/diy/ - Do It Yourself

how do I desolder smd components without hot air gun?

So I'm trying to get my head around schematic layout, I'm not sure which method of grounding is more correct. Or do I just put a ground point on pin 4 of J1 and replace all the GND global labels with ground points? This is one of 8 sub boards hanging off a controller board.

>>2848242
use normal GND power symbol for all grounds.
plus I would recommend direct connection with normal labels, not global labels.

>>2848250
>plus I would recommend direct connection with normal labels, not global labels.
Yup, just realised the error of my ways. Last time I used globals I did not have headers on the same schematic sheet. Cheers anon.

>>2848237
Oven

What are the best feeling toggle switches you've tried?

Its not too difficult to find small snappy ones, but for permanent fixtures I want something that feels satisfying to throw. Kind of hefty, and maybe less stiff.

Unfortunately most product descriptions online do not include metrics of clickyness and thunkiness so I've been struggling to find what I'm looking for.

Any recommendations?

>>2848297
nobody buys toggle switches this way. they are made to a spec and work the way they do to handle the voltage and amperage rated for.

low amp and volt switches will be less "clunky" maybe.

you will have to mess with a lot of switches if you want to be this autistic.

>>2848297
switches are designed to break connections FAST to prevent arching. yes there are switches with mercury in them that make no click noise at all, I think they have been banned though (could be wrong).

bigger switch more voltage/amps bigger clunk more stiff.

also they wear out over time and become less clicky/clunky.

>>2848237
SMD is of the devil.

>>2848235
How do I go about fitting a ring magnet with a bearing in its centre &/or outer circumference?
I want it to freely spin like a turntable with both sides accessible.
I'd prefer to use the strongest N52 but its so brittle, is it too delicate to work?

I've been thinking about mating pic related with 20V drill battery for a simple portable PSU, anyone?

I skimmed the manual quickly but couldn't find any info whether it has low voltage cutoff or not.

Can someone recommend a good KiCAD tutorial?

>>2848237
>how do I desolder smd components without hot air gun?
Use bigger solder tip

I want to do coursera assignment for real.
1. Is it even possible to do on chinesium FR4?
2. What analyzer should I buy? Is LiteVNA better than NanoVNA ?

>>2848237
depends. If it's a cap or resistor, just use two irons. Pin lead just use wick. If it has a pad on the bottom you're gonna need le gun

>>2848313
Is using a nylon spacer or a rubber gasket between the bearing and magnet an option?
Or hell even tape?

>>2848237
https://www.youtube.com/watch?v=Xszfkjoqpbs
here he changes smd leds on an xbox 360, it's very possible with just an iron.

i saw another video of somebody that did this to a controller with just an iron too.

>>2848393
How thin on the nylon spacer?
(for the outside bearing)
How tight on the interference?

Another question: How dumb is an idea to interference fit a metal disc inside the ring magnet centre hole?
Is it feasible to spin the ring magnet with central shaft?
0.001 or 0.0005 interference?

since the /ind/ thread is garbage
does anyone know of a panel mount beeper that emmits a 50 to 500 hz tone? all the ones i find are ear raping 2.4khz, and the electro mechanical ones are not even remotely encapsulated

>>2848345
not the answer you're looking for, but just doing a simple project in it (schematic, layout) while reading the "getting started" and documentation on the website should get you going.

>>2848631
https://www.mouser.com/c/electromechanical/audio-devices/piezo-buzzers-audio-indicators/?frequency=50%20Hz~~500%20Hz&mounting%20style=Panel%20Mount&type=Buzzer%7C~Piezoelectric&instock=y&rp=electromechanical%2Faudio-devices%2Fpiezo-buzzers-audio-indicators%7C~Frequency%7C~Mounting%20Style
just did a simple search on mouser for panel mount buzzers of 50-500hz. the auer signal ones look pretty good, albeit expensive.

>>2848362
GHz projects are not trivial. Start with an HF (3-30MHz) project first, then a VHF project (30-300MHz), and do some serious reading on microwave circuit design before you even consider it
>1. Is it even possible to do on chinesium FR4?
different materials have different loss tangents at those frequencies. learn how to test unknown materials.
>2. What analyzer should I buy? Is LiteVNA better than NanoVNA ?
you can't buy your way to competence, and analyzers at those frequencies are not cheap

fuck this dogshit textbook

>>2848344
Use a BMS board.

>>2848235
why did you not post a link in the old thread
do people really browse the catalog whenever the thread gets old?

>>2848812
I browse the catalog as a rule. Idk how people arent aware of all /ohm/ threads at all times.
Literally just bookmark 4chan.org/diy/ohm

>>2848849
>Idk how people arent aware of all /ohm/ threads at all times.
Not terminally online.
>Literally just bookmark 4chan.org/diy/ohm
Well I do that, but I only ever use that bookmark when the thread dies without posting a new one, or I clear my tabs. A new thread link was posted in the old one, so I didn't bother looking for a different one. Why would I?

>>2848859
>Well I do that, but I only ever use that bookmark when the thread dies without posting a new one, or I clear my tabs.
You have been whining about this for multiple threads now (ie months, years), and it never occurred to you to change your workflow?
/ohm/ is slow anyways, you dont need a tab open all the time. Just check once a day through the catalog like with the news

>>2848865
I think posting the new thread's link in the old thread is a motion that takes significantly less effort than whatever you're suggesting I do, but sure I'll take it to heart.

Also the OP didn't use the new paste I posted. I just updated it to remove the grounding issue, by popular request.

Marco Reps referred to the LT3439:
https://www.analog.com/media/en/technical-documentation/data-sheets/3439fs.pdf
When making an ultra-low-noise DC-to-DC converter. Specifically, it's a push-pull transformer driver with slew-rate limiting, with two low-side transistors to switch a centre-tapped transformer primary. As far as I understand, it does this by limiting the collector dV/dt, and the emitter dI/dt. It apparently smoothly transitions from voltage rate and current rate limiting, so I guess the voltage waveform would be trapezoidal, with an amplitude smaller than the power supply voltage in order to limit current slew. That sounds like it's linearly regulating all the time, which seems like it would produce a fair bit of heat, especially in a monolithic chip with internal power transistors. But maybe it's ok.

I imagine that this will be significantly less noisy than a TL494? I wonder if there is a simple external circuit I can put on the output of a TL494 for the purpose of limiting its slew-rate? It switches its transistors on hard, so you'd need to take that signal and use it to power a capacitive ramp that itself gradually turns on another transistor. If not an op-amp integrator. Two or four op-amps is doable I guess, but it feels a bit silly.

I imagine a common ZVS circuit would be even better at noise? It has a sinusoidal current waveform, but the square switching waveform might induce transients. I plan on winding a ferrite toroid with RG174 coax with plenty of separation between the windings.

This shit sucks and leaks very easily between cap/threads making mess of everything, recommend me somehting similar that doesn't and foreso can be carried in a toolbox

>inb4 just get a locked one
No

>>2848996
pressol brass pipe is the og toolbox oiler, also works for ipa
available in half a dozen sizes

>>2848999
Yeah, wondered something like that could do, though I quite like the idea behind Menda's dispenser. Shame they don't do a mobile version with properly sealed threads and a second cap.

>>2848996
>>2848999
>>2849008
>Lockable for traver
>Spill an leak resistant design

>looks wimply as fuck

https://www.youtube.com/watch?v=2InbSBHqyhc 6:00 >

Wonder how bad all the others are if Menda sucks this bad

>>2849017
>looks wimply as fuck
the pressol? their smol toolbox oiler was made for a bundeswehr contract to mil spec, that thing neither leaks nor breaks in a human lifetime. ebay is full with 60 year old ones like pic

>>2849020
No, mate, meant the plastic Menda.

