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

Maintenance is complete! We got more disk space.
Become a Patron!

/diy/ - Do-It-Yourself

View post   

[ Toggle deleted replies ]
File: 150 KB, 960x540, 1566121271509.jpg [View same] [iqdb] [saucenao] [google] [report]
1720387 No.1720387 [Reply] [Original] [archived.moe]

overheated thread: >>1715180

0. Electrics ≠ electronics. Appliances/mains/sparky stuff to /qtddtot/ or /sqt/. PC assembly >>>/g/
1. Search web first. Re-read all documentation/data-sheets related to your components/circuits. THEN ask. Show your work.
2. Pics > 1000 words. Post relevant schematic/picture/sketch with all part numbers/values/etc when asking for help. Focus/lighting counts.
2.5. State your skill level if asking an open-ended question.
3. Read posts fully. Solve more problems than you create.
4. /ohm/ is an anonymous, non-smoking general.

>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:

>Principles (by increasing skill level):
Mims III, Getting Started in Electronics
Platt, Make: 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)

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:

>Li+/LiPo batteries
Read this first: http://www.elteconline.com/download/pdf/SAFT-RIC-LI-ION-Safety-Recommendations.pdf
>headphone jack noise
Look up "ground loop isolator".
>I have junk, what do?
Get rid of it.

>> No.1720391

first for unodes

>> No.1720392

>Henry W. Ott, Electromagnetic Compatibility Engineering
Thanks fren. Grabbing it on Libgen as we speak.

>> No.1720393
File: 215 KB, 1062x1375, 1563501022448.png [View same] [iqdb] [saucenao] [google] [report]

this thread's digits brought to you in part by the LM387 low noise dual preamp

>> No.1720404

Should’ve used a pair of 12AX7As. Much better opamps.

>> No.1720414

So I have successfully hoarded over the years
>Digilent electronics explorer
>NI Myrio
>Spartan 3E development kit
>Spartan 3 starter kit
>Spartan 2 digilab
>Various other minor shit
I have zero inspiration rn and I feel bad, give me ideas

>> No.1720416

>I have junk, what do?
fuck, that shit's ancient. it's probably worth more to someone else than it is to you. eBay

>> No.1720531 [DELETED] 


>> No.1720549

Spartan 3's are still useful I guess

>> No.1720570

go make a badass video card with an SPI interface or something

>> No.1720693

post ceramic ics

>> No.1720697
File: 61 KB, 403x307, RUS-IC.jpg [View same] [iqdb] [saucenao] [google] [report]

>the virgin ceramic packaging
>the chad metallic packaging

>> No.1720743

>piece of metallic dust falls on your package
nothing personnel

>> No.1720749
File: 349 KB, 1024x683, 888.jpg [View same] [iqdb] [saucenao] [google] [report]

>real chad epoxi blob package

>> No.1720775

Ok real talk, can't they just laser etch a part ID on top of the blob? Not being able to find a datasheet just kills enthusiasm towards modding anything.

>> No.1720784

Clearly you have never designed something to be mass produced. You are always, ALWAYS looking for ways to cheap out on it. That's your job. Laser etching costs $5k in equipment for the factory plus $0.01 more per part? hell no I'm not paying for that.

>> No.1720787

>doing shit to help modders

>> No.1720837

you have to >>>/g/ o back

>> No.1720839

coffee is for muntzers only

>> No.1720890
File: 199 KB, 800x672, decapped.jpg [View same] [iqdb] [saucenao] [google] [report]

I guess it was a naive dream. What kind of chips do they have under those blobs anyhow? Really streamlined single-purpose circuits, or one-size-fits-all one-time-programmable things?

>> No.1720892

almost anything
possible reasons include: because custom or semi-custom (e.g. mask ROM) chip they don't want to pay for a leadframe and overmold, because high volume means chip-on-board saves $thousands or $millions, because they don't want other chinks finding out what kind of chips they're using or how they got them

>> No.1720893

makes me wonder if there are chip-in-blob power transistors, where the silicon's direct thermal link to the vias underneath it is better than a DPAK would be

>> No.1720898

there's also tempco to consider. power devices need special die attach materials (tempco of expansion* vs. cost) and a tightly controlled die attachment process. it wouldn't do to just spit metal-loaded epoxy on it, smoosh the die on top, and hope for the best, except maybe in small-die situations
* this is one of the remaining RoHS lead exemptions

>> No.1720903

Oh I didn't even consider that. I wonder how much thermal expansion accounts for the thermal characteristics of common semiconductor devices?

>> No.1720950

My xmas lights has some 8 SCRs controlled by a blob. Most likely a no-name chinese uC. Calculators have them so instead of bridging two ports you pay 20$ for the most expensive model. Appliances are dipped in yellow transparent blob because of mechanical vibrations. Reasons and chips vary.

>> No.1720974
File: 8 KB, 416x174, 8KR.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.1720986

Hello need some help.
So my laptop charger was no longer working. So I twisted the wire end which connects to the laptop (DC) and it works. But as soon as I remove my hand it doesn't work.
Thanks to /ohm/ I learnt soldering and basic wiring an year ago. Would it be a good idea to slice the misbehaving wire section and resolder the ends?
Or am I better off getting a new charger?

>> No.1720987

>Would it be a good idea to slice the misbehaving wire section and resolder the ends
Yes, so long as you insulate the splice with heat-shrink tubing.

>> No.1721048

Need some ideas.

I have squeaky floors. The basement ceiling is open and I want to screw the boards down to the floor joists. But I can't figure out exactly where they squeak from below.
I CAN figure out exactly where they squeak from above. The floor thickness is about 1.5 inches.

I want to make a mini directional radio transmitter and receiver pair so I can set the transmitter on the floor above, then go below with the receiver and pinpoint within about 1/4 inch where the transmitter is.

I know how to make radio oscillators and simple diode receivers, but I'm not sure about what kind of antennas to make or what frequency to use. I need miniature beam antennas basically.

Any ideas, or is there something like this that exists already that I can just buy?

>> No.1721076

Lol what the fuck, just watch these and fix your shit without all that nonsense:

>> No.1721081

No, I don't want to punch holes through the top of my floor. I have access to the underside already.

The radio thing would have other uses too like locating the other side of walls.

>> No.1721082

Why don't you make something that beats on the wood? Like a motor or something? Much simpler and practical.

>> No.1721084

>No, I don't want to punch holes through the top of my floor. I have access to the underside already.
You understand that the reason the floor is squeaky is because the subfloor is coming away from the joists (or the flooring from the subfloor). There is nothing you can do from the bottom to remedy the squeaking. If you need to avoid putting holes in wood flooring, you need to take up the flooring—and bear in mind there are already loads of nails through the top of your wood flooring. Don’t believe me? Get a spherical magnet and roll it around the floor. It’ll find every nail.

I mean fuck, even if you could do something from below, you don’t need some retard-complicated RF shit. Just use magnets. Or acoustics. Or (holy shit) another person.

But again, you cannot fix squeaky floors from below. The squeak is caused by the flooring or sub flooring’s attachment to the joists failing, not anything purely structural that you can fix from below. Any fasteners you put in the subfloor from below would be just as likely to penetrate through to the floor above, leaving you with not only holes in the floor, but a Burmese tiger pit of spikes.

>> No.1721097
File: 25 KB, 466x466, magnetic stud finder.jpg [View same] [iqdb] [saucenao] [google] [report]


radio waves are difficult to focus in a small arc. instead, use a big neodymium magnet on top (like the ones from old hard disks) and a magnetic stud finder below. i'll react within 1.5 inches of the magnet, and give you centimeter-level accuracy.

>> No.1721107

cheeky breeky

>> No.1721108

>the virgin delta
>the chad wye

>> No.1721139
File: 48 KB, 427x380, E2C8E7D3-1A49-43EA-81EA-72472373E362.jpg [View same] [iqdb] [saucenao] [google] [report]

Are you saying that you can go from virgin to chad just by applying pic related?

>> No.1721143

Perfect. I stopped at 10||5, close enough to (B).

>> No.1721234
File: 1.95 MB, 322x252, Whaaaaaaat LOL.gif [View same] [iqdb] [saucenao] [google] [report]

regarding that pic, I was telling a guy I worked with that I was building a power supply. he asked what I was using a raspberry pie or arduino?

I'm like you know you can build a power supply without any computers or controllers what-so-ever?

He said he had an electrical engineering degree and he had never heard of such a thing.


>> No.1721247

Probably an associates in science from community college and he happened to take some electronics courses.

I can't imagine anybody with a degree that isn't written on a cocktail napkin being that stupid.

>> No.1721319

many years ago, i had a dodgy laptop charger. At first, i used to shape and wedge the power cable to get it to work, but eventually it stopped working completely. It was clearly broken wire(s) near the barrel jack, so i sliced off all the rubber coating leaving just the exposed metal. Like the other anon suggested, I would highly recommend heatshrink. I only had cheap tape on hand, and it kept coming undone and shorting out the charger. Luckily the charger didnt die, and I used it like this for over 2 years until the laptop finally became unusable.

tldr; cutting and soldering will work, but 100% buy some heatshrink. Alternatively, you could set it in resin (but it would be harder to repair if it breaks again). Also worth noting new chargers are around $10-$15 on ebay, or $1 in a thriftshop.
Goodluck anon!

>> No.1721332
File: 593 KB, 716x1028, SimpleLedCircuit.png [View same] [iqdb] [saucenao] [google] [report]

To be fair, arduino's are so cheap now its almost stupid NOT to be using one. Pic related...

Simple circuit to turn an LED on when the Arduino is on. The green wire feeds 3.3v into D12, and in code I check if it is Logical HIGH (This ensures the LED ONLY turns on when the arduino is powered). I also added a 100k pull down resistor to D12 to make sure it is LOW when the arduino is off.
The code is fairly simple too...Basically
>>Read D12 into a variable
>>If (D12 == 1) {D10 = 1}
>>else {D10 = 0}

Pretty straight forward. I check as fast as i can (1milli second) if the arduino is on. If it is, turn LED on. If its not, turn LED off.

If anyone is interested in the code, im more than happy to share

Another bonus is you only need 1 usb cable to run the whole thing!

>> No.1721341

I just threw up in my mouth a little bit.

>> No.1721375

nice bait

>> No.1721387
File: 76 KB, 617x559, vfc.png [View same] [iqdb] [saucenao] [google] [report]

So I'm pretty sick off messing with this circuit... Anyone know why a V-to-F ckt would not respond to the input voltage, and just stay at its max freq?? My breadboarded ckt and ltspice simulation are both doing the same thing. I notice that the summing point of the circuit will float up or down a few 10s of mV when I adjust 0-5V in, but the 1052 doesn't seem to be adjusting the pnp current at all (and is not moving at all). Oscillating at around 1.2Mhz, I have that ring oscillator feedback network at 5k-50pf.

>> No.1721390

>as fast as I can

>> No.1721395

list all deviations from the circuit as drawn, no matter how small

>> No.1721415

is this a tip, or you want me to post them? The problem is I'm not entirely sure what all the waveforms should look like.

>> No.1721442

>arduino's are so cheap
I can build a linear power supply for nothing. Just using found components or pulls from old equipment, all the power supplies I have a built that way. If I really need a specific voltage I slap a cheap chinesium buck convert on it.

>> No.1721450

do you not value ur time anon? it would probably take a whole day in eagle to design the pcb, around $10 to have them made and shipped (for only 10), and you would need to wait a month. Even when they arrive you need to source components and then assemble it.

You can do all this OR just use an arduino, which already has 5v AND 3.3v pins ready to go. You can also use one stack the digital outputs in series to get higher voltages. Eg, if you need 10v, just set D10 to high, and wire it in series with the 5v pin. you can probably get 0-50v+ doing this.

inb4 >>hurp durp, how do you get 2.5v?
use a resistor as a voltage divider

>> No.1721456

fuck sakes lads, he's got us, better pack it in

>> No.1721473

Why not just use a Vackar oscillator? I've gotten a few MHz bandwidth in a spice simulation before

>> No.1721477

I suspect the charge pump at the bottom is fucky

>> No.1721507

You really should add a sense circuit to check if the LED is functional.
If the primary LED has failed it would turn on a secondary LED notifying you of the fault.
You wouldn't want the Arduino to be up and running and not notifying you of the status.

I'm not much on coding so you'll have to do that part on your own.

>> No.1721515

seems redundant

>> No.1721526

Trying my hand at a manhattan audio circuit. How am I doing so far? I'll scrub off that flux later, pretty sure it's only mildly activated if it's even activated at all.

>> No.1721528
File: 1.19 MB, 3264x2448, don't mind me.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.1721538

>he doesn't know about check check engine light lights

for better shielding, cut some more pieces of copper-clad and solder a box together

>> No.1721551

>a box
I hope I won't need that much shielding, it's only audio frequency.

