/ohm/ - Electronics General: Summer Edition

[Anonymous No.2649781]

Old thread blew a fuse >>2645377

>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.

>Incredibly comprehensive list of electronics resources:
https://github.com/kitspace/awesome-electronics
Additional resources below:

>Project ideas:
https://adafruit.com
https://instructables.com/tag/type-id/category-technology/
https://makezine.com/category/electronics/
https://hackaday.io

>Don't ask, roll:
https://github.com/Rocheez/4chan-electronics-challenges/blob/master/list-of-challenges.png

>Archive of Popular Electronics magazines (1954-2003):
https://worldradiohistory.com/Popular-Electronics-Guide.htm
>Microchip Tips and Tricks PDF:
https://ww1.microchip.com/downloads/en/devicedoc/01146b.pdf
>Li+/LiPo batteries required reading:
>https://buster-spb.ru/files/SAFT/li-ion_user_manua.pdf

>Books:
https://libgen.rs/

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

>Recommended Design/verification tools:
KiCAD 6+
Circuitmaker
Logisim Evolution

>Recommended Components/equipment:
Octopart
eBay/AliExpress sellers, for component assortments/sample kits (caveat emptor)
Local independent electronics distributors
ladyada.net/library/procure/hobbyist.html

>More related YouTube channels:
mjlorton
jkgamm041
EcProjects
Photonvids
sdgelectronics
paceworldwide

>microcontroller specific problems?
>>>/diy/mcg
>I have junk, what do?
Shitcan it
>consumer product support or PC building?
>>>/g/
>household/premises wiring?
More rules-driven than engineering, try /qtddtot/ or sparky general first
>antigravity and/or overunity?
Go away

[Anonymous No.2649792]

I want to try upgrading a monochrome CRT from SD to HD. Can you spot any holes in my plan?

My idea would be to drive the horizontal deflection coil with my own custom signal. The horizontal deflection coil normally gets a ~15kHz sawtooth signal which sweeps each line across the screen, from left to right. There are 262.5 horizontal lines per interlaced field, which makes 525 (480 active) lines per complete frame: 480i resolution.
My mod: generate a new sawtooth signal of the right frequency with my signal generator, amplify it, and apply it to the deflection coil. For 1080i (1125 total lines), this is 565.5 horizontal lines per interlaced field (540 active). So ~33 kHz is the frequency I need.

Then, I can presumably apply my 1080i video source (monochrome, so I'll probably use the Y from YPbPr component HD) to the cathode. I don't care about the frame rolling horizontally/vertically for now, so I'm not going to worry about the sync pulses.

[Anonymous No.2649797]

>>2649792
CRTs have shadow masks, you can't get more resolution because there is a metal grid in the way.

[Anonymous No.2649798]

>>2649797
Nvm, it should be fine if it's monochrome. They don't have anything besides phosphorous coating.

[Anonymous No.2649800]

>>2649797
I mean a monochrome one, no color.
But by the way, how does a shadow mask work? The answers I can find on google are very vague (like "blocks the wrong gun from hitting the wrong phosphor").

[Anonymous No.2649804]

>>2649800
>But by the way, how does a shadow mask work?
Electron guns are at different angles and mask only allows a certain spot to be hit. Same principle as parallax effect.

[Anonymous No.2649807]

>>2649804
Thanks, that makes sense!

[Anonymous No.2649819]

>>2649781
How the fuck do I calculate heat resistance of a funny shaped PCB track? I forgot all calculus and numerical analysis.

[Anonymous No.2649826]

>>2649819

print it out 1:1 size.
drag a string along it.
measure string.

[Anonymous No.2649830]

>>2649819
>heat resistance
Just copy shit over from IPC-2221 or any PCB calculator. If you want to do anything that is not conventional then you'll have a lot of fun solving Fourier's heat equation or at least dicking with that PDE in FEM packages.

[Anonymous No.2649842]

I accidentally posted in the dead thread, so I'll post again: Hey guys, I want to repair an old oscilloscope and I found what I believe is broken. Anyone have any experience de-potting these hybrid ICs? Can I just soak it in acetone, or should I be more surgical in my approach? In case if you are wondering, I do have the schematic for this part, but the transistor types have been left out, so I want to see if I can identify them and maybe do a component level repair.

[Anonymous No.2649853]

I have some cheap current transformers (SCT-013-000) for an arduino project that I'd like to test and calibrate. Reading the datasheet they don't seem to have a minimum voltage. So I'd like to somehow make a current source that's around 2 to 10 vac@60hz and can do as close to 100a as possible and pump that into some ceramic resistors on heatsinks. How stupid is this plan and do you have a better idea? Would it be practical to wind my own transformer?

[Anonymous No.2649895]

>>2649853
>Would it be practical to wind my own transformer?

no.
but you could wind your own secondary onto a microwave transformer.
many DIY welders spewing 1000's of amps have been thus made.
if you cant find a microwave in the trash, burrow your mom's then later tell her it died unexpectedly.

[Anonymous No.2649934]

>>2649826
How's that gonna help me calculate heat (thermal) resistance?
>>2649830
>Just copy shit over from IPC-2221 or any PCB calculator.
I have tried using it, but smth tells me it is overly pessimistic.
Plus I want to do smth non-standard like making a hole in the soldermask and partially tinning the thing in order to increase the surface area.
> you'll have a lot of fun solving Fourier's heat equation
Right. But last thing I want to do is solving diff equations in free time from uni.
> dicking with that PDE in FEM packages.
What FEM analysis software would be most retard friendly?

[Anonymous No.2649969]

>>2649853
Get an existing mains transformer, and replace its secondary winding with a couple of turns of thick-ish wire, which you just put a variable/assorted power resistor across. Just respect the VA rating of the transformer, so go for a really low voltage output for high current. That will make it like 0.1V or less for high currents. Use a big microwave oven transformer for higher voltages, but voltage shouldn’t actually matter. Either way I recommend sawing through the secondary instead of swing open the EI core.

>>2649934
I wonder if anyone has used edge-plated slots in a circuit board to blow air though for heat removal? Don’t forget about those SMD ceramic heat conductors, or just making the PCB on an aluminium base. I think PCBway even do PCBs with a thick solid copper layer on one side.

[Anonymous No.2649980]

>>2649969
>>2649895
So rewind the secondary of a microwave oven transformer, got it. Hope it'll be big enough for a few turns on 2 gauge. I want at least a volt to test with.

[Anonymous No.2649996]

>>2649980
>I want at least a volt to test with.

you should be able to get 2-3 turns which works out to be about 2-3 volts.

[Anonymous No.2650001]

>>2649969
>I wonder if anyone has used edge-plated slots in a circuit board to blow air though for heat removal?
Probably somebody did.
> Don’t forget about those SMD ceramic heat conductor
Mmmm Beryllium Oxide. Tasty.
Didn't know those exists.
170W/mK sounds like thermal conductivity of aluminium aint it? or really close to alu.
> or just making the PCB on an aluminium base
I thought about this, but I've punched numbers into calcualtor and didn't get any difference what so ever, mainly because area didn't increase (i.e. copper probably conducts heat so well it is at the same temperature)
Other thing is that current design is double-sided.

[Anonymous No.2650002]

>>2649980
>I want at least a volt to test with.

you should be able to get 2-3 turns which works out to be about 2-3 volts.
you can definitely use under-rated wire if the duty cycle is low.
they're often used as spot welders so they're often used like 1/4 second every 5 secs.

[Anonymous No.2650007]

>>2650001
Hmmm... If I make PCB thin and single sided, I can glue it to the case (maybe even using high thermal conductivity glue)... I am dumb.
Problem is that I dont want to do the single sided board as it is pretty challenging as I'd have to change my nice capacitive dropper to some shitty buck converter and make power company bit more happier due to increased PF, and I don't want to make them happy, they have to suffer.

[porcini No.2650055]

What are some low cost and ubiquitous PLC units? I am usually working with microcontrollers and component level side of things but want to broaden my knowledge to more industrial style systems. My only requirements other than it not costing an absolute fortune is ease of programming, I don't want some shitware that only works on windows xp, just a good representation of what's out there so I can focus on playing around with ideas rather than troubleshooting the tooling.

[Anonymous No.2650070]

>>2649934
>What FEM analysis software would be most retard friendly?
Probably something from ansys. But you will pay for that retard friendliness. It looks like there are student versions available for free. If you want to go FOSS route then you have ElmerFEM and fenics, but there levels of retard friendliness go down.

[Anonymous No.2650075]

I just bought a 120V 1200va medical toroidal transformer to use as an isolation transformer.
Switching it on would sometimes cause my UPS units to trigger due to what I suspected was huge current draw on start.
I borrowed a friend's scope and current clamp and tested it out and got pic rel. Should I be looking at min, max, or pk-pk?
Either way, does Over ~120A of inrush sound correct? If so, is there anything I can do easily to reduce it?

[porcini No.2650079]

>>2650075
toroidal transformers are particularly spicy when it comes to inrush. you can put a basic "soft start" circuit on the input end with appropriately sized resistor and cap, basically you rectum fry your mains into DC, use that to charge a capacitor with a small-ish (around 1-2W since it needs to drive a coil) such that once that capacitors voltage reaches the threshold voltage of a contactor, that closes and bypasses a another resistor that sits across the contactors AC side. Since you don't want the output side of the transformer to be loaded during start (a short would cause the whole thing to catch fire without blowing any fuses lmao) you also run the output through the same contactor so both ends are fully connected at the same time.

a more reliable and easier (don't have to deal with ringing or shit catching fire) solution is to replace all the custom shit with an industrial timer, keeping the contactor and it's charge resistor only, since you have a current clamp you can then just turn the knob to get the best starting characteristics.

generally speaking you want the minimum interval that doesn't cause tripping, since you are only dealing with an unloaded transformer this would be in the 10-100ms range so you could get away with a 1W resistor of 20-40Ohms.

[Anonymous No.2650081]

>>2650075
That transformer is a big inductor after all, so yes 120A of inrush is to be expected. You could use a NTC to limit it
>https://www.ametherm.com/inrush-current/transformer-inrush-current.html

[Anonymous No.2650104]

>>2650075
Toroidal transformers have high inrush current, so you should make a soft start circuit or use NTC or something.
I'd make something like this:
Or you can use a triac with capacitor but idk fuck triac, high vdrop.

Other thing, maybe zero crossing relay?

[Anonymous No.2650106]

>>2650104
>Other thing, maybe zero crossing relay?
eh, dont think it would work for inductive load.
But yeah, all circuits are basically resistor in series with some sort of switch.

[Anonymous No.2650135]

>>2649781
does anyone know of any good brands of infinitely dimmable, natural light floor lamps? figured /ohm/ autists might have a better understanding of this than /diy/ in general. i did some research on amazon but it's hard to filter out chinkshit from high quality. who knew that it would be so hard to find something so simple...

[Anonymous No.2650249]

>>2650135
>infinitely dimmable
Nothing is.
>natural light
Maybe you can just open blinds?
Well, I can tell you that you can get lightbulbs or LEDs that mimic sunlight pretty good. They use Seoul Sunlike LEDs or whatever chinks clones, at 4000K or 5000K. Thats pretty close to natural light.
> i did some research on amazon but it's hard to filter out chinkshit from high quality.
I'm pretty sure that everything is chinkshit these days.
>what do I do
Make your own lightbulb and lamp, it is DIY after all.

[Anonymous No.2650267]

Found this bad boy, its the Palm Tungsten C. I saw the charging port was some obscure proprietary port, so i make a shitty charger on pins 1 and 16 and it worked since a green light came on. Anyway, I left it alone some time, but it wouldn't turn on upon pressing the power button, I also tried to reset it according to the manual but it didn't work. What are some tips on fixing it and if it is even worth it to try.

[Anonymous No.2650269]

>>2650267
Did you change the battery? lmao

[Anonymous No.2650275]

>>2650269
currently trying a battery to buy

[Anonymous No.2650280]

>>2650275
You could use an old vape pen battery to test it. Any 3.7V lithium cell will work. If you're really desperate you could put three AAs in series, just don't try to charge them.

[Anonymous No.2650297]

>>2650280
are you sure that 3.7V will work, according to the part number, the battery is 4.2V, but also this site https://www.palmdr.com/cart/palm-tungsten-cw-i705-battery-p-269.html about fixing palm pdas has a picture of the 4.2V battery, but says its 3.7V. I don't know what to use

[Anonymous No.2650299]

>>2650297
3.7V is the nominal charge, but it's actually 4.2V when fully charged and around 3V when discharged.

[Anonymous No.2650305]

>>2650299
oh, great, so pretty much any battery that is labeled 3.7/4.2 will work

[Anonymous No.2650313]

I have a 12v autopilot controller that drives a motor using 2 wires and a 12v pwm signal in both forwards and reverse by changing the polarity. I want to fit a bigger ram but the manufacturer says that I could burn out the MOSFETs in the controller. Since it's pwm I can't drive it using a relay. What component would the equivalent to a relay but for PWM?

[Anonymous No.2650314]

>>2650305
Yes. Red to red, black to black.

[Anonymous No.2650315]

>>2650313
A MOSFET, just a beefier one. There are some caveats in replacing a MOSFET, but in general you can replace it with a beefier one without much problems, if you get one with same or lower gate charge. Maybe bigger heatsink
Boat autopilots are kinda low quality overpriced devices anyway.

[Anonymous No.2650317]

>>2650315
So could I mount an external MOSFET that then drives the motor directly from the battery? The ap is a heap of shit and I will be replacing with something better but don't have the cash right now.

[Anonymous No.2650355]

>>2650317
>So could I mount an external MOSFET that then drives the motor directly from the battery?
You can, maybe, theoretically, but it would be much easier to just disassemble the thing and replace it there.

[Anonymous No.2650375]

>>2650355
>it would be much easier to just disassemble the thing

this is quite wrong.
anytime you gut a gadget and start playing with its innards there's a substantial risk you'll break something.
particularly if it's an untested solution by a noob.
and even if you're a DIY goddess like my granny.

[Anonymous No.2650391]

>>2650375
>this is quite wrong.
I dont think so. You don't know how are they driving a motor there, you don't know the topology they are using, so you'd have to see wtf is going on there and such.
With dumb disassembly of the device, all you have to do is just to desolder 2-4 transistors and solder different ones.

