But I have a couple of SOPs and one MSOP, will cross that bridge once I get to it.

Got variety packs (20 of each flavour) of both resistors and ceramic caps. Then got 100 pieces each of common values, so I doubt I’ll run out this year or next.
Also got 100 each of BC807/817 and BSS138. Plus some linregs and other transistors, and some of those neat capacitive touch pad detectors.

In the THT domain, I got 100 IR LEDs and 10 TL1838 38kHz IR receivers. To make remote controls. Also 5 D203S pyroelectric PIR sensors. For some reason.

Mixed the two up.

>> No.2035909

I was talking about LDRs

>> No.2035955

You can't sell finished products (outside of exempt categories) which contain restricted substances (e.g. lead, cadmium). You can make them for personal use. I suspect that use for plumbing is under different regulations (possibly national rather than EU).

Who still uses LDRs in preference to photodiodes? AFAICT, most of the existing uses were hobbyists using 50-year-old schematics or who don't know how to use OTAs.

>> No.2035967

>but it doesnt sit right with me that this is the only source for a ready made solution.
Internally iPad's use eDP, basically just DP with power.
Back in the day you'd buy adaptor boards that'd hook up to the iPads display and give you a DP connector with a power jack, maybe even a built-in HDMI converter, that one's cutting to the chase with DP going in with some spare wires for you to feed power to. Really shouldn't be that hard, and if it's all so cheep these days, what have you got to loose?
Maybe plug it into a shit GPU first, just in case.

>> No.2035968

>you can't let it run until it dies unless your inverter has a low-voltage cutout feature.
That's pretty standard with pure sine wave inverters, but either way I'd throw my own voltage monitoring on it, just so I could pretend the whole system's a UPS/Battery and shut down my computers early if needed. Then cut power to the whole thing a volt before the battery might start sweating, give it time to charge back up.
>but AGM shouldn't ever be discharged below 20% of capacity.
Is there a general rule on roughly what voltage that is for AGM?
>That is also a maximum as some should only be allowed to hit 50%.
Sorry, I'm not sure I understand?
Anyway, thanks for the help, appreciate it!

>> No.2035973

Aren’t perfect, and are more expensive. And possibly less linear than a dual matched vactrol with op-amp feedback especially close to either rail. There are almost certainly some cases where the full isolation of a vactrol is a necessity, could even be a decent form of analog feedback for an isolated SMPS. But OTAs are certainly better at high frequencies.

I don’t think anyone other than blinkies use LDRs instead of photodiodes or phototransistors, especially since every LED in your parts box can act as one.

>what voltage that is for AGM
Look at a common discharge/charge curve, if it has an Ah and V axis it should be easy. If there’s just an A curve you’ll need to integrate or do some inference. Or just look at the datasheet for the specific battery, should have a minimum voltage before it impacts lifespan written on the front or second page.

>> No.2035991

Has anyone ever had success asking companies for schematics for their older products?
I'm expecting a "no" and I'm trying to come up with some way I could ask that might get a "yes".

>> No.2035996

don't ask them directly
call them up and ask to speak to an engineer, and then ask for old schematics

>> No.2035997

Not sure about electronics components suppliers. But one guy on YT (Post Apocalyptic Inventor) asks for info on 50 year old german power tools from their manufacturers and frequently gets the info he's after.

What sort of product are you trying to get a schematic for?

>> No.2036005

>call them up and ask to speak to an engineer, and then ask for old schematics
Ah this sounds like a good idea. Maybe it would get me past the person whose job it is to say no.

>What sort of product are you trying to get a schematic for?
An Exeltech inverter from 1997. It's seriously one of the highest quality products I've ever taken apart. Its failure was unexpected and perplexing. the density of the PCB layout makes it pretty hard to reverse engineer.

>> No.2036008

Post pics. I need the next OP image anyhow.

>> No.2036018
File: 1.79 MB, 2736x1824, exeltech1.jpg [View same] [iqdb] [saucenao] [google] [report]

This is it. Exeltech XP125.
Things worthy of note:
All components surface mount, backside of the board right on the heat sink (and it's mostly just ground planes on the back)
TO-220 packages modified for SMT (TO-263 didn't exist back then)
No microcontroller, all discrete logic
Literally 1/4 the weight of my Samlex SA-150
True sine <1.5% THD
Card-edge connector suggests that some of their larger inverters might just have a bunch of these modules plugged into a backplane

Schematics or not, I'm sure I'll end up fixing this eventually because there aren't many of these around.

