/diy/ - Do It Yourself

>>2533439
I'm the BMS anon. Figured out how to do current amplification correctly, everything seems to work out in the sims.
I'm debating if I need a voltage doubler on board so that I can turn on the NFETs fully or are they going to be fine at 7V or 13V or whatever the battery voltage is. Eh probably should be fine

Why does everyone omit the +1 then calculating darlington gain?

>>2533478
Fundamental theorem of engineering: approximation. There is no point in trying to get the exact figure when gain is varying by 400% due to manufacturing process variations. We're not doing science or math, it's perfectly fine to use methods that would make physicists or mathematicians cringe. No need to be autistic about analytical expressions - throw that shit in excel, copy over formulas from reference books/app notes, add some solver magic and then fudge values a bit more with simulator, real circuit will be off by a few percent anyways because of parasitics you didn't account for and model limitations. The trick is to understand when you can simplify without sacrificing required precision/predictive power.

>>2533439
We can smoke in here right?

Anybody want to spoonfeed me and explain LoZ and inrush? Wanted a meter that could do all sorts of shit around the house and got this fuckheug meter.

Seems pretty good with DCA. That was the main isssue I had with my Southwire from Lowe’s, it does DCA but jumps around a lot and is inconsistent unless it’s measuring >10A.

>>2533494
4got pix

>>2533494
LoZ?
Inrush current is when a big fat cap is charging up and will bust your shit unless limited

>>2533497
It’s like HVAC and electric motor stuff, right? Wondering what I would be diagnosing with the thing.

Asking my threadly recommendation about textbooks on SMPS design and on transformers design and manufacturing.

>>2533498
>Wondering what I would be diagnosing with the thing.

It doesn't have an attention whore setting so you dun goofed.

Has anyone heard of this Sperry brand of MM? I got one for Xmas, it is very basic, just 2000 count, only 200ma fused, no true RMS, no C, no Temp, no kick stand, nothing. Guess how much it cost? $50! I almost fainted. what a waste of money. Why is it so overpriced? Apparently it is a Home Depot / Menards brand.

>>2533597
Wow, what a waste of money

>>2533597
>>2533605
The money? It was wasted.

>>2533597
Spend big money at Muhnards.

>>2533459
Check the RDS vs VGS graph, then calculate ohmic losses through P = R*I^2. Then do thermal calculations to see how hot the FETs will be. A simple doubling charge pump might be fine.

>>2533597
Does it at least have autoranging? A decent rubber outer case? A high CAT rating?

>>2533519
Damn? What do I need for that? Greenlee? Fluke?

>>2533597
That’s a somewhat reputable brand of electician stuff I think. As for the price, when you compare name brand meters from the big box stores to the chink brands on the internet, of course it’s going to look bad. You can get a $40 Kaiweets with the features of a $200+ Fluke all day.

$50 sounds about right if it’s auto-ranging. Home Depot and Lowe’s won’t sell you a Klein or Southwire or Ideal with a 10A fused slot until you spend ~$70, and you don’t get True RMS until like $100.

A nice dive into making a compressor circuit:
https://youtu.be/Wag-yTyAxPA

a while ago i made an esp8266 circuit to power cycle my modem as part of an automatic ban evasion script. it was a perfboard nigger rig and the poor little sot223 3v3 linear regulator finally died. i'm redoing the circuit now with a switcher so it won't burn up in its sealed enclosure.
QUESTION: has anyone run an ESP off a switcher? how sensitive is it to supply ripple? i added pads for an optional linear postregulator and shield can but i'd rather not bother with those.

>>2532432
it's listed for each capacitor model individually on their search tool, along with the other useful info like dc bias derating

>>2533651
>>2533678

>Does it at least have autoranging? A decent rubber outer case? A high CAT rating?
Nope, no auto-ranging, plastic case. Cat-III 600V (Is that high?) So I guess it is a $10 meter with a $40 markup for the "brand".

>>2533729
Yea that’s kind of rough. Similar Klein at Home Depot is $35 and $50 will get you the meter and outlet tester and NCV. I guess if you’re in the store and want the cheapest version of a known sparky brand, you gotta pay $50 to get above whatever Menard’s version of Commercial Electric is.

It’s not a $10 meter from a home improvement store though, I’m sure the thing is better built and properly fused unlike the $10 little rectangle generic meters. Bring it back for store credit if you think it was overpriced, otherwise toss it in the car for emergency use.

>>2533745
Yeah I am going to keep it as a backup meter. As for the quality, I dunno, you guys tell me, it seems solid but I don't really have a good reference point. But those humongous 6-30 fuses truly stand out lol. That's where all the money went to! Look how tiny a standard size fuse looks in comparison! Wow it is funny, why the hell do I need a 600V rated fuse on a 200ma current input. I didn't even know fuses were rated for voltage. I guess it's just so it doesn't explode when exposed to high voltage and high current. But I still don't get it how you can possibly drop 600V across it. Oh well. Fancy shit.

>>2533799
It’s the insulation rating for after the fuses have popped. If you accidentally put your meter in amps mode across 240VAC, the fuse pops. But if the fuse is only rated for 50V, current might keep flowing regardless, turning the meter in your hands into a glowing puddle. Have you watched Arc-Flash-Fatality.wmv yet this year?

>>2533705
Generally it’s fine with something digital so long as the ripple isn’t more than 0.2V or whatever. Digital stuff doesn’t care that much. Could always shove a TO-220 linreg in there instead, with a heat sink. Does the ESP just toggle a relay or something? Or does it speak UART to the router board?

Gonna put a can over those components? Also if you’re using wifi, ensure there’s no ground plane beneath the antenna. Non-issue if the ESP goes upside down.

>>2533865
>Digital stuff doesn’t care that much.
i'm more concerned about the wireless performance, i'm sure it'll work since the esp will be literal inches from the router but my concern is more on the general principle.
>Does the ESP just toggle a relay or something?
literally just switches a fet
>Gonna put a can over those components?
yes, i ended up ordering one because it was only ~$1.50 for the frame and can, it's funny how easy it is to jew out over pennies when doing one-off projects
>ensure there’s no ground plane beneath the antenna.
the antenna on this ESP model is located north of the pads, i've done a similar board (but with a linear regulator) for controlling a LED panel and found that you're fine as long as you don't extend the ground plane beyond the esp's pads.

>ground plane
Just drive a car or ride a bike. Dios mio.

>>2533909
>no ground plans
La luz extinguido.

>>2533862
>Arc-Flash-Fatality
Gives a new meaning to Eddie current.

if I had a battery pack that I wanted to recharge using a tp2045 recharge module for each group, if I used a PC ATX power supply to power each module, would this cause a grounding problem?

>>2533967
this won't work
i've drank too much to elaborate

>>2533967
>would this cause a grounding problem?

dont do it!
when you connect parallel things in series you create multiple pathways for short-circuits, fire, and calamity.
you may use (numerous) switches or relays to separate a parallel charge cycle from a serial use cycle.

>>2533967
The negative input and negative output are basically connected, draw it up like that and see what happens.

So, assume I have 450V 10 000 uF capacitor bank charged to 400V. Capacitors have ESR of about 12 mOhm (in total), and ESL of insignificant.
And I have a coil, or really low inductance to allow for more power.
What kind of device I should use for switching? SCR or IGBT?

Problem is that I want to test device with much less mean components before making a real deal.
As for real device, hopefully 2 1/4 inch steel plates would work as energy absorber.

>>2534007
SCR has a benefit of being cheap as nobody really needs it after IGBTs were invented and improved. But IGBTs are so nice as you can turn the off without homosexual tricks.

>>2534007
Are you this anon: >>2530652? As per my last reply (>>2530701) most of the SCRs I've seen look to be designed for mains use, and so aren't rated for the kind of speed you're probably after. If you can dig about through manufacturer lists then you may find SCRs specifically designed for pulsed use, otherwise you may have to go for IGBTs.

>>2533799
You can measure up to 10A thru the leads, that’s what the hole on the left is for. 10A max between the red A-hole on the left and the black hole in the middle.

The real cheap <$10 meters, a lot of the time you can’t even change the fuse if you blow it.

Also manual ranging meters are nice sometimes, quicker readings when they don’t have to jump around between ranges. Doesn’t seem bad, maybe like $20 overpriced.

Finally after a hard few days of design and research i can just sit back, put my feet up and let the bugmen slavelabor do their thing

Hey all, I'm trying to create a power latching circuit that uses a single button but also uses that same button as input.

I've made pic related but would like to have smarter people then me look at it to see if I made any issues before having this made on a PCB.

The switch is a momentary NO-switch that pulls to GND when pushed

>>2534144
>have smarter people then me look at
Are you too dumb to use a simulator?

>>2534148
Well, at least unable to get one working properly beside the examples they provide.

You know one I can best use?

>>2534192
OP needs to be updated to include this information, but most used here is LTSpice. You can also try MicroCap or Qucs. For web based solution that is geared towards beginners you can use
>https://www.falstad.com/circuit/circuitjs.html
as it gives you nice visualizations, but be aware that it only has ideal models of components so it's better to verify in proper spice based simulator with device models from manufacturers. There are a lot of tutorials on youtube about all of them so start there.

>>2533729
>Cat-III
Well that's why. Think of it as a poor man's fluke. It is yellow too. Problem is you most likely don't need CAT-III if you are a hobbyist.
CAT-III meters offer safety features suitable for
"three phase distribution; large building lighting systems and polyphase motors".
>>2533799
>humongous 6-30 fuses
Now you know why. That's CAT-III. They are ceramic filled with sand to prevent arcing. That's an industrial electrician meter not a DIY hobbyist meter. That's why they are sold at hardware stores.

Any recommendations for a good multimeter or do I just go for the cheapest one (~15€)

>>2534038
Not even IGBTs are suitable.... :/
Well, maybe there are, but I am not made out of money.
Maybe I should check SiC or GaN transistors?

In worst case solution is a mercury switch I guess.

>>2534144
Yeah that should work. Though schottky diodes do kinda have a high leakage current. 4.7k is also really small compared to the 1M pullup, if you're worried about quiescent current use I'd definitely increase that a ton.

>>2534217
>That's an industrial electrician meter not a DIY hobbyist meter
>for $50
>with no protective elastomer at all
sure it is

>>2534333
Aneng's AN8009 has been a high-spec DMM for a low price for a while now.