[Anonymous No.2650442]

soo the new HOT varicap transistors.
any good? any with useable range? i think they go up to 221 picofarad which is tiny for audio filtering, which is my interest

[Anonymous No.2650465]

>>2650442
I thought SiC semiconductors were the new world-changing hotness. Hotness… get it?

The varicap diodes have been around for a long time.

Anyway, you can amplify the capacitative effect.

[Anonymous No.2650467]

>>2650442
>>2650442
I thought SiC semiconductors were the new world-changing hotness. Hotness… get it?

The varicap diodes have been around for a long time.

Anyway, you can amplify the capacitative effect.

[Anonymous No.2650508]

I don't get this about the typical scope specs

>Input Impedance 50 Ω ±2.0%, 1 MΩ ±2.0% || 16 pF

Why is 16pF only specified for 1Meg input? Does it magically disappears at 50 ohm?
I thought that the scope input capacitance was due to the BNC contacts being close together and there's not much can be done about that.

[Anonymous No.2650523]

>>2650375
>replace the internal MOSFETs without opening it up
Are you high? The only alternative is buying or making an entirely standalone speed controller, the former he already plans to do, and the latter would be e even more likely to break when done by a noob. Desoldering existing FETs and replacing them with FETs with lower RDSon and equal Qgate has the least guesswork, the only chance to fuck up is in the soldering. Though if such FETs can’t be found, he may be better off trying to add a better cooling solution to the FETs, hopefully internally using a heat-pipe or whatever.

>>2650508
There was some discussion about oscilloscope input impedances last thread, so I’d check that out. If you aren’t the anon from that last thread, that is.

[Anonymous No.2650552]

>>2650523
>The only alternative is buying or making an entirely standalone speed controller,

how about this alternative: wire up 2 external FETS driven by the controller's output wires.
this way, the original controller remains intact, and fully isolated (by big resistors) from any screw-up that may occur on the (easily-to-replace) externals parts.

[Anonymous No.2650594]

>>2650552
> Wire new external mosfets’ gates to the output
Some controllers sense the back-emf through the windings, so that might be a design issue with that.

[Anonymous No.2650626]

Roast my ender 3 rice:

My spare nozzle, thermistor, and heat-sink-clamp are yet to arrive via a seperate package from TriangleLab, for some reason.
Still got dual-Z shit and tie rods and oldham couplers and such coming.

[Anonymous No.2650632]

>>2650552
>long wires
>big resistors
>into a MOSFET gate
Are we assuming it's not doing PWM? If it's just fixed 100% duty cycle then I agree that it's not a bad idea. But the thing has MOSFETs instead of just using a DPDT relay, so I'm guessing it's doing PWM. In which case even without those resistors shooting your switching losses into oblivion, you get:
>EMI from the long wires
>gate wire inductive spiking
>double the gate charge by not desoldering the old FETs and so double the switching losses

[Anonymous No.2650651]

Can someone explain what a "high band" or "low band" crystal is? I'm reading a PLL clock synthesizer datasheet saying high band crystals will produce less jitter. I know it's not the same thing as ppm accuracy since that's used elsewhere in the datasheet.

Google tells me high band means the frequency will be slightly higher than the specified frequency. WTF? Why would crystals use such vague terminology instead of just being labeled with their expected center frequency and ppm error spread evenly in either direction?

[Anonymous No.2650668]

>>2650594
It's quite a dumb unit, it was made in 2006

>>2650552
This was my initial idea, though I'm not sure how to wire the whole thing up, would it be an H-bridge? Where the switches are the FETs? My electronics knowledge extends to A-level physics and not much further

[Anonymous No.2650774]

>>2650668
>It's quite a dumb unit, it was made in 2006
Yep, you can easily disassemble and just install beefier transistors, because in 2023 you can get much better transistors with same gate charge. Unless they have used P-channel shits, those didnt progress much.
As for motor control methods, they existed prob since god knows when

[Anonymous No.2650832]

>>2650651
>Can someone explain what a "high band" or "low band" crystal is?

probably refers to ''double side band, suppressed carrier'' modulation.
presumably a crystal could be labeled at the carrier frequency, but then be actually tuned to EITHER lower or higher sideband.

[Anonymous No.2650935]

Whats something cool I can do with this thing?

[Anonymous No.2650937]

>>2650935
Use it with a 555 to blink an LED

[Anonymous No.2650946]

>>2650937
How many microwave transformers do you think I should string together for that? 5?

[Anonymous No.2650947]

>>2650935
>something cool

https://en.wikipedia.org/wiki/Electrofishing

with 10,000V you can probably fish out an entire lake.

[Anonymous No.2650950]

>>2650947
Maaaan! Water, electricity, and easy fish. That is cool. Thanks for the recommendation!

[Anonymous No.2650987]

>>2650081
Inductive loads themselves do not have high inrush current. A transformer having high inrush current only happens when there's a load on its output, and hence is not a feature of inductive loads.

[Anonymous No.2650996]

>>2650937
cuteee
>>2650935
fry your cock to a crisp

[Anonymous No.2650998]

ok this from moog pedal service manual. it says "clock signals be routed serially, no stubs" what does this mean?

[Anonymous No.2651010]

>>2650996
You're telling me I can cook roosters with that thing? Got a schematic? Will their skin come out nice and crispy?

[Anonymous No.2651085]

>>2650987
I was measuring that with no load on the output though.

[Anonymous No.2651088]

>>2650998
pic related.
your clock should go to each component in a chain, instead of having a main line with branches (stubs).
the stubs will cause weird reflections

[Anonymous No.2651107]

>>2651088
great thanks anon

[Anonymous No.2651158]

>>2650987
Yes, that would be true if it were a perfect inductor, but unfortunately there is core material involved and it can (will) saturate, drastically reducing apparent inductance for a brief moment causing a massive inrush spike even if secondary is open. Another part of the issue is that there can be some residual flux left in the core due to residual magnetism and this will also cause current spike that will ring out over few cycles. Life would be so simple if everything was linear... But reality always has limitations and curves bend in weird ways.

[Anonymous No.2651175]

>>2651085
Oh, not even a rectifier and filter cap?

>>2651158
>there is core material involved and it can (will) saturate
Why would it cause a current spike when first plugged in, but not for every subsequent mains half-cycle? Is it something that only happens with the high dV/dt when it's plugged in in the middle of a half-cycle? If not, I can't understand how the first half-sine would be received in any different way than the hundreth half-sine, unless there's some sort of absolute-value (or square) function going on that causes subsequent half-cycles to not cancel each other out. A rectifier and filter result in an absolute-value reservoir of energy to be filled up that causes a spike when turned on. I guess 2nd order nonlinear effects could also add up though, are there any significant energy sinks in a magnetic core proportional to I^2 or B^2?
>there can be some residual flux left in the core due to residual magnetism
I can imagine this causing a current spike if it's magnetised in the same direction as the incoming half-cycle would push it, but that's only a 50/50.

Anyhow, I guess it's a requirement that you put the soft start circuit before the transformer, not after the transformer but before the filter caps. Good to know.

[Anonymous No.2651178]

>>2651175
>Oh, not even a rectifier and filter cap?
The secondary is hooked up directly to a pair of regular duplex outlets with nothing plugged into them for the test.

[Anonymous No.2651188]

I think I'm about to become China at home.
I can fabricate any type of light I need now. I know where to get good high cri sunlike LEDs, I know where to get high efficient LEDs, and I know where to get LED in between. I know how to make a driver. I know what annoys people the most about light, and I know that at Tj=80C shit would last way over 100 000 hours, so prob capacitors would die first. And I know I'm exempt from regulations like PF so fuck power companies :D.
Well, not any sort of light, I still haven't figured how to calculate cooling, as it is super hard, unexpectedly, as solder mast, all of a sudden, radiates quite significant amount of heat (we're talking 20C drop in temperature compared to bare copper/HASL). Which leads me to believe than nobody in China did calculate that shit like ever, and they all eye-balled it lol.

Other question is, how do I explain people the difference between high CRI, high efficiency LED and shitty common LED lightbulb? I know that painters/artists/faggots suffer from shit light, especially since ban of incandescent bulb. But now the cure is finally here, sunlike high CRI 4000-5000K LED, which is nearly like sunlight.

[Anonymous No.2651189]

>>2651188
I'd first make a decent workflow for reliable and cost-effective designs. Then you can think about marketing once you have a price-point figured out. Some sort of demo with certain pigments that your bulb is better at distinguishing. See the Technology Connections video if you haven't already.

Add some near IR LEDs too, for that circadian rhythm bonus.

[Anonymous No.2651190]

>>2651188
Are there any even remotely decent off the shelf led lights (or even just strips) out there that have good CRI / R9 or do I have to go full autismo like you to get good light?

[Anonymous No.2651194]

>>2651190
Yes. You can get good LED bulbs/fixtures. Including high efficiency LEDs (150-200+ lm/w), or SunLike ones.
But they aren't sold in my country. I could have just imported stuff from EU, but they include WiFi, dimming and other bullshit that makes them 20 euro a bulb which is... Too much.
Chinamen, who makes pretty good clone of Seoul Sunlike, they only sell LED strips and those come with whole problem of picking 12/24V PSU that won't catch fire and overall inefficiency of LED strip (25% are usually lost to heat, which combined with low efficiency of sunlike LED is not great). I couldn't find light bulbs or light fixtures.

In the US, I think you can get bulbs and fixtures with Seoul Sunlikes rather easy. Norb or GE makes them, for $11 which is much better than Eurostan, but they are 110V, and I have 220V...
>>2651189
>Add some near IR LEDs too, for that circadian rhythm bonus.
Circadian LEDs are a good autism specialty bulb too, I suppose. Much easier to source parts too, as all you need is yellow or low CCT LED.
Maybe also grow light, but idk, don't really want to deal with drug dealers.
>Some sort of demo with certain pigments that your bulb is better at distinguishing
That's pretty hard. I'm thinking of metal film resistors, they are impossible to read unless you have sunlight or really high CRI LED, but they are too tiny.

[Anonymous No.2651198]

>>2651175
>Why would it cause a current spike when first plugged in, but not for every subsequent mains half-cycle?
The reason why this happens is because flux is in sync with current in steady state operation (flux lags 90°). If you plugin your transformer such that sinusoidal voltage starts at 0 flux can not instantaneously jump to steady state value (if voltage is 0 then flux should be -max peak, but is 0 because transformer was switched off) and if you calculate exact values it turns out that flux jumps to 2 times normal peak flux, pushing core into saturation causing inrush spike. If you switch transformer on when it passes through zero (90° delay form when voltage is 0 or peak voltage) there shouldn't be any spikes in theory. That's why they use such switching with big power distribution transformers.

[Anonymous No.2651220]

>>2649781
anyone have book recommendations for learning "modern" (1990s and onward) automotive electrical systems and how to troubleshoot them? i know basic electronics as a hobbyist but i'd like a more comprehensive understanding of things like open/closed loop EFI, ignition wave forms, etc.

[Anonymous No.2651227]

In the pictured circuit, how do i determine the proper winding ration between the inductor?

[Anonymous No.2651230]

>>2651227
Looks like an antenna and variable capacitor.

[Anonymous No.2651232]

>>2651230
Yes but what I do not understand is how many turns should the receiver winding be? Does it depend on the antenna input impedance?

[Anonymous No.2651235]

>>2651232
https://www.electronics-tutorials.com/filters/antenna-pre-selector-filters.htm

This is an AM radio?

[Anonymous No.2651236]

>>2651235
https://industrial-electronics.com/measurement-testing-com/rf_design_7.html

[Anonymous No.2651549]

>>2651198
>If you switch transformer on when it passes through zero (90° delay form when voltage is 0 or peak voltage) there shouldn't be any spikes in theory
Ah that makes sense. I guess that also means resonant converters (i.e. no spikes of dV/dt) can use smaller magnetics.

[Anonymous No.2651614]

how does my first solder look?(top right) i know this is baby shit but im really glad i was able to fix my remote control..........

[Anonymous No.2651617]

>>2649781


-I have a bunch of broken tv lcd displays bulked there (mainly from my family breaking them while drunken), but the only thing damaged it's the lcd panel, now since getting a spare display from oems/shops/aliexpress/amazon/mouser/newark, isn't cost viable, it is possible to use the logical and driver boards for tvs using a different panel display? (lets say a display driver from a 50 inches for a 32 one).
afaik, it can, but not work well, since the voltages levels are the same for the lcd control (but for the backlighting who knows), perhaps it will be too dark, or too much contrast (or too few).

One of my brothers has a tv with no rca audio output and wants to connect it to a home teather, so its possible to get the analog speakers signals , solder to a 3.5 jack (perhaps adding a lm/max for audio amplification) and make it able to connect it?
I'm not used to audio signals, so I-m unaware if it will be distorted or the expected levels will blow the speakers?

[Anonymous No.2651631]

>>2651614
looks like shit, start again

[Anonymous No.2651639]

>>2651617
> swap panels
It’s like an organ transplant. It’s better if it comes from the same family. In this case, same product line and time. There is some standardization, i have successfully switched lcds on portable dvd players successfully.

> connect tv speakers to home theatre line in
Yeah, it’s possible, but you want to dampen the tv speaker signal which could be as high as 60 volts down to below 1v.

[Anonymous No.2651649]

>>2651617
>I-m unaware if it will be distorted or the expected levels will blow the speakers?

if the speakers are matched to amp's power, they should never blow.
as for distortion, it'll probably happen.
you wanna use 2 resistors on each channel to attenuate signal by 10x.
or just use 2 trimpots to get variable attenuation.

[Anonymous No.2651671]

>>2651614
Looks perfectly fine. Nice work.

>>2651617
>no rca audio output
Does it have a 3.5mm output? Toslink? HDMI ARC? Bluetooth?
If not, I'd want to bodge into the wires before the built-in amplifier. That should make it a decent line level that any amp can pick up with a bit of volume fiddling. If you tap into the lines after the amp, you'll have a higher signal level that you need to attenuate, and probably also a higher noise level.