>> No.1721561

If I want to build a f-MHz Pierce gate crystal oscillator, how do I start looking for an appropriate CMOS inverter? What switching times are acceptable? Is an open-drain style inverter like the SN74HC05 acceptable?

>> No.1721568

choose an unbuffered inverter for the oscillator, then add a buffer gate on the output. here's a part that does it all in one package and some good reading too

>> No.1721588
File: 10 KB, 1024x768, 3PDT switch arrangement.png [View same] [iqdb] [saucenao] [google] [report]

Shit, just realised that I need a 3PDT switch instead of a DPDT switch. I need to swap the order of two components like this, and I'm pretty sure there's no more efficient method. But if someone has a method using only a couple of transistors I'm all ears.

>> No.1721599
File: 31 KB, 800x367, 1546820341097.gif [View same] [iqdb] [saucenao] [google] [report]


>> No.1721600

pretty shitty pinouts there, but I guess I'll consider it

>> No.1721603

yeah, they're not great, and you probably don't really need the built-in level shifters
https://www.ti.com/product/CD4053B is a better fit for the application I think

>> No.1721641
File: 10 KB, 256x289, MSGEQ7.jpg [View same] [iqdb] [saucenao] [google] [report]

i'm trying to parse analog audio using an arduino, a 3.5mm jack input. using a multiplexed equaliser such as MSGEQ7 would be ideal, but no stores around me have it and if I ordered it online it wouldn't arrive in time for what I need.
so i'm left with using arduino's analog input pins directly. has anyone done something like this? first thing i would need to do is shift the signal +2.5v from -2.5v-2.5v to 0-5v, since the arduino can't read negative voltage, right? any tips appreciated

>> No.1721646

>I hope I don't need shiel-BUZZzzzZZZzzzZZZwowowowowowowZZZZzzzzZZZzzzZZZzzzWOWOWOWO
Mains hum is a thing friend. But it's looking pretty nice. mine usually are ugly as ass

>> No.1721659

cmon, all the wires are either close to the ground plane or shielded. Plus I've got decoupling caps and reasonable impedances everywhere.

>> No.1721662

I've been heavily researching analog signal processing and the design of analog synthesizers, and the more I learn, the more I realize how much of a wicked PITA noob trap they are.
I want to start branching out and learning about DSP and digital synthesizers but I have absolutely no idea where to start. What would I be looking for in order to start learning? Search terms? Tools of the trade? Introductory textbooks? etc.

>> No.1721670
File: 17 KB, 531x274, accircuits-acp25.gif [View same] [iqdb] [saucenao] [google] [report]

Hi guise, theory question. Which one is correct?
>a) -(10e^30j) = -10e^30j
>b) -(10e^30j) = -10e^210j
Common sense tells me b), but I'm not sure. Both are polar numbers corresponding to currents in a three phase system.

>> No.1721684

Nvm, just used my calculator, answer is b.

>> No.1721685

There are 10 billion books on dsp and signals in general. Pic one, go to genlib and have fun.What is PITA? (Aside from the bread).

>> No.1721694

Pain in the ass.
As in, designing an analog synthesizer is a massive pain in the ass. Designing a functional one is even more so.

>> No.1721697

Also I know there are trillions of books on DSP & friends, but there's a reason there are - it's a fucking huge field. That's why I wanted specific recommendations, I don't know if there's a specific subfield or group of topics I should look at first. I really don't know where to start, I've only ever focused on analog circuits until now.

>> No.1721700
File: 279 KB, 1500x1644, Simplified_1V-Octave_Oscillator.jpg [View same] [iqdb] [saucenao] [google] [report]

DSP is a PITA too, lol.

I've been seeing a lot of posts about analog synthesizers, both in /ohm/ and in /mcg/. So I looked through the Roland TB-303 schematic, copied the oscillator circuit, and fudged the values to create an oscillator that:
>takes 2-5V input CV, at 1V/Oct
>uses only common, entry level parts: 9V power supply, 5V regulator, E6 series resistors, only one matched transistor pair vs. two
>Could be crammed into an 8x10 square of prototype board
>simple and reliable enough for hobbyists and beginners to make blindly
I won't promise that it's the best oscillator, but if you need a ton of simple 1V/oct oscillators - like, say, for a polysynth - it should do the trick.

To match the transistors, buy a reel of at least 20 of your favorite NPN transistors (2N2222, 2N3904, 2SC945 etc). The simplest way is to use the diode-drop setting on your multimeter; place each transistor on a sheet of paper with tweezers, since body heat will change their values. Test the diode drop between the base and emitter, then write the value below it. In less than 15 minutes, you should have several pairs that show the same drop, to the nearest millivolt (and several triplets, quartets, sextets, etc.). There may be better methods out there, but this is sufficient for the hobbiest.
*The extra TL072 op-amp can be used to buffer the CV input or the saw output. To see how saws can be converted into squares, PWM, triangles, see http://korganalogue.net/korgms/images/service/ms20/circ1.gif near the top right.
*The bias voltage, set by the pot on the left, is usually 6.2-7.0V in my experience. Use multiturn trimmer pots for precision, and to save on space.
*2.75V is standardized to A2, or 110Hz. Likewise, 3.75V is 220Hz, and 4.75V is 440Hz.


>> No.1721701

Is BEAM robotics dead?
All of the stuff online is from 10+ years ago and the only book on the matter is from 2002. Looks like the whole subject got nuked by arduino.

>> No.1721720

Well, any introductory book will do. I usually HATE the ones with 10000 exercises and really highscool-like because I learn more by building and reading theory. Whatever floats your boat I guess

>> No.1721724

Never got started on it because I have no way of doing the mechanical part of the robots :(

>> No.1721729

I must have misunderstood something conceptual. What's the point of frequency domain filters in software? If you need to filter out certain frequencies, and you originally had a time domain signal and you have already converted it to a freq domain, why can't you just drop those bins with f> threshold (low pass) and reconstruct the time signal without those freq bins < ?? Why bother with sinc in freq domain if you can apply a low pass filter in time domain?

>> No.1721780

Because a finite frequency spectrum = infinite time signal

>> No.1721790
File: 37 KB, 658x343, ac-coupling.png [View same] [iqdb] [saucenao] [google] [report]

>first thing i would need to do is shift the signal +2.5v from -2.5v-2.5v to 0-5v,

super simple to do: just AC couple the signal. choose R5=R6 to make the zero-level signal equal +2.5V.

>> No.1721810

Is there a way to make e-bike controller sound like GTO VVVF (https://www.youtube.com/watch?v=Ya7OcohAKQ0) on some trains?

>> No.1721812

You have 30 seconds to explain to me why a VOM isn't better than a DMM.

Protip: You can't.

>> No.1721836
File: 1.79 MB, 2448x3264, IMG_5977.jpg [View same] [iqdb] [saucenao] [google] [report]

Perhaps in a moment of pure insanity/stupidity, I bought another VX machine from ebay, to frankenstein parts from because my unit was in a state of unfixibility... Here's to another week of checking waveforms, belts, and transistor values..

>> No.1721864


so much work to digitize your parent's sex tape. much dedication.

>> No.1721869

research the early digital synthesizers. DX-7 (and the story of its inventor, in case you get any ideas that people are rewarded for invention), later Synclavier, Roland's LA stuff. the web is better reading for this sort of thing, more colorful
now that you have math, the implementation is the relatively boring part but you'll need to think carefully about how not to lose precision as you write your algorithms e.g. through arithmetic division
also look at the SynthEdit community


input impedance

if your mom really needed to schlick it to dear dead dad, you'd be a good son and help her out, wouldn't you?

>> No.1721873

oh I forgot the Ensoniq SQ-80. if you hunt around you can find the original ROM images for the early synths, since they weren't protected by copyright until the mid-1980s or so

>> No.1721875
File: 1.80 MB, 3264x2448, IMG_5981.jpg [View same] [iqdb] [saucenao] [google] [report]

It became an obsession once I was the finish line and took 5 steps back. I was at the final stages, ready to digitize be done for good but when everything got fucked up, i couldnt rest until i finished it properly
This unit seems to be in decent shape aside from a melted belt which is a bit to deal with but I have extra belts. This huge crack might be an issue so I will have to swap that board with my original board.

>> No.1721880
File: 4 KB, 211x239, e09.png [View same] [iqdb] [saucenao] [google] [report]

>input impedance

>> No.1721917
File: 98 KB, 683x538, fill my crack up2.jpg [View same] [iqdb] [saucenao] [google] [report]


a crack like that is no big deal on a single-sided PCB. if you turn the board over you might find a trace or 2 that's broken. just solder a wire across the crack(s).

>> No.1721918


>> No.1721955
File: 1.49 MB, 2448x3264, IMG_5985.jpg [View same] [iqdb] [saucenao] [google] [report]

its a pretty bad crack lol. Thankfully my parts from the first unit can be used until I have some time to fix all these traces.
Is there such a thing as "PCB glue":?

>> No.1721996

You really only need Cin, R6, and R5 to offset the audio to the middle of the ADC range. The opamp and everything to the right of it isn't necessary.

>> No.1722002


just 3 traces are broken. 2 mins to fix.

>you really only need Cin, R6, and R5 to offset the audio

that's so obvious, didnt need to mention it.

>> No.1722015

How? Because I want this gimmick badly.
I don't care if it is not efficient or what.

>> No.1722037

>PCB glue
Those things are made of glass-fibre reinforced thermoset polymer, so epoxy resin should work pretty well. Might want to specifically look for a resin that can handle soldering temperatures, but I don't think it will be

If recording the audio and playing it back is out of the question, you could try putting hand-wound coils in series with some components to generate some coil while and see what that sounds like I guess. Piezo discs in parallel too. Changing the algorithm used by your motor controller is also potentially an option, be it in software or by modifying/replacing hardware. Key places to put the components will be after the H-bridges, but maybe other places too. Take an FFT of the recorded sound if it helps, ignore everything below 20Hz or so.

>> No.1722059

>Might want to specifically look for a resin that can handle soldering temperatures
the standard Tg for FR-4 is 140°C. doubt there would be any problem with some rando epoxy from the hardware store

>> No.1722067

Ok now how am I gonna fit a CD4053 on there? I'm thinking deadbug either in the lower-left or lower-right. It will be controlled by the switch that's currently at the lower-right (that now only needs to be SPST) but it deals with the buffered inputs from the left, the direct outputs to the right, and needs ±6V from the DC jack.

Another problem I may run into is the fact that my DC supply has a grounded negative output, meaning my ground plane is actually at +6V relative to the actual ground. Because I suspect that my laptop power supply is grounded, this may cause a ground loop, but shoving input caps on my input and output audio jacks would be massive pain.
So I guess I have to get an ungrounded power supply.

>> No.1722138
File: 138 KB, 1253x861, 7v motor.jpg [View same] [iqdb] [saucenao] [google] [report]

Here's something interesting, the 7V motor wouldnt spin during playback, I checked the voltages both at the motor leads and at the board, and it read 7.5V, 0.1V and 2.6V. I assume the ~0.1v is the ground. Could it be possible the motor is just seized up? Seems like an odd thing to happen to it.

>> No.1722159
File: 1.06 MB, 959x719, 1572929389469.png [View same] [iqdb] [saucenao] [google] [report]

This is the inside of a psu. What is all that white stuff for? Electrical insulation?

>> No.1722161

I think it's for rigidity; keeps vibration/shock from breaking loose components

>> No.1722171

now don't you wish you'd built a proper panel instead of just welding the switches on? :^)
yeap, that's a problem, unless you want to generate -12V with an ICL7660 or whatever

Silastic® adhesive insulating silicone foam. provides mechanical support and electrical insulation in spots where it's needed

>> No.1722177

It's called silastic, what the other guys said is correct, though it's also used to stop connectors from coming loose and to provide strain relief so wires don't break from work hardening.

>unless you want to generate -12V with an ICL7660 or whatever
Well it's an audio circuit so having switcher on the board could be a bit of a problem noise-wise, plus the CD4053 can only handle a ∆V of 18V maximum. I could use a 5V supply instead, but that would also require buying another supply, which would be pointless in the face of just buying an isolated 7-12V one. Doubt I'll ever use the full ±6V range.
The junk shop nearby has a few chunky 240V-120V transformers that might give me isolation (and be usable for other purposes), but for only a couple of dollars I should be able to pick up a 12V power supply from them instead. Probably a lower-power one too, as this thing isn't going to use anywhere near 1A.

Ideally I'd like to be able to toggle the grounding somehow, but that would require a mod inside before the power supply and three conductors going to my board, so instead I'll just use my oscope's grounded binding post and chuck a ground lead onto my ground plane somewhere.