>> No.2036070

>All components surface mount, backside of the board right on the heat sink
Oh, interesting. I wonder if they used a thinner PCB because they don't need it to be as structural and it would give them better thermal conductivity.
>TO-220 packages modified for SMT (TO-263 didn't exist back then)
I thought it was D2PAK? Anyhow that's the main reason I buy TO-220 MOSFETs, because they work for both SMT and THT projects.
>No microcontroller, all discrete logic
Now that would be interesting to see a diagram of! If you've got a few minutes, you could write IC/transistor numbers and descriptions on that picture, maybe that alone will give you/us a good idea of how it functions. Draw the isolation slot too. Not that I know the first thing about inverter design, besides just making it a class-D amplifier.
>Card-edge connector
I was thinking that maybe it was just used for calibration or testing or something in the lab before shipping. Since the card-edge connection wouldn't work with those thick wires soldered, and wouldn't carry as much current.

What vintage is it? Those right-angle traces look pre-2000.

>> No.2036078

>I wonder if they used a thinner PCB
Standard 1.6mm.
Same thing, different name.
>IC/transistor numbers
Most of these are TL459 PWM controllers, LM339 comparators, and LM324 opamps. A 14060B ripple counter seems to do some of the sine wave synthesizing.
>the card-edge connection wouldn't work with those thick wires soldered
The power pads are on both sides of the connector, with vias between them. I think they'd carry plenty of curret in a stiff spring contact.
Two of the other edge contacts are wired up, one of them provides a filtered 15v out and the other, I didn't figure out yet.
PCB's dated 1997, and yeah the style of trace routing is totally pre-2000. This thing shows more skillful SMT design than lots of present-day stuff.
Some vias under resistors and traces under silkscreen outlines kinda obfuscate the layout.

It didn't fail in operation, just powered it down and it wouldn't power back up. No fireworks, no current spike, it draws absolutely no current. Plan is to start drawing out the input boost converter circuit and see what's supposed to be bootstrapping it. Hope it's something stupid simple.

>> No.2036092

Same as my kit. A bit old even for 97, though, no? But if they work they work.
>TL459 PWM controllers
I think you meant to say TL594. I've heard that part name a few times, but I'm not too familiar with it. Looks like it might be surprisingly useful, especially with the dual error amplifiers and dead-time control. All with only 46 active transistors, the IC designers back then were true kings of their craft. Should pick up a dozen. While the different output modes are likely very versatile, they don't look too good for driving FETs, not that it's much of an issue with modern FET drivers being small and easy to interface with. Looks like the error amplifier can ride a current sense resistor against one of the rails, which is really good looking.

Likely a good start. There should be some caps as a part of such a circuit that could have gone bad. That or a cooked zener or something. I assume you've already checked for fuses, but there's also the possibility of a thermal fuse wrapped up in something's windings.
Also considering that it's all surface mount, whatever solder is holding the heavy components like inductors, transformers, capacitors, etc. might have cracked. So first off just reflow the high-strain joints and ensure they aren't delaminated.

>more skillful SMT design than lots of present-day stuff
Well they did a very good cramming shit in there, but these days circuits don't need nearly as many passives, even for analog and radio stuff. Everything is far more highly integrated. Plus now we can use 0402s and such instead of whatever that is. 0805?

>> No.2036129

I need a heat conductive adhesion pad.
Anything specific from Aliexpress I should buy?
Thanks frens.

>> No.2036132

whatever you buy, ensure you do the thermal calculations. do they need to be electrically isolating?
also consider thermal epoxies
or just lapping both surfaces or using an indium sheet or thermal grease, which require clamping them together

>> No.2036138
File: 67 KB, 399x351, 1576008378076.jpg [View same] [iqdb] [saucenao] [google] [report]

Nothing fancy currently, just for replacing pads on laptops and similar.
Thanks fren.

>> No.2036151

>Who still uses LDRs in preference to photodiodes?
LDRs have a much nicer curve, you can get useful results feeding the output straight into an ADC. photodiodes have a retarded nonlinear characteristic that saturates if you look at it wrong and take endless tweaking.
Why do synth faggots assume the whole world revolves around them? Such gall to call out others for using obsolete components when your life's work could be replaced with a 10 cent DSP and the sample code from the vendor's SDK.

>> No.2036208

It's a memory... You pass currents according to a certain scheme to change the magnetism on the hysteresis plot, and later you can read the magnetic polarity of the ferrites.

>> No.2036209

I know what it is, I just don't know the model or origin.

>> No.2036211

Anon has IQ problems

>> No.2036220

I'm not sure you can find the model, I don't think it has ever been mass produced or factory-machine produced. Its origin is very probably from a fridge sized computer in the fifties before the age of integrated circuits, something in a lab or for people rich enough.

>> No.2036222

"Model" was probably the wrong word to use in this context.

I've been given some leads, I'll be trying to figure out what, specifically, it came out of.