>>2848235
i'm trying to make an electronic load with an n channel mosfet and an op amp. pic related isnt my schematic but it's similar. i want to prevent the op amp from turning the mosfet fully on as a safeguard against short circuits. this means clamping the maximum gate voltage at some arbitrary value. i haven't characterized my mosfets and i expect it to be variable from FET to FET.... but how can i do this? i've heard low voltage zeners are basically worthless.

I have this PWM controller destroyed and its schematic seems very simple, can i substitute it with the same array of high voltage transistors?
>t. noob

What's a safe temperature to desolder a through-hole component without damaging anything around it?
I want to replace the analog stick on a joypad but the old solder is not cooperating.
How high can I crank the soldering iron temperature to draw it with a wick without ruining the board?

>>2849063
You should at least use a solder sucker. Use the wick to clean the pads afterward. It only needs to be heated until the whole joint is in a molten state. You may need to use a larger tip loaded with solder to aid in heat transfer to the joint.

>>2849063
crank up the temp to the maximum. if you're holding heat for more than like 5 seconds you're doing something wrong. time at temperature is what blisters PCBs.

>>2848237
>>2849063
If the board is free from sensitive components you can use the physical violence method.[The Slap Method for Solder Rework] https://www.sparkfun.com/news/802

>>2849072
compressed air works well too.

>>2849026
The easiest way to prevent too much current going through the MOSFET is a fuse. The second easiest way is to have a current sense resistor below the MOSFET and use an op-amp to provide negative feedback to the gate voltage. Which you're already doing.

>>2849040
Maybe? Doesn't look like any of them need to be matched pairs, but I can't find a datasheet to say for sure. That diagram could well be missing resistors. See what happens I guess.

>>2849063
I wouldn't usually go above 350C, maybe up to 420C if it's a really large bit of copper and I don't have my extra fat tips on me. Add some fresh leaded solder to the joint to help lower its melting point. A Yihua 929 desoldering iron is much more convenient than a conventional solder sucker and it isn't that expensive.

https://www.amazon.com/MG-Chemicals-8341-Clean-Paste/dp/B00425FUW2

This shit says "no-clean" but can I clean if afterwards without hassle if I don't want a flux-ridden board?

>>2849158
Yeah, all electronics fluxes I've seen come off with IPA.

What are some cool projects I could do if I learned to into electronics?
Will I be able to make something useful or something I could give someone as a gift, for example?

>>2849268
an led christmas tree. all blinking and cool.
and you still have time until xmas

>>2848235
I have a spare washing machine motor. What should I connect to it to cap the voltage at, say 5v when using it as a generator? What should I connect to it to be able to regulate rpm when plugging it into the socket?

>>2848237
>take old electric iron
>put it on its back to form a DIY heat plate
>set to whatever temperature works for the solder used
>lay PCB on top and wait
works well enough for single sided smd

>uncle borrow soldering iron for fixing something
>hasnt returned it 20 days later
>buy a new 60w soldering iron, solder wire and flux
>soldering iron is fine but came with 0.001g heavy pig iron tips
>solder wire is labelled as 60/40 but I'll be damned if it has even a single atom of lead
>solder flux is the only nice purchase, it's a liquid type, the first I've owned this kind

if the chink chip decap guy is around: im probably going to start SEMing shit again in the next few weeks.

>>2848344
Are you already set on a specific brand of battery? Ryobi's 18V lithium batteries have a BMS in the pack which would handle the low voltage cut off. Maybe some other brands do too, but I know Ryobi does for sure.

>>2848235
Thanks bro

>>2849669
I think Makita does. Tool Scientist on YouTube has analysed the Milwaukee and DeWalt hardware and protocols and they don’t have the ability for the battery itself to cut off current. And only some can balance-charge, for some reason.

>>2848697
why. It sounds cool

>>2849339
> as a generator
More cap = Less Ripple. Simple as.
Also depends on the load. More Load = More Ripple.
> regulate speed
If ein europe, probably 3000 RPM or 3600 in NA synchronized with mains frequency.
More Load = Less speed. Slippage as.

Am I stupid for thinking I can drive a bunch of DC motors with a TLC5947 and a bunch of p channel MOSFETs? I have 9 motors to drive via PWM and I'm stuck with a raspi which does not have native PWM like the ATmega2560 I wish I could be using.

>>2849669
Well ryobi is probably the most accessible.
You could easily convert a dual charger into a dual (or single) battery *holder* for 36 V out.

It shows you how retarded they are for not putting a USB A/C charge port in every charger for use as a power bank when mains power goes off.

As usual, I have to think of everything.
If you make one and sell it on etsy, I’d buy it.

>>2849723
A rasberry pi is practically a gaming pc compared to atmega. It’s an effing AVR. Does your actual gaming PC support PWM out through the quad display ports on your RTX 4090 Ti Founder’s edition even though you paid $9999.00 for it? No.

Get a cheap-ass microcontroller with PWM and connect it to your Pi via I2C or something.
Then, when you blow up the chip because the flyback diode died, it won’t kill your gaming PC.
Also put it in a socket.
Hard mode: opto isolate everything and put optoisolaters in a socket.

>>2849726
> A rasberry pi is practically a gaming pc compared to atmega.
Which is kind of what the project needs. It'll have a 3.5 touchscreen and a 12 button matrix for UI. It was originally going to be 2560 based but we got a rush of blood to our cocks and the scope of the project blew out.

Seems like a cannon for a mosquito kind of deal though. And the only uC I am familiar with that supports 9 PWM outputs is the 2560. Which would still need a bunch of current drivers after it anyway. Effectively I'd be substituting a $5-$10 IC for a $60+ dev board. Which is silly anon.

I might just breadboard it up and see what explodes.

>>2849723
I guess it would work, if you had appropriate pull-up resistors.

Or just bit-bang the GPIOs into Nch FETs, you’ve got spare cores for a reason.

For my next project I'll be using a shitton of thin low voltage wires, is there a nice and small male/female connector that I could use for dividing long wires into shorter sections? Not sure what's the proper name, but something easily detachable like pic related (XT60) but for single cable. small and light prefered.

>>2849769
Bullet or spade connector?

>>2849769
JST connectors. Small, crimp on, is nice

Is there a cheap machine recommended for pcb milling. Ive repurposed an old 3d printer for this but its just not meant for any loads whatsoever on the toolhead. I still make it work but its a full day to get 1 working pcb.

>>2849795
Not that anon, but are there any JSTs for a single cable?

>>2849799
You can bolt a 5W diode laser on the 3D printer instead. The idea is you spray-paint a black lacquer on the copperclad board and burn it off. Then you wipe off the residue with a wet paper towel and etch the board. Once etched, you can dissolve the lacquer with alcohol. I've done this a bit and it works great for SMD boards, or boards without precise hole drilling. But then I tried to turn my Ender 3 into a drilling machine too, and it just wasn't rigid enough, the drill wandered significantly. By etching drill marks you can kinda centre your drill bits, but it isn't that reliable. Doing solder mask is also doable, but only if you can spread the mask thin enough to not burn the FR4, and even then it smells like shit.