>> No.1722290

switched-capacitor converters are much quieter than those involving inductors
besides it doesn't need to be symmetrical, you just need to get "some" footroom under ground. witness the legendary CEM3340 and its ability to run on +15/-5V supplies

>> No.1722292

Well I still have two more reasons for you. 1: there's likely not room for another IC on the board as well as the 4053, and 2: I can't get a CEM3340 within two days. I've decided to rush myself to finish a project for once in my life.

>> No.1722296

1. ICL7660 is an 8-pin switched-capacitor voltage generator
2. you won't be needing a CEM3340 or the Latvian clone unless you're designing a high-quality synth, just citing its status as an asymmetrical voltage consumer

>> No.1722306

If the motor doesn't spin then it doesn't generate a lot of voltage drop, most of it goes to that resistor. What it's driving? Maybe something is rusty and the motor stalled.

>> No.1722308

Or some bearing somewhere is fucked and it can't rotate. When dealing with motors of any sort remember:
>Voltage drop proportional to speed
>current is proportional to torque

>> No.1722313

thanks, that seems really simple to do

>> No.1722375
File: 654 KB, 1705x1938, IMG_5989 (Edited).jpg [View same] [iqdb] [saucenao] [google] [report]

The guy did say that tapes havent played in years and one one of the belts was melted and this motor controls playback heads so i guess it could be possible it just seized after all this time. I WOULD replaced it with the working one from my other unit but one the screws is nearly stripped and im scared to go any farther. I have a dremel with a disk head but im worried im going to start slicing up wires because of the tight clearance.
I could try to get a single drip of oil into the shaft WITHOUT getting it on the little belt that runs around it, A flexible pipette would come in handy now lol.

>> No.1722376
File: 867 KB, 3264x2448, IMG_5990.jpg [View same] [iqdb] [saucenao] [google] [report]

If i try to move the head it DOES spin just very clunky and slowly.

>> No.1722382

Instead of the disk head try one of the end mill heads and cut a slot that way. Probably easier to control, if you’re destroying the screw anyway might as well do that.

>> No.1722425

the motor is probably fine, I think things are just too stuck and old for it to start spinning so it just stalls. Those are my two cents anyways

>> No.1722426

I mean, you can always take the motor out and test it if it's worth all the trouble

>> No.1722436

Are there any USB mini connectors that can handle around 14V and 6-7W?
Google only gives me results for 12V car adapters.

>> No.1722440

I would take it out but see >>1722375, theres a stripped screw preventing me from removing the solenoid (which is in front of the motor screws)

>> No.1722464
File: 17 KB, 416x212, JISPhilipspozi.jpg [View same] [iqdb] [saucenao] [google] [report]

the tiny punched dot on the screw head marks it as a JIS screw. needs a specific screwdriver.

>> No.1722467


yes, every single one can do that easily.

>> No.1722471

well fuck wish i knew that earlier.
I was able to get a little bit of oil on the shaft, the motor is spinning a little more now, but i think some got on the part where the belt goes because the head is spinning but the belt isnt moving. the only way to get this fixed is to get the motor removed and lubricated properly

>> No.1722475

Ah thanks.
It's just weird how they're all rated for 5V.

>> No.1722476

Anyone who thinks about making another kind of screw shot be shot, no matter where they are in the world.

>> No.1722478

should be shot*

>> No.1722498
File: 183 KB, 600x307, file.png [View same] [iqdb] [saucenao] [google] [report]

I'm making a board powered off a USB type C connector. The other end of the cable will plug into a wall wart, computer, or a phone. I don't need any data, I just want to get as much power as easily as possible. Do I need to worry about pins besides VBus and GND?

>> No.1722507

I mean, if you're just using the connector because it's a convenient form factor rather than actually using the USB-C spec, you can do whatever you want.

>> No.1722513

I'm curious if there's anything I should care about in the spec. Like if putting a pull-down resistor on one of the data lines puts it in high current mode, or if I can't power it from the phone unless I have a pull-down, or something like that.

>> No.1722516

We assumed you meant "double tap" and would second that sentiment.

>> No.1722524

No, none of that old school shit will work. The usb power specification is insanely complicated, you need a custom and expensive chipset/mcu to get the power out of it, and many of them are still buggy.

>> No.1722565

>has way more useful functions
>better resolution and/or don't have to be an ant

>> No.1722579

>270-5 & 270-5RT Extra-High Accuracy VOM - Simpson Electric
DC voltage accuracy is ±1.25% full scale; AC voltage accuracy ±2% full scale; and Resistance accuracy ±1.5 º of arc.
Are they all this trash or did I pick a bad one?

>> No.1722591

i would imagine they are all that bad, though that's kinda hilarious and directly defeats the OP.. Simpson was GOAT in that regard

>> No.1722602

based screw death squad

>> No.1722610
File: 27 KB, 864x622, PowerCircuit.png [View same] [iqdb] [saucenao] [google] [report]

Reposting from the MCU general:

Are there any simplifications I can make on this power supply circuiit?
- I am wondering if it is safe to drop D1
- I think I could drop R4 and connect DC level to pin 3 of the Barrel_Jack_Switch and use an internal pullup. Never tried using this pin before though.
- Anything else a smart anon may notice.

R1 and R2 are required because I don't want to buy more resistors and BAT_LVL will be attached to a 3.3v ADC.

>> No.1722615

D1 serves no purpose. What is DCLVL connected to/does and what is that jack doing? Because depending on what D2 is also useless.
If J1 is to connect another supply and the diodes are doing a logical OR then obviously they are needed.

>> No.1722616

And it won't be 5v, it'll be 5V - Vd

>> No.1722619


diodes D1 and D2 are required so there's no collision between the 7805 and the external DC source which would would cause high currents to flow into the weakest of the two.

but you have the switched jack which can be used to switch between sources, and so the diodes are no longer needed. the supply voltage would normally come from the regulator, but when you plug in a voltage into the external jack it disconnects the regulator.

way to do that is:
pin1 goes to ground of both sources
pin2 is the output voltage to your load
pin3 comes from the output of regulator

when plug is out, 2 and 3 are shorted and so the output is 5V from regulator
when plug is in, 2 pulls away from 3, and so output is external voltage.

>> No.1722620

C1 and C2 are way too big, see datasheet of the actual regulator you're using for recommended values
>safe to drop D1
see datasheet of the actual regulator you're using for recommendations on what to do about back-powering
>a 3.3v ADC.
verify the open-circuit voltage of the battery and ensure the 3.3V limit is met at the worst edges of tolerance

>but you have the switched jack which can be used to switch between sources
only if plugged or unplugged. anon can't just leave it plugged in and switch elsewhere
>pin1 goes to ground of both sources
>pin negative
>current year

>> No.1722634

Is there a current equivalent of the charge pump circuit/voltage multiplier ? (large current at very low voltages)
>inb4 a transformer

>> No.1722645

well, both are the basis of DC/DC converters, which typical require inductors so...
but what's the point? what are you doing? you can't get more power from a circuit than you put in, so for a fixed load, the only way to get more current is by raising the voltage.

>> No.1722649

buck converter

>> No.1722654

Thanks for the help anons.

I should have been clearer about the funciton of this circuit. The objective is to supply 5V from either a 9v battery or a 5v DC power supply. One concern here is what happens if both are active at the same time, hence the diodes were added to prevent current from one flowing back into the other (or into the voltage regulator).
BAT_LVL pin provides the battery voltage level sensing to the MCU.
DC_LVL provides a digital true/false as to whether the DC supply is connected (another reason D2 is requires).

I am using a MCU that has both 3V3 and 5V tolerant pins [and I just double checked that DC_LVL is 5v tolerant].

I like this idea of switching with the jack to disconnect the battery. Ill see if i can make it work.

I'll have a look over the datasheet and fix the caps. Cheers.

>> No.1722660
File: 2.58 MB, 4096x3072, IMG_20191124_183903.jpg [View same] [iqdb] [saucenao] [google] [report]

I want to make a small 120VAC to 24VDC power supply. I took apart a 5VDC 750mA wall wart usb phone charger for inspiration. The circuit seems backwards to me.
I was expecting transformer -> rectifier -> regulator. When I got the thing apart I found two boards;
One has the USB connector, couple diodes, resistors, and a capacitor.
The other board has the transformer, a transistor, a rectifier, a 6 pin controller/regulator?, and some passives.
It looks like the AC lines go directly into a rectifier. The wires to the secondary board with the USB conenctor on it come directly from the transformer.
What black magic are they doing here? Is the IC on the main board running off rectified mains? Fireing the transistor at something other than 60Hz?
The output of the transformer must be single wave rectified and smoothed with the cap. Does this just allow them to claim 750mA peak without having to drive it constantly?
Why not just send clean 5VDC directly to the secondary board?

>> No.1722661
File: 2.50 MB, 4096x3072, IMG_20191124_184019.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.1722675

wew, never saw that coming, only ever heard of it used in jap bikes.

>> No.1722677

welcome to reality lol
your expectation is the old way (though not outdated for sure) of doing it. What you're looking at is an offline dc/dc buck converter (google) and
> Is the IC on the main board running off rectified mains? Fireing the transistor at something other than 60Hz? The output of the transformer must be single wave rectified and smoothed with the cap.
all of those are close to, if not, true. you typically wont find a similar circuit with such a small transformer, because those are made with very high freq in mind

>> No.1722693

Why use this over the older way? Is it because transformers are expensive?

>> No.1722707

Yep, a common flyback converter will rectify and filter mains, which is switched by an IC at numerous kHz into the primary of a transformer. Higher frequency means the transformer can be much smaller. The output negative pin is either left floating or tied to ground (not neutral).

If you want to make your own isolated (or unisolated) off-line converter, there are some fairly common ICs for doing so even on ali. Like the TNY268 or DK112.

>> No.1722720

Thanks. I'll have to look into that. I want to make a device that needs enough juice to uC and half a dozen industrial analog sensors. Will it work well with 24V at around 1 amp? Or is it more of a low voltage thing?

>> No.1722732

>It looks like the AC lines go directly into a rectifier.
welcome to the world of switch mode power supplies

>> No.1722742

Well for sensors, you'll almost certainly want an isolated power supply, meaning having a flyback transformer instead of a run-of-the-mill inductor. I've not looked terribly deeply into this, but you may well need to wind your own transformer to this end.

24W is well into the range that I'd consider using an off-the-shelf power supply instead, or at least one of those loosely shielded chinese power supplies with screw terminals. Unless you want a power supply that makes you a 5V or 3.3V logic rail as well. The kind of bare-bones design I'd be using for a <1W power supply won't cut it for a 24W version, so if you do make a custom-jobbie you'll want to look towards controllers that use an external switching transistor, along with some extra safety features.
Putting your power supply a distance from your control board will usually mean less switching noise radiated by the psu will make it to your micro.

What's using up 1A of power anyway? Heated sensors?

i thought it was "switched mode" power supplies

>> No.1722751

>If recording the audio and playing it back is out of the question,
Nonono. Making motor controller that makes motor whine like a fucking train.

>> No.1722752

Honestly 1A is probably overkill. I'd like to get data from 8 sensors. Each will consume 20mA plus whatever their overhead is. Need some power for the uC, then some for the comms, hopefully zigbee or similar.

>> No.1722756

>I was expecting transformer -> rectifier -> regulator
ok boomer
>why two boards
isolation, size

yes, also extremely inefficient and shit regulation, and xboxhueg and heavy

you can get off-line controllers for any size supply. be advised that switching power supply transformers are application-specific devices, engineered for a specific operating frequency, voltage/current, and feedback mechanism, except for little crappy single-outupt charger things that won't output more than 12V anyway. there's at least one place on aliexpress that do custom transformer winding for not completely unreasonable prices (RANLO) and will do a tiny bit of the engineering for you if you fill out their form and use a chip they know about
there is also software from various off-line controller vendors which will take specs in a wizard format and output a schematic and winding instructions. Power Integrations' PI Expert is one of them but they're complete and total niggers who will b& your account within 24 hours if you're not going to order 10k units next month. they must really like fake email addresses
real talk, life is too short for making one-offs of off-line SMPS

well you'd better figure out what you need, it's a very important input into the design process

>> No.1722759

it's worth thinking about a lower-voltage, higher-current supply e.g. [email protected] or whatever, and using isolated potted step-up converter module(s) to get the power to the other side. how cost-constrained are you?

>> No.1722774

Shit, they were out of 4053s so I got a 4066 instead. It might not be enough on its own but I've still got one of the two poles of the original DPDT to use up. The other pole will be used for toggling the analog switch.

Yeah in that case take my other advice, creating coil whine with a coil in series with the motor, or piezo noise with a piezo disc across the motor (to ground or to other windings). Oh and feel free to put passives in parallel with the coil, or in series with the piezo.

>shit regulation
Well for something with no negative feedback I'd say they're fairly decent.