[Anonymous No.2651688]

>>2651639
swapping power sources board would (theoretically) work? (at least the ones giving the same nominal power rate).
>as high as 60 volts down to below 1v.
this is what I don't know, I was thinking using an oscilloscope but perhaps it's too drastic.
>>if the speakers are matched to amp's power, they should never blow.
but let's say I'm unaware of the levels reaching the line in of the home t heather, and suppose I don't use any regulator, resistor or any method to limit a max input voltage, does not anyway the home theater board would have his own protections? (diodes, a dedicated ic, or whatever).
>you wanna use 2 resistors/trimpots
Thanks I will.
>>2651671
>Does it have a 3.5mm output?
nop, the thing only have the toslink, and bluetooth.
But there are many bt devices out there, perhaps there would be interference, since I don't have the schematics I dunno where are the audio lines pre and after the amps of the tv.
I'm going to try to measure the output levels, since the power received by the speakers should be at least 10x the one expected to an line out/headphones.

[Anonymous No.2651708]

Hello /ohm/.
I'm using this chinky plug/jack combination to charge an ebike. current should be around 3A max.
Will it melt in the long run?
So far it's been working fine.

[Anonymous No.2651709]

>>2651708

[Anonymous No.2651711]

>>2651708
I mean I've installed this jack in around 4 different batteries.
Would be a huge pain if I had to replace them all.

[Anonymous No.2651712]

Forgot to mention it's all 2.1x5.5mm.

[Anonymous No.2651723]

>>2651708
Just use amass xt60s like everybody else in the world please.
This message was brought to you by /rcg/

[Anonymous No.2651724]

>>2651723
I do. Bur not for charging

[Anonymous No.2651733]

>>2651688
If the hifi system has a digital input, you can get a simple TOSLINK to coaxial digital converter. If not, you can get a TOSLINK to twin RCA DAC. That one is going to be more expensive than the modification though. If you do bodge an analogue output on it, I'd add some nice panel-mount RCA sockets to the housing of the TV for a professional job.

>I'm going to try to measure the output levels
Why tap into after the amp instead of before it?

[Anonymous No.2651745]

>>2651724
> not for charging
Huh… weird. The first thing I do is lop any fuckin 2.1’er off and slap on some amass. If I want the 2.1 back for any reason (like i was too lazy to convert the other side immediately) i slap another one on the plug stub. Picrel is the one that happened to be on my desk.

Those are xt30s though… a bit small. If you have fat fingers use the 60s not forr the current handling but for the ease of gripping the little buger.

[Anonymous No.2651757]

>>2649781
test

[Anonymous No.2651770]

>>2651757

[Anonymous No.2651831]

>>2650104
>>triac
>fails short circuit and burns your house down
>nothin personnel, kid

[Anonymous No.2651833]

>>2650442
just use more in parallel lmao

>>2650935
neighborhood-wide PFC

[Anonymous No.2651860]

>>2651770
damn bro.... this is deep af...... stay woke.....

[Anonymous No.2651861]

>be rewireing old lamp with split sockets
>simply have a separete set of cables for each
>now four wires are sticking out of the bottom (two neutral, two live)

How do I join them into one each?
I can't find a connector for three cables (two in, one out)

And idealy it doesn't take too much space since I want the merging to happen in the space under the lamp.

>electro-tape / shrink tube
But is it secure enough? I mean it's 250v

[Anonymous No.2651886]

>>2651745
I rather have two different plugs for charging and load. Maybe I'm just retarded.

Anyways. On amazon where they sell the same chink shit for a huge markup, the max current for this particular jack is rated at 1A.

Does this mean anything? I have seen the same plug on higher current chargers. Should I give a fuck?

[Anonymous No.2651887]

>>2651886
inb4
>reddit formatting

[Anonymous No.2651890]

I have a 3volt lithium battery and two modules that use 3.3 volts each.
Would it be a good idea to make a voltage multiplier and then a two way buffer to power those two things?

[Anonymous No.2651931]

>>2651890
Would depend on the specs of the battery/module/project but I'd probably just wire the battery to the module and see if it's happy. If not then I'd swap the battery for a lion/lipo and build a simple zener regulator for it. Stepping up the voltage only to regulate it back down sounds like an efficiency nightmare which is not good for battery powered stuff.

[Anonymous No.2651935]

>>2651860
frfr nobody's woker than me....... i'm the wokest....... no cap ong not bussin.......................

[Anonymous No.2651961]

>>2650055
Go with codesys. Has a built in free emulator and many cheaper and PLCs (PLCs from Wago, Eaton, Schneider and Beckhoff at least used to be) are programmed with codesys.

If you want it to be really representative of industrial work (most factories are on either Allen Bradley or Siemens) and a physical plc: an s7-1200 is $450 ish and a copy of step 7 basic is $300. The modern Siemens stuff uses Tia portal but the transition is pretty easy. Theres also ekbtool but I’m not allowed to talk about it

[Anonymous No.2651975]

Is 400 samples a second enough to measure RMS of a 60hz sinewave?

[Anonymous No.2651980]

>>2651975
It depends on your other components, mainly your anti-aliasing filter.

[Anonymous No.2652009]

>>2651980
It'll be software based so whatever I want. Any suggestions?

[Anonymous No.2652010]

>>2651961
>Theres also ekbtool but I’m not allowed to talk about it

No one knows who you are, but maybe someone you don't know who sees this post will discuss it as we respect your silence.

[Anonymous No.2652015]

>>2652009
>It'll be software based so whatever I want.
So you're doing a simulation? Or are you processing data? You need to provide more details if you want to get useful answers.

[Anonymous No.2652016]

>>2652015
I'm monitoring current using a current transformer, ads1115, and esp32.

[Anonymous No.2652025]

>>2652016
You don't need to dwell too much on details then because you have enough processing power to use full 860 SPS of ads1115. The benefit of higher sample rates is that you're going to get more accurate readings if signal gets distorted and higher harmonics start showing up. Unless you want to do extreme power saving there is no need to go lower, just use max. sample rate of the device.

[Anonymous No.2652027]

>>2651975
Nyquist criterion says 120 is enough. But if you’re delivering power to something (so it won’t be a perfect sine wave because of load impedance) and want to measure that, it depends on the impedance of whatever you’re measuring

[Anonymous No.2652053]

>>2652025
I'm monitoring three phases using two ads1115's so one of them is splitting half it's rate between a and b phase which is why I'm wondering if 400 s/s is enough.

>>2652027
Monitoring mains coming into my place so it'll be a near perfect sine. Looking into Nyquist criterion now.

Here's the raw data at 3 different sampling frequencies. It seems that 400 is close enough that it catches all the peaks. 250 s/s only catches every 8th peak. I think the anti-alias filter is what I need to extrapolate the data. Just have to read up on which one to use. Thanks.

[Anonymous No.2652058]

does anybody make intel 80186 compatible microntrollers?

[Anonymous No.2652059]

>>2652058
*microcontrollers

[Anonymous No.2652068]

>>2652058
I believe InnovASIC is still making their IA186 which is a clone of AMD's AM186 which is a clone of the Intel 80186.

https://datasheet.lcsc.com/lcsc/2007061530_Analog-Devices-IA186ESPTQ100IR03_C661086.pdf

[Anonymous No.2652073]

>>2650935
Make an EMP, that capacitor can dump a lot of energy

[Anonymous No.2652076]

>>2651614
I'd have said it's fresh from the factory if I didn't know better! You did a very good job for your first solder joint, mine were very shit at first. Always remember, you cannot have too much flux!

[Anonymous No.2652086]

>>2652068
Digikey wants 55 bones for one of those.
I can get a whole laptop from goodwill for $20 with only 10 year old technology instead of 40 year technology.

[Anonymous No.2652113]

>>2652053
>I'm wondering if 400 s/s is enough
Instead of plotting samples with straight lines going between them, put the data stream through a reconstruction digital filter. It should become a nice sine wave with unclipped peaks, even with 120S/s. Though such a low sample rate can easily get messed up with harmonics, so having an anti-aliasing filter becomes more important. I'd want at least 30dB (~30V/V) of attenuation between the passband (60Hz) and the stopband (200Hz), which is only 0.5 decades, so you're looking at a 3-pole conventional filter. Get yourself a high sample-rate capture of the actual current noise on your mains to see what kind of spikes you get. Especially with SMPSs and phase-fired dimmers, inrush current too. If you get a sudden spike of 1000A for a microsecond that would still swamp a sub-par anti-aliasing filter and give you a strange artefact. I'd personally use audio ADCs, since they're cheap and sample at 100kS/s, and have some degree of anti-aliasing filtration built in by virtue of being delta-sigma topology.

[Anonymous No.2652123]

>>2652086
Oh, I thought you wanted to find a supplier. If you're just after nos then you can get them for 50c with $3 shipping off alibaba. https://www.aliexpress.com/i/3256805254629716.html

>>2652113
Those are all probably really great ideas that will take me time to understand. I'm slowly figuring it out. So far I'm synchronizing at the zero crossing then extrapolating a perfect sinwave that matches the current data point. Then I can see how close it was to the last few measurements for sanity and calculate RMS from there.

I thought about using ads1015's for their 3300 s/s but if 120 s/s is enough then at 400+ s/s I'm over sampling by 3x already. Also I'd rather have the 15bit resolution. It'd not be possible to spike the input up to 1000a as the max reading the ads1115 can send, 32768, only equals around 102.5a. The current transformers (SCT-013-000) themselves tap out at just over 100a anyways. Thanks again.

[Anonymous No.2652127]

I'm trying to lay out a schematic in KiCad. When you're working with things like TL074s, the built-in library is wise enough to know there are 4 op-amps to a TL074 IC. Let's say there's another IC that I have to make a symbol and footprint for, that's also a Quad something-or-other.

How can I make KiCad associate 4 of my custom symbols to a single IC footprint?

[Anonymous No.2652129]

great op pic

[Anonymous No.2652130]

i just realized it is a 90V battery right under your ass. oh joy

[Anonymous No.2652147]

>>2652123
The point of a higher sample rate ADC is to make room for a less extreme rolloff between the end of the passband (maximum frequency you want to measure) and the start of the stopband (the nyquist frequency), that way your anti-aliasing filter has less of an issue. It also gives you the ability to measure higher harmonics, which may actually be necessary for better data-keeping. Just averaging out spikes using a low pass filter will not give you a true RMS measurement of current.

>>2652127
Pic related options when you create the symbol. If it's an op-amp then not all the units are interchangeable, since you have the power input to include as a different input. So you're best off just copy-pasting the op-amp symbol into each one. I think so at least, I've never done it.

[Anonymous No.2652158]

>>2652147
Well I do have a passband filter of sorts already in place. The current transformers have a working frequency range of 50 - 1KHz so I don't really need to worry about anything happening sub-millisecond. I believe their inductive resistance also acts as an anti-aliasing filter too right?

[Anonymous No.2652163]

>>2652147
I didn't even see that. Thank you.

[Anonymous No.2652187]

>>2651745
>>2651886
Most BMS systems have seperated load and charge contacts

[Anonymous No.2652249]

>>2652129
>this died for ipods and phones

[Anonymous No.2653006]

>words words words words words
where the FUCK are all the schematics?

[Anonymous No.2653012]

>>2653006
We only do big brain things here. Schematics are trivial matter left as an exercise to the reader.

[Anonymous No.2653069]

I'm having some trouble with a circuit of mine; pic related. It's a precision full-wave rectifier followed by a diode clamper that feeds into a model of an SSi2164 VCA. The gain of the VCA is -infinity when the control voltage ("B") is 3.3 volts, and then the gain is 0 when the control voltage is 0 volts.
I need the response to be non-inverted (higher control voltage -> less attenuation), and I want the VCA to remain off when there's no signal present. To address these issues, I made the rectifier inverting, followed by the diode clamper shifting ground up to 3.3 volts. This effectively "corrects" the control response.

The issue is, I need the clamper to output a steady 3.3 volts if there is no control signal present, to keep the VCA off. Initially it's at 3.3 volts, but steadily drops about 20 millivolts every second. How can I make the clamper maintain that voltage? Am I going about this all wrong?

[Anonymous No.2653101]

>>2653069
Why the high-pass capacitor? If you're wanting to add a DC offset, do that with op-amps. That cap seems to be the obvious source of drift. Well that and leakage current through the diode, for which adding a pull-up resistor after the capacitor would be a decent idea.

[Anonymous No.2653119]

>>2653101
>Why the high-pass capacitor? If you're wanting to add a DC offset, do that with op-amps
that was the idea, to add the offset with an op-amp, using the diode clamp. But the clamp calls for the capacitor, which is why it's there.

I forgot I could just bias an op-amp to make a DC offset. I'll do that instead. Now just to figure out the lin->log amplifier.

[Anonymous No.2653142]

>>2650651
I will say, one is higher frequency than the other.
>source?
I made it up.

[Anonymous No.2653171]

>>2653119
this gives me a second question. My PCB is going to have a 3.3V rail, and I need 1.65 volts to bias the op-amps for the voltage offset. There will be 16 identical copies of that circuit, each needing 1.65 volts biasing.

What's the smart way to practically implement that? The obvious way is to use a resistor divider but something tells me there's a smarter way to do it

[Anonymous No.2653232]

>>2653171
You mean, 16 seperate circuits? Or 16 circuits all on the same board? If they can share a voltage reference then that's going to be easier.

If there's no DC component to the current going in/out of this middle ground rail, then you can use a resistive divider with large capacitors on it, which serve to decrease the source's effective impedance at higher frequencies. But depending on the current, even with just AC you may find you get unacceptable ripple on this rail, or need stupidly large capacitors or stupidly small resistor values. And judging by the diode circuits I doubt this assumption holds in the first place.

You're likely best off just buffering a voltage divider with a spare op-amp. Adding some output capacitance can help take some of the load off the op-amp, but it isn't necessarily stable. The same goes for adding multiple op-amps in parallel. I'd look for an appnote on these methods. Adding some input capacitance is good for rail stability and noise immunity. If you need even more current capability, then it isn't out of the question to use transistor buffering circuits in conjunction with op-amps (beware of crossover distortion) or to try to use a high-power op-amp or dedicated power amp for buffering your ground rail.

Don't forget to consider the common-mode input range of your op-amps with respect to your supply rails. Chances are you'll either have to add high-pass capacitors before each input, or reference all your signals to this 1.65V rail instead of 0V. In other words, turn the whole system into a ±1.65V one, where the new 0V / ground rail is your intermediate buffered rail. Personally I'd recommend this method.