I've got myself a 2nd hand original Snapmaker router/laser/3d printer and am modifying it to work as a PCB drilling and lasing machine, or just to route the boards without bothering with the etching process. There are even spring-loaded cutters for cutting away a solder mask layer without damaging the copper beneath it.
There are plenty of chinky desktop CNC machines out there, I'd look for reviews for PCB usage. With 3D printed brackets you could probably make your own machine from 4040 extrusions and lead-screws, doing so might be more cost effective than buying one.
I know there are all-in-one PCB fabrication machines used commercially for rapid prototyping, but AFAIK they're really expensive. Maybe you can track down a 2nd hand one.

>>2849803
not smol and not for tiny wires but fastin faston is another popular connector (for automotive)

>>2849699
>they don’t have the ability for the battery itself to cut off current. And only some can balance-charge, for some reason
Wow, who signed off on that? I guess systems that have always been lithium can get away with putting the shutoff in the tool, while systems that support other chemistries have to put the shutoff in the batteries that need it. But no balance is just retarded.

First time removing an IC and installing a DIP socket. I do not understand why every tutorial I've seen drowns everything in flux, I did just fine without any.
Took me more to align the pins of my cheap shitty AliExpress EPROMs than installing the socket did, ended up putting my chip in another socket and double stacking it on to the installed socket. Does some sort of a chip pin alignment tool exist?

>>2849820
>Does some sort of a chip pin alignment tool exist?
It's called a "pin straightener".

>>2849820
>Does some sort of a chip pin alignment tool exist?
You can usually just bend them against a the edge of a table to straighten them

What's the best option for a high power 100-200uH 5-10A inductor? I am finding stuff like in the picture but Rdc is too high on the order of 100-200mOhm. That's probably because they are tiny and use very thin wire. Would it be better to buy a suitable core ring, large enough, with high permeability, and wind a few turns of very thick wire?

>>2849860
The best thing to do when you need an inductor for high current switching is to wind your own. Get some thick wire and torroid core

>>2849860
Your typical junk PC ATX or old TV switching power supplies have several high power inductors and transformers with heavy gauge wire and toroid cores. Pull, measure or modify those parts for fun and profit.

>>2849863
>>2849864
But how do I measure the saturation current? I guess I could just test them and if they work they work.

>>2849873
Lots of web info on doing just that.

From TEXAS INSTRUMENTS:
Waveform Audit: Is Your Inductor Saturated?

https://www.ti.com/document-viewer/lit/html/SSZTAV7

>>2849873
FWIW, an inductor test circuit from another discussion.

>>2849772
> bullet or spade?
So, what I do is I just melt or hammer out the metal pins on the female and male XT60s connectors, then cupsolder the wires and heatshrink the ends.
Voilà! Bannana jacks that are compatible with XT60s connectors.
This is handy when you’re in the field and you need to, say, switch the polarity for whatever reason.

>>2849860
> high power
Just like capacitors, in power applications bigger is usually better.
I turn on p2p wired shit from the 50s and the wax has melted off the caps…. Works fine.

You have to understand that most commercial products are not just build down to a cost, they’re built down to a minimum cost assuming a warranty lifespan. That was true even in the 50s.
When I hear “i’m a professional engineer” i’m thinking they’re doing the absolute minimum so they don’t get sued or fined.
For DIY you can always do better.

>>2849820
The pins of stanard dips are flared out slighhtly when you buy them and it’s a bugger to get them into sockets so we have some of picrel lying around.
As far as I know, it’s called an “IC Insertion Tool.”
Ahh, the 80s were a wonderful time.

>>2849892
What's the idea here? You feed PWM of a certain frequency and the inductance will have a certain impedance and you can monitor the expected current via R1, and as you increase duty cycle at some point the inductance will begin to drop and the current through R1 will increase sharply? And at that point measure the current and figure the DC current based on the duty cycle?

>>2849716
it's 90% self masturbation over derivations and basically no worked out examples; completely useless for teaching yourself anything

>>2848710
You mean a battery with BMS? Doesn't all of them have some kind rudimentary BMS but the tool itself takes care of cutting off.

>>2849669
Most likely Parkside as they are the cheapest brand and as I'm buying it solely for portable battery projects the tool lineup or color doesn't matter.

>>2848235
anyone ever try DIYing one of pic related? $1400 is a bit steep. i know that the key here is rise/fall time...

>>2848297
Carling. They feel the best by far. The ones in your pic feel like shit.

>>2849993
fuck im retarded.

>>2850016
It's expensive due to being a calibrated and certified measurement instrument, guaranteed and warrantied.

I cant understand electronic length. The Wikipedia just made it worse. How can the antenna be wound up? how can width count as length? Am I just too stupid?

>>2849820
>Does some sort of a chip pin alignment tool exist?
I 3D printed one.

>>2849918
EC3 and EC5 connectors have pins that are seperate from their connector housings in the first place, but the anon asked for connectors for "thin low voltage wires". XT60s are definitely not for thin wires.

>>2850025
>How can the antenna be wound up?
The impedence of air is significantly higher than the impedance of the wire, so current is constrained to the copper path
>how can width count as length?
Antennae dont have to be wires. In the extreme case, an antenna can be a 3D block of of metal, in which case all major axes, diagonal, linear, or even more complex reflection loops (think like the Windows bouncing logo screensaver), can be treated like an antenna path of their own.
>Am I just too stupid?
We're all stupid in the eyes of god

>>2849993
100us is a 10khz sampling time, that's literally nothing
Theres probably a sensor for an arduino out there you could buy. What do you need it for?

>>2850025
An antenna with the proper length, like a half-wave dipole, will couple to both the electric and magnetic fields of the radio wave. It will have inductance and capacitance that resonate at the radio wave’s frequency. A shorter dipole will have less inductance, and so couple more to the electric field than to the magnetic field, and will also have a higher resonant frequency. It will have to be compensated with a series inductor for optimal energy coupling at the radio wave’s frequency. Conversely, a longer dipole will have more inductance, and so couple more to the magnetic field and have a lower resonant frequency. It can be corrected for with a capacitor.

Antennas other than simple dipoles and monopoles can be expressed with this notation too, as I understand it just means if they act like antennas with higher or lower resonant frequency than they’re being used at. An AM loop antenna for example is basically an open core inductor, so it will couple more to the magnetic field, and so act like an electrically long antenna compared to the wavelength. The actual length of the looped up wire can still be significantly shorter than the wavelength, “length” is figurative once you move away from simple dipoles and monopoles.

>>2850030
> XT60s are definitely not for thin wires
XT30s exist

>>2850154
Well I already mentioned the EC3, which is a similar size to XT30, but is actually usable as a single wire connector. And I wouldn’t consider 16awg as “thin” as opposed to what a small bullet connector would take, or a JST.

>>2850030
>I 3D printed one

waste of time
you can just press pins down on tabletop
even empty can of spaghetti-o's if you cant afford nice table like mine

>>2849820
Alignment tool? Nikka, just press the pins gently against a flat surface to align them. That's it.

>>2850216
>you can just press pins down on tabletop
Yeah but with this little mijigger the spacing is always perfect to slot right into a breadboard or IC socket. Using a table involves more guesswork. Took 5 minutes to download the model and put it on my printer, will save me more than 5 minutes after a few years. I also printed some resistor lead-forms, not just for the standard 0.4" pitch for a DIN0207 resistor, but for the narrower 0.3" pitch and a 0.1" pitch vertical. I should make one for DO-35s too, both 0.3" and 0.2". Damn is it a pain when you bend them 1mm too wide or thin and you have to force them down.

>>2850056
>What do you need it for?
dicking around with engines, like viewing compression waveforms or fuel pressure trends. i'm not a mechanic so it's not critical for it to be accurate.