>> No.1722936
File: 44 KB, 640x500, 92.gif [View same] [iqdb] [saucenao] [google] [report]


>> No.1722944

Just curious about it desu

>> No.1722951
File: 2.30 MB, 3264x2448, manhattan board.jpg [View same] [iqdb] [saucenao] [google] [report]

Manhattan update, it's turning more into point-to-point with a little deadbug in there too (good luck reading that part number lmao). Still only half way there, got the complex analog circuitry to solder next. Also turns out superglue doesn't adhere to solder.

I got a second-hand ungrounded/2-pin power supply too, so ground loops are no longer an issue.

try a CD4052?

>> No.1722965
File: 11 KB, 456x358, diode.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.1723004
File: 113 KB, 1078x822, 7805.png [View same] [iqdb] [saucenao] [google] [report]

Is that diode to offset the output voltage by a diode drop? Pretty clever, and I imagine the current going through it will be significant enough to make a difference.

Not sure if the diode D1 is necessary though, pic related is the schematic equivalent.

>> No.1723019

Ω-meter between pins 3 and 2: 4.6kΩ.
About a milliamp goes back into the 7805.

>> No.1723063
File: 43 KB, 268x300, old_machine-268x300.jpg [View same] [iqdb] [saucenao] [google] [report]

My relatively old washing machine has a power indicator light that stays on when switched off, but flickering in a lower intensity state.
This light should go off. It isn't even a led, it looks more like a old incandescent indicator light. I guess is some kind of abnormal induction in the circuits through the house.
Is there a trick or some electric component I can use to make it go completely off?

>> No.1723070

pull the plug

>> No.1723071

I think it could be also related with the fact that the led lights in the house show a little lightning (or ghosting) in the dark when powered off.

>> No.1723072

Im playing with linear hall effect sensors and magnets for a physics experiment. Right now its dumbed down to just track a magnet on a pendulum. The issue is that Im getting voltage drift. How is this normally accounted for? The drift is slow and sometimes it increases or decreases depending on some capacitors.

>> No.1723092

In the specs it says this thing can be powered from a 3.6V lipo battery
But i need to know if i can just slap for example 3x1.5V nimh battery in series and power it from that? (around 4.8V tops)
But i am too retarded to understand the data sheet, can someone confirm it will work?

>> No.1723095

I googled the component into which the battery wire goes and it came up with this
https://lcsc.com/product-detail/DC-DC-Converters_BL8026CB5TR_C77972.html and it says voltage max 7V, so does that mean that i can connect any DC power source that is 7V or under and it will be fine?

>> No.1723105

Neon indicator, is your house grounded properly?

>> No.1723107

you just discovered free energy, enjoy your billions of dollars

>> No.1723122

Right, now I remembered these are neons.
Grounding in my "house" is poor.

>> No.1723220

why do crt's draw left to right then reset the beam back to the left(hsync) instead of just drawing to the right aswell?

>> No.1723221

next time do surface mount :^)

frequent zeroing

sounds good. always leave margin. exercise caution if using a long power input wire https://www.analog.com/media/en/technical-documentation/application-notes/an88f.pdf

>> No.1723227
File: 107 KB, 1280x720, 1564750040815.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.1723228

>official /G/ certified toaster(capital g)
gets me every time

>> No.1723249
File: 136 KB, 639x665, good thing yandex works at least.png [View same] [iqdb] [saucenao] [google] [report]

the link is only about gears and your image gets me some pretty interesting results on google
moving the beam across the screen takes high voltage and you still need to time it right to avoid misalignment
wouldnt it be better to just draw continually without any blanking?

>> No.1723280

How do I get into electronics? Can I build myself a microscope?

>> No.1723281

> Can I build myself a microscope?
>open cheapest webcam you can find
>take out camera
>pretty good zooming tool
>take camera lens
>put it backwards
>now it's a shitty microscope in need of a base and light
I've managed to see some algae cells this way

>> No.1723286


>> No.1723291

>frequent zeroing
If something is always in motion how would you zero it if the motion is not predictable.

>> No.1723298

Is dumping a ROM chip as simple as just hitting the address lines and recording the output for each? I’m looking at an AMD clone of an Intel 2716.

>> No.1723311

wew lad
two reasons: to scan both directions you'd need to create controlled horizontal ramps in both directions instead of just one. all that circuitry would need to be duplicated and aligned in every camera and accounted for throughout the signal chain, also, unlike the inkjet printer example in previous Pic related, the vertical ramp generator runs continuously so the return line wouldn't be straight
iow, KISS

sure. at worst you might have to clock ~OE and ~CE too

1. peak detection is one way
2. don't use absolute values, i.e. add more sensors

>> No.1723355

the electron beam has effectively no inertia, where's the benefit?

>> No.1723369 [DELETED] 


a bigger hurdle is that you'd have to feed the data time-backwards on the left-right scan. and these things were developed before digital storage, so storing and pulling backward data is no easy task.

>> No.1723371


a bigger hurdle is that you'd have to feed the data time-backwards on the right-left scan. and these things were developed before digital storage, so storing and pulling backward data is no easy task.

>> No.1723378

mine actually has a lower-case /g/, no lols for you

high-pass filter?

That isn't an issue at all, the only reason why a single line could be considered drawn time-backward or time-forward is because of the direction it was scanned in the first place. So you'd just need to have the cameras and recording equipment scan every second line in the other direction.
Though I don't know how analog broadcasting video cameras worked back in the day. The other problems about having two ramps and the like still apply.

>> No.1723397

Trying to solder to aluminum foil for a quick capacitive sensor. Copper foil isnt sold near me. Ive tried the vinegar trick but its not working> Any tips?

>> No.1723398

>So you'd just need to have the cameras and recording equipment scan every second line in the other direction.
So you are kicking the problem to the camera, rather than solving it?
You're a True Engineer, anon.

>> No.1723409
File: 62 KB, 856x685, Analog_20191125_2001.ts_snapshot_00.52.444.jpg [View same] [iqdb] [saucenao] [google] [report]

Good news, I was able to get the motor out and oiled, and some tapes playing. despite a few more adjusments left, the picture looks about as good as the first unit, AND i am able to record them to a capture card with no horizontal skewing like the first unit.
Its all downhill from here, Time to digitize some tapes

>> No.1723440

based Al Capone's Vault opener

>> No.1723458

You need to remove the thin oxide layer. Try rubbing some steel wool over the spot and quickly apply solder with flux.

>> No.1723465

your need for expedience does not sway chemistry, but if you insist, sand lightly under warm vaseline then try tinning the alfoil
or try soldering to the other side
or use an alligator clip
or get some copper flashing instead. comes off of reroofing jobs all the time

>> No.1723479

>So you are kicking the problem to the camera, rather than solving it?
It isn't a problem at all, you just do the same two-way scanning on the video camera tube as you do on the CRT tube in the other end, no other modifications required.

>> No.1723492

>aDdiNG cOmpLExITy iSnT mOdIfIcAtIOn
is there a point to this shit-tier what-if game

>> No.1723493
File: 14 KB, 230x216, PALNTSC2.jpg [View same] [iqdb] [saucenao] [google] [report]

>no other modifications required
for the sake of the future of intelligent life, you must destroy your brain. leave no trace of it. it cannot be allowed to spread!

>> No.1723495

Jeez guys, it was more of an alternate history thought, rather than a "we should do this now". Not like analog television standards are relevant now-days anyhow. When defining the first TV standards it would be plausible for them to have decided on a two-way beam scanning system, as it wouldn't have required some sort of analog data buffering to get time-backwards data like >>1723371 originally implied. As I said in >>1723378 there's still all the other issues (like those mentioned by this guy >>1723311) to contend with.

Though time-backwards analog data buffering could be interesting if impractical to attempt, like using an electrostatic/magnetic disc or cylinder with contra-rotating read and write heads, along with an erase head thrown in somewhere.

>> No.1723503

>it was more of an alternate history thought, rather than a "we should do this now"
no, we shouldn't have done this back then, either, when components consumed all that much more space, energy, and money
>it would be plausible for them to have decided on a two-way beam scanning system
plausible, but as soon as you start thinking in terms of the practicalities, how do we make this bidirectional-scan thing happen, what is the baseband signal going to look like, etc., you'll find yourself duplicating what was a simple sawtooth generator once for each direction, adding a dual triode to decide which direction the beam scans, painting the left and right edges of the screen unevenly (phosphors aren't linear), all for absolutely NO advantage in the final product, and you'll very quickly find yourself looking for something simpler and more regular

>> No.1723529

I agree completely. I was mainly arguing against the time-backwards data postulation.

>> No.1723570
File: 1.07 MB, 3264x2448, too many wires.jpg [View same] [iqdb] [saucenao] [google] [report]

Almost done now, just two more coax jumpers for passing the left and right audio channels from the input to the output when the TRRS mic is what's being processed. The existing coax jumper is for passing the mic from input to output when it's the left and right (combined as mono) channels being processed. The processing is using a circuit I posted a while back, I'll post more info about it once/if it works.

On a side note, I can see why people prefer to make their own PCBs, getting my fat K tip in there can be a bit difficult. I can also see why people like to use silicone-insulated wire, and why no-clean flux is often preferred. Well a little RMA on the board for less than a week won't kill anyone.

>> No.1723571

Generating a perfectly balanced triangle wave is fairly difficult especially with early vacuum tube technology. If the positive slope is 1% faster than the negative slope the even and odd lines won't align properly and the picture looks like garbage.

>> No.1723589
File: 518 KB, 1200x803, jim-williams-thermometer.jpg [View same] [iqdb] [saucenao] [google] [report]


thx for that. it clearly shows that point-to-point airboard is superior to Manhattan style.

>> No.1723621

Well it's an audio board so I thought the extra shielding would come in handy. Plu all the knobs need some sort of mechanical fixation and I don't have anything structural on-hand other than FR-4.

>> No.1723643
File: 93 KB, 493x307, pease.jpg [View same] [iqdb] [saucenao] [google] [report]

>mom can we have point to point airboard?
>we have point to point airboard at home
>point to point airboard at home:
Epik mene

>> No.1723896

Not sure if this is the right thread but can someone recommend me a first time soldering iron?

>> No.1723907
File: 232 KB, 502x400, 2019-11-27_09:51.png [View same] [iqdb] [saucenao] [google] [report]

How about a DIY Hakko soldering kit. Got one for cheap and never looked back.

>> No.1723911

Some will call that ugly. I call it beautiful.

>> No.1723914

any vendor recommendations?

>> No.1723920

Nowhere specifically. Just buy off the store on Aliexpress with the highest ratings.

If you have the dough, buy the one with the case as well. Should save you time.

>> No.1723925
File: 1.55 MB, 3264x2448, CLIPPPPING.jpg [View same] [iqdb] [saucenao] [google] [report]

This or the pre-made T12 station that comes with its own 24V power supply inside the case. Also the OLED-based controller has many more options.

>I thought the extra shielding would come in handy
It wasn't enough, it's noisy like you wouldn't believe. But I'm ok with that, because it's a distortion box.

>> No.1723940

Alright, the noise was on the virtual ground itself, by slapping a cap from ground to the +6V I eliminated it entirely. Feels good man, now I just have to find that TRRS extension cord I had lying about.

>> No.1723950

What would happen if a mains transformer has no secondary winding? "Small transformer 120v to 12v, say i removed the secondary"

>> No.1723954

It would become a 120V rated inductor with plenty of impedance at 60Hz. Make a filter with it maybe, or use it as a big electromagnet.

>> No.1723957

I like rounding, personally. show FETs

just a yuuuge inductor across the mains, about the same as if you kept the secondary open

>> No.1723959

The spartan boards only work with Xilinx ISE, now very obsolete. The Zybo and Myrio should work with Vivado 2019 fine.
> Chad 74xx voltage regulator has entered the chat
Unless it's a weird voltage, COTS is going to be cheaper, faster, and usually better than whatever you can come up with from old parts.
I did something like this once for a project. It worked for a few seconds, then promptly caught fire...

>> No.1723967

Diodes can be plugged into the side for soft clipping, while the intermediate amplitude can be turned up enough to hit the rails for hard clipping. However the preamp doesn't seem to have enough range to reach this for quiet audio sources (microphones), the sound just stops being audible once the pot is turned up enough. I suspect this is due to a DC bias remaining (some signs of this happening at the output are visible) but there's a blocking cap between the microphone bias and the preamp input, and a 1M resistor from the preamp input to ground, so I'm not sure what's going on. I'd make that a 100kΩ, but I'm not sure how much that would drag down the microphone's output. Perhaps these 100µF caps are too large a value for this? I could try swapping instead to some 1µF ceramics.

Also I forgot to put a DC-blocking cap on the TRRS mic output, so I'll do that soon.