>> No.2036233

>Is there a general rule on roughly what voltage that is for AGM?
There are charts out there that should show voltage vs. capacity for AGM. Use google images to find one (I'd dig one up by my internet upload speeds are dead right now.)
>Sorry, I'm not sure I understand?
Think of a battery like a gas tank in your car. You fill it up to full and you can drive until its near empty before refilling. AGM batteries don't like being used past a certain point. Like I said earlier, some deep cycle batteries can do down to 20% capacity left (about 1/5 of the gas tank remaining), and others only like going down to 50% capacity (gas tank 1/2 remaining).
Discharging too much from the battery can risk damaging its internals. How much you can discharge exactly should be listed by the manufacturer.
This is another reason why Lithium Ion was such a big deal because they can be used down to 5% of total capacity with no problem.

>> No.2036241
File: 2.35 MB, 4032x3024, 910A80C5-0997-4955-A696-B0F21A066D89.jpg [View same] [iqdb] [saucenao] [google] [report]

So I’m looking and isn’t that the 120v side? If I jumper that it’s essentially not testing the relay at all, but directly supplying AC to the compressor right? I have some low voltage wire, and it’s gonna just burn the wire, but if I jumper the bottom two pins wouldn’t that essentially put the relay “on”? Or am I missing something

>> No.2036242

>retarded nonlinear characteristic that saturates if you look at it wrong and take endless tweaking.
wouldn't you just assume a safe operating range (350mV to 600mV) for diodes, then design your pre/post amps accordingly?

>> No.2036243


yeah, it bypasses the relay and tells you if the compressor/fan are working. if they are working, that gives you a lot of options, because they're the heart of the system. all you need other than those 2 is a thermostat, and you have a working fridge. and mechanical thermostats can be found on old fridges, etc.

the rest of the circuitry is just modern bling. they're nice you dont NEED any of it.

>> No.2036245


and the wire in the pic isnt gonna burn. it's more than adequate.

>> No.2036249

we're not trying to impugn the virtue and honor of your maiden daughter, just want to know who besides synthfags hasn't replaced LDRs in their designs yet, and what the others have replaced them with. when $4 worth of hazardous material, manual assembly, and (important) quality control can be replaced with a $0.75 H11L1 opto and $0.25 of silicon and passives, the $3 saved is ample to deal with the supposedly suboptimal electrical qualities of the arrangement, especially if you were going to feed the output side into a microcontroller anyway

photodiode signals are in the microamp range. muh LDR fag probably can't into current mode

>> No.2036269
File: 55 KB, 256x517, ldr_pricing.png [View same] [iqdb] [saucenao] [google] [report]

>$4 worth of hazardous material
u wot?
Not a photo sensor. You realize they have uses outside of synth faggotry right?
Obviously it can be done, but why would you bother with that pain in the ass when you can just use the superior component to start with? Daddy won't let you since you might swallow it and poison yourself?

>> No.2036276

>5 cent LDRs
don't care, show me your cost breakdown for the assembled and tested isolator, aging curves, etc.

>> No.2036282
File: 429 KB, 1920x800, Blood.Machines.S01E03.Chapter.Three.Tracy.1080p.AMZN.WEB-DL.DDP2.0.H.264-NTG.mkv_20200903_232132.150.jpg [View same] [iqdb] [saucenao] [google] [report]

Do you know a way to reduce coil whine of a power supply?
All I find on the internet is "Change your PSU or get a sound proof computer case"
But my coil whine isn't from a computer but from a plasma lamp
Can I replace the coil for a better one?

>> No.2036331

I guess this threads gonna die so I’ll be posting in >>2034505

I did a test on the relay by putting 12v across the processor side, it makes the relay click and gives continuity across the incoming/outgoing L1/L0 pins. So that leaves the transistor right?

>> No.2036462
File: 185 KB, 1771x1169, relay-bypass-schematic-1.2.png [View same] [iqdb] [saucenao] [google] [report]

I want to make this but latching so it doesn't demand current and the relay will stay put on power down.
My guess is switch the FTR-B4C for a latching and re program the ATTINY to send a pulse instead of a constant current and to power cycle every power on.

Am I correct?

>> No.2036469

Now I HAVE to get this.
>student discount requires you to use your professor's information and his email...
God, I hope they don't contact my professor or it'll be awkward using his information...

>> No.2036470

I try to read every post on this thread even though 99% of the time I don't understand anything, should I keep doing it? I feel like I'm learning a little.

>> No.2036484

browsing 4chan is never productive, stop justifying it to yourself

>> No.2036489
File: 35 KB, 462x430, 1597625798211.jpg [View same] [iqdb] [saucenao] [google] [report]

thanks for the wakeup call, I'm going back to my sedra microelectronics textbook

>> No.2036500
File: 10 KB, 600x228, proxy-image.png [View same] [iqdb] [saucenao] [google] [report]

Got all the materials for 100 vactrols for about $6. As a hobbyist the labour costs have never been an issue, and the toxicity isn't a problem either since it's objectively less of a hazard than leaded solder. Or even lithium ion batteries. OTAs and H11s and mixers are at least 5 times the price. Some mirror-style analog optos might come close, but they require an op-amp each. When someone makes a monolithic mirror-style analog opto+op-amp I'll go for it. Picrel.