If you need an exact reference value (e.g. for a lin/log converter) you may find that a voltage divider based method won't suffice. In this case you could either use a TL431 or similar based compensated reference diode and buffer it for your entire ground rail, or use one separately for just the more temperamental circuit.

[Anonymous No.2653257]

>>2653232
I've been posting in here about my project intermittently the past few months, I'm designing and implementing a musical vocoder for my senior capstone project.

When I say 16 circuits, the vocoder is going to have 16 bands. Each band has a modulator subcircuit (pic related, but this schematic is outdated now), consisting of a bandpass filter that feeds into a precision full-wave rectifier, that then feeds into the (new) biasing op-amp, which then feeds into the control pin of a VCA. All 16 circuits will be on the same board, but it's looking like it'll be a large board (or a double sided board, not sure which yet. It depends on if abusing jumper resistors is a bad idea or not).

So, there's going to be a 3.3 volt rail for the microcontroller. Then, I need to bias the biasing op-amps with 1.65 volts. Since it's just going to be biasing the op-amp's noninverting input, it's not like loading is going to be much of a concern.

[Anonymous No.2653322]

>>2653257
Assuming you don't also have a negative voltage rail, will the audio signals be oscillating on either side of 0V? If so they'll be outside the input range of op-amps. That's more the point I was trying to get across. Op-amp input range is one thing falstad doesn't really simulate.

[Anonymous No.2653336]

>>2653322
It's dual rail, +-12 volts; though what you're mentioning isn't an aspect of op-amps I'm very familiar with. I'll look into it.

[Anonymous No.2653341]

>>2653336
>It's dual rail, +-12 volts
Oh, then you'll be fine, I was misunderstanding you. I guess you only need the 1.65V rail for handling signals that are going in or out of the microcontroller. If that 1.65V rail will only be handling the non-inverting inputs of op-amps for adding DC offset, a simple voltage divider will be fine. Though adding a large-ish capacitor to couple it to your 0V rail is a good idea.

[Anonymous No.2653350]

>>2649781
where to get started with what?

[Anonymous No.2653376]

>>2653350
>where to get started with what?

getting in touch with your inner ohm.
to increase happiness, confidence, and sexual potency.

[Anonymous No.2653451]

Think this layout will work?
Im trying to generate HVDC to charge a bank of capacitors then blow shit up.

[Anonymous No.2653453]

Is it just me, or is this a terrible amplifier circuit? There's an hFE factor of difference between the inverting gain and the non-inverting gain. The example schematics show them using the same size resistors on both inputs, but when I try simulating that it doesn't work for shit. Because it's current-mode I can't see any feasible method for getting negative feedback to suppress the open-loop gain in a way that isn't dependant on both hFE and resistor ratios, but maybe I'm a moron.

[Anonymous No.2653463]

>>2653451
This is the full circuit I was thinking of putting together.

The capacitors I have a rated for 2600vac so I was going to put 5 in series. Hopefully that circuit will put out ~11kvdc after the diodes. The isolating transformer will keep the high voltages away from me and I can wind the voltage up slowly with the variac.

[Anonymous No.2653470]

>>2653453
>There's an hFE factor of difference between the inverting gain and the non-inverting gain.
Where? You just have difference of input currents on the second transistor's base. Then you have top part for sourcing current and bottom for sinking.
>>2653463
It will work fine, but you need to be really careful around this thing if you don't want to get fried.

[Anonymous No.2653472]

>>2653463
>>2653470
Only thing that you should add is resistor before capacitor bank because applying 11kv to discharged bank will pull a lot of current and could fry your transformers. Value of resistor depends on transformer current rating and internal resistance of the bank.

[Anonymous No.2653477]

>>2653470
>You just have difference of input currents on the second transistor's base
I don't understand. Is the first transistor + diode with "current mirror" written on it meant to somehow act like a transistor with an hFE of 1? Instead of just drawing it as a current mirror?

[Anonymous No.2653479]

>>2653472
Nice. I didn't include a current limiting resistor because I was planning to bring the voltage up slowly withe the Variac. I think I have some 100ohm 50w resistor could add. Will give this a go next weekend.

[Anonymous No.2653480]

>>2653479
how many kva can your variac handle?

also watch that you get the secondaries hooked up with their polarities the right way around. you could cancel the voltages out and have to flip a mot or two around

[Anonymous No.2653491]

>>2653477
>meant to somehow act like a transistor with an hFE of 1?
Yes, since it is a current mirror it has by definition a hFE of 1 (current in = current out). They drew it as a diode to emphasize where the input current flows, it's exactly the same as current mirror with transistor that has collector shorted to base. What's important is that drop across diode becomes equal to drop across base emitter junction and if PN junctions are matched then currents will be equal. The only bad thing about that image is leftmost I_in+ arrow because it should be pointing down across diode as base current is negligible. As is drawn it implies that +input current is flowing into base which is wrong.

[Anonymous No.2653531]

>>2649781
I am looking for pic related IC or reference design. Basically i "just" want to put a nvme SSD to a sata+power connector.
Adapter like these seem to exist on Aliexpress (2.5 -> m.2 M), but i would like to design my own due to size constraints. Unfortunately i can't find anything on Mouser/Digikey, so i am assuming to be too stupid to know what i am exactly looking for.

[Anonymous No.2653555]

>>2653531
You just want power from the SATA connector? No IC needed for that. Just split off the 3.3V and route it to the M.2 slot.

[Anonymous No.2653556]

>>2653531
I see now. Check the config pins in this link:
https://pinoutguide.com/HD/M.2_NGFF_connector_pinout.shtml

Your BIOS may need modification to enable PCIe lanes for NVMe connectivity.

[Anonymous No.2653561]

>>2653555
>>2653556
No, I would like use sata connection on the main board and get a adapter to use an nvme/pcie ssd card on that. Just like pic related. Yes, I know it will be slower, if that's the question.

[Anonymous No.2653627]

CAN Bus question... hypothetically lets say I'm a truck driver that is really annoyed with all the bullshit safety sensors and beeping shit that I can't disable. Also hypothetically, lets say I want to change the governed speed (trucks are limited to certain speeds via the company or the leasing agencies insurance requirements). Is it possible to send messages to the ECU to reconfigure its settings through the CAN bus?

I have several breadboards, a few arduino UNOs and the arduino CAN bus shield. I have been programming as a hobbyist for years (javascript/c#/python/c/c++) and have a little experience with lower level assembly type stuff.

So question is, can I send messages to alter settings in ECU through CAN? If thats not possible, would it be possible to intercept messages from other nodes prior to them going through the arbitration process and then spoofing values to the ECU? (like, if speed sensor wins arbitration, I intercept the packet and alter the value such that it is always below the governed speed, despite how actually fast I may be going to prevent the governer from engaging?)

How would you go about this in this hypothetical situation?

[Anonymous No.2653634]

>>2653627
If you have to ask you're not ready and it's best to live this shit alone. Last thing you want is for your arduino board to glitch out and completely confuse ECU that would result in catastrophic failure.

[Anonymous No.2653647]

>>2653634
>assuming my collision mitigation system doesn't already glitch out and cause catastrophic failure

This thing mistakes bridges on a grade as if they were a stopped car. Come winter, you have ice and under load this is a nightmare waiting to happen. There is also a roll sensor that calculates it should apply brakes... in the apex of every sharp curve on I-77 in west virginia mountains. The dealership tech don't even know what the CAN bus is. They just plug their volvo software into OBD and if there are no engine codes, they say its good. If there are, they clear the codes and tell me to drive and see if they pop back up. There are literally zero people in this industry that know how these machines operate on the road aside from idiot drivers that weigh 300lbs and don't know how to even log into their own email. You either DIY or nothing changes. You say I'm not ready. Okay, fine. Let me ask you this: when will I be? What will I need to do to be ready? What do I need to know? Where should I start? Tell me wise one. Where did you start your journey of debugging AEB systems for large trucks, the advent of which hasn't even begun to happen in consumer vehicles? (they have only been prevalent in the last handful of years, so this is cutting edge tech)

If this thing freaks out as-is and causes me to jacknife into a schoolbus I'm going to prison and there isn't a soul in my company or industry for that matter that would have the expertise to take the stand and say "did we test the AEB sensors and calibration?" prior to the verdict. So yeah, I will never be ready until I do it, which is precisely why I have my questions

[Anonymous No.2653655]

>>2653647
You're hilarious. Godspeed with your mods.

[Anonymous No.2653667]

>>2653647
Those systems are just too complex to do some "arduino CAN bus hacking". I wouldn't be surprised if messages are encrypted or there is some validation/authentication mechanism that only trusts data after a correct response to a cryptographic challenge. Your best bet is to try unplugging things and see if that disables just the subsystem or whole vehicle. For example what happens to AEB if you disconnect radar module? But again this is a bad idea because you have no idea how system will respond and could make things potentially worse due to unforeseen bugs that not even manufacturer has tested for.

[Anonymous No.2653677]

>>2653647
Different anon, but you could theoretically spoof the sensor data to bypass at least some of the automated shit.

[Anonymous No.2653678]

>>2653667

You cannot get away with simply unplugging modules, and the reason is that there is telemetry software of some kind that the carrier (dispatchers, safety personnel) have access to that shows if/when these systems go offline. If they find out its been tampered with, its grounds for termination and flagging of your DAC report (trucker credit score) that pretty much ends your career.

Now, in the last few years of driving I've been exposed to two different trucks... the freightliner cascadia and the volvo vnl860. From carrier to carrier, their fleets are packaged differently and a lot of these fleet vehicles are using aftermarket 3rd party telemetry systems, but it is my understanding that most of what gets sent to the carrier/company is data that is already required by the federally mandated ELD. The FMCSA has compliance requirements for logging listed on their website, and a lot of it is key on/off, braking events, overspeed events, and gps logging stuff, the bulk of which I assume is tied into the CAN bus system. Me and another driver have ruled out the possibility of speed calculations based on GPS... their logs are much much more precise, however they do correlate with GPS location when flagged.

Anyways, to answer your question.. in the freightliner if you disable the radar module you will get a check engine light as well as a dialog pop-up on the inst cluster. Your cruise control will be disabled entirely (also happens when your radar unit is obstructed). This dialog will keep reappearing, obstructing the menu system entirely as well as the other readouts, and the service engine light will trip. Its simply not worth disabling, not to mention you will get a phone call from safety asking you to get it serviced immediately. On the volvo, you can block your radar unit and you only get a dialog on the inst cluster ONCE at keyON. You can clear it and it stays gone and you still have cruise (albeit no adaptive cruise). I have mine blocked currently

[Anonymous No.2653679]

>>2653677
yeah I was asking about doing this here >>2653627

[Anonymous No.2653690]

>>2653678
Fucking hell, they probably don't even have this level of surveillance in China. Even if you mange to somehow spoof CAN bus data and GPS all it takes is one slip up and misconfiguration which will probably give you a lot of legal headaches if they find out (not just your loicence, but also criminal charges - computer fraud, illegal modification, FCC violations...). You really don't want to mess around with Mr. Goldberg's goyslop logistics.

[Anonymous No.2653696]

>>2653690

I mean, I get your concerns and trust me, I've thought about it a lot. I've debated with other drivers. I've weighed my options here.

I want you to imagine a scenario:

>be you
>get into your car
>"man I really really want some Takis. Guess I'll head to the store!"
>you pull out of your driveway
>"BRRRT" *car beeps loudly*
>"FUCK"
>side collision sensor false positive bc of your mailbox
>heartrate is about 190bpm
>start cruising down the street, calming down
>kid on bicycle crossing the street at a stop sign
>car slams on brakes
>FUCK
>turn onto the main drag
>put turn signal on to change lanes
>"HONK HONK"
>thats the lane departure warning that triggered because your signal terminated prior to fully entering the lane
>car adjacent to you gets in front of you in your lane
>"BEEEP BEEEP BEEEP"
>There is less than 300 feet of following distance so your front radar collision sensor is causing the beeping
>"SHIT, let me disable this shit"
>fumble through your cluster menu
>Its already on the lowest setting possible
>decide to hop on the freeway
>its a marked 25mph curve but you've taken it at 55mph plenty of times, no problem
>halfway through you feel the brakes engage
>BWZZZAAAT
>its your roll sensor
>your phone rings
>"Hi, its John from Geico. Everything okay??? Seems you triggered the roll sensor for the third day in a row..."
>"y-yes everythings okay"
>"Well... actually its not Sir. We keep having these events and we're gonna have to put you into a higher risk bracket. I'm just trying to save you money here... right now you're looking at a $100 monthly increase"

How far away do you think this is?

[Anonymous No.2653701]

the future is now, faggots
>https://www.nhtsa.gov/press-releases/automatic-emergency-braking-proposed-rule

[Anonymous No.2653708]

>>2653480
I have a 5kva varic I'm not too concerned about blowing any of my circuit up. The thing I'm most worried about is hv feedback into my house.

[Anonymous No.2653712]

>>2653696
>How far away do you think this is?
2030. They have been pretty open about their goals for a while now. Don't forget your CO2 credit deductions because system has detected above average engine load. Also
>Hi, its John from Geico. Everything okay??? Seems you triggered the roll sensor for the third day in a row..."
>John
would be more like
>HELLO SIR!? *call center noises in the background* DO YOU HEAR ME, THIS IS KUMAR FROM GEICO. YOU HAVE JUST TRIGGERED A ROLL SENSOR ALERT SIR. *heavy breathing* PLEASE SIR ARE YOU OK SIR! LISTEN TO ME SIR! DUE TO REPEATED EVENTS, *static noises* -UR SMART CONTRACT HAS BEEN VIOLATED SIR! ADDITIONAL FUNDS WILL BE TAKEN FROM YOUR DIGITAL WALLET SIR.
It's going to get a lot worse.
https://www.youtube.com/watch?v=UrEUzKTt7j0

[Anonymous No.2653724]

>>2653696
Not far.
My car is over 10 years old, and the last few times I took it in because of the service light, it turned out to be bad sensors.
The last one, they told me it was too hard to get to to be worthwhile and they suggested putting electrical tape over the area so I don’t see the light. They tested it though (it was coolant pressure) and the car was actually fine.
Actually the last one was someone “stole” my oxygen sensors, so that cost a lot more to fix (vs. just straight-piping it) when they took my cat.
Of course my tire inflation sensors don’t work any more, so they’re always on. That’s a given.
Everything seems like sensor/computer issues already.