>sampling time
the response time of the mechanical device inside of the transducer is more important. like how quickly the diaphragm moves in response to changes in pressure. unfortunately the chinkshit transducers on aliexpress don't really bother to characterize this.

>>2850275
yes, and 10kHz is literally nothing. that's within hearing range. any material that isn't "muffled" in sound response should give a useable measurement, so like a hard plastic membrane or something

Need help. Plugged some wires in power tool battery. Multimeter said 18.8v. Plugged wires into buck converter caus I wanna see 12v. When putting positive wire in battery last big spark appears. Wtf am doing wrong

>>2850326
Parkside 20v battery (18.8) and XH-M401 DC-DC Step Down Buck Converter by the way. I did not have a load connected. Guess i dont know electronics basics enough. What went wrong bros

>>2850326
Sometimes just arcing can cause harm to sensitive electronics, where a pure resistive load would be fine.
You need a better switch.

Here’s one of those things I went 50 years without realizing. Those cheap $1.00 home electrical light switches at home depot or lowes? They work well for minimizing arcing BUT they’re only rated for 400 loads! Not 1500W like you would expect. So they don’t work for very long.

>>2850355
* 400 Watt loads, that is.

>>2850355
Thanks man for the answer but I dont understand at all. See picrel for what I did. Just two components and two wires. What I expected was that the step down converter could handle the battery attached. I imagined I would insert the multimeter on the other end. turn the little screw, and mess with the output voltage. When I then find 12 something volt I can then think of the next step.

But all that happened is I first inserted positive obviously and then negative and then immediately big spark when touching the negative terminal on the battery. Must be a simple mistake on my part in the plan?

the bastard in question

>>2850355
Maybe I get it. You're saying this is volatile stuff and I better build a switch in so that I can gently turn on the circuit instead of holding the wire and the negative terminal which have a huge electrical potential difference so close that they apparently arc (spark and flash happening)

>>2850362
It doesn’t matter if positive or negative first,
If the dc/dc converter had a capacitor on an inductor in the input section it could have magnified the arc and “presented” itself as a short—if only for a millisecond.
The arc itself contains high frequency voltage spikes along with electromagnetic radiation (you can hear them on the AM radio) that can kill proximal electronics.
Put in a beefy current limiting resistor of some kind.
That’s always a good idea when playing around with something for the first time. Or make a resistor. I.e. use a car light bulb or something between the positive battery terminal and the DC/DC converter.
Also be aware that battery packs can supply like 50 A pretty easily, that’s enough to briefly arc-weld something. As you demonstrated.

>>2850368
Yes. Paddle switches are good at not making arcs, although they cost a bit more.
Penultimate is a solid state relay. No arcing whatsoever.
Ultimate is ssss - sold state soft start. On on the ‘make’ connection, it slowly allows current into the load device over a period of, say, 1 to 5 seconds.
You see this on big audio amplifiers… they have huge capacitors that look like short circuits until they charge up so we soft-start them.

I know this is a pain, but it’s just one of those things. Kind of like switch debouncing in the /mcg/ channel.

>>2850362
I think the issue is just that you have a big spike of inrush current going into your circuit when you plug it in. It should work fine as-is, though that spike isn’t necessarily good for your capacitors. I’d look into adding a soft-start circuit with a resistor that gets bypassed by a MOSFET once an RC circuit gets charged up, choose a MOSFET that can handle a 24V gate voltage if possible. That’s like 4-5 components. Adding a choke is an option, but if you have a transient load on the other end of the buck converter it wouldn’t be a wise idea.

>>2850362
>big spark when touching the negative terminal on the battery

- the converter might be DOA (dead on arrival)
- the spark may be normal; does it work otherwise?
- the converter was sitting on some metallic object causing a short
- you connected the battery to the output side

>>2850420
>valtage
Love the care they put into their documentation.

>>2849892
I picked a "real" part and the simulation does show the current slope that looks like saturation. But it seems to be because of a stupid high series resistance: 100uH 7 ohm? And the current slope looks like a low pass filter. How do I tell a low pass filter due to high series resistance from an actual saturation due to loss of inductance?

i have a fairly basic common-emitter amplifier, using an f_t=45Ghz BJT. the collector resistor is 220 in parallel with the input impedance of the next stage, which ideally is very high. unfortunately, modelling the connection to the next stage as a transmission line means that at high frequencies the collector sees the 220 in parallel with z_0, which is 50 as standard. this seriously hurts the slew rate of the amp. in a previous stage i solved this same issue by putting the ac load at the input of the next stage (i.e. on the other side of the transmission line), but that only worked because the ac load for that stage was ~50 anyway.

i could fix this by changing the 220 load to 50, which might be okay. i could also thin out the trace to 220, but i would like to avoid that since im not 100% sure what the tolerances of JLCPCB are.

is there any other way to increase the z_0 of the trace? putting a 170 resistor in front of the line technically works but it forms a voltage divider which is not workable.

>>2850474
By measuring the heat dissipation in the inductor. ESR heats it up, saturation just changes reactance so doesn’t dissipate heat. You could also duplicate the circuit with an identical ideal inductor with that much ESR and compare the two. Or read the spice model file.

>>2850495
You could add a matching network or transformer, but I suspect using a 50Ω resistor would be best. Any reason to use 220? Adding a second transistor as a common-collector might be an option if using 50Ω will mess with the input impedance.

>>2850514
>You could add a matching network
i think the collector would still see 50 ohms, no?
>Any reason to use 220?
for gain. A = 220/11 = 20. i need a gain a tad over 40dB, so for two stages i could go 50/4, but degenerating an emitter with a 4 ohm resistor doesnt seem very stable. i could try 3 stages at 50/10 but that might degrade the signal too much. plus, like you said, using tiny ass emitter resistors makes the input impedance drop.
>Adding a second transistor as a common-collector might be an option
i was going to do that anyway, but the issue was the transmission line. i ultimately decided to just put the follower as close as physically possible to the collector and just hope it works.

>>2850520
>i think the collector would still see 50 ohms, no?
No, the transmission line should see a 50R source, and the amp should see a 220R load, if you design the network properly. Though that only works for just one frequency.

>>2850533
can you describe what you mean by a matching network? because im imagining pic rel.

>>2850533
>>2850534
oooh wait you mean this dont you

>>2850535
Yes. I think a CLC T network would work after your amplifier, especially since it removes the DC bias after your amp. Though a low-pass matching network might be preferable if you’ve got any harmonic distortion to squash.

>>2848313
How do I go about fitting a ring magnet >>2848313 with a bearing mechanism in its centre &/or outer circumference?
same solution should work both inner & outer
I want it to freely spin a ring magnet like a turntable with both sides accessible.
I'd prefer to use the strongest N52 but its so brittle, is it too delicate to work?

Pic illustrates 2 problems.
[1] using a bearing mechanism..
Can we use additional ring magnets as substitute?

[2] Hi-RPM (12k rpm+)
Without wobble.
Interference fit of 0.005mm
using mild-steel.

>>2848235
How do I go about fitting a ring magnet >>2848313 with a bearing mechanism in its centre &/or outer circumference?
same solution should work both inner & outer
I want it to freely spin a ring magnet like a turntable with both sides accessible.
I'd prefer to use the strongest N52 but its so brittle, is it too delicate to work?

Pic illustrates 2 problems.
[1] using a bearing mechanism..
Can we use additional ring magnets as substitute?

[2] Hi-RPM (12k rpm+)
Without wobble.
Interference fit of 0.005mm
using mild-steel.