>>2534385
Spark gap or thyratron are maybe the only options yeah. But I'd have a look at pulsed power circuits (especially nitrogen, ruby, and Nd:YAG lasers) to see what parts they use.

>>2534395
OK how about "a budget electrician meter".
I am not really sure what the market niche is for these kind of meters. They are CAT-III yet are too cheap to be used for serious electrical work yet they are relatively expensive and lack some basic features to attract the average hobbyist considering the abundance of cheap, feature rich chinese models on the web.

>>2534385
Okay, ones you can actually buy are 180A cont max, and peak is just 500A.
>>2534395
>Spark gap
That would require high voltage, which would make it illegal as now it would be classified as shocker.
>But I'd have a look at pulsed power circuits (especially nitrogen, ruby, and Nd:YAG lasers) to see what parts they use.
Maybe. But I don't think it would be powerful enough.
Only application that deals with same power levels is probably a train traction.
In old trains they used resistors and contactors (maybe? would work, if i steal one somewhere)
In newer trains it is thyristor or IGBT, which have similar power requirements, but problem is that they work in audible range, and I need about 10 uS max turn on time.

But then, I can connect multiple thyristors in parallel with suitable balance resistors... Plus probably I can get more cheaper weaker ones, which would be faster.

College guy reporting in, is it normal that i feel like i didn't learn shit? Like i want to make a circuit to connect a UV sensor to the arduino, but i don't know what to do, what it needs unless i straight up copy an existing one.

>>2534410
Did they teach you how to look up datasheets on Al Gore's ww-dot-AOL?

>>2534411
That i've learned by myself.

>Al Gore's ww-dot-AOL
i missed the joke

>>2534414
>i missed the joke
Hell came to Earf. In the before-hell times, old farts and non-computer nerds would habitually butcher website URLs before they were shortened to simple domain names. teepee semicolon apostrophe dubya dot slash geocities slash 78df78yh384t7fiousdhiuf.hpnl

They were pretty good with the fax machine though.

>>2534414
P.S. Al Gore claimed the entire internet was his invention that he engineered by himself with no help while he was a senator shitbag from Tennessee. He calls himself "the father of the internet" and Oprah Winfrey agrees. If you disagree, it's because you're racist. Disagreeing with Oprah is racist and you should be ashamed of yourself.

>>2534407
>That would require high voltage
You can make a 400V spark gap, just make it like 0.2mm wide.

>>2534410
You just need to keep grinding, you'll start understanding things eventually. There are no shortcuts.

>>2534410
what year? if you are not a freshman, you should be able to. didn't you have labs with that kind of assignments?

>>2534424
Problem is that you lack control. I thought to use voltage to adjust energy so projectile doesn't get sucked back in, but with spark gap that would be hard, and spark gap would make it harder to implement any full-auto mode.

>>2534426
Problem is the huge wall at the start

>>2534427
I did the usual stuff, low pass filter, high pass filter, yada yada, i don't see how that helps me telling me what the sensor needs and how, i literally look at the circuit (pic related) and i didn't do anything with this arrangement.

>>2534433
Did you learn about comparators, what they do, how they work? Or a bjt transistor as a switch, high low? How to bias it etc. Next step is to think about how a particular sensor/transducer works. What is changing, voltage, current, resistance, etc. And then try to build your own simple circuit to bias a BJT or a comparator so that it is normally off but turns on when the sensor is activated so it sets your arduino pin accordingly. Try it and post your result and we can help you further. Do not copy random shit you find on the web unless you understand how it works and just need a reminder.

what's stopping me from yoinking the audio amplifier IC out of some old stereo system, driving a 60 Hz sine wave through it, and using it as an inverter

>>2534462
Inefficient as fuk

>>2534465
what if it's class D

>>2534441
>Did you learn about comparators, what they do, how they work? Or a bjt transistor as a switch, high low? How to bias it etc.
yes

the sensor is a GUVA-S12SD, so current will be changing depending on the uv index, but with my limited knowledge, i don't get why i can't just slap a non inverting amp circuit on and be done with it.

>>2534466
Might work, but still, there are purpose built IC for this task, that come with all protections and shit.
Because output part of class D amp has a lot in common with proper inverter.

>>2534462
>what's stopping me from yoinking the audio amplifier IC out of some old stereo system .... and using it as an inverter

if you yank the chip out, you then have to replace all the circuitry (and power supply) you've discarded.
that's dumb: better to keep the amp intact, and just feed in the waveform you like.
i've done this myself, using a 500W amp to drive incandescent lamps, to create a zero-work ''color organ''.

>>2534468
>current
So you need to convert I to V, so you need a TIA, it is a standard building block, are you familiar with it? Are you familiar with photovoltaic vs photoconductive modes?
Here are some basics
https://www.allaboutcircuits.com/technical-articles/understanding-photovoltaic-and-photoconductive-modes-of-photodiode-operation/
> why i can't just slap a non inverting amp
So now you know why, you just need to play with a TIA setup and see how much output voltage you get and tweak the resistor values etc. Also check the datasheet what mode is recommended, photovoltaic or photoconductive.

>>2534533
Shit, now it's all coming to sense, thanks your comment helped a lot, now i can fully understand the circuit above, the sensor operates in photovoltaic mode (datasheet), so that's why it has the capacitor and resistor on the resupply, the rest if the standard non inverting amp setup.

Now, i just don't understand the capacitor i see in every circuit of this type between the op amp positive supply and ground, i'm sure it's something fundamental but i can't make sense of it.

>TIA
Transimpedance amplifier? Never head of, but it makes sense now.

>>2534553
>something fundamental
i think it's just a filter cap vcc to ground, pretty standard

>Transimpedance amplifier?
Yes, it is an opamp configuration to convert current to voltage. Used with photodiodes and all kinds of transducers that convert whatever they measure to current.

>>2534428
>>2534424
Actually, I can just connect capacitors in series and get higher voltage. If I have 30 kA to deal with at 400V, I will get much less if I go 800V, for example.
That said, charging circuit has to get a bit more complex.

>>2534569
>i think it's just a filter cap vcc to ground, pretty standard
I found some stuff on it meanwhile.

I think i'm set for now, just need to fiddle around with the resistor values like you said, thanks a lot

I'm trying to make a light panel from an old monitor. The backlight is CFL, and when I turn it on it keeps flashing, looks like it's turning on and off at around mains frequency with an audible click each time. Shorting the on pin and the dim pin to the 5V pin turns the panel on, adjusting the voltage on the dim pin through a pot does indeed change the brightness, but the flashing persists. Caps on the board look fine and there are no scorch marks. There is a pin named "DCR" that seems to be left floating on the PSU board and nothing of the sort seems mentioned in the datasheet of the driver IC anyway. What could be causing this issue? Are the tubes shot or could it be some safety tripping and everything resetting?

>>2534573
If it’s a coil-based friendship delivery system we’re talking about, that means multiplying your number of turns and making them of thinner wire. The calculations are up to you to decide if the resulting turns and gauge is sensible, though I’d recommend 3D printing some sort of solenoid winding jig.

>>2534602
CFL or CCFL? Some part of the driver circuit will be funky, should be repairable with the right equipment. But that’s high-voltage power circuitry, you’d need a meter and probably also an oscilloscope that can handle those voltages, personally I’d just swap the fluorescents for LEDs.

>>2534635
>If it’s a coil-based friendship delivery system we’re talking about,
Lol. I legit wonder how many ppl in the world have such device, because in some countries it is not necessary because you can easily acquire conventional variant so you don't see much posts about it on the internet, but in some it is the only way, but nobody in right mind would post about such device in possession because it might end up with anal orifice enlargement.
> that means multiplying your number of turns and making them of thinner wire
I think most important thing is to match inductance to thyristor di/dt value first (with 400V I need around 25-30 kA, which is impossible to reach with given parameters of chinesium KP1500A hockey puck type thyristor, but with 1000V this current would be just 10 kA which is more reasonable and I think thyristor can actually achieve). Then figure out if it is possible to wind a solenoid with given inductance and force requirements. Maybe also running some FEM simulation, or idk...
>The calculations are up to you to decide if the resulting turns and gauge is sensible
I think with energy we're dealing with, losses on average thickness wire would be small, and resistance might be required in order to protect the thyristor.
>though I’d recommend 3D printing some sort of solenoid winding jig.
I don't have a 3D printer, but I think that 3D printed parts would be nice because it will allow to make a proper geometry of other stuff, not just some shitty PVC pipes and 1kV lose wires

>>2533439
I don't know if this is the right place to ask but:
Kirchhoff told me that net current into and out of a system is always 0. This is fine when I have circuits like picrel left. But what about static charge/discharge? picrel right. Isn't current just flowing charges? What if there is only one charge moving? What about when I get zapped from static buildup? Was Kirchhoff wrong?

If Kirchhoff is real, how can switches work?

>>2534657
Switches don't conflict with Kirchhoff. Please elaborate on the problem. I don't see it.

>Switch off/open
Current going in is 0. Current going out is 0.

>Switch on/closed
Current going in is 1. Current going out is 1.

>>2534661
If the switch is open, the resistance if infinite, which is, as I understand it, undefined. The counter example is the need for pull-up or pull-down, which is done only to create a defined state.

>>2534662
What in the world are you trying to say? I don't see what biasing a switch with a pull has to do with the conversation.

>>2534664
What I mean is, a simple light switch doesn't need a pull. Why?

>>2534662
Using pulls for creating a defined state is to create a known voltage. Not to control the current.

>>2534665
You don't put pulls on switches that supply power. You put them on devices reading voltage.

>>2534668
Good explanation, but then what is the voltage in a circuit with an open switch? Maybe it doesn't matter for something as simple as a light switch, but it does matter for the theory.

>>2534672
The voltage won't matter because the light runs off current. In picrel V will be equal to NEU since V=IR and I=0 so the voltage difference across the light with finite resistance is 0

>>2534652
You can also get ferrite filament. Probably.

>>2534654
Everything is capacitively coupled to its surroundings. Static charge is just charging up that capacitance. For charge to be applied to one thing it must be removed from something else, that transfer of charge is a current.

>>2534662
Undefined voltage but defined current. In practice you can just assume arbitrarily large resistance and it won’t matter. The undefined thing is just an artefact of the mathematics, not some sort of break in reality.

>>2534672
It’s floating. Undefined isn’t a break state, it’s a valid one that’s hard to calculate around.