[Anonymous No.2653789]

Got a pair of chink SSRs from Amazon (not absolute bottom-of-the-barrel ones, but close) and am using them to create a PID-controlled toaster oven for baking PCBs. These SSRs are notorious for being underrated, sometimes with "60A" units being shipped with 4A triacs. Tested one SSR with a 1A incadescent bulb and it didnt get warm or detonate when switched on. The heating elements are 5A each. One SSR and element are wired up now, just working up the nerve to switch it on.

Not trying to blogpost, but i'm posting this here so that when the news reports "Mentally challenged man found dead in house fire" you can all get a good laugh out of it

[Anonymous No.2653854]

>>2653789
My digital panasonic toaster oven uses tiny relays, i assume they use zero crossing switching. I use it for everything, so those have switched the quartz elements tens of thousands of times.

[Anonymous No.2653918]

>>2653854
>ZCS relays

[Anonymous No.2653971]

>>2653789
Cool stuff. Gonna make your own phase-fired or otherwise mains-synchronous controller? Or just use an off-the-shelf PID controller? Going the former way means it’s a lot easier to program in a temperature-time curve. I’m doing that for a reflow plate. Though maybe those PID controllers have a serial port or whatever for automating them.

[Anonymous No.2654088]

>>2653491
Ah I see, the sim works properly now. Interestingly simple circuit, though I doubt using it with resistors is better than using a long-tailed pair.

[Anonymous No.2654093]

What did he mean by this?

[Anonymous No.2654116]

>>2653463
You can double the voltage if you rearrange those diodes a bit.
>440µF @ 11kV
post a pic of the cap that's presumably 50 times the size of >>2650935's cap and could evaporate a dog

[Anonymous No.2654120]

>>2654093
Chop a nigger in the head

[Anonymous No.2654218]

What values should I have in each input measuring with a tester?

[Anonymous No.2654221]

>>2654218
That’s a high voltage for an ink jet cartridge. Must be european.

[Anonymous No.2654240]

>>2654218
family values

[Anonymous No.2654307]

Anyone here worked with dwm3000 drivers for stm32? for some reason, the library loads itself thrice? once for each device that they support? if that's what's going on it's eating all of my FLASH.

[Anonymous No.2654325]

so whats Q a function of in an inductor?

[Anonymous No.2654327]

>>2653971
Right now i'm feeding it unsynchronized PWM from an STM32 at a slower frequency than mains; the heating elements have so much thermal mass that the ~5Hz strobing isn't noticable. I had originally planned on building a zero-crossing detector to synchronize the micro for phase-angle control, but the chinks lied to my face about the SSRs being non-ZCS and my sketchy solution seems to work, so i'm not gonna bother.

>>2654325
interwinding capacitance, winding resistance, inductance

[Anonymous No.2654329]

>>2654327
>interwinding capacitance, winding resistance, inductance
idk what happened but i self made a small inductor that has gargantuan Q. id like it to have less. but not less inductance since thatll shift resonant frequency.

[Anonymous No.2654331]

>>2654329
Q of an inductor is equal to 2*pi*f*L/R. If you want to reduce it add series resistance.

[Anonymous No.2654334]

>>2654331
aight ty

[Anonymous No.2654366]

>>2654327
>unsynchronized PWM from an STM32 at a slower frequency than mains
Oh that works too.

[Anonymous No.2654532]

I successfully solved this puzzle (13dbm = 20mW into 50 ohm = 1Vpp) but holy shit why do they do that?! Measuring signal amplitude in some meme units instead of volts. Why??

[Anonymous No.2654565]

>>2653678
Start by sniffing it to see what is going on. I successfully used an arduino to both spoof modules and read engine data on a nmea2000 (basically marine CAN bus) network. You would think it would be super secure but they probably rely on the fact that there aint too many truckers capable of fucking with it. No offense.

Find a convenient connection point and just watch. Aint gonna hurt nothing just to look.

[Anonymous No.2654600]

how come my voltage inverter isn't working?
0 volt input at base should mean 12v output at 'inverted' net

[Anonymous No.2654614]

>>2654565
sniff the bus

[Anonymous No.2654625]

if i use a tranny as an inverter which then drives a relay coil, do i need to put a diode between relay coil and tranny?

[Anonymous No.2654650]

>>2654532
>Why??

dbm's let you add/subtract each stage's contribution.
addition/subtraction is easier than multiplication/division.
just like with 13-year-olds, you can do them in your head.

[Anonymous No.2654652]

I am making a custom male to female adapter cable. The female end is a loose part with no casing around it. I want to make a DIY casing around the female end so that it is more like an adapter cable you would buy in a store, ie. like the ends on a USB extension cable. What can I do to DIY this? Also, is there a name for the casing around a cable terminus that is specific enough to use for internet searches? I struggle to find anything that deals in this topic, and maybe it's because I don't know what to call it.

[Anonymous No.2654658]

>>2654652
>to use for internet searches?

there are some connector types that are generic enough to have for sale individual parts like
- connector enclosures
- connector housing
- connector boots

if you dont have such a connector, you can either 3-d print it, or hand-shape it from sugru.

[Anonymous No.2654661]

>>2654532
Imagine getting filtered this hard. You've got a long way to go before you can enjoy your first big boy signal generator. Looking at the world through a logarithmic power perspective makes life so much easier. There's also convenient shortcuts with practical considerations when everything is in terms of dBm. Sometimes shit will get expressed in terms of dBuV and I fucking hate it.

[Anonymous No.2654668]

>>2654652
>>2654658
Backshell is the other term, usually in conjunction with D-sub style connectors. If you can't get such a thing, I'd try 3D printing one. Either out of solid plastic with a TPU strain relief, or wholly out of TPU. It's also an option to 3D print a mould to pour casting silicone or casting urethane rubber into, instead of using TPU. You can also use a metal spring for strain relief.
I've used blobs of epoxy adhesive for this purpose once or twice before, but the result is a bit shitty and comes loose after a few years, and it lacks the elastic strain relief. Sugru is a better idea.

What type of connector are you using? If you haven't decided then going for a D-sub is what I'd recommend. You can key them uniquely by cutting off one pin and gluing it in the corresponding socket. It might even be possible to swap a pin for a socket if you're lucky, to keep the full pin count.

IDC ribbon cables are possibly ok.

[Anonymous No.2654669]

>>2654532
Because they’re retards. Also, non-retards also put thhe smaller number first (e.g. 1 μV to 1 V) so you know it was written by retards, too. There are other typographical errors in that snippet too. Most likely you have a chinese clone where the editing was done by indians—all of whom were non-technical.

This is obviously just the tip of the iceberg, so you’d be better off just measuring it yourself.

[Anonymous No.2654709]

does this layout have any chance of actually working IRL? i'm taking a brushless DC motor controller for fun. nothing really high power, 200-300 watts maximum @ 10 amps. i tried to minimize bootstrap loop area, and used polygon pours wherever possible. i'm worried about switching noise dicking up the analog circuitry due to the op amps being where they are. i'm also lead to believe i should be using thermal vias to improve the FET's cooling. should i just spam these on the PH1/PH2/PH3_motor and VCC polygons?

stackup: signal, gnd pour, vcc pour, signal
U1/2/3: IR2184 gate drivers
U8/9/10: current sense op amps (on the back side of the board)

[Anonymous No.2654711]

>>2654709
oh and should i be using polygonal pours to join the gate drive resistor/diode in parallel or is my ugly ass routing sufficient?

[Anonymous No.2654721]

Soldernoob back again. So I sanded off my tip and very quickly got rid of the oxidation (though not all of it, see the black in pic). I turned it on and put it in tip cleaner powder and put solder on it. I could get it to melt much faster and easier now, but it still doesn't spread out onto the tip and keeps balling up.

I wiped it off with a wet sponge and brass wool, but how am I supposed to know if I tinned it correctly? I think I'm just removing the solder I put on rather than spreading it out on the tip, because it just looks the same after I just sanded it, and isn't shiny.

[Anonymous No.2654724]

>>2654721
You aren't supposed to sand the tip. The core is copper and it's plated with iron. The iron is plated with nickel-chrome. When you sand the nickel-chrome plating off, the tip won't wet and molten solder dissolves the iron and copper.

[Anonymous No.2654727]

>>2654724
It was already fucked. I'm only seeing if I can continue using it.

[Anonymous No.2654730]

>>2654727
You've got to keep the tip clean while in use, and tin it before powering off so it's protected from oxidation. A temperature controlled iron is also a must for normal tip life.

[Anonymous No.2654731]

>>2654709
Looks decent. I usually put the like FETs in the same direction, not flipped relative to one another, but it hardly matters. Also you can ditch the trace between the two drain pads per FET, allowing you to run other traces or fills between them, if you want. Yes, spam thermal vias on the PHx polygons. On the VCC polygons too. Consider putting corresponding polygons on the reverse side for better heat sinking.

The current amps shouldn't give a shit about noise, since they have really low input impedance. Maybe just load the output of each amp down with 1k-10k (read the datasheet for an idea of output current vs distortion).

If you want to measure bipolar current you should tie one of the VREF pins on each amp IC to Vcc, which is a requirement if you want to do regenerative braking. If you don't already have these parts lying about, it's an option to use a DRV830x or TMC6x00 3-phase gate driver IC. It's also an option to add a TMC4671 FOC pre-driver before your gate drivers, which is what I'll be doing in the near future with any luck. FOC is more important for low-speed operation. Though the TMC4671 wants phase inline current-sensing, not low-side current-sensing, so you'd have to use different amp ICs like the INA240Ax or the INA241xx.

Also what feedback method are you using? Hall or back-emf or some sort of encoder?

[Anonymous No.2654735]

>>2654731
>Also you can ditch the trace between the two drain pads per FET
is there a way to make KiCad shut up about them being left unconnected?
>Also what feedback method are you using? Hall or back-emf or some sort of encoder?
all 3 kek. this is just a demo board to learn how BLDCs work and to try various things. i plan on trying out six step trapezoidal commutation, sinusoidal PWM commutation, and FOC.

i have a magnetic angular encoder (MPS MA330) which supports ABZ encoder, UVW hall sensor emulation, and absolute positioning. i also have 5V hall sensors so the PCB will have a 3.3V/5V level shifter to talk to the MCU (hidden from my screenshots.) and then i also have a BEMF setup with an LM339 zero crossing detector, also hidden from my screenshots.

[Anonymous No.2654748]

>>2654709
I would use signal, gnd, gnd, signal as stackup as it's the best from noise perspective. You don't generally need internal power planes unless you're trying to power some big digital ICs that take 100s of amps of current. Each signal via, especially those who carry fast edges - digital signals, should also have a GND return via close to it to prevent noise from propagating in resonant cavity formed between gnd planes. There are just 2 basic rules for routing signals that could generate noise
>vias always come in pairs (signal + gnd)
>never break return path, if there is a signal on top there should always be return GND under it. If you break it you're going to make fields spread everywhere and have noise problems.
Watch https://www.youtube.com/watch?v=icRzEZF3eZo and other videos to fully understand how you should route everything. Everything will make sense when you start looking at how fields propagate.

If you have the space separate the parallel traces as much as possible as less noise will couple between them. For FET thermals you need to find your heat losses and then apply
>https://fscdn.rohm.com/en/products/databook/applinote/ic/power/linear_regulator/to252_thermal_resistance_information_an-e.pdf
As quick approximation you'll have about P = 0.0075 Ohm * (10A)^2 * 0.5 = 375mW of static loss, and then you also need to determine switching losses which are a bit harder to do properly, but if you approximate with linear waveforms P = 0.5 * V_max * I_max * (t_r + t_f) * f_sw = 0.5 * 15V * 10A * (40ns + 20ns) * 50kHz = 225mW. You can sum that together and round that shit up to 1W to be safe then choose appropriate copper area based on previous app note.

[Anonymous No.2654755]

>>2654661
>drop 1dB
>lose 20% of your signal
haha yes, such useful, very intuitive

[Anonymous No.2654757]

>>2654748
thanks, will review this.

maybe a dumb question but for static loss calc, why did you multiply by 0.5?

[Anonymous No.2654762]

I wanna make resol vbus to uart converter, however I don't quite understand schematic provided by manufacturer. Is my interpretation of picrel correct and VCCL has to be ~2V with resistor values in the schematic or am I missing smth?

[Anonymous No.2654768]

>>2654757
0.5 is there because I assumed 50% duty cycle between top and bottom transistor so average power is half of what continuous through one transistor would be.

[Anonymous No.2654779]

>>2653531
>>2653561
Shameless bump.

[Anonymous No.2654808]

>>2654762
>http://danielwippermann.github.io/resol-vbus/#/md/docs/vbus-specification
That's the shittiest documentation that I've ever seen. I think VCCL is assumed to be 3.3V, which would make sense since comparator is of open collector type. Levels for mark and space are 2,25 V and 2,00 V respectively, which means 0.25V of hysteresis. Vref is 1.65V. Packing R4, R3 and R9 into thevenin equivalent gives R_eq = R4 || R3 + R9 = 13.4k, with V_th = R3 /(R3 + R4) * V_in = 0.34*V_in. Divider ratio changes because of positive feedback. When output is pulled low

V+ = R6/(R6 + R_eq) * 0.34 V_in = 0.32 V_In

when high use superposition

V+ = (R6 + R8)/(R6 + R8 + R_eq) * 0.34 V_in + Vccl * R_eq / (R_eq + R6 + R8) = 0.32 V_in + 0.18

As Vref is 1.65V, switching happens at V_in = 1.65/0.32 = 5.16 and 4.59 which is strange and means I messed up or schematic is messed up. But then again if we ignore feedback (hysteresis) for a moment V+ switches at 1.65V (Vccl = 3.3V). This means it needs 1.65V/0.34 = 4.8V on the input which is outside specified range. You would have to change R1 to 2.4k. Then switching points will be at 3.5V and 2.94V which is closer to what they want. But I still have no idea what levels are supposed to be space and mark because somewhere it says 3.25 and 3V. Best just to measure and then adjust R1 as necessary.