So I tried making an LC low-pass filter, then added some secondary windings to the inductor and shorted them out. I expected the effective impedance to the signal to increase at high frequencies, but it actually went down. I tried this with a toroidal choke, and with a mains step-down transformer. My intuition for this circuit is obviously inadequate. Looking at a sim of a low-loss transformer, the corner frequency increases without a significant change in Q factor, unless I use a non-trivial resistance across the secondary winding, and the corner frequency shift doesn't seem dependant on secondary inductance.

I guess it's similar to a current transformer, eh?

I'm trying to think of a failsafe system for a microcontroller based device. I have a MPC23017 driving a bunch of solenoids, I need the ability to fire its reset pin to de-energise them if either the firmware loop stops looping, or the microcontroller itself shits the bed and stops sending signals.

I initially thought of dedicating a GPIO to a simple transistor inverter but this will only catch the controller losing power. If it freezes the GPIO will likely remain high, preventing the failsafe from triggering.

I'm considered using a PWM signal at a relatively high duty cycle via an electrocap to both smooth the wave out enough to be something close enough to DC for the transistor, and to hold the capacitor at sufficient charge to keep the transistor saturated during the off cycles of the wave. Microcontroller freezes? Clock stops, PWM stops, cap discharges, BJT shuts the gate, reset gets triggered.

Anyone got a better idea?

>>2850670
>firmware loop stops looping, or the microcontroller itself shits the bed and stops sending signals.
That's called a watchdog timer

>>2850676
>That's called a watchdog timer
Well fuck me, isn't that just a delicious item to find out about. TY anon.

>>2850420
No made sure to check which side is input and used that. The spark scared the shit out of me and made me doubt every electrical knowledge i have (little). Will try again soon maybe once I understand all the stuff that was written

Can somebody explain to me why a single 3 watt AA battery connected to a high voltage generator cant kill you? With around 100 volts it should have 30 milliamps of current, which can cause heart fibrilation, no?

>>2850688
The real world is full of inefficiencies. No boost converter will be 100% efficient, and tiny cheap ones are notoriously poor - they won't be able to maintain that much current. Also at 100v, the resistance of the outer layer of skin will probably block a good chunk of it unless it's pierced/wet/breaks down.
It is certainly possible to get a lethal shock from a small battery-powered device, as the many accidental deaths from tasers show (a device designed NOT to kill), but it takes specific circumstances.

>>2850514
>>2850474
Never mind, I am retarded, I realized the saturation is when the current shoots up not down, i.e. inductance drops, and what happens in this graph is a typical LR series wave form because of the high ESR (probably an SMD part). I doubt it is possible to simulate saturation in LTSpice with their lump models of real parts. I don't think there is anything related to the ferrite properties in the model.

>>2850726
Here is what I am talking about.
It probably shouldn't matter much for power inductances since their ESR is typically very low <0.1ohm, so the unsaturated current can be approximated by a straight line Vt/L and it would typically saturate long before it starts curving due to the ESR.

would these little fuckers work in a hard vacuum? i know alkaline batteries are usually too leaky for this sort of stuff but apparently metal lithium cells are better sealed so im not sure if these would be any better

>>2850828
Whatcha makin bro?

>>2850828
Any battery with a liquid electrolyte is suspect, those are lithium-iron-disulfide cells and have the normal range of volatile organic carbonate electrolytes like you’d find in an 18650. I’d be looking for solid electrolyte batteries, maybe ceramic. Liquid metal or molten salt would have low enough vapour pressure to contain in a cell, in case you wanted to make a mercury battery like what was inside the VT shell. Sulfuric acid has a pretty low vapour pressure, so maybe something like a lead acid battery could be vacuum capable, so long as there’s never very much water. If this is vacuum chamber work we’re talking about, then every battery in existence probably offgasses something awful.

Can someone give me a ballpark of how much it cost to order a custom rp2040 board.
I have something that I wanna sell around 100 of that uses 2 picos and Im already ordering boards for this. So I was considering adding the rp2040s/components with the assembly to the board instead of the footprint of 2 picos.
Thing is if its going to be like triple the price of just slapping on 2 picos Id rather not. I have a hot plate but I also dont want to value my time at $3/hr. I could do like $10/hr.

>>2850891
>Whatcha makin bro?

vacuum chamber to degauss the jooz

>>2851152
Does it have a wooden lid and swimming pool?

>>2848235
how were incandescent signs like pic related animated? it would be trivial to do this with relays or solid state relays and mcu-driven timers, but i doubt this is how they were done "back in the day".

>>2851170
Think 'pinball machine'.

>>2851170
Rotating cams+switches were common for cycling.
Cheap, fairly reliable, and easy to service.

does it make sense to do pic rel with inductors? i.e., put a couple inductors in series to compensate for the frequency dependent inductance of non-ideal inductors?

>>2851285
that many capacitors doesn't make sense. iirc it's just an electrolytic + a ceramic to account for the EL's ESR for high frequencies

>>2851328
>iirc it's just an electrolytic + a ceramic to account for the EL's ESR for high frequencies
not enitrely sure how putting something in parallel with a leaky cap could possibly fix it.

>>2851285
The major non-ideality of inductors is ESR. You can't fix that by putting them in series. The next-most significant one is equivalent parallel capacitance, putting them in series doesn't help for that either. You're better off just designing one inductor with thicker windings or larger spacing between the windings. Furthermore, inductance is proportional to turns squared. So two seperate inductors with X total volume and Y turns (and so a similar current rating) are going to have less inductance than a single inductor with the same volume and turns count X and Y. And the less total turns you need to reach a certain inductance, the thicker you can make those turns, and so the lower the ESR.

>>2851330
>leaky
He said ESR, not parallel resistance. A capacitor with capacitance C, series resistance ESR, and series inductance ESL will act like a snubber/filter circuit. If you plot its impedance with respect to frequency, you'll see a standard -10dB/dec slope until it tapers out due to themESR, before going up again due to the ESL. Pic related. Ceramic and film capacitors and other low resistance caps will often spike to ESL before the ESR even effects them. I find X2 mains caps to be particularly good.
By stacking capacitors of different sizes and constructions in parallel, you can broaden this curve, maintaining lower impedance at a wider range of frequencies. Many datasheets recommend 100nF MLCCs at each power pin for high frequency bypass, with a larger 470nF-10µF MLCC that's connected to all of them for lower frequency bypass.

>>2851285
tried simming this but it doesnt seem to work the way i think it does. magenta line is impedance and green is inductance (in uH). this is two inductors in series, a 33uH one with an SRF at 20 MHz and a 470 nH one with an SRF of a little over 1 GHz. theres a dip in inductance around 170 MHz since the inductors see the parallel capacitances of each other. damn, back to the drawing board.

>>2851438
>The major non-ideality of inductors is ESR.
im trying to block RF so thats not an issue for me.
>He said ESR, not parallel resistance.
whoops my bad

>fancy heterojunction BJT
>B-E built-in potential: 1 V
>B-C built-in potential: 0.4 V
soooo i guess this thing can just never be used in saturation?

>>2851458
>B-C built-in potential: 0.4 V
I don't understand this

>>2851475
i assumed its the diode drop from the base to the collector. a difference that large guarantees that saturating the transistor puts the base-collector junction DEEP into forward bias. its absolutely killing my switching speed, by like 3 orders of magnitude.

>>2851477
Wait, so when you turn on the transistor, the collector voltage drops so much below the base voltage that current can flow through from base to collector?