>>2534677
I still don't see how the moving of static charge is consistent with Kirchhoff's Current Law. If I get zapped that could be electrons built up on my skin traveling to ground. So while I'm being zapped there is current entering my body but where is the current leaving my body?

>>2534677
>undefined is valid
So what happens if you go and measure the voltage? The voltage also matters for other effect. If the voltage is high enough, it could break through the air and create an arc.

>>2534680
We are talking about home line voltage across a light switch made for said line voltage. Nothing is going to arc

>>2534687
If the voltage is going high enough, it would arc. But I realize the air resistance would add another resistor to the theoretical circuit.

>>2534677
>You can also get ferrite filament. Probably.
I don't think it would help here anyhow tho, because magnetic field pretty much goes where it should

>>2534688
The electricity in your light switches isn't arching whenever they sit there in the off position. What makes you think that happens?

>>2534694
It was just an example. In truth, any voltage potential creates an electric field. Arcs happen if the resistance is low enough to cause the electrons traverse the electric field. Obviously I know that light switches just work, and that there are no arcs. I'm just trying to understand the theory. I think it's legitimate to come up with contrived examples, if they serve to test/explain the theory.

>>2534696
I would simply say a switch in the open/off position is a resistor with infinite resistance and in the closed/on position it is a wire with no resistance

>>2534700
Would you say that measuring the voltage would add another circuit (the one of the instrument to measure it) to the theoretical circuit? So the infinite resistance is just theoretical fluff, that doesn't happen in theory, because even the dry air between the switch's contacts has a defined resistance?

>>2534706
Yes, it does make a new circuit, but one that is negligible. But you don't need to measure it to know that the voltage will be equal to the other side of the light bulb.
The infinite resistance is to make it possible to model the circuit. Otherwise you would need to consider the 3D resistance of all air, insulators, variances in the conductivity of the wires, imperfections in the supply voltages, electromagnetic noise coming from space, your local radio station blasting noise into your circuit, your neighbor's bluetooth sending noise into the system, etc. Just use a model that gets the job done well enough. The little thing's won't matter in almost every case

>>2534711
>Otherwise you would need to consider the 3D resistance of all air, insulators, variances in the conductivity of the wires, imperfections in the supply voltages, electromagnetic noise coming from space, your local radio station blasting noise into your circuit, your neighbor's bluetooth sending noise into the system, etc.
But a theory has to consider this. Reality will certainly be like this. A theory is not a simplification, only applying it is. You simply say this and that is negligible. What I have a problem with is that saying an open switch has undefined something, while you can obviously design a light switch that uses this undefined state, and it just works.

>>2534715
But it is defined:
>>2534676
>In picrel V will be equal to NEU since V=IR and I=0 so the voltage difference across the light with finite resistance is 0

>>2534680
Practically speaking, there's capacitances everywhere. So even when you turn off a switch, it's going to hold its last potential, or otherwise be swayed by local dE/dt.

>>2534693
>I don't think it would help here anyhow tho, because magnetic field pretty much goes where it should
Might help for concentrating the flux a bit, or maybe for shielding your electronics from magnetic noise.

>>2534736
>even when you turn off a switch, it's going to hold its last potential
No it won't

>>2534737
Assuming there's no load to act as a pull-down/pull-up that is.

this could be cool for plating pcbs but i doubt it will work on vias and holes since technically the middle of the holes is just fiberglass, not metal so the electrolyte wont stick to it
https://www.youtube.com/watch?v=G-PtnwtOR24

>>2534830
http://turtlesarehere.com/html/through_hole_plating.html

>>2534635
Yes, CCFL. Using LED strips was my last resort. I kinda want to try and repair it for the sake of it. My scope can't handle the lamp's hot side, but I should be able to see what the driver IC sees without danger.

I shouldn't mix flux types when I'm looking to buy solder/flux right?
I have a rosin core solder wire but shopping around for flux and there's an abundance of no clean flux and barely any rosin flux

>>2534892
All electronics flux is rosin or synthetic resin based, including no-clean

>>2534573
>Charging circuit
You could go full grug and use an incandescent bulb to limit inrush and a comparator to switch a relay on when the current falls to an acceptable value. Elegant? No but piss easy.

>current-controlled inductor
now this might be poggers

Has anyone here done any fun projects with these?

>>2535098
i put one in my pee hole once because there wasn't an explicit warning on the package.
DO NOT PUT IN PEE HOLE

Can i make a transimpedance amplifier (TIA) with a TL071CP?

>>2535036
Is it just saturation based? You can do the same with a normal transformer, with the downside that the change in control voltage is going to induce a voltage in the inductor. Not a big deal if your control frequency is much lower than your signal frequency, but that dual circuit seems to cancel that out so is superior if it matters.

But yeah, they look way nicer to use for resonant stuff compared to varactors.

>>2535149
Yes.

>>2535098
that's an oshpark board

>>2533498
pretty much every motor with a capacitor or not will have an inrush current on start that will be many times the data plate amps for less than a split second. This meter will be fast enough to catch it on start. Inrush isn't always useful but it did tell me when a bearing was going out on one of my cooling tower fans.

>>2533439
Today, I have drawn up for all you lovely gents this new invention. The Plexi-Duck-Box (patent pending). It is for all my new devices to be contained in a fast, cheap, easy-to-build, waterproof enclosure.

The only problem is the waterproof part. So I was hoping you would be so kind as to fix my fatal flaw. I want a cheap container that's easy to build and size that can go outside indefinitely. So not submerged, but exposed to rain. And it needs to be able to have wires and buttons.

Thank you, God bless

>>2535219
>The only problem is the waterproof part.
screen door + flex seal (jew tar)

>>2535219
Tupperware? You can seal any holes in it with rubber grommets or silicone. But it might be prone to UV Degradation. PVC pipe and endcaps works too, using PVC cement to be waterproof, though getting a button to seal around a curbed surface may not be easy, and making it serviceable necessitates figuring out some sort of gasket and clamping mechanism. Not impossible though, having a strap or wire applying tension end-to-end, with a thin layer of casting silicone poured into a loose endcap would make a resealable lid. The wires coming out making it look like a bomb is a plus.

Is clock jitter to a DAC/ADC just some audiophile thing? Is it actually audible? How do you prevent it?

>>2535219
Entomb everything in epoxy like the youtube fags. Put a small sign inside that says nigger.

>>2535231
it manifests as broad-band phase noise or errant harmonics in the signal, depending on if it's random or periodic. It also effectively reduces the resolution of the signal. It's for similar reasons that we use windowing functions when taking the fourier transform of a signal on a spectral analyzer.
Depending on how shit the clock jitter is, yes it can definitely be audible. You decrease it by getting discrete high-quality ADC/DAC ICs with high slew rates, and by supplying them a clean clock signal. Use a precise clock source like a PLL, don't use something like an RC oscillator subject to all kinds of variation. Avoid anything that would incur parasitics like long traces or poor multilayer layout. This will prevent your clock from getting skewed in transit.

>>2535239
I’ll be converting the I2S from an ADC directly to S/PDIF on the same board, so no long traces to care about. I think I’ll need to use a Si5351 for the multiple synchronous clock sources anyhow, which is PLL-based, so I think that should be good.
>inb4 ovenise it

I’ll have to look into the qualitative clock stability requirements.

I have a BSEE with a specialization in communications
are there any EE jobs that don't mandate I spend all day sitting down in an office? Like, can I get to be the person who climbs the big comms towers or something
the thought of being sedentary most of my life is dreadful

>>2533967
I think I will just use 2x of these (it's real though finding 20 pin connectors).
The thickness is awg 18 which is probably going to be warm at 15a, but it's probably the best thickness I can find for a 10-20 pin connector.

>>2535293
should probably mention this is for a 10s battery for a ebike.

>>2535297
Just buy a 10S charging BMS, they shunt away current while charging to prevent overcurrent, so all you need to do is provide a constant current to the whole setup. Can also just use two 5S charging BMS boards in series. You determine that the whole pack is charged once the voltage is up to a certain level for an amount of time.

>>2535279
look into fiber, but climbing poles is a monkey job
did you know that until recently they had actual monkeys picking coconuts (for human consumption) in thailand?

>>2535307
I should, but if I go with my approach, I will spend $30 on a the cables (super pricy, but a 1/3 is from shipping, so if I buy more stuff from digikey, maybe it's $20), and $5 on the TP2056's, and I have a ATX supply.
If I buy a BMS it would cost $35, and even though I could use a cheap $5 boost converter to feed the power to the BMS, I don't really trust it, also I have heard about how the cheap power supplies can do odd things, like surge stuff (I forgot), so to prevent that you need a better power supply, and that would cost $40 (I would get the DP50V5A, and power it with the $5 boost converter).

>>2535342
>DP50V5A
*and note I was thinking of getting the buck-boost version, because the $5 boost converter might damage the $40 power supply, but that version is $20 more, and I could also just buy a 150watt 42v power supply for $20 on digikey (and I don't think the boost version has polarity protection).

I sure hope the solder sticks to these random screws, and that i can just melt off the insulation on the inner strands.

Where do you guys get cables and wires with specific connectors and lengths? I know there are lots of solder-ready connector+wire prefabs available and you can buy common lengths of popular cables/connectors (Dupont 2.54, jst xh, etc) in packs of 10 or 20 retail, but if you need, say 500-2000 pieces of custom length jst xh 8p, it seems you're either stuck with diy or you need to deal with an industrial custom cable fabrication house which will scoff at your puny order (or charge an arm and leg).

Basically I want a jlc/pcbway style service where I can select a connector type and length and have X pieces custom made (by machines so cheap) and shipped to me asap.

Does this exist? If not, should I create it? Would you use this service, anons?

>>2533967
Look at your own diagram and search your heart anon, deep down I think you know that this can't work. To charge the battery you need to produce a supply voltage higher than the cells in series, or charge just a subset of the cells at a given time.

>>2535392
500-2000 isn't a puny amount, I am sure you could go to alibaba and you will have enough orders to get past the MOQ, and optionally get a test sample, and then you will make your order.
But I don't know anything about B2B, but if a business doesn't mainly work with north america, you probably need to buy an agent to handle communication because the people you work with probably don't know any english.

>>2535396
Well what I'm proposing is precisely to make this entire process simple and turnkey for other engineers and hobbyists who want to order as little as 1 custom cable and up to thousands. Ideally the service would be physically located in north America to minimize fulfillment time. Since the gentrification would be highly automated, cost of labor would be less significant

>>2535401
I've looked at datasheets of molex connectors and I have seen the machines you could buy to mostly automate production, but it will obviously cost an arm and a leg.