Nobody in their right mind should use LDRs for light sensing though. So outside of vactrols in somewhat obscure low-frequency analog electronics, there isn't much of a reason to ever use LDRs.

>Can I replace the coil for a better one?
Yeah, or maybe pot it in something. So long as you respect its thermal characteristics.

Ask a question if you don't understand something. And watch the youtube channel recs.

>> No.2036510

Let me clarify that a mirror-style analog opto like that isn't a replacement for every use of a vactrol, but it's good for a lot of it. Might be a way of making a proper mixer with one.

>> No.2036514
File: 50 KB, 850x431, nand.jpg [View same] [iqdb] [saucenao] [google] [report]

Can someone highlight the actual transistors for me? I don't actually get it from the arrows. Which ones are the P types, and what are their boundaries?

>> No.2036539
File: 1.38 MB, 2320x1740, 20210220_165343.jpg [View same] [iqdb] [saucenao] [google] [report]

>whatever solder is holding the heavy components like inductors, transformers, capacitors, etc. might have cracked
You were close. It was this inductor, provides Vcc to all the other chips. It wasn't a cracked solder joint, one of the wires under the encap actually failed. Was intermittent when I measured it, I got it back into whack just by heating the leads until, I assume, the solder under the encap reflowed.
It's disappointing, though, because I know it's going to fail again. It's gotta be hanging on by a thread.
Must have got knocked while I was moving it around.

Now I've got to decide, do I desolder it from the board again and try to measure its inductance so I can find a replacement? Another thermal shock like that might kill it forever.
I wonder if I could measure it in circuit.

>> No.2036545

That's the main problem of surface-mounting of heavy components.
>one of the wires under the encap
Do you mean one of the traces on the PCB, or one of the magnet wire strands inside the inductor? Or the joint between the two? If it's the inductor, definitely replace it. If it's a trace, do your best to scratch off the solder mask and solder a solid jumper across the crack.

How closed is the core? If it's open, I'd consider replacing all the inductors with toroids anyhow. See if you can fix the replacement inductor to the nearest mounting screw, like by gluing it to a support pillar and using a longer screw through it and into the aluminium back. Longer wires won't be an issue with an inductor, unlike a capacitor.

>> No.2036552

>Do you mean one of the traces on the PCB, or one of the magnet wire strands inside the inductor?
One of the magnet wire strands inside the inductor. Not too far inside, just where it connects to the lead. After I desoldered the inductor, I noticed it making intermittent contact when I pushed down on one of the pads with the probe.
There's no markings on it except those three colored dots (they say 105, maybe?) but I think this is the only off-the-shelf inductor on this board. All the other magnetics here are custom.

>> No.2036579

I'd do your best to measure the inductance before it goes kaput for good.

>> No.2036588

>Look at a common discharge/charge curve
>voltage vs. capacity for AGM
Thanks anons!

>AGM batteries don't like being used past a certain point.
>and others only like going down to 50% capacity (gas tank 1/2 remaining).
Ah, I gotcha, I'd try to keep the discharge as minimal as possible, I know car batteries really crack the shits if you cycle them too hard, I thought deep-cycle AGM would be a little happier, but still, if some are unhappy going down below 50% then I'd try to stay the hell away from ever getting near to that.
I'd like to think a nightly cycle of going down 20% should make them last forever, but I'm guessing if I wanted that I'd really need something fancier.
Still, they're cheap enough that I'm willing to risk it.

>> No.2036590
File: 449 KB, 850x431, 1587041309358.png [View same] [iqdb] [saucenao] [google] [report]

latching relays have two different ON states and two different ways to actuate them. single-coil units are designed to actuate one or the other state according to the polarity of the coil pulse. double-coil units have one coil for each desired state and the user chooses which one to pulse. I recommend a double-coil unit, so that you can simply duplicate the existing coil driver. you will need to repurpose an existing pin, perhaps by removing the muting circuit or the muting jumper and the associated code. you will also not be able to determine the existing state of the bypass unless you somehow monitor the state of the contacts or record each coil pulse to EEPROM, so pulsing on power up is your best option. RIP if your battery dies mid-performance and you can't even play clean

the integrated product would be expensive. that's four dice in the package now, and the wiring and lead frame get more complicated. I mean all you really need is the 5 cent op amp, one resistor to set gain and a $3 dual-photodiode matched opto like LOC110, which also happens to be fast (-3dB near 32kHz) and linear (80ppm). imagine the studio and stage applications

it's been a while since I tried to read a chip, but I think it's like pic. honestly not sure what's going down on the PMOS side

I wouldn't measure an L in a power supply application in-circuit, as it could pump energy through a diode into the Vcc bus and the rest of your circuit becomes an L
does 10x10^5 nanohenries (1mH) sound reasonable for an input filter in that low-current, high-noise situation? it sounds reasonable to me. that looks like a standard SMT inductor base from the olden days, measure it with calipers and see if digikey has the size in their selector list

>> No.2036600

>Ask a question if you don't understand something. And watch the youtube channel recs.
Got it, thanks bro!