>>2851478
yea, if the collector voltage drops to within 1 - 0.4 = 600 mV of the emitter for this particular device (assuming im understanding "built-in potential" correctly).

>>2851479
Wouldn't this not be a problem if you weren't saturating your BJT? I thought for high-speed applications you never wanted to saturate them much in the first place, hence the invention of the Schottky transistor. Which kinda does the same as your problem in the first place, which is odd. Add an emitter resistor and call it good?

>>2851482
>I thought for high-speed applications you never wanted to saturate them much in the first place
yea i think im learning first-hand why thats the case.

i have a project where i need to do current sensing and using a quad op amp for space saving is attractive, so i thought to use an lm358... except its input offset is completely unacceptable at ~3 mv. i would need something on the order of tens of microvolts or less input offset.

is externally trimming the offset at all viable -- like does it work in practice to the extent that i would need it to work?

obviously i could just get an op07 (or something more modern) which has very low input offset but i would prefer to stick with a single quad op amp if at all possible.

>>2851568
well i tried setting input offset externally on an op07 and it worked.

>>2851614
how low were you able to get the offset trimmed?

the more i think about this (thermal effects, etc.) the more i think that i might need to re-assess how i'm doing this.

>>2851621
I don't remember exactly, about a few uV but I also have a basic DMM not a 6 count bench meter so that was good enough for me. Also if you use a microcontroller you can do it in software so thermal stability is more important than the offset.

Necer really soldered SMD except for some 2 pole components like capacitors LEDs etc.
Having a lot of trouble soldering this ESP32 to the board.
I flood the area in flux, run my TS80p at 320ºc, and run 60/40 lead solder across the connections.

Looks good untill I check for continuity and find a few shorts, so I solderwick it off and start again.

Any tips? this is driving me mad

>>2851633
Instead of wicking it off, clean your tip and run over the joints until the tip is loaded with solder. Clean the tip again and repeat until the bridges are cleared. Add flux throughout the process.

I like the idea of this solar mains hybrid lamp but the whole thing seems over-engineered. Also, why is the current sensed only from the first of 9 LED strings? It's like a disaster waiting to happen.
https://www.edn.com/solar-mains-hybrid-lamp/

>>2851651
I tried this and saw a little improvement, only managed to do one side with this technique. Thanks for the tips, will practice.

>>2851752
It might also be that your solder sucks, so consider getting some Kester 44 or MG Chemicals flux-core 60/40 or 63/37.

>>2851568
LM358 is a dual op-amp, the LM324 is the equivalent quad. Consider a chopper amp like the COS8554, it's pretty cheap.

>>2851633
That WROOM-32E module doesn't look very flat to the board. You may want to remove solder from the pads so they sit flush, before flooding with flux and applying solder. A bump caused by one solder joint can stop the pads touching. When the pads are touching, flowing solder onto the base pad should make contact with the castellated pad above it, making a joint, but if they're too far then this blob of molten solder might not reach. Well even then, the castellated vias should be possible to apply solder to directly at the same time as applying solder to the pads below. Ultimately this comes down to having a good tip geometry, in order to apply both molten solder AND enough heat flux into the board and the module at the same time, without taking so long that it burns away the solder.

>>2851727
Why would you want to power an LED lamp while the sun is out in the first place? Anyhow, that LED driver will probably be a constant-current driver anyhow, so the feedback isn't really necessary for protecting the LEDs. The current feedback is likely to pair with the voltage sensing from the solar panel to approximate an MPPT algorithm. Assuming it's a switched mode LED driver. Without MPPT tracking, by drawing too much current from the panel when it has less sun can cause the voltage to drop too much. The MCU is also monitoring the panel voltage so it runs off solar if both it and the adaptor are turned on, by telling the adaptor to go into standby until solar power is no longer available. This is solely a cost-saving measure, and really only makes sense if you're using a lot of power since LEDs are pretty efficient.

Can a pager battery explode by shorting it remotely? Sounds like bullshit. How do you do that and even it was shortened the explosion wouldn't be so deadly. There must have been something else there.

>>2851850
All the media are talking how there have been explosive charges in them.

>>2851850
The pagers were produced in Taiwan. The shipment was intercepted and diverted to Hungary for modifications before they were sent to their final destinations.

>>2851727
You could probably get 90% the way there with a analog circuitry. Big solar converter doesn't want you to know but the difference between maximum power point and constant voltage is not a whole lot.
Either a bank of comparators that switch strings over to mains when the voltage drops below certain thresholds or if you use a constant current linear regulator you could just regulate the power from the solar panel and let the mains take up the slack.

>>2851862
Sounds like an awesome electronics project.
>what do you do at work?
>umm not much just some modifications to pagers

>>2851875
This was the work of Clock Boy!

>>2851850
There's no question it's a hidden explosive charge. Even the most touchy of batteries don't just detonate like that, no matter what abuse you put them through.

How am I supposed to adjust resistances when potentiometers are so damn expensive

>>2851886
lol forgot about the clock boy. except he would probably be on the other side.

I have a fairly modern turntable. Probably 2010's. Piece of shit. I want to slow it down, slower than 33 rpms, without fucking up the motor. Can I solder a potentiometer somewhere to give the turntable a variable speed? If so, where would I look to solder such a potentiometer? It's an Ion TTUSB, for those wondering.

>>2848235
I want to use a couple of 18650 in parallel to charge an MCU. What will happen if I hook them up via TP4056 each like picrel, instead of hooking them all up to one TP4056? Is it safer/better?

can somebody help me identify a SMD switch part number? On the board it says S601, it has a little rubbery momentary press part in the center.

>>2851906
TP4056s are just charging controllers. By running two of them connected to the same 5V supply you gain the ability to charge the system twice as fast, since they can only handle ~1A or so by default, though a switching charge controller would be a more tactful solution to this. Just running a single TP4056 will charge it fine. Protection-wise, you may want to connect a seperate DW01 protection board to each cell, if you have protection chips that are built into your TP4056 boards, then I guess using two seperate TP4056 boards isn't a terrible idea. If they're also boost converters though you might run into some strange issues. Best off running the MCU directly at battery voltage if it's 5-3.3V tolerant, or running the MCU off 3.3V from a low dropout linear regulator like the HT7333.

>>2851895
i was going to tell you to take the chinkpill with LCSC but most of their pots are trimmer style (screwdriver) rather than plugin/through hole knob style.

>>2851895
>so damn expensive
Wait until you see the price of motorized slide pots.
Wanted 8 to 16 of them for a project. Don't think I'll ever build that unless I figure out my own way to add motors to them.

>>2851862
>The pagers were produced in Taiwan. The shipment was intercepted and diverted to Hungary for modifications
Actually not, the pagers were supposodley licensed to a Hungarian company, who built them here. Only problem is that the Hungarian company has 1 employee, the owner, who is some dumb bitch. The company HQ is one of the flats on the 8th floor of a residential building. It's obviously a shell company. Fucking kikes doing their dirty business in my country, but I swear to God if some sandnigger blows himself up on the metro while I'm omw to work, I will personally deliver the mail bombs to the Jewistani consulate. In minecraft of course

>>2851996
>25 dollarydoos each on Adafruit

>>2851996
>>2852011
Pricey for sure, surprisingly so on AliE too. I would set up ebay alerts for Music Gear known to use such sliders and try to salvage some. Or you can try buying a 100 piece lot or samples from Alibaba. $400 for 100 is expensive too but you could flip the excess to break even or for profit.