>>2535404
>to mostly automate production

you sound like some helpless Gen X girl.
order spools of wire, connectors, pins, and a $40 crimper.
then spend 6 hours cutting, stripping, crimping.

>dood! i havent worked 6 hours straight my entire life. except Fortnite, once.

>>2535342
>TP4056s
Again, you need isolated power supplies for each TP4056. Your circuit won't work because the negative inputs and negative outputs are connected together, so you'd short all but one of your cells completely. Check out the TP4056 datasheet to see for sure.

>If I buy a BMS it would cost $35
$5 on alibay. Read the first rated reply here:
https://electronics.stackexchange.com/questions/184962/how-to-charge-a-18650-battery-pack-with-bms
Looks like going for a higher voltage supply and a current-limiting resistor or linear current regulator (e.g. op-amp and darlington, LM317, etc.) is the easiest way to go by far. It should be possible to get a switching current regulator, so long as it can handle the voltage. You may be able to use a boost converter with inherent current limitation, but they tend to get pretty hot when running at high voltage ratios. Again, check the voltage range.

>>2535404
I've built a few custom fabrication machines before and could probably build one for this too. It likely wouldn't be as good as what's commercially available but it doesn't have to be.

>>2535413
not only is that a big waste of time, the whole point I'm making here is that there's potentially a business opportunity here to solve this problem for pretty much every engineer in north america, much like PCB fabrication/assembly services have become the de-facto go-to for prototyping and small batch fabrication.

>>2535429
>there's potentially a business opportunity here

i seriously doubt that.
we had a project that needed a few hundred cable harnesses.
our ''buyer guy'' has a bunch of contacts in the industry and he was able to get 3 local quotes for the job in no time.
and this isnt even a city with any kind of serious tech industry.
the rolodex can still beat google in some fields.

>>2535430

What kind of price per unit are we talking about here? Your needs (and budget) for professional projects may not be characteristic of the entire market. My guess is, those local shops are not going to do batches of 5, or 10, or 20 or 50 for that matter, at least not without the buyer paying a hefty premium for setup/overhead. The economics are comparable to PCB fabrication; there are also local PCB shops, but JLC & PCBWay are still doing massive business.

I'm talking about simple cables like this: https://www.amazon.ca/Terminal-Extension-Extruder-Charging-Connector/dp/B0B87NPPGG in this case, a 2m JST-XH 4P (male on one end, female on other) sells for $4 USD. There might be a 1m version available, but there's certainly no 1.1m, or 1.2m, or 1.15m, etc. Of course, anyone can buy a longer cable, cut it down to size, and re-solder it, but those extra steps are really undesirable for many people, especially at scale.

>>2535434
Also, when I say it's a potential business opportunity, I'm not saying it's a billion dollar company. In a way, it's the opposite - the reason I think it's interesting is because I think it's a small niche and could be addressed by a small operation that grows organically.

>>2534873 here.
I found the rail powering the CCFL driver and energized it with the bench PSU and it worked fine. This confirmed my hunch that the supply wasn't happy with an open circuit, so putting a resistor between the 5,1V rail and ground fixed it. CCFL light isn't great temperature wise, but hey it's free.

>>2535187
Now I think I get it. Little fan runs at like 0.3A but inrush caught it at closer to 0.7A on startup.

Now I have to learn what LPF and LoZ are for.

Happy with this guy so far. I almost got the little $50 Uni-T with the sideways screen that everybody likes because people said it’s really good for a clamp at lower DCA readings. I paid $80 for a Southwire from Lowe’s to get a DC clamp meter and that thing bounces a lot so it’s hard to get good readings below like 5A, and it maxes out at 400A. Always need to zero it and it still bounces around. This big Uni-T seems way more stable with DCA. Pic related without hitting the zero button, and the Uni-T sits at that value while the Southwire ranges from like 0.18A up to 0.25A sitting here on the bench.

Those Southwire meters were rebranded something, saw the same thing on Amazon as Reed and Amazon Commercial and none of them are really cheap either.

Can the average DMM reliably measure picofarads? BTW what's the Aneng AN8009 of 2003? Anything better, 20000 count or whatever, for a similar price? Is autoranging a meme? Would you rather wait or switch the dial 100 times?

I want to power a 1W power LED from an 18650.
Problem is the LED is rated at 3.6V 350mA max and the battery fully charged sits at 4.1V so I can't connect directly or I will fry it.
The quick and easy way out would be to get a fat resistor and be done with things, but I was wondering if there's a smarter way to do it.
>inb4 just get a buck boost with current limiting

>>2535561
>Can the average DMM reliably measure picofarads?
No. Measuring in pF range requires some effort and is not as simple as hooking up DMM and reading the value. Waving your hands around, standing near, or farting in the other side of your room will disturb measurements unless you do it properly.

>>2535574
>farting in the other side of your room will disturb measurements unless you do it properly.
Can confirm. Sharted a 1uF spike in the summer of '92.

>>2535565
You have 2 choices
>linear way
>switching way
Because you don't want to do switching, you can do a dumb resistor solution or build a constant current driver with transistor and opamp.

>>2535565
I don't understand why voltage matters. Does it have an internal limiting resistor so it draws 350mA at 3.6v? Or it simply states that Vf=3.6v at 350mA?

>>2535565
>fat resistor
also, it doesn't need to be fat at all. it needs to drop the difference 4.1-3.6=0.5v 2 ohm 1/2 w is all you need.

>>2535610
>Or it simply states that Vf=3.6v at 350mA?

it's a point on the VI curve.
of some significance, coz it's the max safe current.
except 3.6 is not an actual number you can expect, but an average of many units.
in stats, they call it a normal.

>>2535615
Yea I get that, which is why I thought it would be unwise to directly connect an LED to any kind of power supply or a battery, even if it is 3.6v. You are relying on the fact that Vf is exactly the same as your battery voltage, then you will get the max current but it is risky. If it is higher, it won't turn on, but if it is lower, you will burin it, correct? Which is why I thought you always need a limiting resistor. Except for the high power LED the value may be inconvenient and also the batter voltage will drop so you'd need some kind of a constant current driver.

>>2535618
>If it is higher, it won't turn on,
Wait but Vf=3.6v is for I=350mA max current, so it if Vf is as low as Vf it would still turn on just the current will be lower? I thought the datasheet should also specify Vf as a turn-on threshold and the minimum current. Say, 3v at 100mA or whatever.

>>2535623
Ah here it is. Just like I thought.

>>2535098
A memredistor in the wild? Huh
Didnt know you could just buy them

Whoa, why am I measuring 8V hot to ground?? Sounds like free electricity since only hot to neutral is metered?

>>2535640
I mean neutral to ground

My dad tried to scavenge the motion sensor from a halogen work light and put it on a led light. Now it's just always on.
Judging by the way he had tried to wire it he is basically incompetent, what could he possibly have done to the sensor that would make it always on?

Anybody know a high-k dielectric that's cheap and would be easy to shape into a flat ring/washer shape?

>>2535642
>>2535640
Kek
Your house grounding is bust anon, this is usually seen in old houses. You might get a wee shock from the chassis of heaters and other things

>>2535663
Afraid can't help you much with that but would love to know what you're making/playing around with

>>2535663
Manganese dioxide? Are you making weapons, Anon?

>>2535667
>>2535670
I'd like to make a modified version of a lifter based on an old experiment: http://jnaudin.free.fr/html/elghatv1.htm

Supposedly Biefeld and Brown only used air as the dielectric in their experiment later on when they were trying to improve the thrust-to-weight ratio.

>>2535582
Switching would be nice but kind of defeats the point of making something cheap out of leftover parts and salvaged cells if I have to pay 5 euros per current limiting buck/boost. I am open to trying to DIY a switching regulator myself but I lack inductors and most importantly a scope.

Anyways, I have some LM317s, would they be of any help in this?

>>2535614
Fat by my standards, most of my resistors are puny 1/6W ones.

>>2535673
I'm thinking spin-coated aerogel for the ring.

>>2535685
Look up LM317 constant current but you would still need a current sense resistor. 1.25v/0.3a = 4 Ohm, about 0.4W. A 1/2W resistor would get really hot, so you would need a 1W resistor.

>>2535685
>Anyways, I have some LM317s, would they be of any help in this?
No, because it requires minimum 3V drop across it to operate, so you would only get 1.1V for LED when battery is fully charged.

It goes same for any linear solution, you need at least 1V drop for regulator so LED won't be on for long before battery discharges. It's probably best to just stick to resistor or spend a few € on converter IC.

>>2535698
I'm pretty sure that's outside of my budget.

>>2535565
if you don't want a buck boost with current regulation converter because of the size, you could get a much smaller solution by using a premade regulator board designed for recharging phones using batteries, and then turn that 5v into 3.3v using AMS1117-3.3
only problem is that you need 6v, so you need a 2s configuration, and it's not much cheaper than a buck boost (but I haven't checked the prices of 10pc batches).

>>2535711
*my bad, that chip is 3.3v, but you can use a adjustable buck converter if you care about efficiency.

How would one DIY a quantum computer?

>>2535561
Pretty sure you need kelvin sensing for that.

>>2535640
>>2535642
I think Electroboom or someone did a video on that. It's not free energy, it all comes from the hot in the end anyhow.

>>2535650
Literally anything. Put it backwards, put it with too high or low a voltage, cooked it while soldering, damaged it physically, used it to drive too high current a load, whatever.

>>2535565
>>2535703
A BD139 only has ~0.5V maximum drop from collector to emitter when saturated, it's feasible to make a regulator with that and a RRIO op-amp that's pretty low dropout. You'd just have to use it in the common-emitter configuration, and use a sufficiently low current sense resistor (e.g. for a 0.1V drop), possibly requiring a current sense amplifier if the op-amp won't run well that low. Using a power MOSFET would probably doable too so long as you avoid linear thermal issues. The only issue with common-emitter and common-source topologies for linear regulation is it's harder to get them stable, but since an LED is such a simple load and the feedback path is so direct I think it will be fine, perhaps just with a small cap from inverting input to output.

>>2535711
*if the led is current regulated, you can use NSI50350AST3G, but there might be other chips that are like this which may be better.