>> No.2036604

Honestly if you perform with batteries you deserve it

>> No.2036609

Wait, really? That's so fucking interesting. Why can the nmos and pmos share a common gate?

>> No.2036614

Thanks for all the info on relays, how did people do this without micro controlers? Some kind of flip flop transistor? Might make my project simpler.

>> No.2036624

>four dice in the package
Can the op-amp not be integrated onto the photodiode die? Or do they not use silicon for the photodiode? Also at 3 buckeroos each that's not really in the territory of using them everywhere.

As a snythanon of sorts, I plan on having a control voltage from my vactrol PLL, that I send to near-identical vactrols on VCFs. That way I can have filters that ride up and down along with my frequency. Maybe also some open-loop VCOs for a chorus effect.
My diy modular synth rack pinout is rather complicated, has a bunch of pulse and CV and audio lines all running parallel. Hence why I bought materials for 100 vactrols, among other things.

>> No.2036630

A "T flip-flop" is what you're after. It toggles on a clock edge. You have a few options for making one. Firstly is getting a JK flip-flip IC. A JK is just an SR with T functionality when S and R are both high, best of both worlds. Such logic ICs are 14 or 16 pins, kinda bulky, but it should work. Then there are mini ICs like >>2035638, presumably there's a dedicated T-FF but a D-FF makes a reasonable T by connecting NOT(Q) up to D, with possible disadvantages. Finally, there's probably some decent analog methods using schmitt triggers, though I can't speak for pin-count.

But I have a further question: why use a relay instead of an analog switch? Like a CD4051/4052/4053. If isolation isn't an issue, an analog switch is going to be easier, cheaper, and smaller. If you're only using one or two of the internal switches, you might even be able to use the third switch as logic to make a simple T-FF.

>> No.2036634

it's just a conductor. it only needs to convey an electrical potential to produce an electric field across a channel

mechanical push-on-push-off foot switches. after that I've seen CMOS flip flops, and iirc DOD pedals used a CD4007 triple complementary pair to build their own flip flop and do the output switching

>Can the op-amp not be integrated onto the photodiode die
sure, if you want to throw it all out of whack with photocurrents lel. maybe you can make them exactly symmetric, but I don't think that's precisely possible

>presumably there's a dedicated T-FF
the TFF essentially doesn't exist outside of FPGAs and academia
>possible disadvantages
do tell

>> No.2036640

There aren't weird wave effects from sharing a common conductor?

>> No.2036644

>do tell
Don't know them off the top of my head, just I remember watching a YT video where someone found that a looped D-FF would work for some cases but not others. I think I remember which video it was, FesZ Electronics' guitar pedal base design video 1 or 2. I'll go through it now and see what it was.

>> No.2036651

Ah the reason was it needs a fast edge to trigger on. While he was triggering it on a debouncing capacitive ramp, which was too slow for the feedback-based circuit to work properly. He should have just flipped his debouncing circuit to get a sharp edge where he needed it, but he used a JK instead.
Shouldn't be an issue here at all, so long as you set up the debouncing properly.

>> No.2036663
File: 1.39 MB, 2320x1740, 20210220_195402.jpg [View same] [iqdb] [saucenao] [google] [report]

>does 10x10^5 nanohenries (1mH) sound reasonable for an input filter in that low-current, high-noise situation?
It's not an input filter. It looks like a boost converter topology, and over most of its rated input voltage range it does boost the input voltage to 15v, but it also bucks. it's rated to operate up to 16.5v and is happy up to 17v which is stellar because it'll run on a 4S Li-ion pack.

Does 1mH sound reasonable for that? I'm not sure. The damn thing runs now and I'm disinclined to desolder that inductor again.

>> No.2036672

>why use a relay instead of an analog switch?
I've seen one too many 3PDT fail and they're expensive as hell, my idea was to use a relay + optoisolator with a 35ms muting circuit to avoid the audio pop of the latching relay.

>> No.2036680

no more than they would sharing a wire outside the package. FET action is a local phenomenon

sure, that'll getcha on any flop without a Schmitt clock input, esp if your bypassing game is poor

I don't imagine it's an extremely high frequency switcher, given the vintage, so it's credible

>> No.2036687

>I don't imagine it's an extremely high frequency switcher, given the vintage, so it's credible
I think you're right, 105 in nH for a value of 1mH.

>> No.2036688

What's a decent power supply that won't kill a idiot (me) and has at least two voltage outputs?

>> No.2036696

What are you powering? You could use a laptop PSU brick or ATX PSU with buck/boost converters.

>> No.2036701

Just simple breadboard circuit stuff, at most, about 20V?

>> No.2036705

>the audio pop
I wonder if you could avoid that by sending a capacitive ramp into the 4053 input instead of a proper signal edge.