>> No.1723973

that all depends on the microphone
oh, you mean diodes
>1µF ceramics
you misspelled "built-in microphone" :^)
remember f=1/(2*pi*R*C) and grab a large greencap instead (or harvest an X cap from some defunct device)

>> No.1723981

hello /ohm/agicians I am currently doing EE in college but they don't offer courses in IC design. I want to get into chip design especially analog chip design and would be nice if you had any books to recommend, books on digital IC are welcomed too !

>> No.1723989

Yeah maybe. This is one of those cases where audiofools use tantalums, no?
>built-in microphone
I've never had that issue, and my X caps are too valuable to waste on this sort of low-voltage circuit, I do mains stuff too often. Not to mention that this pin will either be being fed buffered audio or connected to a literal microphone anyway, so it shouldn't matter. I heard that ceramics' more pressing issue is their voltage-dependant capacitance, but I can't see that mattering anywhere except high-Q filters. Also these might be greencaps, I'm not sure. They're resin-dipped, but feel too small to be anything but ceramics.
Actually trying to maximise the SNR of a ceramic cap as a mic could be interesting.

I know but I don't know the output impedance of my microphone. Swap caps now, reconsider later.

>> No.1724022

Well replacing that 1M with a 100k removed that DC bias issue, now the preamp can use its full range and get more use out of microphones. But with the preamp using that much gain there's a very noticeable whine. So my next issue is finding out where that's coming from and stamping it out. So in all likelihood, more caps. The scope isn't really helping as there's a bunch of multi-MHz noise superimposed atop everything, so all I know is that there's ~2.5kHz hitting the input to my preamp. Probably from the PSU again, I may test with my grounded supply just to be sure. Keeping my scope probe floating, of course.

>> No.1724048

you should have picked a better school. meanwhile, git gud at semiconductor physics and also read this

lol no, they use axial film caps. regular designers use nonpolar elcaps or radial film caps, if they don't just say fuck it and ceramic

>> No.1724067

Ok now I've made a few small optimisations but none of them are removing this noise. It's about 2mV from the PSU and it's being amplified by the preamp. When I'm using a source with greater amplitude (i.e. phone at 1/4-3/4 volume) I don't get the noise issue at all because the gain of the preamp is so low. So I'm going to make a battery-powered pre-preamp. I have a rail-to-rail op-amp lying about somewhere along with a 3V coin cell in a holder, though I'll use one of my remaining OPA2134s if it can handle the low voltage. See if I can get the voltage up to about 20mV pk-pk. Probably gonna deadbug it, but it does need at least one more audio socket, and a TRRS one at that if I want it to handle my earphones. But for now, baby steps.

My other option is busting open the power supply and slapping an LC filter in there, the output noise is definitely a capacitor charging ramp. The grounded PSU's noise isn't nearly as bad, but either way I'm getting random noise by having the high gain in the preamp. If I get similar random noise with the external pre-preamp (or is it preamp-amp or amp-preamp?) I'll cover the circuit in a cage of grounded solid core copper and maybe even cover that in alfoil.

>> No.1724068

kiwi or straya?

>> No.1724071

obviously irish lolcat

>> No.1724085

>Yeah in that case take my other advice, creating coil whine with a coil in series with the motor, or piezo noise with a piezo disc across the motor (to ground or to other windings). Oh and feel free to put passives in parallel with the coil, or in series with the piezo.
I want motor to make such sounds, not some coil or piezo.
I wonder, if motor drivers on train have some sort of output coil to increase inductance?

>> No.1724087
File: 395 KB, 2048x1536, IMG_20191127_125109.jpg [View same] [iqdb] [saucenao] [google] [report]

I am making a smol tanning bed for curing solder paste on my PCBs.
I really hope 30W will be enough

>> No.1724090

Hanz cameznind, the guy that made the 555 has a very good introduciton into IC design.

>> No.1724100

>I want
too bad. grow up and put the babby noisemaker by the motor or fuck back off to where you came from you childish fantasist

>curing solder paste
wait, what?

>> No.1724101

>Hans Camenzind

>> No.1724103

>wait, what?
i meant solder mask

>> No.1724104

oh like a post-developing cure? should be plenty. I'd be a little worried about parallax if you were trying to use that little nail dryer to expose the board tho
btw that's the sort of situation where copper tape comes in awesome

>> No.1724107

it's not a nail curer i ordered a pack of uv leds and glued them to some aluminium bars

basically the way it works is simple, just etch your pcb and then smear solder mask on it to protect the copper and to help with soldering, then you put it in the uv chamber, solder mask exposed to UV light hardens very fast and boom you have a professional grade protected pcb ready for soldering

>> No.1724108
File: 56 KB, 483x409, when programming socks arent enough anymore.jpg [View same] [iqdb] [saucenao] [google] [report]

>it's not a nail curer

>> No.1724109

I mean the thing I made isn't from a nail lamp

>> No.1724113

and isn't isopropanol a misnomer?

I wouldn't worry about parallax, if he's got a transparency squeezed between the board and a piece of glass there shouldn't be any noticeable blur. It's not like he'll have it up a couple of mm, and he won't need trace-by-trace accuracy, just a rough border about each pad where ±0.5mm should be more than enough, and where a little trial and error would prevent covering any of the pads.

>> No.1724116

fair enough. just saying, it *could* be, if you had Simpsons hands

yes it is
I suppose it depends on what you're putting on there. unless spin-coated, it doesn't seem like it would be a good match for SMT

>> No.1724118

pretty much, but i won't really be etching or using the cover foil to do the pads anyway, my workflow is simple
>mill the traces
>smear uv mask on the pcb
>place the uv lamp on top to harden
>mill the solder mask from the pads

>> No.1724163

why were girls hotter in the 80s?

>> No.1724174

>isn't isopropanol a misnomer?
IUPAC can suck my dick. Everyone knows what it means. It’s just the order of operations they prefer for affixes means the -nol comes after iso-, so isopropanol would be the -nol form of isopropane, which isn’t a thing.

But we all use it to describe the iso- form of propanol and everyone knows what we mean.

>> No.1724187

>tfw you just call it rubbing alcohol

>> No.1724195

Watching some EEVBlog to learn proper soldering technique, and he mentioned something about there being different surface mount "pad" sizes for a given component size (0805, 0602).
Specifically, he mentions there is a "Least", "Normal", and "Most" pad size (smallest to largest). I was looking for the dimensions of these, but wasn't able to find them. Anyone know the reference/standard for these pad sizes?

Time-stamped video for reference.

>> No.1724196

>He doesn't drink IPA

>> No.1724253

It was their beard and hatred of not understanding the proper value of simulations.

>> No.1724296
File: 42 KB, 389x480, 2b4d9026b95eb23123051e7b5ed10246.jpg [View same] [iqdb] [saucenao] [google] [report]

Lol...no. I love tubes but even a "shitty" 5532 is better.

>> No.1724303
File: 62 KB, 396x566, 644288439de841ab4d83a06375876479.jpg [View same] [iqdb] [saucenao] [google] [report]

You also have a dog shit lm324 that's slow. Make pcbs; a laser printer, pcb paper and expresspcb will make your life easier and I guarantee you'll be happy with the results.

>> No.1724307
File: 640 KB, 1500x1125, IMG_20191127_221840.jpg [View same] [iqdb] [saucenao] [google] [report]

Vintage motorbike capacitor guy from last thread here.
I grabbed a few 1000uF capacitors from my old as fuck capacitor jar.
Tested them with my power supply to replicate the bikes voltage and measured the output and voltage was nominal and when I cut the power I got one or two seconds till discharge.
So I replaced the battery in my bike with this and it ran poorly, under low revs,stalled, and my lights blew.
So I tried again with 400uF and same result.

Output voltage was exactly the same as battery, and in check with what the service manual specified.
Wheres my mess-up?

>> No.1724313

Post schematic of what you are doing and what you hoped to accomplish.

>> No.1724331
File: 26 KB, 858x551, drawing.png [View same] [iqdb] [saucenao] [google] [report]

Battery is what I removed, yellow is where I spliced.
I've tried a few capacitors with similar result and results in testing enviroment was nominal.

>> No.1724332

Also, this is only what I imagine my 70's bike to be wired like.

>> No.1724378

I dunno family, maybe motorbike batteries work as voltage regulators and the lights can't handle 8v

>> No.1724394

this anon gets it

better?... or worse?

>in an audio application
everyone laugh at his pain

>output voltage was exactly the same
you are dealing with amps-per-second beyond your comprehension and you don't even have the tools. seriously just find a trickle charger and keep a small SLA battery in one corner of the shop. gell cell should be fine for a tester, which is what you need, right?

>> No.1724402

>you are dealing with amps-per-second beyond your comprehension
Ok, I apologize for my raving ignorance, I understand amps and have a general idea, but nothing specfic, 400uF should give me 400amps second capacity but does that mean it sends all those 400 at once down the line independant of what the components are drawing? Sorry if I phrased that weird.

My objective is to get my bike on the road as I can't source a 6v SLA battery locally without ordering online.

>> No.1724422
File: 843 KB, 863x1390, file.png [View same] [iqdb] [saucenao] [google] [report]

Is there a recommended soldering iron, or does it not matter what I get?

>> No.1724424
File: 25 KB, 480x320, Heat-Sink-On-Triumph-TR6R_large.jpg [View same] [iqdb] [saucenao] [google] [report]

this. they'd try to balance alternator output against bulb/coil load. city riding would drain the battery, highway riding would boil the battery.
having a zener diode on a heatsink was LUXURY in the 70s.

>> No.1724425

You could solder with cheap ones fine. Although just watch out how hot the temperature gets, you might burn out heat sensitive components.

>> No.1724428
File: 36 KB, 630x411, 400 microfarads.png [View same] [iqdb] [saucenao] [google] [report]

>400uF should give me 400amps second capacity

>> No.1724429

I actually meant that but I was stupid enough time type the comment out, calculate how many zeroes, and forget to type them in

>> No.1724441

tell us what manufacturer/model/year it is, so we can work from the schematics on the microfiche instead of playing twenty goddamn questions.

>> No.1724453
File: 3.84 MB, 4000x3000, IMG_20191128_014148.jpg [View same] [iqdb] [saucenao] [google] [report]

You're right I should've started by that but I didn't think it would be relevant.
Here's what I could find

>> No.1724455
File: 1.87 MB, 3502x2626, IMG_20191128_014720.jpg [View same] [iqdb] [saucenao] [google] [report]

Part 2

>> No.1724458

Oh are you using one of those spring-loaded end-mills like I saw some other guy do in a video? As a method of carving away the solder mask without touching the pad underneath I mean.

I've already got the shit to do inkjet PCBs, just not with me at the moment. Considering designing some sort of LCD lightbox once I have a significant disposable income.
>they're making fun of my cheap-ass IC choice
I don't care about speed for an audio application, unless it's the higher gain-bandwidth-product that comes in handy at lower frequencies. What are the (noise) specs I should be looking for in an audio op-amp anyway? CMRR? Input offset/bias? PSRR? Crosstalk?
Also I can easily just pull my LM324 out of its socket and pop in an LF347 or something along those lines.
But for the pre-preamp I don't have the luxury of picking a JFET op-amp. With a cursory glance the LM6132 looks to have worse stats than an LM358, but again I don't really know what I'm looking for here.

the T12 chink station is serving me well, KSGER brand.

>> No.1724498
File: 39 KB, 1000x667, 6n6.jpg [View same] [iqdb] [saucenao] [google] [report]

ok so it's a 1970-73 honda CL100? why does every cunt want to do this guessing game shit
there's no voltage regulation for DC (which only drives coil, indicators, horn and some instrument lights. head and tail light is unregulated AC)
assuming you don't want to *add* voltage regulation to this system, it will need a 6N6-3B flooded cell, exactly this thing and only this thing. not fucking sealed/AGM/motobatt/lithium/nanotube. the flooded cell soaks up overvoltage by boiling, and this kills the other technologies.
topping up the battery electrolyte is a regular maintenance item. a plastic vent tube attaches to the battery and drapes down to the road, so that gas does not build up in the battery box.

we can talk about switching to regulated 12V, it's pretty cheap to do with a new coil, new horn, couple of new lightbulbs and a chinese pitbike's reg/rec. although you still wouldn't be able to run an ignition coil from tiny capacitors.

>> No.1724522
File: 12 KB, 640x640, MFG_ACS781LLRTR-050B-T.jpg [View same] [iqdb] [saucenao] [google] [report]

Nowadays we have teeny single-chip hall effect current sensors, and we have teeny single-chip high side load switches.
So somebody tell me WHY in the FUCK do we not have single-chip programmable circuit breakers? Why are we still using thermoelectric polyfuses like it's the 20th century?