Does nobody in the world manufactures small (5-10") CRT screens like picrel? I can't find a new one for the life of me, I can only find old ones from the 90s at insane price

>>2535779
What price is insane to you? You can find one on ebay for $80.
https://www.ebay.com/itm/224370681880

>>2535779
Somebody call John Titor.

>>2535779
Modern doorbell video telephones are still sometimes made with CRTs. But they’re side-emitting specimens with somewhat small screens, don’t know if they go above 4”. Also probably all black and white, for which you could just use a replacement oscilloscope screen + driver.

>>2535838
He needs it to point a video camera at connected screens with a seed pattern in between in order to perform analogue fractal decryption. Only a full analogue system has low enough latency to complete the pattern in a reasonable period of time, without inviting the scrutiny of NASA. The resulting visual fingerprint is required to unlock the Roswell UFO’s online portal, allowing you to access the anal records of all recent US presidents.

>>2535779
Nope. There are idiots online that will repeat hearsay about india or china still making them, but that's a load of bullshit.
There's a place that still refurbishes CRT military equipment, but you can't afford them.
You or someone else is going to whine and cry when I say this (someone always does) but the color CRTs that exist now are all that ever will.

Hello Ohm this might be out there but I want to get into radios, oscillators and filters. I need a test jig to test passives that will work with through-hole as I don't plan on using smd components yet. The only one I really see is this https://www.walmart.com/ip/NanoVNA-Test-Board-Kit-VNA-Test-Demo-Board-Vector-Network-Analyzer-Test-Board/2786332485?wmlspartner=wlpa&selectedSellerId=101268731
I was thinking I already have the stuff I could just make a test jig myself. The images though don't really show how exactly each pad is wired to the smd pads if they even are through the whole grid. I also cannot see the back of it either. So was wondering if someone has seen a design or might know of something that I could use to design up in kicad that would be great. Looking to have maybe max a two pole filter on it if possible. Though if possible more within reason.

>>2535993
>walmart sells vector network analysers
what the fuck

yeah I'd look into making impedance matched coaxial-style traces in a PCB to solder components to (not only existing passives but hand-wound inductors too), to make as universal as possible. Then order a bunch of them. Having one board with a bunch of dotboard on it like your link may not be great for impedance matching. I'd just have a few set layouts using standard DIN0207 resistors, 5mm disc caps, and whatever inductor footprint suits your fancy, in something like a double-pi filter layout. That way you can add wires and components in there as you please to make a variety of different filters. Will probably need 2 or 3 different board layouts (or multiple on each board).

I got some li-ion charger modules with DW01 protection IC. I looked up datasheet and found these
>Overcharge detection voltage: 4.30±0.050
>Overdischarge detection voltage: 2.40±0.050
That doesn't look like good values for protecting li-ion batteries.
Am I intepreting it right?
Is this yet another planned obsolescence shit?
Is there another IC with proper voltage threshold (or even customizable?)

>>2536044
just bootstrap your own overvoltage and undervoltage circuit to the system

Is it possible to treat normal through holes for header pins as though they were castellations?
I want to keep a board as flat as possible but I need to use a 3rd party board that I can't reverse engineer or modify. There are no components on the other side that could get in the way.

>>2536056
yes i do this all the time do either:
1. make SMD pads that go under each hole and are wide enough such that they stick out from under tge pcb by about 1-2mm. put your iron along the edge of the pcb and the pad , feed solder, it will "fill up" the through hole.

2. make through hole pads in the exact some positions, except make the hole diameter later, put tonnes of solder in there

these will both work, ive done 1. extensively, the only problem is you need to presold 4 corners so theres a small air gap between the 2 pcbs. hope that makes sense, though if theres nothing on the other side just do 2.

>>2536022
>what the fuck

Walmart's website is basically just bootleg Amazon. Most of the stuff on there isn't actually sold or shipped by Walmart.

>create a pi filter (CLC) with C=100nF and I=22uH
>pass a sine signal @ 104KHz
>instead of drop off, seeing a gain of 24dB
>holyshit magic_voltage_doubler.png

Welp remember to include the source and load impedance on any kind of AC filter type simulation or you'll get some dumbfuck results like me

>>2536044
That is actually pretty bad.

>>2536095
Yeah, the high Q of not including a load impedance does tend to produce rather peaky AC analyses.

man i wish jlc had the 3 day shipping for five bucks.. the only downside of ordering there is i have to wait two weeks for my pcb to arrive
but for anything that isn't just a small simple pcb making it at home simply isn't worth it

>>2536095
Or just say fuck it im rich and buffer everything

>messing with custom li-ion circuits
Aren't you afraid you might burn down your house?

>>2536127
why? lipos are braindead easy, literally all you need to do is not go over max current or voltage for the cell

>>2536129
I suppose that's enough to avoid overcharging them, if you got the right values, and actually limit them correctly. But you forgot the min.

Guys, how do i crimp the cable? Is it with some specialized tool? I tried using a plier.

Also, is the cable too thick?

>>2536147
yes you need special crimper pliers that supports the size of those crimps

>>2536022
Well I do not have any particular board design in mind. At the moment I am not sure what kind of layout would be suited. I am beginning in this area and I honestly don't expect to get any impedance matching. I just want to get relative numbers and learn to use my cheap little VNA I got. That walmart board is at least a design I can start with but I am not sure about all this wiring and stuff. I mean I know the component leads will cause errors in measurements so I figured traces on the board but that may limit what setups I can do.

If I guess is correctly, I can turn step up converters into variable constant current drivers by doing pic related mods.
Am I correct

What are these SMD resistors that look like TH but with no leads? I've only seen them in a couple of different DMMs. Is there a reason that they are used in DMMs specifically? They have 5 bands, so they don't seem to be particularly precise.

>>2536286
Fusible resistor

Today I miniaturized a boost converter. Did I do well?

>>2536290
>Fusible resistor

naw, nigger, he's talking about components injured in combat: war amputees.
legless or leadless resistors.

>>2536292
>Did I do well?

you've completely ignored the 3-rd dimension.
by stacking the components, you could make them fit inside the handle of a sharpie.

>>2536293
yes, this. what's the point? they take up as much as space as TH resistors of similar power rating.

>>2536292
The big pad on the PCB is a heat sink. The switcher IC dumps heat through the tab.

>>2536296
Part of the problem was the height, otherwise I wouldn't have bent the output cap.

>>2536302
There's fuckall current going through it so it should be fine.

>>2536292
Now the two caps are missing?

>>2536286
MELF or as our professor called them MILF resistors.

>>2536312
>can't count to 4

>>2536312
>Now the two caps are missing?
Original board has 2 SMD electrolytics and 2 SMD ceramics. My board has the exact same number but they're THT. I don't know what you think is missing.

>>2533496
*beep**beep**beep*
>Off.

>>2536325
I don't know what the original board looked like. You desoldered some components and used them on the new "board". I thought the electrolytic caps were big enough to reuse, so I thought you dropped them completely.

>>2536292
For what purpose?

>>2536292
Chink PCB was already bad enough. Now your converter is only going to boost noise and EM interference.

>>2536330
>I thought the electrolytic caps were big enough to reuse
They were too big to reuse. What's on the original board were some that I soldered in hoping I could save enough space to make it fit. It wasn't enough.

>>2536332
Has to fit in a tight space to provide 9V from a lithium cell

>>2536334
I'm getting stable 9V out of it even under some decent load so it should be fine.

>>2536342
I'm curious.
Microphone battery?

>>2536344
IR thermometer.
9V blocks are getting expensive so I thought I could make it run off an 18650.

>>2536347
Negro, you alright? rechargeable 9v batteries exist, they even recharge through micro usb.

>>2536350
>go to /diy/
>tell people to "just buy product"
I sure hope you're getting paid for such posts.

>>2536355
I'm telling people to not waste their time, especially after just mangling a buckboost board onto stripboard.

>>2536360
>noo you can't just resolder a boost onto perfboard with smaller footprint while also learning a bit about its design
The one time I don't take the cop out option of "just wait 3 months for more shit from china" I make some turbo autist seethe. Guess I should do this more often.

Also
>I'm telling people to not waste their time, especially after just mangling a buckboost board
>after
Wow, galaxy brain move there.

>spend 30 minutes soldering 555 circuits
>accidentally short output with probe
>chip blows up
>didn't use sockets because forgot to buy them
how's your day going anons

>>2536218
Technically yes, but in practice you’ll find that to pump any significant current through your load, the RI^2 power dissipation in your current-set/sense resistor/pot will get really high. Because the input voltage to the feedback pin will equal the reference voltage during normal operation. It’s common practice to use a much smaller resistor (milliohms) and add an op-amp differential amplifier with a gain of 100 or so. Just ensure that your op-amps can sense and output close to he ground rail. This is what those LM2596 modules with more than one trimpot do, though they’re bucks.

Another modification is to add external error amplifiers And an external reference voltage that you can adjust by hand (or via a DAC) for current or voltage control. Especially for a buck converter, this lets the voltage or current output be lower than what would equal the reference voltage.

You can also have both current and voltage feedback by using diodes to combine the signals from both a voltage divider and a current sense amplifier, such that the higher of the two voltages will enter the feedback pin. It’s also an option to use a diode to implement overtemperature protection from a thermistor voltage divider. This kind of diode-combining circuit is already included in the TL494 switching control IC. It also has accessible error amplifiers and a fixed reference, that you can either use or replace with your own.

>>2536044
pretty sure the tp4056 will prevent overcharging beyond 4.2v, also note that the circuit isn't designed to work with a load while charging.
https://www.best-microcontroller-projects.com/tp4056.html

>>2536393
Yeah but the 2.4V overdischarge protection threshold is way too low.

>>2536393
The TP4056 circuity is just a concstant current/constant voltage source with specified cutouts. You can use it while charging, so long as the current to the load is less than the charge current specified to the board/circuit.

>welding some nickel to lithium cells
>one weld gives a shower of sparks
>one electrode slipped off the nickel and blew a crater into the battery's ass
Fuck. I can see some blue shit at the bottom of the hole. How likely is it that this shit will blow up?

>>2536439
>damaged lithium cells
Extremely likely?

>>2536444
I want to know if I damaged something crucial of if I just blew a hole in the conductive surface so it's no biggie.
Like, if it was going to blow then I'm pretty sure it would have done so already since it was close to fully charged.

>>2536445
All I know is that they're very susceptible to damage. I'd keep it in a place where it can blow up any time, without causing further damage.