>>2852013
>$400 for 100
Hmm, not too bad. Per-unit this is more in line with what I'd expect, a damn shame they can't be had in low quantities for that price.
I had considered going the route of buying in bulk and reselling most to cover it, but it's a bit steep upfront for something that might be crap. Still, not a bad find.

>>2851797
Just received these from Amazon, ordered before I saw your comment.
I'm from Spain so my access to stuff is limited. I'll definitely look for those you say.

Are these considered good?

>>2852024
I looked further and found one that allows smaller orders. $5/ea for 16 with shipping. Costs nothing to ask the sellers for a quote. But if these prices are real it raises the question about why they aren't already being resold on AliExpress or in western stores...

>>2852030
The brand is fine, but that's lead-free solder which is more difficult to use with a regular iron because it melts at a higher temperature. If you can get leaded solder that would be best, but if you can't, then use a larger tip with more thermal mass and increase the temp on the iron.

Is it ok to run 2 stepper motors with single driver as in left side of pic related, than using 2 drivers?
I am going to use TMC2100 and want both the motors to rotate at same time and degrees

>>2851895
Use digital pots. They're cheaper and you just need some pushbuttons to achieve 60+ dB adjustment ranges.

>>2852045
I think the motors have much less torque if you connect them in series like that.

>>2851895
>$1
they should be $0.50 wtf

>>2848235
Could someone recommend an oscilloscope and soldering station for a beginner? Used is fine - something that I can grow into without breaking the bank

>>2852057
By the time you've set up a microcontroller and all the extra bits to control a digipot, you may as well have just bought a conventional potentiometer

>>2852059
can increase the voltage. They are designed to handle 5v. the controller is designed for more than double. I wonder if this setup will mess with micro stepping or silent stepping of the tmc2100

>trying to put together an order to build a circuit
>it uses some obscure JFET that is discontinued as a through-hole component

I get the impression that everything is moving towards surface-mount, but what is the equivalent of making a stripboard prototype using surface mount components? Surely you're restricted to using custom designed and etched PCBs?

>>2852267
>but what is the equivalent of making a stripboard prototype using surface mount components?
Custom PCBs are as cheap and fast as they'll ever be. The chinese fabs are basically sponsoring electronic hobbyists at the moment. But there is always breakout boards for smd's if you want the more traditional breadboards and stripboards.

>>2852267
>Surely you're restricted

dont call me shirley

>>2848235
Lads, I need some help with my car's radio (will post this to /o/ as well tomorrow, am too shitfaced to deal with their 'tism rn).
I've got this BMW Multi Radio (6910475) in my E46 which has been acting up for years ever since I had to jump start it in 2018. I assume the other bloke's starter-generator thingy blasted too much juice, whatever.
The radio used to be a on-board computer as well with a lot of options including a CD-Navigation system in the trunk, but ever since that jump start, I've been left with the on-board-computerless options a standard radio would give on an E46 (Audio, [2nd one I forgot], Outside Temperature, Turn off Screen) and I can't cycle the on-dash screen anymore with the indicator-stalk (stuck on range estimate).
It also has had problems with the audio which are as follows: If the radio is turned on WHILE a cassette is seated inside it (say I forgot to eject the aux-cassette I use to play audio earlier), the radio will literally just die over and over again (screen fades to black and comes back to life for a split second) with a loud and annoying popping sound from the speakers until I hold down on-off button for the radio which turns the radio off (it checks for on/off before it checks for a cassette I guess) after a few cycles. If I eject the cassette, turn on the radio, press the cassette button, insert the cassette, the cassette reader works. The same thing also happens when I try to tune the bass/treble/fader/balance while the cassette is inside the radio.
Trying to fix this, I bought a used replacement unit just to realize that the same head unit exist WITHOUT the computer features, (only 1 instead of 2 connectors).
Anyways, after that I took the (first) sucker apart because I thought some capacitor or something might've blown, but the twat is pristine inside.

Now, before I shell out ANOTHER 80€, what could be the cause of this? EEPROM erased?
There's no separate computer I could replace from what I can tell.

>>2852282
You need to enter anti-theft radio code.

>>2852293
Thought about that as well, but BMW models E46 and onward do not use nor have anti-theft codes for OEM radios so it was never delivered with one in the first place, not that there's an option or prompt to enter a code anyways.

Thanks for the thought though.

>>2852295
Check the fuse contacts and fuseblock for corrosion. If the radio is switched by a relay, remove it, clean the socket and replace.

>>2852297
Thanks, I'll check tomorrow.

>>2852267
> that everything is moving towards surface-mount
Everything already moved to surface mount.
Now things are moving to flip chips, bare die, and pellet. As the 1 cent on epoxy and the 1 cent on the leadframe carrier is cutting into private capital’s profits.

>>2852282
Get it into that boot-looping state and measure the voltage rails on the PCB with a multimeter. I'm guessing some PSU section of the board is working suboptimally, and can't handle enough current to supply two things at once. Could be capacitors, but switching and linear regulator ICs that might get hot are also suspect. You've potentially sent a strong voltage spike through the radio with the jump-start, and that would have hit the 12V input electronics first, so anything that's directly connected to the battery voltage is what you're looking into.

>>2852282
Rip that fucking garbage out and replace it with a 3rd party radio that is (a) better than the OEM crap and (b) infinitely re-replacable.
Just like you got rid of your keylock wheel nuts the first time you were stranded on the side of the road, unable to get them off to change the tire.

>>2852267
Could be that almost any jfet will do.
Especially where there is negative feedback or used for switching.

Negative feedback was invented because transistors have a beta that can vary by 100% between devices.

>>2852267
An exact match isn't necessary, in this case I think you'd just need to match the R_ds vs V_gs curve. Maybe the V_ds vs I_d curve too. Even if you use a different JFET, you could change resistors until it works, assuming the control voltage can go high enough. But I have no idea what the transconductance specs in a JFET datasheet refer to, so it might not be trivial. There's probably a guide somewhere to finding equivalent JFETs for VCAs and such, or you can try Octopart.

Also are 100nF electrolytic caps really used for audio circuits? 100nF ceramics exist and would work better for power rail bypass, and 100nF film caps aren't that expensive for signal use.

>>2852336
I don't think there's negative feedback in this case, they're each being used open-loop as VCA, VCO, and VCF.

>>2852368
Thanks anon, I have actually been continuing to read more about JFETs over the last few hours, to try and understand how I might substitute a different component (I'm still noob).

I agree that it will be important to match the parameters of a replacement part as the JFET is being used as a kind of proportional control. What I'm wondering is if I can breadboard sub-sections of that circuit (like just the VCA or VCF) for testing purposes before I actually commit to soldering the whole thing. I guess resistor values are the thing that will need changing.

Could be a good learning exercise or a complete nightmare, not sure which.

>>2852374
get a bunch of these, cut out a row of 3, then swap in JFETS until you get the sound/effect you want
JFETs vary a lot anyways, so even if the part was the right number it might still not have the right specs. given how inconsistent they are in general, in price, specs, and availability, I'd say learning how to swap parts is an essential skill if you ever work with them

>>2852374
>What I'm wondering is if I can breadboard sub-sections of that circuit (like just the VCA or VCF) for testing purposes before I actually commit to soldering the whole thing
Absolutely, I'd even consider breadboarding the entire thing. Though I'm the kind of person who only ever prototypes the bare minimum workings of a circuit before designing and soldering up the supporting circuitry. Only to find I made a silly mistake or two.

You could also solder it up with these: >>2852376
But if you never find the right JFET, or need to swap out a bunch of resistors often to get the circuit working well, you'll waste more time.