>> No.1724524

I need to read 4 analog inputs as fast as possible. Preferably reading getting around 4 samples for each all within the span of 1ms.
I bought an ads1256 which can read 1 input in about 32us, but if I want to cycle through once its about 900us. Besides buying multiple boards, (they are around 20 each) what are some options. I was thinking that if I were to build my own breakout board for the ads1256 I could have multiple ics on the same circuit and since the ads1256 chip is around $2~4 my project would be a lot more replicable (I plan on making several). The fear is that Id end up with a lot of noise if I made my own circuit because I dont really know much about that besides following designs the manufacturer sometimes gives.

>> No.1724525

Fail conditions I'm assuming.
But I'm interested in this hall current sensor, how do you have to arrange the wire with respect to the sensor to obtain a good reading?

>> No.1724533

you don't. The current path is part of the chip, it's that large metal bit on the bottom. that lets these little chips have surprisingly high current ratings. that 6x6mm chip in the pic is rated for 100 amps continuous.
you do have to be careful of nearby magnetic fields, that's all.

>> No.1724536

>The current path is part of the chip
Ah, so not whatever those DC amp-clamps use. I'm wondering if those ones even need a clamp at all.
>nearby magnetic fields
Surprised the thing doesn't come in a metal can, but I guess they don't suit modern assembly techniques.

>> No.1724539

I took apart one of those clamp meters, it had the hall effect sensors inside a slot in the metal ring of the clamp. So the magnetic field around the wire would go through that clamp and pass directly through the sensor.
I think they only need to do that to compensate for the fact that you can't get the sensor precisely located with respect to the wire.
Having that fixed in a chip like this allows them to not have to circle something around the wire, which allows such a teeny thing to handle insanely high currents.
"nearby magnetic fields" are more like interference from other high current traces.

>> No.1724575

>accidentally drop your VOM
>your meter movement is now completely borked
nothing personal, kid
>accidentally drop your DMM
>at worst you scratch the case
>with a decent case protector nothing at all happens to it

Also as some other anon said:
>input impedance (on voltage ranges)
..and furthermore:
>min-max memory
>frequency measurement
>capacitance measurement
>diode junction test
..and so on.

Go ahead and keep using your antique, though, if it makes you happy, boomer.

>> No.1724578

>witched-capacitor converters are much quieter than those involving inductors
Not necessarily. High frequency switchers are not only easy to filter, but use smaller inductors and capacitors so they have an overall smaller footprint on your PCB.

>> No.1724581

But higher frequency noise radiates easier. Like those LED signs that annoy hams.

>> No.1724620

It sure seems to be. I wish there was more info about it out there, I really wanted to do the advanced walker robots, but didn't have the skill at the time BEAM was popular.

>> No.1724625

My 2-pin micro USB sockets finally arrived, now I can shove them in every damn project I want.
>inb4 I should be using type-C instead

>> No.1724631

power semiconductors tend to fail short-circuit

900µS < 1ms, doesn't it?
>a lot of noise
probably. a 4-layer board would alleviate almost all noise concern, but that cost might make it less reproducible depending on how many "several" is
look for app notes. manufacturers love to help you make them look good. borrow tips from https://www.analog.com/en/technical-articles/layout-for-precision-op-amps.html and https://ez.analog.com/cfs-file/__key/communityserver-discussions-components-files/324/TechDay09kr_5F00_hpa_5F00_Track2_5F00_1_5F00_Precision_5F00_Analog_5F00_Designs_5F00_Demand_5F00_GoodPCBLayouts-_5F00_JohnWu.pdf
might look into other devices like the MCP3903
>$2 each
if you're lucky, and hopefully you're not getting factory rejects or recycled chips

sorry I forgot the implied "all else being equal". there is no inductive spike in a switched-cap converter because there is no inductor (except for mostly avoidable parasitics, which are orders less than a typical switcher inductor)

>he leaves D+ and D- open

>> No.1724634

>D+ and D-
Will leaving them open result in PSU noise, or just being unable to draw the full 2A from some sources? Because I'm fine with 500 or 100mA. Fuck, I'm fine with 5mA, I'm not building some sort of amplifier here. Any circuit that isn't driving a load or running at some umpteen MHz that draws more than 1mA should go back to the drawing board. I'm willing to push that to 5-50mA for indicator LEDs, perhaps a few hundred mA for obscure sensors, vacuum tubes, or other peripherals.

>> No.1724638

What do you guys use to store components? I need something thar my cat can't get into. I wanted one of those desk cabinets with the little plastic drawers, but my cat would probably just rip those out and scatter everything. Something resembling a tackle box but with containers long enough for resisted and shit would be nice. Can't seem to find one though.

Also does anyone else have a cat who won't fucking leave your shit alone? I've been cucked by my cat and forces to hide away in my bedroom with all my components because if he sees a wire he just makes a beeline to go chew on it. Should I leave a live wire for him to shock himself with and learn a lesson?

>> No.1724639

Cheap tackle-boxes for me, but I don't have enough of them. Instead I write labels on my china packaging and fill them with components. Got one for resistors, one for capacitors, one for LEDs and lasers, one for buttons and pots, etc. All of these live in a carboard box under my bed. It's a far sight worse than a proper organisation system, but it's basically free and still has a fair bit of optimisation as far as taking time to find the part I need. Except for resistors, which are kinda a lost cause at this point. But c'mon, 12 values per order of magnitude, 5-7 orders of magnitude? That's like 70 individual compartments, where each one needs to be 60mm in at least one direction. I have belts of common values (10k, 100k, 1M, plus a few others) in more convenient places.

>> No.1724640 [DELETED] 
File: 477 KB, 1600x1200, DSCN0555.jpg [View same] [iqdb] [saucenao] [google] [report]

You're cat's an asshole.

>> No.1724642

He's still under a year old, hopefully he'll grow out of it.

>> No.1724643

>Oh are you using one of those spring-loaded end-mills l
good springloaded bits are expensive little shits, but it is impossible to use a regular endmill, because the copper is too thin and mills aren't precise enough to handle that

>> No.1724649

What are those bits meant to be used for? First I saw one I assumed the guy had machined it himself in order to do PCBs. I can't imagine they're only for use milling away solder mask, but is there some other industry where milling a soft layer off something hard needs to be done in extreme precision?

>> No.1724650

they are used for quickly engraving metals

>> No.1724651

I don't really understand how a spring-loaded bit would help for that.

>> No.1724658

You're doing gods work here,I actually found a 6N6 for 10€ in a boat shop a town across so I'll be making my way there today.
If be up to switching to regulated 12v, my main concern was having to mess with the dynamo.
So my problem here was introducing unregulated voltage to the system? I had no idea the battery had that function and now it sounds silly to substitute it with simple condensers that can't soak up overvoltage.
So if I where to use reasonably sized cap to keep the bike alive while running id need something to soak up excess voltage or regulate it.

>> No.1724672
File: 129 KB, 800x800, 1571530977420.jpg [View same] [iqdb] [saucenao] [google] [report]

they mean you won't be able to run data over them, silly

tackle boxes are good for some things but get expensive if you have a variety of values
I use Pic related for my standard chip sized SMD passives, usually one book per one or two package sizes
for R and C that's back-stock or doesn't fit into a book, I have allocated Digi-Key shipping boxes, usually one bag per package size/tolerance combination. my Ls and my MCUs are all in one of the LCSC front-flap flat boxes
diodes and transistors are in small ESD bags per part number, organized in ascending order by current, separated according to polarity, inside a small shipping box with added dividers
for ICs and misc shit I use zipper bags, one per value or part number (sometimes I mix part numbers where it saves space and/or eases storage, like all my LQFP48 STM32-alike ARMs go into one waffle tray inside one ESD bag). those are organized inside 1 L zipper bags more or less according to categories (MCU peripherals, linear, power, pin headers, pin sockets, big LEDs, small LEDs, etc). those are in turn placed inside shipping boxes, usually the small Digi-Key size or so
all those sit on bkac powder coated wire shelving

>> No.1724675

>tfw my wife's son jumped on the keyboard

quicker, rougher indexing

>> No.1724682

If i use one of those internet to 230v convertors what range do they have? Also do things like transformer stations stop the signal?

>> No.1724686


it's called Powerline. they're good for your whole house, and a couple of neighbour houses. the transformers on the poles outside will attenuate it a lot, but some signal can get through. for good speed, you gotta make sure the 2 units are on the same 120V phase. this is critical: some sockets will work very badly, some will work fine.

>> No.1724687

>device that needs enough juice to uC and half a dozen industrial analog sensors.

5volt is often enougth. Read the datasheet about the specs, replace the sensor WHO needs 24volt with something else if possible. The USB powersupply you disassemled would be good enought.

>> No.1724689

I've had my cat since I was 12. He was there for my teenage high voltage experiments. he learned very early on to avoid the workbench.

>> No.1724699
File: 16 KB, 301x225, STOP_PLC.jpg [View same] [iqdb] [saucenao] [google] [report]


>> No.1724700

arduino faggots go back to your containment board

>> No.1724756

Where do I get started on NeTwOrKs? I hate the fact that I can design an IC but have NO clue about how to make networked devices.

>> No.1724757

dead batteries. that is all.
My primary meter isn't an analog one, but my backup is.

>> No.1724758

>id need something to soak up excess voltage or regulate it
that's what >>1724424
>a zener diode on a heatsink

>> No.1724759

>Anyone know the reference/standard for these pad sizes?
What software is he using? what are you using?
Kicad standard footprint library has just regular and "handsoldering" smd footprints, I've never needed any more specific than that.

>> No.1724761

can i use capacitors of any capacitance for capacitive coupling?

>> No.1724764

by this i mean if it's alright to overkill. some guide i'm following calls for 100nF but i only have stronger ones

>> No.1724767

IPC-7351 or a revision thereof

coupling what to what, at what load impedance, and what frequency range?
you can figure the half-power frequency of an RC filter by fc=1/(2*pi*R*C)

>> No.1724768
File: 39 KB, 467x364, 1567760894112.png [View same] [iqdb] [saucenao] [google] [report]

connecting a non-amplified 3.5mm audio jack to a microcontroller, it needs to be shifted +2.5v into the 0-5v range
i'm really not sure what the frequency range and load impedance would be, do you have any ideas?

>> No.1724783

Thank you! Having trouble finding free versions of it, but found an old version for like $35.
Might consider buying it if none of my friends happen to have it.

>> No.1724805
File: 26 KB, 1242x747, high-pass filter.gif [View same] [iqdb] [saucenao] [google] [report]


find a high-pass filter calculator on the interbutts. plug in your component values to see what frequencies will be accepted or rejected. so, in your example, i use 50K (coz you have 2 paths of 100K each to the power supply) and 100nF, and it tells me it'll allow frequencies above 32Hz to pass. this sounds fine for audio.

i ignore the 5K resistors coz they have no business being there in the first place. if you wanna protect things from other things, try something around 470-ohms.

as for that other 5K from 3.3V, that makes no sense in your context.

>> No.1724859

you need multiple filters in a row, since they aren't magic, the filter boundary is a slope not a wall

>> No.1724900

What circuit do I use to convert the capacitance of a water level sensor to a dc voltage?

>> No.1724903

#include <arduinoFFT.h>

>> No.1724934

most of these channels are total shit. why is this list still being shilled in /ohm/ OPs???

>> No.1724939

OP here. Everyone who made it to the list got there by blowing me on the regular. If you wish to remove some names, your mouth will have to take up the slack. You game?

>> No.1724964

Nice non-argument you've got there. Any actual constructive criticism to add?

>> No.1725017
File: 28 KB, 1024x889, 1550617148536.png [View same] [iqdb] [saucenao] [google] [report]

try this one. add series/parallel capacitance to tune

>> No.1725018

he's only whining bc it's imperialist passover day and his parents won't take him to GameStop for Black Friday

>> No.1725040

Isn't that just an oscillator? I can see it working with a filter after it and basically a diode AM envelope detector (or some other sort of FM demodulator circuit), but wouldn't the ESR be too high for it to work properly? I'd use an LC oscillator instead so it's less prone to having its frequency disrupted by ESR. I'd then use a CD4046 PLL as an FM demod to turn the varying frequency into a DC voltage.
Fuck I wish my CD4046s actually arrived, because they're 3 months late and I don't think they'll ever arrive. Thanks Yanwen.

>> No.1725042

>900µS < 1ms, doesn't it?
True but I wanted to take a few more samples so that I could account for any noise/jitter in the inputs.
What effect would a mosfet have on an adc. With a mosfet I was thinking I could connect 4 analog inputs to 1 pin on the adc. and then use a mosfet to switch which pin signal is connected.

>> No.1725045
File: 21 KB, 466x466, cell.jpg [View same] [iqdb] [saucenao] [google] [report]

Has anyone had a play with these very cheap solar cells? they are under 10c ea in quantities of only 100, and even cheaper in larger amounts. I've had a good look at reviews and videos of them, and they seem to match the rated output (short circuit, but sellers mention this in the specs). Even under load, they are still very good.
My only concern is housing them, since they are wafer thin (0.25mm) and extremely fragile. Proper solar resin is expensive, and it ends up being cheaper to just buy premade.
The only thing i can think of for these, would be to secure them to something rigid, like plastic board, and only put them out when you need to use.