>>2536445
Bury it in a bucket o' sand strapped to hand trucks equipped with a smoke detector so you can wheel it out of your house at 3 AM in your underwear. lmao

>>2536439
If it's just surface damage and no lasting heating happened, it's fine. The important part is the electrolyte and electrodes, if neither were damaged then the battery isn't going to explode for no reason.

Made a water detection circuit to drive a relay/pump for this bomb ass cyclone. Used bare copper pcb that I routed down the middle to act as switch for mosfet driving the relay. 12v supply. Tested great, super easy, but in practice one side of the pcb corroded to no longer conducting in just a couple hours. What do?

>>2536559
>bomb ass cyclone
Is that zoomer speak for "Flowbee haircutting system"?

>>2536559
float switch?

>>2536581
Fuck float switches. They always get stuck on or off. This is replacing one.

Electrically I guess I could use a pulse on a long ass period to check for water, then latch it on if so, not sure how resetting would work though. Could also use dpdt relay on same timer to alternate pos/neg sides.

Chemically using something like stainless would probably be a good idea. Nothing seems ideal though, electricity and water is a corrosion losing battle.

I assembled extremely simple and easy to build amplifier with 2xTDA7056A. I first tested it with bottom tier micro stereo system speakers I picked by the trashcans and soon used it for a few weeks every day when visiting relatives with some random 90s Philips cheap but still better than the trashcan speakers. It's not event bad.

>>2536591
Maybe you could use a jewish lizard space laser beam and sensor from a mouse.
>>2536596
Noice. You gonna solder it to protoboard or leave it on the breadboard?

>>2536559
>one side of the pcb corroded

Yeah, that's how electrolytic corrosion works. There's a reason this kind of sensor isn't really used where it's expected to get wet regularly. Float switches or ultrasonic sensors are the norm for fluid level sensing for a reason.

>>2536599
I'll make a board for it and put it in a case for the sake of practice and a memory. The cold and distant yet still logically flat but lacking punch in every frequency range -sound of 3W amplifiers designed for TV speakers, soundbars or such on a breadboard circuit struggling to drive 90s household low-to-midfi speakers is something I'll never forget. Not because it's bad but because despite being bad it's still decent in a way. Also first time making a circuit board, hope that UV oven for drying nail jobs works like someone on the internet said it should and at least third attempt is good enough to use.

>>2536559
Make a capacitive water sensor by covering the raw copper in epoxy. Hook it up to a 74HC14 (or 74LVC1G14 if you're feeling fancy) with a sensible resistor to make an RC relaxation oscillator, and use a 555 to detect when it goes below a certain frequency. Might be too low a capacitance to use another 555 to make the oscillator, but it might be fine. A CMOS one would likely work.

You can also prolong the corrosion by using it with a really high value pull-down resistor (1-10M) to switch a MOSFET that switches the relay. Plating it with tin or nickel is also a good idea.

>>2536630
If you or a friend has an MSLA 3D printer, you can use that with UVtools and your gerber files to photomask the board directly.

>>2536292
Cool noise transmitter anon

>>2536648
Thank you. Im going to give the resistor a try first. This is a couple times a year use case that doesnt really have space for a circuit board.

>>2533439
bros, what are the precautions to take when working on old electronics? I heard that cotton gloves are a thing since some capacitors might discharge on you. Anything else of these sorts to consider?

>>2536680
You could also just bodge an ATtiny13 or similar to do the oscillating and frequency detection, no PCB needed.
As for the resistor method, pic related.

>>2536690
That's more of a high-voltage thing rather than an old electronics thing. For old electronics, keep in mind that traces and pads on phenolic PCBs are more delicate than fibreglass ones. Especially on single-sided boards without plated through-holes.

>>2536381
>>2536218
So, as long as my feedback resistors can handle the current (I knew that trimpots can't), it's good to go.
Just wanted to make a Portable 12v light supply, as cheap as possible and to run it on 18650

Thanks

>>2536746
so if I calculate right, for 12v led with mt3608 and sx1308, that be extra loss of
(feedback voltage/driven load voltage)*100

aka (0.6/12)*100=5%

for 24v led, it'll be 2.5%

and as long as my (now shunt) resistor can handle the wattage, it'll work.

Not bad.

>>2536292
couldn't you have just sawn off the capacitor parts you didn't need?

>>2536648
>epoxy
Wouldn't the corrosion have a similar effect as it insulates the copper, or is that different?

I am not an EE I don't know much more than how to use a multimeter and how to solder. But I have an led on a pcb that does not light up when the others around it do. This led has test points and a gnd pad nearby which will light up the led when I'm continuity testing. Any ideas on what might be the problem/why this happens are appreciated because I don't know much here and googling gives me somewhat related results but no answers to my question.

>>2536852
ah yeah nah this shit is fucked hahaha don't worry about it. looking harder this thing has been liquid damaged and I don't have the tools or knowledge to fix it.

>>2536852
are they all supposed to light at the same time? You can try checking continuity between what is powering that led, along with that LED's connection to ground/controller pin.

>order a digital panel mount voltmeter/ampmeter from ali
>current displayed is off by a factor of 3
>voltage doesn't even display a valid number
Thanks, Chang.

>>2536969
>buying chink shit

>>2536969
>factor of 3.
did they just fuck up a sense resistor? Are there empty pads on the back?

>>2536972
>paying 10$ to get it locally
And I'd probably be getting the exact same shit.

>>2536980
Don't know and at this point I don't really care. I wanted it for the voltage readout, current readout was just a bonus.

Is this shit worth attempting to fix? Bearing in mind, this is an old desk lamp, that I could probably replace for $60.

I have a VERY rudimentary understanding of electronics. I've fixed small appliances before by just poking around with a multimeter until I've isolated the break in continuity and then soldering on a new fuse or whatever, but never on a PCB this crowded. I also notice at least two burnt out components (don't know what), so before I spend any time of this, could somebody who knows what they're looking at give me an assessment of whether it's worth my time?

Thnxbbs

>>2537015
Financially may not make sense, the minimum order+shipping for those components might exceed the cost of replacing the lamp. Can you just replace the power supply?

>>2537017
I'm not sure, it's just a 13W compact fluorescent light with a single inline switch on the 120V cord. The PCB is wired to the switch. I guess it being a fluorescent light makes it more complicated (is that big yellow thing ballast?). Just looking for a direct replacement for the whole switch-pcb assembly hasn't yielded any results. Don't even know how old the light is, it's this: https://www.worthpoint.com/worthopedia/tensor-vision-max-desk-lamp-17386-535224408

>>2537029
That's just a ballast then.
>>2537015
It even seems that you can find ceb-1a 13w ballasts:
http://sanseitec.com/en/product_xq.aspx?ID=88&nTID=1
though any 13 watt ballast will do.

>>2537015
It's a nice looking lamp. As someone suggested, replace the fluorescent ballast. Or you could convert it to LED. Time consuming and all, but a neat project if you have the ingenuity and time.

>>2537038
>>2537046
Thanks, unfortunately after some searching I can't actually find a ceb-1a ballast to buy (Canada). I can find other 13W 120V rapid start electronic ballasts, (like this https://www.bulbsdepot.com/bd13-sl27t.html)) but the listings all specify compatible lamp types and don't list my bulb. If they're listing (or not listing) specific types, I'm assuming that means that it matters?

I have a 2 pin, twin bulb (CFT13W). And the closest I can find listed is CFQ13W which are quad bulb (found NEMA designation guide here: https://www.lightsearch.com/resources/lightguides/cfllampid.html).). So for now it seems I'm SOL, and would be better off just replacing it :(

>>2536817
Epoxy creates an insulating layer, no corrosion could happen through it. The issue in particular here was galvanic corrosion, running current through the copper and water caused the anode to dissolve, so if no current flows into the water you won't get that.

>>2536985
>And I'd probably be getting the exact same shit.
Yeah but you'd be able to walk back into the store and swap it out for another. Also more like $15-20.

>>2536960
They are yeah. 3 color leds and the problem leds either are off or on the wrong color. 2 when they should be on 1. Lots of the test points around the liquid stain appear, when I looked at them under a magnifier, to also be damaged/scraped off by the previous owner in what I'd imagine was a super shit attempt at repair. Thanks for the tip anon I'll try that and hope I follow the trail to the culprit. Would be easier if it was just a simple pcb instead of this massive multilayer monster with paint over the topmost traces.

Are all fluxes no clean? Spent a while trying to find non no clean flux because I hear theyre harder to clean but ended up buying a couple of them and they all ended up saying no clean when it didnt even say anything about that on the product description page

>>2537122
"rosin flux paste"
MG Chemicals 8342
Kester TSF-6522
Caig RSF-R80-2

>>2537101
>you'd be able to walk back into the store and swap it out for another
Where I live the stores give you warranty only if you have an electronics engineering degree. Don't have degree? You can go pound sand.
I wish I was making this shit up.
If I come across the warranty paper I'll post a pic for you guys.

>>2537122
Why would you want to make it harder? I don't know if non no clean has better qualities, but if you're not looking for these qualities, you probably don't need them.

>>2537138
>Where I live the stores give you warranty only if you have an electronics engineering degree
Do you not have some sort of consumer guarantees act?

>>2537173
It's EU so of course there is. If I had to guess they added that so they can claim you connected your shit backwards and they're off the hook.
Either that or random electronic parts don't fall under the consumer electronics law that forces a 2 year warranty on everything.

>>2537181
>the consumer electronics law that forces a 2 year warranty on everything
Yeah I was meaning more of a general consumer guarantees thing that applies to every good and service. Basically, if it objectively doesn't function like it should, then you should be eligible for a replacement or return.

Tunrs out toner transfer on proto board doesn't work as good as on full copper

>>2537191
Interesting idea. Maybe spray the whole side with white paint first. Then figure out some alignment setup.

>>2537191
Go to your room and think about what you've done. No dinner for you, and I don't want to see you until morning.

rate my reverse current protection for usb
i think maybe for reverse polarity protection i'd need another p-chan in antiseries

>>2537194
Yeah two problems I figured are the paint grid that makes uneven contact and the fact that I put the board on a carboard box that bent with the pressure. For the alignment I'm thinking of placing dots on the four corners and pass needles or something in the corresponding holes
>>2537201
Please anon no, at least I didn't fuck it up too much both soldering and making the schematic

something is wrong with my car. pretty sure theres a loose connection or a short somewhere. Ive already done all sorts of testing, now im going to bite the bullet. im getting flickering of the lights and some internal lights here and there arent working.

what im going to do is go over all the possibly offending wires and try to find this 'short'

whats the best way to do this? i have a multimeter to check continuity(sp), that in combination with just physically testing connections, but also just running my fingers over the lengths of wire to see if i feel a part thats missing shielding and maybe making contact with the chassis.