[Anonymous No.2654831]

>>2654711
You should always use polygonal pours everywhere you can to maximize your FREE COPPER that you’re paying for.

Unless you’ve got a HV clearance or capacitance issue.

[Anonymous No.2654847]

>>2654808
> a lot of calculations
Also, you’d immediately guess it would be something like 3.3 V or 3.5 V just glancing at it because it’s a UART output.
If in doubt, just breadboard it and raise the voltage slowly until it starts working.
Sometimes I do this just for lower power operation, then When I find the lowest voltage it runs on, I just blip it up by 5 or 10 percent depending on wire length and voltage drops and whatnot.
Finding the upper limit is usually destructive to the device itself, and/or what it’s plugged into.

[Anonymous No.2654864]

>>2654847
>3 equations that you can read from schematic directly
>a lot of calculations
It's not that much work to do this analytically. And you're also assuming that circuit/values are correct which seems not to be the case in given example. There is no reason to skip quick back of the envelope calculations which verify your basic assumptions. If you actually want to git gud at analog design I recommend watching
https://www.youtube.com/watch?v=RDJkmgEKLk0
You'll be surprised how easily you can solve circuits that seem so complex at first, but with few theorems, that not many people know about, you can just fly through the problem with very little effort. Nobody told us about this even in my master's EE course.

[Anonymous No.2654905]

I am building a custom male to female adapter and the female end is being built with a loose part that normally mounts to a PCB. I want to make a wrapping around this end, so that it is like a store bought cable, with a proper casing around the ends of the cable, but I am struggling to find resources on how to make one. Is it possible to DIY a cable plug casing/wrapping, and if so, what tools and resources do I need? Also, what is the wrapping/casing at the end of a cable usually called? I've been searching but i cannot find any references online to what I want to do. I suspect I am missing the key word.

>>2650267
take it apart and clean the buttons and contacts for starters. might just be dirty.

[Anonymous No.2654926]

>>2654669
That's an official HP manual for a top of the line signal generator from the 1980's. Chinese or Indians didn't exist back then.

[Anonymous No.2655016]

>>2654905
>what is the wrapping/casing at the end of a cable usually called?
One end has a strain relief on it and the other end is a plug and housing or socket and housing.

[Anonymous No.2655021]

not sure exactly what the amp rating is for this fuse on my dead fan motor.. 250v @2A (hence the F2?)

[Anonymous No.2655045]

>>2654730
I can't get it to tin properly though, even after sanding it shiny.

[Anonymous No.2655049]

>>2655045
wow, no shit

[Anonymous No.2655076]

>>2655021
Maybe it's a thermal fuse that goes open-circuit when it surpasses 115°C?

[Anonymous No.2655082]

What are some of my options on this?
I tried cleaning it what I could, one of the components came off though and another came loose but I was able to stick it back down with solder
I'm sure many traces are damaged though.
It's the main board for a thinkpad from 1994

Is this stuff even important either? The connector is for the IrDA, and the torn off one was the ram standby battery which leaked and won't be needed anymore.
The laptop turns on but does not post fully

[Anonymous No.2655103]

>>2655082
It's impossible to know that there isn't corrosion beneath components, unless you have a schematic with which to test continuity. I'd desolder all the components in the area, then clean the thing with citric acid and a toothbrush. I'd solder bodge wires not just across broken traces, but through crusty vias.

[Anonymous No.2655107]

>>2655103
Yeah I know I minimum have to desolder everything but Sheit idk how to reestablish microvias

[Anonymous No.2655146]

>>2655049
Yeah, okay, then how do? What am I doing wrong.

[Anonymous No.2655154]

>>2655021
yea if def is (and did in this case) and i'm trying to replace it but I am not 100% on the amperage rating of the fuse. I am trying to figure out the same for this one from a burnt out DC adapter.

[Anonymous No.2655168]

>>2655154
That one looks like it breaks at 130℃

[Anonymous No.2655183]

>>2655154
>I am not 100% on the amperage rating of the fuse.

amperage is not relevant, only temp.
typically they'll use 10A-capable fuses in things that draw considerably less.

[Anonymous No.2655201]

>>2655146
previous post >>2654724 said you probably fucked the plated coating with the sanding. short of re-plating it yourself, i don't think the tip is salvageable
if you really want to be ghetto, maybe try sanding through the iron coating down to the copper base. idk if the iron layer takes solder well, but copper should pick up solder easily

[Anonymous No.2655202]

>>2655201
Iron takes solder well enough that it was the main material of the "jam it in the fire to heat it" kind of soldering iron. Copper conducts heat better but dissolves too quickly for long tip life.

[Anonymous No.2655205]

>>2655202
Thats why you use savbit solder, it has copper in it already.
Note that Oatey® Silver Lead Free Plumbing Wire Solder also has like 3% copper in it. You could try that to tin your tip which should offer a bit more protection, and higher heat tolerance (melts between 420 and 460 degrees farenheight). Also easy to obtain.

[Anonymous No.2655208]

>https://www.falstad.com/circuit/circuitjs.html
In the 555 square wave example, what's the point of the 1M resistor?

[Anonymous No.2655225]

>>2654721
Is this a blue turd iron?
Just bin the iron.
You need plumbing flux to tin well oxidized copper btw.
>>2654724
Plating is iron, not nickel-chrome. Solder doesn't dissolve iron, but it does dissolve copper pretty fast at lead-free temps. At normal solder temp - meh.

[Anonymous No.2655228]

>>2655208

see end of page at
https://electronics.stackexchange.com/questions/600758/555-timer-with-50-duty-cycle

[Anonymous No.2655232]

>>2655228
This is a good description. If only the CMOS 555s had as high a current sourcing/sinking ability as the original, then I'd recommend them a lot more.

Though of course falstad doesn't simulate those finer details of the 555.

[Anonymous No.2655249]

>>2655228
Oh, that makes sense, thanks.
And I'm already using a CMOS version (TLC555) anyway.

[Anonymous No.2655251]

is anyone here from ##electronics on IRC?

[Anonymous No.2655263]

>>2655251
>##electronics on IRC?

yeah, 20 years ago.

[Anonymous No.2655264]

>>2655251
>##electronics on IRC?

yeah, 20 years ago.
my nick was 1N914.

[Anonymous No.2655270]

Okay, I am completely retarded, but please help me.

I want to take a car seat with active ventilation (the button is on the seat).
How do I know where the control unit is located?
How do I connect the connector attached to the seat to a small 12V moped accumulator?

Also let's hypothetically, the seat is also heated but the button to turn it on isn't on the seat, how do I go about connecting it?
Thanks.

[Anonymous No.2655283]

>>2655270
>How do I know where the control unit is located?
Follow the wires. The controller may be in the seat, or it may be under the seat nearby, possibly shared with an electric seat-moving controller. It may be centralised near the ECU, but I kinda doubt it.
>How do I connect the connector attached to the seat to a small 12V moped accumulator?
Look up the pinout and pray you don't fry its CAN bus.

>the seat is also heated but the button to turn it on isn't on the seat, how do I go about connecting it?
Either find a connector to the controller where a button should be connected and plug it in, find a connector where a controller with button should be plugged in and plug it in, or find a connector where a controller with button should be plugged in and just wire it directly to 12V. That last option may well bypass existing protection circuitry and/or temperature feedback, so isn't really advised. Shoving a 45°C thermal switch inside the seat might be a passable hack.

[Anonymous No.2655296]

>>2655283
Thanks a lot!

[Anonymous No.2655433]

So i am at a dead end.
Bought LEDs without checking what current needs they have and they run at a weird 0.135mA at 36V. So ether i construct something with a LM317 for each LED or i go with an adjustable power supply. But the only common ones go up to 30V. So i just shelved the project for some time.

[Anonymous No.2655441]

>>2655433
Maybe they’re designed to run on rectified mains when you put 4 of them in series. Don’t forget a
1 MΩ resistor. And other such safety precautions.

[Anonymous No.2655457]

>>2655441
Doubtful as this is not america and they run at normal PSU voltage.
This whole project started with me wanting a high CRI aquarium light with no spyware channel dimming(at least on of of individual channels). I then scaled it down to just a single channel. As such i already have a PWM controller. I then scaled it down to just a high CRI reading/painting light without digital dimming need.

[Anonymous No.2655494]

>>2655433
>0.135mA at 36V
What kind of LED is that? Did you mean 0.135A?

I think simplest way to drive an LED is a simple transistor current limiter circuit as you don't need precision that much. But I'd go with some sort of driver from aliexpress.

[Anonymous No.2655511]

>>2655494
Meant 0.135A.
Am pretty sure to have settled on a BCR 421 or BCR 420 based limiter. But that would i need to design myself. Wich would also partially solve my power deliver pains as i previously just soldered directly to the cob. Wich they strongly advise against and was a pain in the ass. So i stopped halfway there. Will probably design some PCB where the cob would be connected to the driver or ground via a spring

[Anonymous No.2655531]

hello, I need help, I solidered a cable for my headphones to an 1/4 and now I'm getting:

>Left Audio outputs on Left Side
>Right outputs on Both Sides
>Stereo output on Right and gives weird tones on Left

can anyone help me with this?

for Context some weirdo put a DX12-4pin aviation female plug into the DT990 that I have and it has an additional cable for Video as far as I know

[Anonymous No.2655550]

>>2655511
>Am pretty sure to have settled on a BCR 421 or BCR 420 based limiter.
Well, they basically do the same but less shitty.
> But that would i need to design myself.
It should be simple, you can probably even make a PCB with dremel and diamond ball thingy.

Other thing, chinks are overly optimistic about specs, so instead 135 mA you'd prob want 100 mA.
If it is a normal LED (just one type of chips), you can definitely run them at even lower current without CRI penalty.
But if it is weird sunlike chip that has violet and blue chips, you have to bias them properly or else CRI is shit... Well, 90 instead of 97.

[Anonymous No.2655553]

>>2655433
If making a mains circuit isn't an option (you can buy the mains voltage current regulator ICs from LCSC, probably also aliexpress), I'd get a constant current LED power supply capable of ~40V, and put all the LEDs in parallel. Naturally, the current setting of the PSU should be N_LEDs * 0.135. Putting ~10Ω in series with each LED should help to balance the currents, which only drops your efficiency by 4%.

>>2655531
Try swapping right and ground. Never trust the work of the previous guy. If the colour coding is wrong anyhow, i's not out of the question that the yellow wire carries audio and the video wire is one of the other three.

[Anonymous No.2655646]

What's the best approach to both attenuate and amplify a signal? Say, a signal generator chip outputs 0.7v to 1v pp depending on frequency, but I'd like to convert to a 0.1v - 3v pp range. One approach is to use a voltage divider for attenuation followed by an op amp with a fixed gain. But that restricts the op amp BW unnecessarily. Another approach is to use a relay or a switch to switch between 1:1 and 1:10 attenuation followed by an opamp with a variable gain controlled by a pot. What do you think? I am not considering any fancy VGA options.

[Anonymous No.2655649]

>>2655646
use a VGA

[Anonymous No.2655654]

>>2655646
Whenever I want to chop .7v off a signal I just put it through a diode junction from an actual diode or a transistor. For .3 v I use a germanium one.

[Anonymous No.2655745]

>>2655646
differential opamp with 0.7v as the reference

[Anonymous No.2655835]

Are there encoders in the form of a switch? Like it would have position 1,2,3,4 and output a 2 bit number on each position. What is the name of what I'm looking for?

[Anonymous No.2655879]

>>2654748
is this any better? stack up top to bottom is: signal, gnd, gnd, signal.

vcc polygon is top/bottom layer vcc. i would extend this upward/downward to capture the other 3 phases, with plenty of via stitching. ph3_motor polygon is top/bottom layer. i would extend this to the right, with plenty of via stitching, and connect with the 3 wire terminal block to connect with the motor.

4 green traces on the left will continue to extend left to an off-board interface for the MCU.

2 pink traces are power rails, left one is +15V for the gate drive IC. right one is +3.3V for the current sense op amp.

[Anonymous No.2655882]

>>2655835
rotary code switch

here's one that sorta matches your spec:

https://www.mouser.com/datasheet/2/26/PT65%20Series%20Rotary%20code%202018-1314458.pdf

there's a version with a spindle. these are 4 bit but you can ignore the upper 2.

[Anonymous No.2655884]

>>2655882
thanks

[Anonymous No.2655898]

>>2655879
Could you move yellow traces to the top layer? It's better from noise perspective to keep inner layer as solid GND without any breaks if possible. This is especially critical under traces, you should always try to provide solid GND path that goes directly under trace. Otherwise this will happen
>https://www.youtube.com/watch?v=LdEHJXs5xpw
Each via that is used to change layers should also have a GND via nearby. For example via near R5 should have GND via near it. For your case these rules aren't that critical because signal rise times are not that fast, but it doesn't cost anything to make sure signals propagate properly and don't stray everywhere.

[Anonymous No.2655901]

>>2655425

[Anonymous No.2655908]

>>2655898
i understand your concern but moving the yellow traces to the top layer would bisect the VCC pours on top/bottom layer. this layout is more challenging than i expected...

[Anonymous No.2655909]

>>2655553
Thank you it works perfectly now, I added some extra wires and finished it half nicely with some shrink wraps

[Anonymous No.2655911]

>>2655908
If you're worried about FETs not getting enough current, 7.2mm trace can carry 10A no problem (use kicad's calculator to get widths). You can route power and only use pours for thermals. But yes, this is challenging and there are tradeoffs and you have to decide what's more important - noise or thermal/power handling. You shouldn't worry too much about it though, if things don't work out the first time you can always do another revision, custom PCBs aren't that expensive nowadays.

[Anonymous No.2655955]

>>2649781
I currently have a machine tool wired to a wall electrical box using SOOW, which is all inside the electrical box too (as individual wires) and the machine.
would it be going overboard to redo it all with conduit and MTW? or at least conduit leading to a plug that is code to accept SOOW which then goes to the box. I just learned SOOW is not code when used in conduit situations (with elbow fittings for instance)

[Anonymous No.2655998]

>>2649781
anyone covered what the deal with UV (385nm) LEDs is?
I have a powerbank that supplies a single LED of any other color just fine but the UV one I need to "jump start" with another color to get the power flowing and then it works as expected.