For those wondering, they are .5v cells @ ~400mA short circuit. 50mm * 20mm * 0.25mm. Also worth noting, they work out to be roughly $1\Watt, so only good for small or custom shaped panels.

Any ideas, opinions, etc on these?

>> No.1725052


>> No.1725067

>I'd use an LC oscillator instead so it's less prone to having its frequency disrupted by ESR.
it's a water level detector, not a 5-digit capacitance meter. it'll need to be calibrated anyway. KISS

>> No.1725073

>a few more samples
but D-S ADCs already oversample
>what effect
they're non-linear voltage-controlled resistors, roughly. better use proper analog switches if you care about parts per million
I don't think you'd gain very much by duplicating circuitry already inside the chip
anyway, if I needed more ADC, I'd start looking for better parts that aren't much more expensive. think about why bare chips cost 1/10 what the assembled breakout boards cost, and possible reasons that ratio is so far out of line with other chinkshit

>> No.1725076
File: 321 KB, 600x400, 1571434744795.png [View same] [iqdb] [saucenao] [google] [report]

>mfw EEVblog has better EE memes than 4chan

>> No.1725088

But ESR changes depending on water salinity, so an RC oscillator could also need to be calibrated often. If the ESR of the capacitor is significantly lower than the in-circuit resistor this won't be an issue I suppose.

Any fun examples?

>> No.1725090

Hvem sa det, din jævel?!
So I have to think about everything?
3.3v or 1.8v or 4.2v?

>> No.1725093

>try to make preamp
>doesn't work at all
>just shits out noise
I feel like a retard, half because it's just a simple non-inverting amplifier, and half because I soldered it already. Back to the breadboard for me.

>> No.1725095

I recognised the bottle Jaycar carries ;)

>> No.1725096

post pics.

>> No.1725097

no i delammed pads on this shitty FR-2 perfboard it's ugly

>> No.1725098

post ugly.

>> No.1725108

And now I realise that my input was lacking a proper ground reference, but more than that I'm getting a shit-ton of noise when looking at the relevant amplitude range (1mV), and this is now a proper circuit on a breadboard with a variety of power supplies, both battery and SMPS. Decimal SNR kinda shit, the noise is about 10mV pk-pk. I'll try a different op-amp I guess, maybe these chinky LM358s aren't up to spec.

>> No.1725111
File: 345 KB, 612x450, 1557537876672.webm [View same] [iqdb] [saucenao] [google] [report]

true, it does. assuming that salinity wouldn't vary significantly in the sort of application where capacitive sensing would be appropriate. if absolute value is a requirement a mechanical switch would be better ofc
webm related

>> No.1725113

And now I'm getting a surprisingly stable ±0.8V sinusoidal oscillation that's dependant on my virtual ground capacitance. I don't understand.

>> No.1725114
File: 192 KB, 550x730, 1575005202035.jpg [View same] [iqdb] [saucenao] [google] [report]

it's an amplifier. it's supposed to oscillate
post the schematic already

yes. yes, you do. it's called engineering

kinda hoping the next OP uses this

>> No.1725115

wildly guessing, I bet you're loading your virtual ground. throw a buffer on it

the LM324 is an early low-power opamp. not designed for especially low distortion. I'm pretty sure crossover distortion isn't the kind you're looking for

>> No.1725117
File: 5 KB, 640x400, buffer.png [View same] [iqdb] [saucenao] [google] [report]

>post the schematic already
Ok here's the current one. The other op-amp is configured identically and with the same input. Oddly enough, putting a 10kΩ pullup on the output works, though it does drag the output way up near the +3V rail.
Yes this is a nightmarishly simple circuit to be having such a problem with. Technically the thing should be able to handle supply voltages as low as 1.8V, so the coin cell shouldn't be the issue.

>virtual ground
Now that could have been the case previously, the caps on it were only 1µF, but now it's still happening and I've got no virtual ground at all. And putting 1mF across the battery doesn't change this noise much.

>> No.1725120

>software pass filters on fartduino
you are adorable

>> No.1725121

ok but what are you driving with it? opamp stability tends to be lower near unity gain than at higher gains. try it in the inverting configuration
>way up near the +3V rail
that's not good. it's almost as if the opamp isn't even trying. maybe you need a lower voltage part

>> No.1725122
File: 1.20 MB, 3264x2448, it never ends.jpg [View same] [iqdb] [saucenao] [google] [report]

>what are you driving with it
Not driving anything but my 1M scope probe. It's running off my function generator.
>opamp stability tends to be lower near unity gain than at higher gains
I didn't know that. I'm trying it at a gain of 10 now, with a pair of 10ks and a 1µ as my virtual ground. The oscillation is gone, and so is all the gain, it's acting very strangely now, pic related being input and output respectively.

I've noticed that in all my experiments the output impedance is acting really low, because just touching the output is causing it to quiver, while touching the input does nothing at all.

>a lower voltage part
I'll definitely consider it.

>> No.1725123
File: 1.53 MB, 3264x2448, das buffer.jpg [View same] [iqdb] [saucenao] [google] [report]

And this is the buffer at work.

>> No.1725125

Ok screw this coin cell horseshit, I'll just use a linear regulator or two to stamp out that power supply ripple. Saves the trouble of having to turn the thing off and replace the battery.

>> No.1725126

your rail splitter should ideally have one cap to each rail. if you're still running in non-inverting mode, the smaller resistor in the - input divider still has to carry current
even more ideally, you would be buffering that rail splitter actively
also put several tens of pF across your feedback resistor to get that spoopy oscillation down
the MCP6002 looks like a good choice for this sort of thing. a bit slower but it's not like you really need the MHz gain

>> No.1725151
File: 42 KB, 626x521, question.jpg [View same] [iqdb] [saucenao] [google] [report]

For one project I'm currently working on I made pic related (top pic). But I would like to do it all through just the USB cable, but the LEDs can suck up to 3A.

Will I burn my PC-port or the thin USB wires if I try bottom pic? If not, what about the floating voltages? Do I need to connect the ground to both power supply and PC? Is it a problem the 5v from the data line not the same 5v as the one on the plus wire?

>> No.1725158

USB data lines are referred to USB ground. you have to connect that thru no matter what. VBUS is entirely optional and you can splice it in wherever
>burn the thin USB wires
it's absolutely possible. they only need to pass 500mA under rated conditions
if it's just WS2812 why not put the controller closer to the computer and run power and WS-data over the longer cable?

>> No.1725160


It's a MIDI music instrument with lots of inputs aside from the LEDs. My idea was integrating everything inside of it, so that it's basically plug-and-play. The final version will kind of look like a guitar so I wanted to make it with just one cable so that the musician can move comfortably while playing.

The computer is to process the data for the DAW and send the LED signal. There is no arduino I want to integrate everything in the PCBs so the more I take out the more bulky it'll be outside.

>> No.1725177

any multisim users know how to output a table of component power dissipation for an entire circuit? it's annoying calculating the power for each one individually.

>> No.1725188

ah. sounds like you need https://www.usb.org/document-library/usb-power-delivery
and probably some specialized cables too

>> No.1725232

WHat sort of wires do headphones use? I need some super soft wires that bend easily like a string and can carry up to 1A, is there any such thing?

>> No.1725233

Headphones? 1A? That seems extreme.

>> No.1725244

microphone cables, usually
but your expectations require management

>> No.1725245


I see, it seems some USB-C cables supports USB-PD too. So if I get a cable that supports power delivery I can theorically do it? Cool, thanks.

>in order for a USB-PD compliant source to advertise capabilities greater than 3A (or up to the full 5A limit of the spec) the Type-C cable must be an Electronically Marked Cable Assembly (EMCA) and support SOP' packets

>> No.1725246

theoretically. USB-PD seems to be a bit up in the air rn and you kinda need a whole MCU on each end just to negotiate power
I suggest you try to avoid needing more than 3A. maybe you can turn the voltage request up and buck it down in your device. or use an LED control scheme a bit more power-efficient than WS2812B

>> No.1725249
File: 66 KB, 1181x596, main_board_WIP.png [View same] [iqdb] [saucenao] [google] [report]


3 amps would be worst case scenario (every LED white and max brightness), still, I want to prepare it for the worst case scenario.

What about USB 3.0 cables? I kept reading about it and some people say 1.5A, others 3, others 5... I guess the 1.5A is the charging port not the cable itself.

As for the LEDs, I either the WS2812b or something similar like the SK6812 because of how the program works. And I don't think any of those types of LEDs have much of a difference consumption-wise.

For now I'll stick with the two cable solution for the prototype, until I advance more on the PCB-version.

>> No.1725278

I meant more than 3A through the cable. power is conserved across step-down regulators. if you can convince the upstream to send a higher voltage you will reduce the current and the resistive losses in the cable
not all cables are equal. you'd have to do your research on that one. still have to negotiate with the host, maybe you can get away with just a static request in your device descriptor but I haven't into USB 3.0 yet
good idea for now. USB-PD might be more fully baked with more bugs shaken out in a few months

>> No.1725294

Anyone here use the stcube ide? Thinking of switching from my vscode setup.

>> No.1725343

Can I use a LM358M IC op-amp thing to make a small amplifier for testing beginner synth circuits? (555/556 and Hex Schmitt trigger kind of stuff)?
I've been using a 10watt guitar amp, but I wouldn't mind having a little speaker set up. Maybe even have the amp feed into an oscilliscope.
Thank you for your help. I'll keep searching in the mean time.

>> No.1725350

>have the amp feed into an oscilliscope

no problem.

> a little speaker set up

no way unless you use headphones. not enough power. you probably want an LM386 which will give you like 0.2 watts into 8 an ohm speaker.

>> No.1725353

Thank you. Everything seemed to point towards the LM386 being the right choice, I just happened to have the 358 around tonight.
I copied this guys thing (https://lowvoltage.wordpress.com/2011/05/21/lm358-mic-amp/) and just sent the output to a 1.5" speaker rather than the Arduino, and it works alright, but could be better. Thank you, again.

>> No.1725376

I refuse to believe that computers aren't made to get worse to non-usable in 2-4 years of use. I know software and games get shittier and heavier by the year but they still run the same programs but worse. Goddamit, we have ICs that cna run for 40 years without a hitch and this consumer electronic crap starts slowing down every fucking time reeee

>> No.1725381


there's your path to billions: write an OS that gets faster with time instead of slower by, for example, downloading more ram and more cores.

>> No.1725416

programmer time is worth more than processor time

>> No.1725424
File: 20 KB, 728x421, R4 in LM5085.png [View same] [iqdb] [saucenao] [google] [report]

So I'm trying to make this buck converter running in continous current mode work but I'm having difficulties selecting the R4 resistor value as shown in the picture.
The datasheet explained how to calculate the maximum ripple current but nowhere is it mentioned how to calculate the minimum ripple current Ior(min).
Does this mean I'm supposed to guess?
My Ior(max) is 1,4A and my inductor value is about 13,4uH.

There are no diagrams to get any values either.
I'm starting to believe that I am suppose to select Ior(min) myself.

>> No.1725433

This is some of the jankiest circuit design i've ever seen but if it works i guess fuck it haha.

>> No.1725494

It looks even worse at the moment, and I've still got more wires to solder onto it.

>> No.1725503

cellphones aren't computers, they're crystallised malware.

>> No.1725506


it's an illusion

9 out of 10 times, Marxian analysis cuts to the root of the problem
>except for that time resonant molecular bonding couldn't be true

>> No.1725586

Preamp retard here. Since I'm going to use a 5V linreg from +6V to -6V instead of a 3V coin cell I'll be able to use an OPA2134 instead, probably. Maybe an LM317 if I want a bit more range (since the input range on the 2134 is "V- + 2.5" to "V+ -2.5"), but I suspect those have worse ripple-rejection than an L7805, plus they need more passives and I'm already possibly dealing without a ground plane to solder to.
So I have two design queries. First of all, should I buffer my virtual ground with one of the two op-amps? It probably means less required capacitance and might give more linearity, but after what was happening yesterday I'm not so sold on it. In any case it's what I've done with op-amp 4 of the LM324.
Second question, should I make a new -6V ground plane for this preamp circuit? My coin cell circuit was going to be on a bit of perfboard and sitting off the main board on some detachable jumper wires, but now I'm not so sure if that was a good idea.

>> No.1725595

Can any anon recommend me a decent soldering station for under €40?

Live in Euroland.