>> No.2036709

Connect a laptop brick to something like that, or buy a programmable bench PSU.

>> No.2036713

Gotcha, thanks mate

>> No.2036717

You're welcome. You have the option of connecting multiple boost converters to one PSU, and there are cheaper modules that do the same thing if you don't need the LCD display. That ebay module is just an example to get you on the right path. Happy hunting.

>> No.2036727

I'm new to all this electronics stuff so I was watching some videos to familiarize myself with lab equipment and I came across this video:
This guy shorts(?) like fifty times and gets shocked like one million times, am I going to be okay? I feel like I'm going to die two days into using these tools from simple mistakes.

>> No.2036733

>This guy shorts(?) like fifty times and gets shocked like one million times
That's his gimmick for clicks. Just don't be a retard and avoid HV until you're more familiar with the quirks.

>> No.2036736

>That's his gimmick for clicks.
Okay thank god, I thought I was going to have to wear goggles and a full suit or something.
What's that?

>> No.2036747

HV = high voltage. He's playing with mains voltage in the video, which you should also avoid until you have a better understanding of the dangers and how to mitigate them.

>> No.2036756

Sorry, what's considered high? I'm thinking like over 100 volts? And yeah, I'll avoid the hell out of these high ones.

>> No.2036762

Anything over mains potential (110-220V). It gets to a point where corona discharge/arcing through insulation becomes a problem.

>> No.2036768
File: 94 KB, 849x568, 1590314653166.png [View same] [iqdb] [saucenao] [google] [report]

the nice thing about these monolithic FETs is that it is a bit easier to tie the body to the substrate than to the source, like the CD4007 in pic, for more headroom and less distortion esp into low-impedance loads

extra-low voltage has been defined in most commerce to end at 50Vac or 120Vdc. at ELV there is "essentially" no risk of heart fibrillation to a healthy person

>> No.2036787

I've never seen an IC like that. What are they intended to be used for, analog circuits like a JFET? I once saw someone sending an analog signal into a CMOS logic IC for some sort of biasing. Since the three pairs are wired differently, I imagine they have slightly different uses? Also you can apparently use a 74C04 as an op-amp, in case you can find one of those.

>> No.2036792

>Some mirror-style analog optos might come close, but they require an op-amp each.
I've never worked with optos, so I'm pulling this out of my ass.
But couldn't you just scale the input with a resistor divider, then take advantage of the BJT nature of the phototranny to make a high gain common emitter amplifier output?
That'd be like 4 resistors plus the opto

>> No.2036796

around +/-40V I think is when you should be careful

>> No.2036823

anything i should consider before buying something that connects 9v batteries to a breadboard? noob here

>> No.2036828
File: 20 KB, 423x232, idk stuff.png [View same] [iqdb] [saucenao] [google] [report]

the datasheet provides connection diagrams for 3-input NAND and NOR gates, among others. kinda fun to think about the Apollo Guidance Computer in that context. I figure the 4007 designers were pin-limited as much as anything
>74C04 as an op-amp
just an inverting amp, no operations provided. the inverter in the CD4007, on the right in previous pic, is specifically cited as suitable for a crystal oscillator or amplifier. it is similar to unbuffered versions available now in smaller packages and faster families, 74xxx1GU04, which I keep around for the same general reason

the duals are for high linearity applications
sub-unity overall current transfer ratios from side 1 to side 2 are common in optos

>> No.2036834

Wait, what happens if you get an opto with a 3-terminal output, and build a BJT amp circuit with it? Does changing the DC into the LED part of it change the gain of the amp, or just change the output bias? Could just use an opto's phototransistor instead of a collector resistor of a normal CE amp I guess, that's a singly-balanced mixer / VCA with isolated input right there, and pretty cheap too.

>50Vac or 120Vdc
I'd stick to half of that. At least that's when you can no longer feel anything with moderately dry hands.

>pin limited
Then why not use dedicated logic chips instead? They surely had them back in the Apollo days, no? I doubt mass-production costs would have prevented dedicated logic ICs but allowed a jack of all trades but master of none like the 4007.

>> No.2036840

why do most breadboards have two receptacle lines along the power line but mine only have one?

>> No.2036842

Post pics so we cal laugh.
Real talk though they should have three, V+, V-, GND.

>> No.2036849
File: 2.70 MB, 1729x3791, 062E1ACA-E31D-47A8-9182-C7D16ACACB2C.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.2036854

Run all the V+ lines to the top, and all the V- to the bottom. If your IC orientations are correct, that's the natural way to connect them anyhow.

>> No.2036915

>implying manufacturers give a fuck about convenient pinouts

>> No.2036918

Well he's breadboarding with DIPs, so we can assume the parts are somewhat old and popular. Any dual or quad op-amp or comparator does this, any 7400 does this, and a lot of 4000s do this too.
Fuck the 555 though.