[Anonymous No.2656003]

>>2655998
Vf (other diodes) ~= 1.7V
Vf (UV) = OL

seems like the UV relies on avalanching?

[Anonymous No.2656005]

>>2655998
>"jump start"
connect in parallell that is

[Anonymous No.2656054]

I'm trying to make a PWM controlled current source using a Howland current pump.
>picrel is how far I got
How can I improve this? Are there any equations I could use to minimize the time needed a current is reached when PWM changes?

This is a repost

[Anonymous No.2656070]

Why did F&T become smol, anons?

[Anonymous No.2656110]

>>2656070
Better dialectic oils.
Too bad we’re not alowed to use PCBs any more. That was good shit. Also, lead paint was really good too.

[Anonymous No.2656112]

>>2656003
UV LEDs have a forward voltage of 3.5V or so, which is generally too high for a DMM to measure. Try connecting the thing to 5V with a 1kΩ series resistor, and measure the voltage across it then.

[Anonymous No.2656173]

I want to create a switch for mains voltage without exposing any 240V wires (for safety purposes). How do i accomplish this?

[Anonymous No.2656206]

>>2656173
>How do i accomplish this?

hard mode: one single-gang electrical box, one switch, one cover plate, two wire clamps that fit knock-outs in the box.

easy mode: one inline lamp-style switch

[Anonymous No.2656208]

>>2656206
sorry i mean a voltage controlled switch

[Anonymous No.2656215]

>>2656208
>sorry

replace switch by relay.
search by 2 things:
- voltage and current ratings of the contacts to match your load (e.g. 240V @ 10A)
- coil voltage rating that matches your control voltage (e.g. 12V @ 0.1A)

[Anonymous No.2656226]

>>2656215
my point is that those relay terminals are exposed

[Anonymous No.2656229]

>>2656112
Thanks. I heard it's because of (quantum) physical properties so that was comforting - no reason to hope for technical improvements!
I'll try a latching transistor pair with a dummy "kickstart LED" to circumvent the minimum load requirement of the powerbank but perhaps a well placed cap could do the job as well? I have a small collection of jellybeans but the UV LEDs are more rare in my collection so rather would not like to experiment with them.

[Anonymous No.2656230]

>>2656226
>exposed

they kinda have to be in order for you to make a connection to them.
kinda applies to just about voltage controlled switch.
but when you place the whole thing in a plastic box, you're safe and sound.
you can get alternately get a board where there's a lil more isolation.
but it still needs a box.

[Anonymous No.2656232]

>>2656229
...oh yeah, I tried with a white kickstart LED and that almost worked. I can see it pulse ever so slightly with every polling for a load from the powerbank (~2 Hz) but I guess not enough current gets through to unleash its full potential.>>2656229

[Anonymous No.2656237]

>>2656229
Never heard of an LED having that kind of behaviour before, it's almost certainly a stupid circuit problem rather than some sorta quantum bullshit. I'm guessing your power bank doesn't detect enough power drain from the UV LED in order to turn on properly, since it doesn't turn on very hard with just the unboosted ~3.7V from the internal lithium battery. Test it with a USB output of a computer instead.

Post a pic of whatever circuit you're trying to use, there better be current limiting somewhere.

[Anonymous No.2656284]

>>2656215
> matches your control voltage

Is there a standard for control voltages that control relays and/or solid state relays? Also, is there a connector associated with said standard?

I have a bunch of these kind of things, and sometimes need to control them with pwm, thermostats, motion sensors, cds cells, timers, etc.
For whatever reason I always put the relay, contactor, scr, triac, ssr, etc. in a separate enclosure from the control for flexibility. For 50 years or so I’ve been using RCA phono jacks to connect the control signals, but they’re all different voltages and whatnot and I’ve blown some of them up by plugging the wrong one in. As I get older, I’m making this mistake more frequently, so I want to standardize before I’m too old to rebuild everything every time I zap one.

[Anonymous No.2656314]

>>2656284
>Is there a standard for control voltages

yes.
all relay coils are rated for 12Vdc.
except for the other 90% which arent.

>is there a connector associated with said standard?

there are more standard connectors than there are relays.

[Anonymous No.2656335]

>>2656237
>Never heard of an LED having that kind of behaviour before, it's almost certainly a stupid circuit problem rather than some sorta quantum bullshit.
https://forum.arduino.cc/t/why-forward-voltage-of-uv-led-is-so-damn-high/503669
>I'm guessing your power bank doesn't detect enough power drain from the UV LED in order to turn on properly, since it doesn't turn on very hard with just the unboosted ~3.7V from the internal lithium battery. Test it with a USB output of a computer instead.
Yeah, it works well with a wall plug. It's a way to kickstart it automatically with the powerbank that is the issue now :) Final application is as a child's night light lamp.
I'll try something with transistors when I'm back at home. Of course I use a 100ohm resistor on the negative side, got two LEDs each with its resistor in parallel. Not much to show off really.

[Anonymous No.2656342]

>>2656110
>Also, lead paint was really good too.
it really was! after the banishment of lead paints a handful of boats in the local dock capsized due to the new paints raised the center of buoyancy

[Anonymous No.2656348]

>>2656335
how about an ldr or photodiode circuit that activates a transistor to short out the rails (with an appropriate series resistor) when the light is detected to be turned off?

also uv led for a nightlight sounds really bad for circadian reasons

[Anonymous No.2656349]

>>2656342
>highly poisonous paint
>but it was good because some ancient ships were designed for it!
Come on.

[Anonymous No.2656359]

>>2656349
Barnacles doesn't stick to it as hard either.

[Anonymous No.2656367]

>>2656359
Can barnacles be electrocuted?

[Anonymous No.2656542]

I'm building a power distribution box to use at my property. The main power line coming into the property is 3-phase (400V) on a 25A circuit breaker. I think it means that I effectively have 10kW at my disposal (400V*25A=10000W).
My electric box is going to have two 3-phase sockets and four single phase sockets connected to a 3P circuit breaker and a 1P circuit breaker accordingly. My question is what amperage should the circuit breakers be? I'm not sure if my calculations are accurate. I assume that I can safely use a 25A 1P breaker (230V*25A=5750W), but I'm not sure about the safe amperage for the 3P breaker. If my calculations are correct, I'd only be left with 4250W for the 3-phase sockets, which would only allow me to install a 10A 3P breaker (400V*10A=4000W).
Am I making any sense at all?

[Anonymous No.2656568]

How do I disconnect julet connectors that are stuck together without damaging them? Please help.
t. newbie

[Anonymous No.2656569]

>>2656568
pee on them

[Anonymous No.2656572]

>>2656542
Main purpose of circuit breakers is to protect the cables so that they don't burn up from to high of a current running through them. Cable downstream of breaker determines breaker amps.
>I assume that I can safely use a 25A 1P breaker (230V*25A=5750W), but I'm not sure about the safe amperage for the 3P breaker.
If you fully load this single phase you can't use 3 phase power anymore because any additional current on phase that already draws 25A will mean that you're pulling more than 25A and breaker will trip eventually.

You actually have 3 * 230V * 25A = 17.25kW of power available.

You can use 25A 1P breakers as well as 25A 3P breakers provided that:
>cable current rating (cross section) exceeds 25A with all safety factors
>cable length is less than maximum permissible which would prevent your breaker from tripping at short circuit (you can ignore this if it's only a few 10s of meters)
>sockets are rated at 25A

[Anonymous No.2656576]

>>2656568
I'm with the other anon. How strong is your pee stream?

[Anonymous No.2656592]

>>2656348
It was enough just adding a green LED with a high value resistor to make it dimmed. Didn't find that satisfactory yesterday but today I just want to move on.

[Anonymous No.2656599]

>>2656572
Ok, that clears things up somewhat, except this part:
>You actually have 3 * 230V * 25A = 17.25kW of power available.
Why 230V? I have been using 400V in my calculations because the electrical company supplies 3-phase power.

[Anonymous No.2656618]

>>2656599
You have 3 phases of 230V phase to neutral power. Each phase can provide max. 25A. By taking phase to phase voltage instead, you can get your 400V 3 phase power. It's the same thing just used in a different way.

One thing to keep in mind that if you use that 400V it means that you have 400V across the load or what is known as delta connection. In this case current in load is sqrt(3)*load current. So if you had 400V at 25A load current you would actually pull 43A on line which is way too much. Your max current in 400V 3 phase mode is actually 25A/sqrt(3) or 14.43A. If you calculate how much power is that P = 400V * 14.43A * 3 = 17.3 kW (there is some rounding error because of sqrt(3), if you put in the root it comes to exactly 17.25kW). If you don't want to make things complicated just stick to phase to neutral mode as you can simply think of the 3 phase system as 3x1 phase that returns through neutral.

[Anonymous No.2656628]

>>2656618
Oh damn it's starting to click, I'm understanding it now. Thank you for the clear explanation.
The interesting part is that I requested the electrical company to install 14kW power at my property, but it seems like I got a little extra. Maybe they use a different type of formula or something.

[Anonymous No.2656634]

>>2656628
While they do install 25A main fuses, your contract probably states that you're allowed to take max. 14kW. Anything more and you would pay extra if you were an industrial customer. For residential they probably won't charge you if you pull full 17kW, but you'll have to check how they handle this with your distributor. Then there is the whole can of worms with reactive power which is probably another reason for 14kW figure. Killowats measure real power, all previous calculations assume resistive loads so there is no reactive power. If you take that 14kW and assume power factor of 0.8 (electric motors) then you actually get close to 17.5kVA of apparent power.

[Anonymous No.2656641]

>>2656592
Remove green led and keep the resistor.

[Anonymous No.2656736]

>>2656576
>>2656569
Not very strong. Should I try first pee in the morning? Should I eat pineapple or something?

[Anonymous No.2656779]

got this old car battery charger that stopped charging
design is very simple with only a transformer, diode rectifier and this weird blue unknown part wich seem to be the fault (burn marks inside)
anyone know what it is ? heat fuse maybe ? can i just bypass it ?

[Anonymous No.2656780]

>>2656779

[Anonymous No.2656789]

>>2656779
Brown and Blue mains. Euro jobbie.
Yeah, a fuse can be self-thermally tripped as well. Still a fuse. Some of them reset after the bimetallic strip cools down and it snaps back to it’s original connected state.
Well, technically, I fuess even wire fuses are thermal ones-hot enough to wire

[Anonymous No.2656791]

>>2656779
>can i just bypass it ?

you can bypass it with another fuse, unless you like living dangerously.
often the input current rating is written on the information sticker on the back of the unit.

[Anonymous No.2656801]

how are you handling the fact that the entire field of electronics is about to change?

room temperature 0 resistance conductors, holy fucking shit
Do you have any idea what this means you dumb fucks?

V=IR
V=I*0
No fucking voltage drop, no heat loss, you can have a wire thin as the human hair which can carry literally infinite currents, billions of amps, for across infinite lengths of cable with 0 voltage loss
holy shit i my cock has never been this hard

[Anonymous No.2656807]

>>2656801
It's most likely overhyped academic fraud, as is the custom nowadays. Wait for others to replicate the results.

[Anonymous No.2656808]

>>2656807
then why is it coming fro a highly reputable quantum research center?

[Anonymous No.2656809]

FUCK
nevermind, it's fake
https://youtu.be/QHPFphlzwdQ?t=486
moving on
fuck the bugmen for raising my hopes up

[Anonymous No.2656810]

>>2656808
That says a lot about quantum research

[Anonymous No.2656811]

>>2656808
For all I care it could also come from some office located between curry shop and call center in Mumbai. It wouldn't change a thing.

[Anonymous No.2656849]

>>2656801
>No fucking voltage drop, no heat loss
too bad the same doesn't apply to semiconductors. good luck making cold-running cpus

[Anonymous No.2656854]

Why do people build their EMP devices like this (the top example) and not like this (the bottom example)? Wouldn't the bottom example have longer range?

[Anonymous No.2656862]

does this XL6008 SEPIC converter layout stand a chance at working? my peak current draw should be below 1 amp so i'm not worried about thermals, just the layout itself. i tried my best to minimize the return path -- i think? not sure about the exact # of layers (the PCB will have other components downstream) but it will feature at least 1 continuous ground pour layer.

[Anonymous No.2656878]

diy electronics is a form of obsession. i love it but i also hate it. i no longer think of simple things, like taking pictures of flowers and birds and go for a walk or camp by the lake. i spend all day studying theory and selecting parts on LCSC and dealing with paralysis through analysis. autism is kicking in. but hey, at least i don't have to fix the drywall in the garage since i convinced myself i have no time for that.
i am thinking one day i will move to CA from the midwest, buy a tiny shack for 1kk usd with no basement and no room for a radio shack and i will let myself free and i will hike all the day long and grill shrimp. i just realized that life is too short. i can spend three months optimizing my shopping cart on LCSC.

thx for reading my blog

[Anonymous No.2656883]

>>2656862
I would add more GND vias to GND pin of the IC. Then maybe move C18 and C20 to the top so that they would be at the same level as D10. This way high frequency noise gets to ground as soon as possible. The most problematic loop is from SW through C14/C15, through D10 and then through output caps C18, C20. Did you also consider using SMD output/input capacitors? Pic related is difference in loop sizes. It's a small improvement, but as is layout would probably be fine.

[Anonymous No.2656888]

>>2656878
You should really lean about the art of engineering approximation. It's too easy to fall into scientific mindset of trying to understand things completely and apply too much theory.

[Anonymous No.2656893]

>>2656888
well i always start that way, high level details etc, but i can't make progress and get stuck at some point thinking about how the maxwell equations drive electrons in my circuit and then i read on the theory of the laplace transform and it takes 6 months and then i forget everything including the details of the project i was working on.

[Anonymous No.2656896]

>>2656883
thanks, that makes a lot of sense. i do have some 0603 10uF capacitors on hand (C16/C17 are this size and capacity) but i'm not really saving space relative to electrolytics. i'm not exactly sure of the input/output capacitance requirements but most designs seem to use between 100-470 uF.