>> No.1725600

936 clone.
it will come fitted with a stranded aluminum power cord. after an initial function test this cord should be removed and destroyed. at this time it would be appropriate to confirm ground continuity from the wall to the front panel metalwork and the tip of the iron, and to retrofit an internal fuse if it has been omitted.

>> No.1725606

>should I buffer
>with one of the two op-amps
whatever works. an LM321-tier op amp might be enough to hold a ground plane steady if you give it a small pull at the output to keep the output driver out of its deadband (like you did with the pullup before, but it'll actually work when you have more room to pull lightly with the higher supply voltage)
>ground plane
lel, that would be a V- plane, wouldn't it?

>> No.1725615
File: 304 KB, 566x415, 1563175725950.png [View same] [iqdb] [saucenao] [google] [report]

Ahh, something like this then: https://hobbyking.com/en_us/soldering-station-with-adjustable-heat-range-with-eu-plug.html?wrh_pdp=7

Looks interesting enough, thanks!

>> No.1725620

KSGER or Quicko T12 chink station. Comes in a variety of models: ones with an internal SMPS or ones with a simple 12-24V DC input socket, ones that come assembled with a case or ones you build yourself, aluminium cases or plastic cases, plastic or aluminium or stainless handles, advanced OLED or simple 7-seg GUIs, etc. In all cases, I'd recommend the former. Notably, the OLED controller has a UI that allows you to set the standby temp, boost temp, boost duration, timezone, temperature reference method, and I think it even has an alarm in there too.
Personally I'm going to drill a hole in the back of mine and route a DC barrel jack to the output of the SMPS, so I can easily use it off mains or off a battery.

>give it a small pull at the output
Wait, so I should deliberately pull a few hundred µA or even a mA or two from my existing virtual ground buffer on the main circuit? It makes some sense, as linregs often want some sort of load for stability, which I might also want to do if my preamp takes too little current. Feels good to not have a power budget unlike when I've worked with battery-power or capacitive dropper circuits. I assume it would be ideal to have these "load" resistors going both to V+ and V-, no?
Reminds be a bit of this negative resistance some guy used to let a low-current precision op-amp drive a higher current load, for some reason.

>lel, that would be a V- plane, wouldn't it?
Well I figure it's the best choice for "shielding" as it's the most stable rail as far as the rest of the circuit is concerned. But maybe the buffered ground is a better choice, I'm immediately chucking it through a DC-blocking cap anyway. My desire is to stop any ripple from getting carried along with the preamp's output into the main circuit, but even more important is to stop the power supply ripple before it gets to the input of the preamp, because that's why I'm building the preamp in the first place.

>> No.1725629

>from my existing virtual ground buffer on the main circuit
yes, if you use a shit-tier opamp with poor crossover distortion specs, it can help keep teh output loaded so that one output drive transistor remains on at all times. that way it doesn't have to take the time to turn on in case of transients hitting your VG output. if you use a better buffer than the 324, you might not need to bother
using one to each rail would defeat the purpose
>for shielding
if it's a double sided board, why not both? srsly tho it probably doesn't make a whole hell of a lot of difference which you use as long as it's good and electrically solid

>> No.1725631

>using one to each rail would defeat the purpose
Ah I see now, it's that class-B-style crossover distortion, and I just need to pull a bit more current than the peak current output is imagined to be. Considering that it's likely to be highest when I'm using ~32Ω earphones (two ears in parallel to be 16Ω) with 10-50mV pk output, that's 0.3-1.5mA peak. So a 2kΩ pulldown to -6V should do the trick. Then again, some of this circuitry might be disturbed by such an addition, so I'll have a proper look at the circuit to ensure that nothing bad will happen, and to ensure that it's not too much current to stray from linearity or significantly affect my open-loop gain.
Also the entire thing is designed to make distortion in the first place, so I might just do so on the ground buffers.

>it probably doesn't make a whole hell of a lot of difference
That's probably true. What I choose will determine how compact the circuit can be though, so I'll consider it when I somehow find a piece of appropriately sized copperclad. I currently lack a hacksaw, so this might get messy.

>> No.1725634

Quite fragile to work with, make sure to use flat tinned wire to solder to the top/bot contacts

Also if you want to pot them DO NOT use UV curable epoxy (the cheap stuff that smells crazy bad anyway). I made a small panel (4x1) last week and the thing worked for a couple of says before going kaput, suspect the resin ruined reacted with the surface in a bad way and now the solar panel give ~0.3Voc instead of 2Voc

>> No.1725640

>hacksawing FR-4
>tin-snipping FR4

use non-corrosive neutral cure silicone instead

>> No.1725655

blink_led.ino poster pls go

>> No.1725680

>implying arduino isn't an IDE
none of that rubbish here, m79. libopencm3 + make 4 lyfe

>> No.1725682

+ gdb + blackmagicprobe

>> No.1725686


>> No.1725694

it tells you right there in the pick you uploaded
>where Ior is the minimum ripple current, which occurs at the minimum input voltage
Go look back thru the datasheet
It says your L choice primarily determines your ripple(max), but I'm pretty sure that just means worst case, but roughly what you should expect. Go back in, reorganize L equation to find Ior,min but using your lowest input voltage (ie replace max subscript with min). Then use that Ior value for MIN R4. THEN, use Ior,max for the R4 eqn and you have the range for R4.

>> No.1725704

So just a tip, I don't think you want to use this fella in unity gain(while you are troubleshooting). I just forgot but the phase margin is ASS, like 33 degrees! Which is considered only conditionally stable.. A gain of 10 should put you much closer to 45, what you want as a min.

>> No.1725709

Can aliexpress tier true RMS meters be trusted?

>> No.1725710

can som1 recommend books on computer architecture, asm, logic controllers etc? I want to design my own architecture with my own arithmetic

>> No.1725723

Not really the topic for this thread but yes
I can recommend my college lit
You plan on fuckin with HDL i hope
Computer Organization and Design is good for theory. Easily pirated with a google search
HDL is kind of its own fucky thing and I don't have any great resources for it, but any book teaching you HDL will work. If you have experience coding, I'd go Verilog, and SystemC looks cool but you may be hard pressed for resources. Avoid VHDL at all costs if you like sanity

>> No.1725727

you might be surprised how effective a simple single-stage RC filter can be on power inputs. if you'd really love to enjoy life, try an LCL tee filter
also, maybe put termination resistance on your input

as long as you don't ask them for the true RMS
EEVblog has a cheap multimeter spreadsheet, might be worth the look

Computer Architecture: A Quantitative Approach is sort of the sequel to Computer Organization & Design. CO&D helps you ask "can we?", CA:AQA helps you ask "should we?"
>own arithmetic
cool, I've been waiting a long time for an INTERCAL processor core

>not really the topic for this thread
yes it is
>Avoid VHDL at all costs if you like sanity
look, mane, as Amiga pioneer Dave Haynie once pointed out, some people like having fat women sitting on them while their hands are tied behind their back

>> No.1725740

thanks anon, I asked in dpt and got referred here, perhaps /mcg/? I'm a total noob, but I'm a math guy. working through SICP and C and haskell, but I want a ground up understanding. I'll be taking some intro compsci classes, but I'm a self study guy

>> No.1725751

Check out Clash if you're into Haskell. I just made a simple project with it, it's pretty nice.

>> No.1725762

comp sci ≠ computer engineering. if you want to into HDL, get used to everything running in parallel, REAL parallel, ALL the time

>> No.1725810

that makes sense
thank you

>> No.1725822

You mean the COB (chad-on-board) package

>> No.1725848

I prefer a VOM when working in high RF environments. DMMs tend to glitch out.

>> No.1725869

Wait, phase margin for a particular op-amp? Where's this in the datasheet? Does this still apply to the OPA2134? Will this cause problems with my virtual ground buffering?

I'd need a power resistor for that, right? Sounds like a bit of a pain.
>LCL tee filter
How about a CLC pi filter? The only thing stopping me from doing so is the fact that there's no room in my circuit for it (the DC socket's terminals are being used as the main V+ and V- connection nodes), but I can avoid this by just cutting open the adapter's cable, or possibly splicing an extra DC jack and socket together with the filter in between. Oh also I'm a bit wary towards using either my normal axial inductors or my more expensive mains toroidal chokes, so I'll likely have to wind my own toroidal(s), which is never fun. At least now that I have a scope and function gen I can easily measure inductance.
>termination resistance on your input
Already have one, both a ~5kΩ pullup on the pre-preamp mic and a 100kΩ to 0V before the first main op-amp. I could drop that value to 10kΩ but I'm not getting any problems with that input in the first place.

Also shit I might need to add another switch to my circuit because I need an optional bypass for the mic preamp when using the aux-in instead of a mic.

>> No.1725889

Can a CS major become an embedded programmer? If so, how?

If I want to do embedded would computer engineering or electrical engineering be a better choice?

>> No.1725896

CS is a complete joke of a degree if you want to do anything other than academia. Sure it'll look good on your resume but it's not going to teach you anything worthwhile. But then again school isn't for learning, it's for credentials. Pick whatever you want it doesn't really matter. An EE can easily get a programming job with thag degree, but a CS couldn't get an engineering job.

>> No.1725981

How can one post be so based?

>> No.1726014
File: 289 KB, 2196x1384, 2000 Watts Induction Heater Schematic-2.png [View same] [iqdb] [saucenao] [google] [report]

I'm trying to build an induction furnace using an LC tank circuit with a target frequency of about 20 kHz. The guide I'm following doesn't list the magnetic permeability of the inductor cores used so I bought some with a value of 2500 and used 21 turns of 1.2mm copper wire. The capacitors are 470 nF each. About what frequency could I expect with this? I'm a noob at this so help is appreciated

>> No.1726067

>power resistor
depends, how much current do you plan for a fuzz box to draw anyway? I suggested T over pi becaues pi performance is dependent on your input impedance, which I suspect is pretty low. besides, a little sag might be audibly pleasing

>school is for debt so that (((employers))) can hold it over your head and threaten to fire you for capricious, unjust reasons
fixed for USA

>> No.1726105

what's wrong with f=1/(2*pi*sqrt(L*C))

>> No.1726114

inductance varies with the square of the number of turns. your core's datasheet should have listed an A(L) value, which is usually in nanohenries per turn squared i.e. a core with a specified A(L) of 200, wound with 3 turns, will yield 1800nH of inductance. or you can do it the hard way and estimate the core cross-section and magnetic length and all that happy horseshit (equations to be found out there on the web somewhere), but I wouldn't if I didn't absolutely have to

>> No.1726116

Wind and measure, wind and measure. Not like you have to get it spot on assuming that's a proper resonant circuit, as the addition of a piece of iron or copper or whatever to the middle will cause the inductance to change anyway. Actually, if you are using magnetic cores, doesn't that mean your heating coil will be seperate from the coil you're currently talking about? I figured they all just used their heating coil as their only coil and picked the caps to make it work with a high Q factor and a reasonable frequency, but I guess that would mean very large caps or a very high frequency.

>> No.1726150

Hey guys, question (might be a bit stupid): With your typical wall wart/power brick power supply, are those typically just voltage supplies and the current value on them is a current limit?

That is, where a wall wart says it puts out X number of mA at that voltage, it's not gonna blow up a device that says it takes half that amount of current at the same voltage?

>> No.1726162

>might be a bit stupid


>value on them is a current limit?

yes, in the same way that a chair rated for 400 pounds doesnt REQUIRE you to weigh 400 pounds.

>> No.1726164

Hah thanks. I guess my point of confusion could be summed up as not being sure whether the power brick was the chair or the fatty in your analogy.

>> No.1726167


the point is the values are maximums, not requirements.

>> No.1726188

gj you just put /g/ on suicide watch

>> No.1726205

is there some rough way t calculate heatsink size for certain watts? for example i have lm317 and i need to dissipate 2Vx1A of power so 2W
How big does the heatsink have to be if it's the standard aluminium

>> No.1726246

the math isn't that hard, but you seem pretty stupid, so use the "small custom designed heat sinks" calculator on http://www.giangrandi.ch/electronics/thcalc/thcalc.shtml

>> No.1726259

I'm not sure how to use that equation with the two inductors on either side of the circuit

>> No.1726380
File: 6 KB, 225x225, images.jpg [View same] [iqdb] [saucenao] [google] [report]

Apply it to the (single) working coil you left out in your circuit fragment. The (two) supply chokes (why both L1?) play no role, they are much larger than the resonant working coil. The chokes are usually 2 toroids.

>> No.1726401

Alright, I think I understand now. Thanks for the help!

>> No.1726517

>The (two) supply chokes (why both L1?) play no role
Wait, if those aren't in series with the working coil in order to change the resonant frequency, what do they actually do? If they were just there for feedback there'd be no need for them to be so large.
Also jesus I really want to rearrange that schematic >>1726014 to be more symmetric and less of a clusterfuck.

>> No.1726767

go make a badass video card with an SPI interface or something

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