>> No.2036946
File: 10 KB, 466x355, 9V battery holder with switch.jpg [View same] [iqdb] [saucenao] [google] [report]

>something that connects 9v batteries to a breadboard?

if you can solder, you can pull the top off a dead 9V battery, and use is as a battery clip, since they're symmetrical.

this is ugly but it's good functionally because you can make a very solid clip this way. many many chink clips you buy are just pathetically weak. instead of plastic, they use cardboard as the base, the soldering is awful, the wires are too thin.

an enclosed battery holder with a switch, as in the pic, is ideal.

>mine only have one?

it forces you to have + on one side and - on the other, which is inconvenient but creates clarity, which helps prevent accidental connections.

>> No.2036949

that makes sense, god those are awful

>> No.2037120
File: 158 KB, 1010x852, 1611470325540.png [View same] [iqdb] [saucenao] [google] [report]

>Then why not use dedicated logic chips instead
CPUs were still being built out of discrete transistors when the CD4000 series came to market. the need filled by this particular chip was a semi-committed transistor array, out of which more complex CMOS logic and signal processing functions could be built semi-discretely, without incurring the speed penalties of unnecessary transistors or multiple stages of gates, right side of pic related. compare the CD4007's connections to a schematic of a CMOS gate and you should see a correspondence right away
you younguns and your demands that everything be cooked to order, baka
>They surely had them back in the Apollo days
>made entirely of 3-input NOR gates
I could imagine the semi-discrete version might have had better properties than a fully monolithic logic gate wrt transient abuse and cosmic ray upsets, but I am just speculating

convenience is in the eye of the beholder. it's also situational

those look pretty old, to put it nicely

>> No.2037142

I'm looking for some schematics for a decent binaural microphone that could use two PANASONIC WM-61A electret capsules, possibly 9v battery powered or whatever would get me lowest possible internal noise/hiss/buzz and could output to a regular stereo jack and be plugged directly into computer for live stereo broadcast.
My googling led me through some odd audiophile sites that left confused about something that I imagine should be pretty straightforward. Any pointers appreciated.

>> No.2037165

>My googling led me through some odd audiophile sites t
Because it's snake oil. Binaural shit is, at best, the beat frequency between two sines, and at worst, random asmr shit.
If you want abstract drones to vegetate out by, try >>>/mu/, they'll hook you up for free. >>>/g/ will recc you a decent set of headphones for less than your next paycheck.
Fantasy is all well and good, but don't spend any money on these cranks.

>> No.2037167

>Binaural shit is
in stereo

>> No.2037175

I was just curious since it does look straightforward at a glance, just hooking up two mono mics together as a pair, but I got to see some circuits and the discussion got really muddy really soon (aforementioned abstract drone territory).
Long story short what I'm after is a cheap mic that doesn't hiss for internet radio and can be used through simple jack interface.
Searching for those led me to binaural recording since it seems low background/own noise is pretty important for that.

>> No.2037196

a microphone is like Italian food: the recipe is simple, but the appeal of the result depends on the details, like ingredient quality and the diligence and attention of the cook. shielding of the finished assembly, the capacitor dielectric, the quality of your phantom power source, the type and length of cable, etc. make a difference to the usability and sound quality of the finished product. even the shape and venting of the enclosure matters a lot
if you need something to use and sound good, buy it, especially if you don't understand everything that's going on

>> No.2037203

the thing is I tried a good amount of mics and sent them back because I found them to be noisier than a 25 year old panasonic lavalier I've been using so far and I'm not looking to buy any more intermediate hardware that could make it sound better or splurge on some professional gear for what I'm doing only as an occasional thing

>> No.2037204

Two ordinary microphone circuits, sharing parts when possible, connected to two microphone inputs is basically all it is. Hardware design aside. But that's a problem in itself, as many microphone inputs on computers and such lack stereo. So look for a stereo sound card, or get a dank I2S ADC and make your own with a native USB MCU like an STM32F4 or G4 or whatever. Making your own PCI/E card is also an option, but I've never looked into it enough to know if it's practical or not.

>> No.2037218

with that much variance you're not going to do better from scratch, and you're only going to demand spoon-feeding on what you can tweak to make it work with your system
why not just use a software noise gate to kill the hiss when you're not talking, as is commonplace in studios everywhere

>dank I2S ADC
why not a dank USB ADC? TI (formerly Burr-Brown) has a ton of good parts

>> No.2037253

Oh they make those? With dual channels? With the MCU method you can do all sorts of digital filtration and oversampling and whatever to get a nicer signal, but the simplicity of a USB ADC might be well worth using instead. Especially if you can set it to a high sample rate and just do all the required DSP on your computer. How do the old Burr-Brown ADCs compare with the most modern I2S ADCs? Mainly in terms of resolution and sample rate, though linearity is important too.