This is going to take fucking forever. Is there a better way to do this!?

>>2537206
is that an alibay ad to dc converter board? brave move, but they do look rather handy and cost-effective.

>>2537210
It is, but I was thinking it would be better to put it off board or at least get better isolation from the rest of the circuit.
What I'm trying to make is basically an emergency light powerred by a lithium battery since I got a bunch of cells laying around

>>2537215
If you scrape off a row or two of copper pads seperating the HV and LV sides I'd call it good. So long as you have a good plan for the housing.

>>2536731
Dig. Thats my circuit now, but pulldown is just whatever was sitting on my bench. Literally soldered across mosfet legs, the whole shebang heatshrinked to the cable. Inline circuitry as I like to call it.

>>2537237
The smaller the resistor, the more current flows between the plates, and the more galvanic corrosion you get. Chuck it in a container of water and see how it fares. I wouldn't do it with anything lower than a few hundred kΩ, and would highly reccommend using something larger than 1MΩ.

>>2537211
not an expert at electronics or automotive electrical by any means but i would first buy access to a schematic. checking the relevant grounds would likely be quick and easy. if grounds are good, then start understanding the schematic and measuring voltage drops across different parts of the wire with a long set of multimeter leads with piercing tips.

>>2537244
ok, i do have a schematic from the maintenance manual(s). It helped and I already did hunt down all the grounds and checked their connections/tightness.

I think at the electronics shop (so fucking awesome we have one here) they have a tool thats has like a circle on the end so you dont have to pierce the wires. I'll check if its not too expensive and it can target the wire sizes/voltages im looking at.

Thanks!

>>2537092
As I think you have already established, the original is a rapid start ballast. This can be confirmed by the photograph that shows only two output leads. A rapid start ballast for a single lamp will have just 2 output leads. A rapid start ballast will have a more or less constant current into the lamp. The exact power delivered will depend on the lamp. for instance, more voltage will appear across a longer lamp. More voltage at same current gives higher power, which is why the ballast can be used with the variety of lamps of differing power listed in the ballast specs. You may have more ballast choices than it seems. I could be wrong, but I think a sure thing would be to find a ballast with 2 output leads having a compatible power level when used with a compact lamp like you have.

>>2537206
Resin 3D printer works well, ~50 micron resolution out of the cheap ones, fancier ones now push <30 microns.
Scale your shit appropriately and bring it into UVTools, slap your board down into the vat, expose it.
Depending on your printer, it might be easier to stick the board directly to build plate with some hot glue.
Works as solder mask, works to add a silkscreen, pretty versatile process.
If you pick up a spare vat cheap, you can use actual UV solder mask, but don't do it in your normal vat.
It's a massive bitch to clean out, it'll tint or otherwise fuck with your resin for actual printing.

>>2537358
A note unrelated to the replacement problem: many 4 pin lamps can also be used with a rapid start ballast. In this case the pins at the lamp ends are shorted together and tied to a single ballast output. Not doing so will create uneven current in the lamp electrode and shorten the life of the lamp. Also, in general, use of rapid start ballasts will shorten lamp life when compared to programmed start ballasts that preheat the electrodes before ignition (two wires per lamp end). A rapid start ballast can have more than two leads if it is designed for use with more than 2 rapid start lamps, but the lamps still connect to only one ballast output on each end.

>>2537363
Eh, I've seen mixed results putting normal resin on a PCB. Using one to photomask of dry film photoresist is probably easier.

>>2537417
Easy as fuck

Based. No rest for the bugmen.

>>2537358
thanks for the reply. One of the issues I was having was finding a ballast with only two output leads, then adding another variable in wondering about only running a single lamp from a 2-lamp ballast. At that point I realized that for less than $20 I could get an LED that plugged straight into the same mount as my bulb, and all I'd have to do would be to bypass the ballast (and thus also not have o worry about building a new enclosure for an oversized replacement ballast).

Your post did pique my interest on fluorescent lighting though, and I found this cool web 1.0 site with a treasure trove of info, including all different kinds of lighting, not just fluorescent https://donklipstein.com/f-lamp.html

didn't think I'd be late for bed reading about germicidal ozone bulbs

>>2537441
I'm not sure to what extent an LED replacement for a fluorescent tube light will run without a ballast. As far as I know they're designed to work with a ballast, but I may be wrong. Be interesting to know the optimal method here.

>>2536852
>I am not an EE I don't know much more than how to use a multimeter and how to solder.
to be fair, neither do most EE's

Need to power a gutted 12v cordless drill motor. What can I buy that will just plug straight into a 110VAC outlet and vary the speed in one item? Most things I'm seeing are either only a power supply or only a speed controller.

>>2537537
>in one item?
Don't know about that, but you need an AC-DC converter and a speed controller (PWM/H-bridge). You could leave the AC-DC converter at the wall, put a load on the end of the power wires, and adjust the output voltage to compensate for drop in the cable. Then put the H-bridge into an old battery case and put a twist lock power plug with strain relief on it.

>>2537545
This is assuming the power cable isn't very long or small gauge.

What is an impedance? Can you give me the explanation you would use to describe it to your grandmother. Thanks.

>>2537553
Resistance relative to alternating current, and the reactance of non-resistive components in the circuit.

>>2537553
https://electronicsclub.info/impedance.htm

>>2537562
Is this a good description: resistance, but for AC instead of DC (it depends on the signal frequency)

>>2537575
It's a very basic definition, but it is correct.

>>2537448
That old website is a treasure trove of old lighting technology info. I read for some time, too.

IDK the LED product you mentioned. If it says 120V input, then I guess it would be fine. Just don't plug a fluorescent tube into that socket again!

BTW, an LED power source is often referred to as a driver. A driver is basically a DC power supply of either constant voltage or constant current, depending on the requirements of the load. If the LED assembly has a spec of, say, 24V, then a constant voltage 24V driver is needed. If the spec is for a current, then a constant current driver is needed.

>>2537566
Retarded definition, your grandmother would not understand. Says depends on capacitance and inductance, but two lines below proceeds with the same formula as for resistance. Fuck off.

>>2537562
>>2537575
Retarded definition, because in tutorial to making PCB with DC current there was an impedance of 50 ohms, while you talk about AC. Fuck off.

>>2537553
Bump

>>2537627
What is an fuck off? Can you give me the explanation you would use to describe it to your grandmother. Thanks.

Does anyone have any experience with the Signalhound spectrum analysers? I'm particularly interested in the lower frequencies, it states the range as 1Hz to 4.4Ghz. which is much wider than the entry-level benchtop SAs which are also twice the price.
I'm looking to do something like emc testing but not quite, looking for any kind of emissions from electronic devices. Am I along the right lines with this equipment?
Sorry for noob question and sorry if this is the wrong thread. Thanks.

>>2537553
anything that impedes the flow of current

>>2537553
it's when your peepee can't stand up

any Phil's lab watchers?

any suggestions on a scope that's cheap for someone who may use it once or twice a year? I don't know anyone personally that has one that I can use when I get the time to sit down and prototype shit

>>2537678
Feranek -> any random youtuber -> shit -> urine -> phil's lab

>>2537553
I hope grandma is familiar with differential calculus and laplace transforms. Impedance is ratio of laplcae transform of voltage to laplace transform of current. Grandma also needs to be familiar with complex numbers, because in general we talk about impedance when dealing with steady state problems where we can abuse euler's formula to greatly simplify the problem and even use graphical methods which are retard friendly. To further simplify concept of impedance we can tell people that it's like resistance but one that depends on frequency. We like keep quiet about the phase though, because it only confuses people due to expanding problem to second dimension.

>>2537690
Sam Ben-Yaakov -> Feranek

Feranec is 2h of
>ooooh
>I didn't know that
>so wait...
>look at this amazing proprietary simulator that you'll never be able to afford
For PCB design just pirate Bogatin's books from libgen, it's all you need really if you understand your fundamental theory. That jewish professor has a lot of great lectures on all kinds of power electronics. If you really want to become a circuit elder-wizard and breeze through them like a chad that not even multiple feedback paths can stop then watch
https://www.youtube.com/playlist?list=PLc7Gz02Znph-c2-ssFpRrzYwbzplXfXUT
It's a lot better than spending listening to Feranec for 2h repeating same concept several times for 10 minutes while guest shills a product.
>WOOOOW so amazing

>>2537708
Who the fuck need fundamental science shit unless you are a neckbearded school teacher? Just learn how to do the job from Feranec.

>>2537718
How are you going to design your PDN if you have no idea about your components and how various parasitic elements determine your performance characteristics? What if you have to design a composite amplifier because there is no opamp with needed specs? Which MOSFET do I pick for my converter? Why is my amp ringing? Why is step response as slow as a dog dragging its balls on a carpet?

As soon as you can't copy shit over from datasheet/app notes you'll be on your own and no amount of seethe will spare you from doing neckbeard science shit.

>>2537730
>As soon as you can't copy shit over from datasheet/app notes you'll be on your own
Sure, right after apocalypse when I'm the only soul left alive and the internet does not work.

if i buy a clamp meter, do i still need to keep my cheapo-multimeter? Im not doing anything really serious, and it seems that a clamp meter already performs 90% of the multi's functions

Should I add multiple castellated holes dedicated for ground on my breakout board or would one be ok? It has a 2.4ghz rf chip on it.

>>2538095
What features does the clamp meter lack? I think it's usually nice to have two meters if just to make simultaneous measurements. Like current and voltage. Or voltage at two different points.

>>2538138
So long as all the RF stuff is properly contained with cans and vias and such, you should be fine. The castellations for ground shouldn't make a difference either way. Though for convenience's sake I like to see one or two grounds on each side. Maybe multiple VCCs too.

can someone design a LED flasher for me?
I have some jars of strontium aluminate rocks and I want a small LED powered by a button battery that flashes once every 5 minutes or so (ideally this delay would be somehow configurable) that I can glue to the jar lid. that's pretty much it

>>2538168
This is actually harder than it seems because of button battery requirement. You can't just use something like 555 timer because it will drain your little battery in no time. What kind of LED are you working with? Easiest solution would probably be to just add a small microcontroller with very low sleep power consumption. It all depends on what kind of efficiency/operating time you want.