[Anonymous No.2656898]

>>2656893
One project at a time. Write a short outline of each project, and a checklist to track progress. Combine your bills-of-materials and order passives in bulk. Dedicate a fixed amount of time X days per week and log your hours against your checklist. Document everything to reduce debugging time.

Adderall / cocaine helps.

[Anonymous No.2656902]

>>2656896
Ceramics will generally have lower impedance and will shunt more high frequency noise compared to electrolytics. Even better if there are more in parallel with big via stitched fill to ground. Only issue is loop stability and if you don't want to take chances stick to what datasheet recommends. In worst case things could begin oscillating.

[Anonymous No.2656905]

>>2656893
Build your prototypes quickly, and fix the issues that present a problem later. Easier to separate theory from reality in a shorter time that way. Do like 20 iterations of your PCB.

[Anonymous No.2656907]

>>2656801
You know, copper is about 99% superconductive for all applications.

[Anonymous No.2656914]

>>2656893
>thinking about how the maxwell equations drive electrons in my circuit
That's physics. We don't do that here. We don't need to do that here:
https://www.youtube.com/watch?v=_ivqWN4L3zU
>and then i read on the theory of the laplace transform
We don't do that here. You just use table of transforms/CAS systems and leave theoretical things to mathematicians. At some point you just have to accept to let things go and accept that it would be impossible to learn about everything. Even if you start to really understand how laplace transform works, then what? Your whole reality collapses because you find out about Gödel's incompleteness theorems? Lay off the hard math before it's too late. That's reserved only for turbo autists and even they get fed up with everything (like Perelman).

[Anonymous No.2656920]

>>2656898
>>2656905
feature creep is a real thing. i typically want to build "a thing", say, a signal generator or an inductance meter, or some other meter. and only have a vague idea about the specs and then discover that the parts i ordered are not really good for the job and then i want to add more features or improve the specs and add buttons/switches and there is no room and i have to redesign everything. and then i start thinking if i really need to build the damn thing. i may be better off buying it since i will never be able to match the specs of a decent device. but if i buy all the equipment, what the hell am i going to build?? i could buy anything.

[Anonymous No.2656922]

>>2656914
>Even if you start to really understand how laplace transform works, then what?
a bit of a dopamine kick, i guess. that's the pretty much the goal of everything we do.

[Anonymous No.2656924]

>>2656920
c'est l'autisme

[Anonymous No.2656927]

>>2656922
Yes, but you have to optimize your dopamine kick/effort ratio. It's not exactly healthy to loose sleep over some abstract concept so you can be happy for 5 minutes and pretend we're not actually incredibly retarded. It's almost as if we're cucking ourselves by felling proud of another man's achievements. Could you come up with laplace transform on your own? I know I wouldn't. I'm just too stupid and lack brain capacity to do anything of that caliber.

[Anonymous No.2656934]

>>2656914
https://www.youtube.com/watch?v=rp8hvyjZWHs
>>2656893
Math problems require Ballmer peak to understand.
>>2656888
Engineer is a jewish scientist.

[Anonymous No.2656941]

>>2656920
Are you studying for a degree, or are you farting around as a hobby? You don't have to get everything perfect if you're not held to company standards and gov regulations. Don't start a fire or cause a catastrophic failure of any integrated systems, use isolation where needed, ensure signal integrity/noise mitigation, and balance the thermal load on the PCB.

[Anonymous No.2656950]

>>2656941
Fire is a feature. I was making an LED light, and well, I forgot/miscalculated voltage over resistor. Hm, I guess light emitting resistors count as a light source too.
Catastrophic failure is also nice.

[Anonymous No.2656951]

>>2656927
>on your own
That applies to even the very basic electronic building blocks. Oftentimes I want to solve a problem and come up with a design on my own without looking up a ready made solution. But there is no way I'd come up with anything smart and useful that has been known for decades, sometimes even centuries. I'd try to reinvent the wheel only to realize that I didn't account for a bunch of things and it is not gonna work. I mean some people swear they "reinvented" some basic building blocks, like current sources, or buffers, or certain kinds of amplifiers or transistor configurations etc. without realizing they existed. That would be awesome.

[Anonymous No.2656955]

>>2656941
>Are you studying for a degree
I have an EE degree but I never worked as an EE since there's more money in coding. Coding is easy but could be mind numbing. That must be some kind of an unconscious remorse that motivates me to refresh my memory and re-learn things. Plus I want to learn some antiquated analog stuff like radio circuits which I never properly learned in college since it is not 1970's anymore. At some point everyone realizes that digital is easy. Analog is hard. Duh.

[Anonymous No.2656965]

>>2656950
Look at the bright side, you have a cheap electric ignition device for explosive holiday celebrations.
>>2656955
>some kind of an unconscious remorse that motivates me
Motivation is the important part.
>digital is easy. Analog is hard.
Maybe you're just more familiar with digital circuits. Things get weird with RF, both analog and digital.

[Anonymous No.2657028]

Are power supplies with a coax output connector common?
Found one in my garage. Such a hack work. It is mounted on the garage door opener and screwed onto the AC outlet cover so it wouldn't fall off due to gravity and an ugly coax cable runs all the way across the ceiling to the legacy coax box. And the best part is the power supply is dead. It is 12V 500mA. WTF is that all about?

[Anonymous No.2657040]

>>2657028
It's a DC power injector.

[Anonymous No.2657064]

Quick question, I'm fixing the TV of my parents and I've found one of the culprits to be an IGBT with the part number RJP63K2 8WB 1E55. I've found an Amazon listing with "RJP63K2" in the same form factor TO-220F (not sure about the F) but without the other numbers after it. That should work as a replacement correct? Anything I should be aware of?

[Anonymous No.2657067]

>>2657064
>TO-220F
The tab is encapsulated in plastic.
https://en.wikipedia.org/wiki/TO-220#Variations

[Anonymous No.2657074]

>>2657067
Thanks, the one I'm replacing has no exposed metal parts either.

[Anonymous No.2657240]

>>2656854
Magnetic loops are inductive antennae, while short dipoles are capacitive antennae. You need an inductive component to the circuit for it to oscillate properly anyhow, so using the inductor as your antenna is easy. If you use a short dipole (e.g. significantly smaller than half the wavelength) you have to worry about matching components for maximal transmission strength, which is a pain.

>>2656878
>i can spend three months optimizing my shopping cart on LCSC
you are like a baby.

[Anonymous No.2657282]

>>2656955
analog's way easier than digishit wtf are you talking about

[Anonymous No.2657293]

>>2656801
I'll believe it when I see it. Extraordinary claims require extraordinary evidence.

[Anonymous No.2657374]

>>2657293
okay give me a single proof that room temp superconductors aren't possible

[Anonymous No.2657377]

fucking get in here lets goooo
http://electronicschat.org/

[Anonymous No.2657392]

I got this 500W AC motor from an old blender, I’m not too knowledgeable about them but it looks like a brushed motor to me. What’s the cheapest kind of speed controller I could use for this, with a control knob? I don’t need exact rpms just want to dial it down to like 20 or 50%

[Anonymous No.2657412]

>>2657392
>want to dial it down to like 20 or 50%

a simple diode rated 10A will get you there.
if you must have variable control, try a lamp dimmer.
if it blows the triac, replace with a beefier model.
they're typically only rated for 300 W.

[Anonymous No.2657438]

>>2657377
>slower than /ohm/
>need to have client running 24/7 because IRC boomer tech
>:P XD :D
>namefagging
>nobody actually talks about electronics

[Anonymous No.2657442]

>>2657412
Thanks, but 500W dimmers are €40 ish, would something like this work?
I’ll be running it for maybe 10-20 seconds at a time

[Anonymous No.2657445]

>>2657442
multitap transformer

[Anonymous No.2657461]

are there any EC11 style encoders (encoder + switch) with long threaded sections? i have to mount them to a stupid thick enclosure that doesn't leave enough threads for the nut to grab onto. the shaft length it's fine, but the threaded portion needs to be like 10-12mm deep. i did some searching on mouser/digikey but i'm not finding much.

[Anonymous No.2657501]

Can anybody identify the connector that fits the footprint south of the SATA DOM? Dell Wyse 5060 (N07D)

[Anonymous No.2657523]

>>2657442
>would something like this work?

it'd work if the specs were real, which they're not.
they use "china watts" so you divide by 7.

>running it for maybe 10-20 seconds

if it's for occasional use, you can just put various gadgets in series.
things i'd use: toaster, electric drill/saw/sander, 300W halogen bulbs in parallel, 500W projector lamp, 1000W toaster oven, 1200W portable heater, 1800W hair dryer.

[Anonymous No.2657524]

>>2657501
>identify the connector

it says PCIe right above it, meaning PCI express.
could be used for an NVME SSD or wifi cards, or other stuff like a MP4 accelerator.

[Anonymous No.2657531]

>>2657461
>long threaded sections?

very rare.
on wood or plastic, you use a forstner bit to hollow out room for the encoder.
of course you dont have one.
but you might have a spade bit, or a dremel (and some patience) or a sharpened spoon (and a free week).

[Anonymous No.2657532]

>>2657524
It's connected to the PCIe bus, but the connector is (I think) a FPC/FFC type with a flex cable that attaches to a PCIe x1 or x4 slot on a separate PCB. I want to connect an external GPU riser directly so I can use the m.2 slot for another SSD.

[Anonymous No.2657533]

>>2657531
it's a cast aluminum enclosure. i don't have the tooling necessary to mill it thinner.

[Anonymous No.2657554]

>>2657533
Mount the encoder behind the plate, and attach the mount to the plate. Then use a shaft extender (huh huh) to couple it to your knob (hehe).

[Anonymous No.2657601]

Is there a contradiction?? Which is it??

[Anonymous No.2657614]

>>2657554
kek. i might try to make a 3d printed adapter since long thread encoders apparently do not exist.

[Anonymous No.2657681]

I need to use one of these big 12v LED buttons send a signal to an ESP32, do I seriously need to hook the button into 12v DC and put a level shifter in between them or is GPT being retarded? Can't I just hook the button to a 9v battery use a buck converter to pull everything down to 3.3v or something? It's not like the button is sending serial data and I only need this thing to work for an evening so I'd rather use a 9v battery and buck converter than a 12v power bank and a level shifter like GPT is shilling me on

[Anonymous No.2657687]

>>2657681
Are you planning on controlling the LEDs through the arduino or are they going to be constantly on?
Also ignore anything GPT tells you, it's the definition of "confidently wrong".

[Anonymous No.2657691]

>>2657687
LED constantly on.

[Anonymous No.2657696]

>>2657691
Then you wire them exactly like normal buttons. One contact to GND, the other to IO, with internal pullups enabled. The LED connects directly to your 12v line (assuming it has a built-in resistor.)
A pic of your buttons could also help.

[Anonymous No.2657715]

>>2657681
what the fuck are you on?
they work like normal buttons and the led works like a normal led. sometimes they have a series resistor. jesus christ man

[Anonymous No.2657746]

i have a bs170 mosfet.
can someone please design a low output impedance, high input impedance voltage buffer for me

[Anonymous No.2657822]

>>2657746
no

[Anonymous No.2657850]

>>2657746
Google "common drain amplifier". Then think for the next few hours about how to bias it properly. The simple answer is probably just a high-impedance voltage divider providing a DC bias directly to the gate, and a high-pass cap to inject the signal. Even just a cap and a resistor to the positive voltage rail alone (no corresponding bias setting resistor to ground) should work so long as your voltage rail + signal amplitude is lower than the FET's maximum gate voltage. Then just add a high-pass stage on the output. Pic related. Resistor and capacitor sizes are up to you, they determine the impedances and output power rating.

[Anonymous No.2657876]

>>2657850

[Anonymous No.2657920]

>>2657850
Man, there should be like an archive of references for common motifs like this. I shouldn't have to derive the impulse frequency of the transfer coefficient every time I wanna switch a light or some shit

[Anonymous No.2658009]

The equivalent circuit for a ferrite bead looks misleading. Rpar (R1) is not really a resistor, is it? More like a frequency dependent resistor? But somehow when I use a spice model for a ferrite bead I get the same Bode plot as the discrete components of the equivalent circuit in the pic. Someone is cheating somewhere or I totally miss the point. Perhaps the bode plot just shows the same shape since it doesn't matter much what happens at the low frequencies? Basically, the difference would be in the behavior of the parallel resistor. In the proper model its resistance would grow with frequency up to the SRF and a discrete resistor would obviously show the constant resistance. But perhaps the difference wouldn't be noticeable in the Bode plot.

[Anonymous No.2658023]

new thread
>>2658021

[Anonymous No.2658025]

>>2658023
too early, summerfaggot

[Anonymous No.2658029]

>>2658009
That parallel resistor is there just to flatten the resonant peak. Frequency dependence comes from inductor and capacitor. At low frequencies L2 is short and the whole impedance is equal to R2. As frequency rises L2 increases its "resistance" (impedance) and curve goes up. At some point impedance of L2 is much larger than R1 and total impedance is dominated by R1, so curve is flat. Then C1 comes along and starts dropping impedance, effectively acting as a short at high frequency.

[Anonymous No.2658037]

>>2658029
Well but the parallel resistor is called an "Rac" as opposed to Rdc which is a separate small DC resistance of the inductor. So it is some special kind of an "AC" resistor and here's what its resistance looks like. It appears that it does depend on frequency. But I am not sure how that would matter if it was just a regular resistor with a constant resistance over the entire frequency range. It would be dominated by the inductance at low frequencies. And yes it would flatten the curve since it reduces the Q. But I am still not sure why bother with a ferrite bead, if one could just use LR or LCR in parallel. LR would probably suffice since it would still have some parasitic capacitance. Wouldn't it behave just like a ferrite bead and attenuate high freq noise?

[Anonymous No.2658044]

>>2658037
>LR would probably suffice since it would still have some parasitic capacitance. Wouldn't it behave just like a ferrite bead and attenuate high freq noise?
It would, but it wouldn't be as practical. Ferrite materials used in beads have complex permeability. Normal inductors store energy in magnetic field, ferrites dissipate it so you get parallel resistor for free together with increased inductance because of core material.

[Anonymous No.2658238]

>>2658044
multi-channel DAQ

[Anonymous No.2659215]

>>2658023
still not archived
faggot