>> No.2037272

Couldn't you use two mic preamps and feed the output into the line input of the sound card?One mic to left channel input, the other to right channel input.

>> No.2037275
File: 80 KB, 739x859, pcm2903c.png [View same] [iqdb] [saucenao] [google] [report]

>Oh they make those? With dual channels?
the PCM29xx range has many options. own mic bias circuitry (a resistor) may be required

that's one way to do it, might improve matters for OP, might not

>> No.2037299

Assuming the sound card has stereo mic inputs.

>> No.2037316

>feed the output into the line input of the sound card
>line input
Use the line input (stereo).

>> No.2037435

Those do look good. Might pick up one or two myself, especially considering I just bought a few 12MHz SMT crystals. Are 12MHz crystals (and using PLLs to boost them up to 96MHz) common for USB peripherals? I had to get these for my CH340Gs.

>> No.2037537
File: 28 KB, 400x500, 51KuoK8baCL._SX398_BO1,204,203,200_.jpg [View same] [iqdb] [saucenao] [google] [report]

What's a good microelectronics textbook?
Everyone tells me that Sedra book is the go-to (see picture) but I'm wondering if there are better options.

>> No.2037555

How do I look for a motor to use as a generator?
Nothing huge, or powerful I just want to play around with seeing how much energy I can store in small flywheels.

>> No.2037561

For starters, don’t use an induction motor. BLDCs are good but need rectification. Alternators too. Alternators have the advantage of being able to change their Kv while they’re running, effectively changing the drive ratio of the system. Brushed DC motors are the cheapest and simplest to use, which is start with as a proof of concept.

>> No.2037735

Siglent 1104
Function Generator:
What else do I need? I just want to make electronics but apparently a oscilloscope is vital to learning and function generators are just helpful in testing.

>> No.2037736

I folded all my BB datasheets into my TI folder a while ago, so I can't talk history very well, other than that they were used on the higher-spec sound cards of the late 1990s and well-reputed. Creative had their own proprietary designs of course

>12MHz (and using PLLs to boost them
quite common
48MHz is the usual multiple but others are not unheard of

>> No.2037765

>my TI folder
You're on a higher level than I. I've just got one folder, though I'm thinking about categorising it like [ADCs, 7400s, op-amps, transistors, etc.]. If only to make it easier to find what I'm looking for.

Those USB audio interfaces are 16 bit and 44.1 (or 48) kHz, which IIRC is CD quality. I don't suppose it's worth going higher?

>> No.2037786

>What else do I need? I just want to make electronics
what kinds of electronics? digital things? analog things? microcontroller things? big things? smol things? fast things? slow things?

oh, it's not comprehensive, just datasheets or databooks I grab and want to keep handy. only about 3500 files, oldest is AVR appnotes from 1999. if I were really on a higher level I would turn Elasticsearch or something loose on the whole documentation repo
hmm, I didn't notice the pleb-tier specs on those bois but I should have. 48k sample rate is probably fine for podcasting but I'd prefer more resolution if there is any to be had, because headroom makes a big difference in post-production. the STM32F1 as I2S-to-USB bridge sounds better now. STM32CubeIDE might even auto-generate the better part of the code, lel

>> No.2037796

>what kinds of electronics? digital things? analog things? microcontroller things? big things? smol things? fast things? slow things?
I'm looking at the roll list and I want to go for a digital clock, a voice changer, led cube, temperature controlled fan, and a useless machine..

>> No.2037812

>just datasheets or databooks I grab and want to keep handy
Same here, I only download one when I use/buy/want to obtain a part, or see a particularly interesting part. Plus a few other technical documents I'm likely to use often.
>only about 3500 files
I don't even have 200 of them, but this is from a mainly analog audio guy. Unlike the digital or power electronics realm where you'll look for more specific and more optimised parts, I'm just using the same ~20 parts that are probably 40 years old. Make that ~30 now that I've jumped to SMD. That said, there are probably plenty of appnotes I should have downloaded, but those can get a little big on the ol' laptop SSD.
You wouldn't happen to have generic logic-series specs, would you? Like what I've named the "74HC CMOS logic specifications" that apply to every 74HC IC. They're kinda hard to search for, and I'd like to have a few more. My 74HC one also has some unrelated datasheets bundled in with it, for whatever reason.

>48k sample rate is probably fine for podcasting but I'd prefer more resolution if there is any to be had
The 2902 datasheet mentions "oversampling digital filter" once, so I wonder if it doesn't have built-in higher ADC sample rates with in order to get a higher effective resolution. But there's no other mention of it. I assume the 2906 is the same, the datasheets looked cut-and-paste. Only difference is probably somewhere in the noise specs or q-current.
Doesn't the F4 have internal DSP capabilities? Might make it less latency or something idk. Better if you want to add any digital filters yourself, if nothing else, with not much of a price difference.

>> No.2037815


Name (leave empty)
Comment (leave empty)
Password [?]Password used for file deletion.