>>2538095
>if i buy a clamp meter, do i still need to keep my cheapo-multimeter?

of course, dummy.
clamp meters are geared toward high-voltage, high current measurements.
a proper bench meter will be able to read microamps and millivolts in both AC and DC for all your small-signal needs.

>>2538168
>flashes once every 5 minutes or so

the LM3909 would be ideal for this but it was discontinued a long time ago.
you could try your luck getting chinked on Alibaby, or build an equivalent like this one.
change the 100uF cap to increase time delay.

>>2538168
>flashes once every 5 minutes or so

the LM3909 would be ideal for this but it was discontinued a long time ago.
you could try your luck getting chinked on Alibaby, or build an equivalent like this one.
change the 0.15uF cap, or 10M resistor to increase time delay.

>>2538228
great, thanks!

can i drive AC motor with a square wave if i basically fake sine with it?

finishing up my bachelors in electrical engineering this spring and i still feel retarded when it comes to electronics. any tips?

>>2538442
Just keep doing electronics. Degree is only a tutorial of basics, learning never really stops afterwards.

>>2538442
Forget everything you know. Go innawoods and eat hallucinogenic mushrooms with the raccoons for 40 days and nights until Moses leads you back to the promised land. Only then will the path forward be revealed. Good luck, kid.

>>2538168
CMOS 555?

>>2538140
> think it's usually nice to have two meters...
good point

>>2538211
but a simple clamp meter should suffice for lower voltage/amps through car wiring right?

>>2538479
wouldn't that turn the led on for half the time? I'm after a sub second flash then a minute wait

>>2538605
Pretty sure you can set whatever pulse width you want with a 555.

>>2538479
Would you need to use a transistor to drive a bright white LED at 3v? I have some glowing figures that I'd like to build something similar for

>>2538140
>I think it's usually nice to have two meters
I too prefer my girl to be on the taller side.

>>2538605
Use a diode and resistor series pair in parallel with the 10M resistor. Actually you’d get half the 555 current use for the given 10M resistor by using an approximately even 50% duty cycle, then using a series DC-blocking capacitor on the led (parallel bleed resistor required).

>>2538660
Read the TLC555 datasheet, from what I remember they can’t source as much current as the NE555. Best use a logic-level FET.

>>2538660
Here's one worth checking out

>>2538168
Would a single LED at the top of the jar charge them all or just the top surface?

>>2538715
be better off with this:

>>2537536
>to be fair, neither do most EE's
As long as you mean fresh EE grads, you're right

>>2537708
>Sam Ben-Yaakov
I thought i was the only one who watched him. Absolutely chad, explains everything there is to know about SMPS. Though as a hobbyist I find the math can be slightly overwhelming at times

I have a weird problem with a multiplug: After like 2 years of usage, its power switch stopped to work, ok I replaced the switch. Then after like 8 months, the new switch stopped to work.
I recently saw a repair video about old CRT screen and the guy installed 2 capacitors on both wires of the switch, claiming it will protect from any "electrical arcs". Ok why not trying this.

Is it a common thing to do to protect switches ?

>>2539071
Yes, adding a capacitor will reduce arcing, but it's usually combined with a resistor to form a snubber or arc suppressor.

>>2539086
Sorry to be noob, but I then asked to ChatGPT, with aim to request the good µF value, and the bot said me to avoid this.
I then tried to argue adding a resistor and it insisted the same to avoid this :(

>>2539097
It's ChatGPT not engineering GPT, so of course it will only give you bullshit answers. Usual standard values are 100 ohm, 1/2W resistor with 100-200nF capacitor. They even sell them already combined in packages like in pic related.

>>2539125
thank you. Can I call this a "RC" Filter ?

>>2539129
yeah it's a type of RC filter

>>2539129
In this case it would be more appropriate to call it RC snubber or if you're autistic
>resistor-capacitor voltage transient suppressor network

>>2539086
Are the arcs caused by inductive spikes through the chokes? Something tells me common-mode chokes don’t have that problem, and it’s more likely to be a normal choke or load inductance. Do chokes designed to be lossy even give big spikes?

Is low voltage that much harder to measure than low current? Most DMMs can easily measure low current down to a fraction of uA, even the cheap ones can measure down too 300-500 nA. But voltage wise they are about 1000 worse and can only go down to like 200-500uV. Are there meters than can measure 10uV or less?

>>2539269
You can feed a current through a known resistor, be it 1mΩ or 1MΩ, for a voltage in any range you care to be able to measure. Say, 2.5V peak for an ADC with a 2.5V reference. Then even if your buffering op-amps and/or ADC have 10mV of input offset, you're still getting only 0.4% error. But if you've got a low-voltage, you need to amplify it since an ADC measures voltage, at which point any offset error will also be amplified. Also the current through the meter leads into the input impedance may be significant enough to cause further voltage offsets than just the amplifier alone. Chopper amps and maybe even coulomb counters may be applicable here, but they're a lot more finicky than a cheap <1% resistor.

>>2539317
So I could just have 0.1% resistor and an INA??? current sensor set up in a x100 or x1000 mode and then feed the output to any DMM in the mV range and that would be pretty accurate? Why don't DMMs have an INA inside to do that for me?

>>2539321
By INA you mean instrumentation amp? They already probably have them, but to measure with a 12-bit ADC, FSR of 2.5V, LSB of 400nV, that's a voltage gain of 1.5 million. 123dB. Getting that extreme kind of gain with less than 400nV input offset would be pretty damn tough.

>>2539343
Yes TI has a bunch of INA*** chips, either general purpose instrumentation amplifiers or current monitors, including zero drift, choppers etc.
I don't understand how you ended up with 1.5 million gain. I thought if the measured voltage is 1uV, you amplify it to 1mV and then feed into the DMM in a 1mV or 10mV range and it will do the rest. Where is the 12-bit ADC? An external ADC or an ADC inside the DMM? Also how did you get 400nv? 2.5/(2^12) = 600uV?

>>2539404
>I don't understand how you ended up with 1.5 million gain
Shit I used E-12 as nano, but that's pico. In reality it's
2.5V/(400nV * 2^12) = 1500
Assuming you wanted 400nV as your smallest discernable value, then the full-scale range at that range would be 1.6mV. So to step 1.6mV up to 2.5V, it's a gain of 1500, but that's basically as precise as a 1mV range that meters often have anyhow.
>Where is the 12-bit ADC?
The one inside the DMM. Go for a 60000-count meter if you want to measure nV better.

>>2539245
In case of common mode choke load current is flowing in both (opposite) directions canceling out magnetic field it produces so it won't cause a rise of voltage when current is removed and magnetic field collapses. Only magnetic field produced is from common mode noise which will flow freely even if you turn off the switch because both power and GND paths conduct it. Normal chokes on the other hand will increase inductance of supply current loop. Chokes usually have low Q so voltage won't rise as much, but it will still contribute to total inductance. The biggest contributors to arcing are inductance of load and inductance of cables.

>>2539486
>inductance of cables.
Is that really more significant than the 10s of µH of inductance of chokes? Even though they're low-Q materials at high frequency, I think they can act as decent inductors at low frequencies. Though I'm not very familiar with different types of magnetic cores, nor am I really sure what "frequency" an inductive spike is.

Can you not just make a noise-suppressing inductor by making a seperate shorted winding? Probably shorted by a resistor, maybe a capacitor too. Or would you be too limited by coupling coefficient or saturation?

>>2539489
>Is that really more significant than the 10s of µH of inductance of chokes?
There is a lot of cabling in your electrical installation. Wall socket is not a perfect voltage source with 0 inductance. Even if your power supply cable is only 1m, there is 20-30m of wiring in the wall + whatever else going all the way back to transformer. For EMC testing standard is usually to assume 50uH of cable inductance.
>Can you not just make a noise-suppressing inductor by making a seperate shorted winding?
For DC it would probably work, but it would also give you a big spike on turn-on because it would momentarily act as a transformer with shorted secondary which is likely why this is avoided. It's better to tweak the core and make it act like a resistor at high noise frequencies effectively achieving the same thing.

I'd like to design a switch mode power supply to convert 120/240V AC to 5V dc. I would like to use a flyback topology, but i don't know how to design for frequency, power, current, inductance values, capacitor values, etc. Is there any good video or book that goes over how to design it? I'm not interested in designing the transformer itself, just buy all the parts (controller, caps, transformers, etc.) from mouser and build a PCB.

>>2539601
I had an idea like that too but gave up since most of those transformers are custom made for a specific device and they are tiny and not really serviceable unlike linear power transformers. There's plenty of info on the DC to DC part, but building a complete 120AC to 5V DC power supply is a different story. There are also "offline converters" that work directly with the mains voltage.

>>2539601
99% of the time controller datasheet will have whole step by step design process included, if not manufacturer will provide application notes on design. Go to your favorite distributor's website and pick a controller that works for you then just follow instructions in the datasheet. If you want more theoretical approach then watch Sam Ben-Yaakov's flyback videos on youtube.

>>2539615
You can get all required transformer parts: EI core, coilformer and clips. It's perfectly doable winding them by hand and there are step by step instructions on calculating required number of turns/wire diameter. But yes, you most likely won't find pre-made transformer that you can just pop in so making a custom solution is required.

Is this the hot? I think it is the hot.

>>2539634
It's 800 degree the hot.

>>2539601
>I'd like to design a switch mode power supply to convert 120/240V AC to 5V dc.

why?
every home has 10 unused phone (etc) chargers that already do that.
if not, you can buy an arm-full at the thrift store for $5.
if you're gonna expend some effort, make it unique and fun, like an orgasmatron or an upskirt camera.

I didn't realize that shipping may or may not be taxable depending on your state. Otherwise I would love the sellers who lower the price and jack up shipping to compensate. Which makes me wonder, if shipping wasn't taxable, wouldn't all sellers basically charge $0.99 and like $20 shipping to compete with each other? That way they wouldn't have to worry about the returns, too. Not sure if there are laws against that.

>>2539700
wait so you buy a $1000 scope for .99 and pay $1000 in shipping and zero tax woohoo, but then it is DOA so they will refund you 99 cents?

>>2539621
>If you want more theoretical approach then watch Sam Ben-Yaakov's flyback videos on youtube.
Cheers.

>>2539950
>>2539950
>>2539950
NEW THREAD