I'd buy an assortment of JFETs that might work, maybe on alibay, but you can never really trust parts that aren't the absolute cheapest BC547s and LM358s. I got fake TL072s once, and my TL431s were frustratingly rebadged TL432s.

>>2852378
>I'd even consider breadboarding the entire thing

Under what kind of circumstances do circuits behave differently on breadboards compared to a soldered PCB - does it only make a difference for super sensitive or high-frequency applications?

>>2852382
Only makes a difference where:
>breadboard connection resistances are significant compared to currents
less than 0.5A should be perfectly fine
>breadboard capacitive reactances at the given frequency are significant compared to resistances
shouldn't be an issue at audio frequencies even with 1MΩ resistances
>you still have gunk on your leads from where you pulled them out of the strip of resistors/diodes
clean or trim your leads before breadboarding
>you have one of those breadboards with power rails that don't run the full length of the breadboard
check this first, it's a classic mistake

Sometimes your component leads or solid-core jumper wires are excessively loose, especially in well-used or cheap breadboards. In this case, you may want to add a layer of solder to them, or bend them in a corrugated fashion to make better contact, or put multiple leads in a single hole.

Do I need anything besides soldering knowledge to make a mini pci-e male to hdmi female adapter? Im thinking it would be as simple as getting a mini pcie to hdmi male-to-male cable, carefully desoldering the hdmi male end and marking pinout then soldering to an hdmi female port. What wire would I need in case I mess up part of it and have to rerun a connection? Am I missing anything else? I just wanna have a cable I can leave plugged into my t440p hanging out the bottom for an egpu.

>>2852368
> I don't think there's negative feedback in this case
I admit I just skimmed the schematic, it looked like two were used as switching and one was piggy-backing on the op-amp’s negative feedback.

You usually see negative feedback in op-amps because the gain is insanely high.

>>2852386
>hdmi
Yeah it's gonna be a PITA. HDMI plugs have pretty tiny pads on them. Modifying a cable specifically means finding a socket that you can solder wires to (all the ones I've seen are PCB-mount), solder a bunch of tiny wires to it, and somehow provide adequate strain relief by potting the resulting cable end in a mould. You'd have an easier time taking an HDMI cable with a socket already on one end, and splicing the two together. That way the strain relief issue can be solved just with some good glue-lined heat-shrink. Though you would need to insulate each solder splice, I'd likely use liquid electrical tape for that.

What is the point of a mini PCI-E to HDMI adapter in the first place? Is it being used for an HDMI video signal or something else?

>>2852389
Well most of the op-amps are being used with negative feedback, either way the JFETs aren't inside negative feedback loops. Though it's possible that Q1 is being controlled with negative feedback via U1-A and U1-B, though the whole schmitt trigger and diode circuit makes no sense to me. Q7 and Q8 are definitely being used open-loop.

>>2852386
they sell male to female hdmi adapters, anon. but lets back up for a second:
>I just wanna have a cable I can leave plugged into my t440p hanging out the bottom for an egpu.
yea thats not how that works. any piece of hardware with an hdmi interface is either a video input or a video output. soldering a different connection onto it does not magically switch it from one to the other. and even if it did, how are you even gonna use the gpu? am i understanding this correctly, you want to connect an EGPU to your thinkpad with an HDMI cable? EGPUs work over PCI-e via thunderbolt, you cant just plug in an hdmi cable and have the thing turn on.

The traditional flow battery uses four different aqueous oxidation states of vanadium to make up its half-cell reactions, meaning the electrodes act inert, and all the battery's chemical potential energy is stored in the solution. This allows the solution to be stored outside of the electrochemical cell and pumped into it. This could work identically with only three oxidation states, though the cell potential would be lower and hence less energy dense.

A more recent idea is an iron flow battery, whereby one half-cell reaction involves iron 3+ reducing into iron 2+, with the other half-cell reaction involving iron metal oxidising into iron 2+, when discharging. The opposite happens when recharging. The first half-cell is purely aqueous just like the vanadium battery, but the second is not. So I wonder, is there any density reason to pump the iron 2+ solution in and out of this second half-cell? In order to scale up the system, the iron electrode will need to be enlarged. I'm guessing that before it's worth adding seperate tanks to the system, the iron electrode will get stupidly large, unless the density difference is that significant. But if not, then why wouldn't you make flow batteries with other metals as they're pulled in and out of solution?

I also notice the difference between this and a lead acid battery. These flow batteries require an ion barrier to prevent the anolyte and catholyte mixing because the electrochemical energy is stored as a potential between these solutions. While a lead acid battery stores most of its electrochemical energy inside the electrodes, and charging produces the same soluble product at both electrodes, so it can avoid the ion barrier.

I was looking into the history of batteries, and found that they really just used unfired ceramic (i.e. terra cotta) as an ion barrier, and then cardboard. Which would surely all leak slowly. Only recently have proper ion exchange membranes been possible.

>>2852434
Fair to assume Im a sperg wanting this to just magically work but Im not that far yet. I dont know why the exp gdc egpu is built to use hdmi connectors for its own cables it just does. Pic related plugs into a Pcie x16 slot that also has connections for power and then you just slot the gpu into the x16 slot and start the software config from there.

>>2852400
Again no idea why they made their cables work this way I just think zhang had extra stock or an opportunity for cheap components and they repurposed the connectors. Thanks for letting me know how much of a pain it'll be. I'm gonna use the cable full length and just deal with the inelegance.

>>2852479
It's high speed differential pair signalling, and it's using an HDMI cable as a cheap (PCIe-x1) cable to connect your mini PCIe slot to the full length PCIe slot. Some other cheap ones use a USB3 cable.

>>2852490
Yeah, it’s a good idea. Cisco routers used hdmi for the inter-router connection bus for a while.

>>2852490
>>2852491
Why HDMI and not one of the other differential signalling cables?

>>2852513
I'm assuming it's due to cost and availability.

>>2848362
>chinesium fr4
at frequencies above more or less 1.50GHz you'll have to start being careful with trace width tolerances, the microstrip to coplanar transmission lines from pads and a myriad of casuistries that will fuck with your project. You could use fr4 made in china if the seller is reputable enough to trust that the substrate and copper thickness are what he's selling with a reasonably narrow margin of error.
>litevna or nanovna
if you're asking these questions then you dont know what you are doing

>>2849158
no clean for some reason doesnt mean that it doesnt leave residue, it means that you dont need to clean it since its non corrosive and not conductive. You can clean it it wont cause any problem, but if you really dont wanna clean whatever i recommend you water based flux like this one from jbc https://www.jbctools.com/fl-15-flux-in-bottle-15-ml-product-150.html

>>2849769
use wagos

>>2852479
>Fair to assume Im a sperg wanting this to just magically work but Im not that far yet.
you were right, and i was wrong, and you were very polite about it. not many people have the restraint to do that. thank you, anon.
i think a female/female adapter ("coupler") would probably work for what youre trying to do. if nothing else, this one on amazon costs $1.25 and i would definitely try doing that before splicing a new connector.
https://www.amazon.com/Micro-SATA-Cables-Female-Adapter/dp/B06WV8J6X1
>>2852535
NTA
>casuistries
thats a cool word, ill have to remember it.

>>2850828
we send satellites with lithium-ion batteries, you'll be fine

>>2852539
no clean flux is a low grade conformal coating, you can even buy it as spray lacquer to preserve pcbs for prolonged storing

>>2851930
s601 is the designator, it doesnt mean anything regarding the part number. Look for tactile switch with a round button actuator style

>>2852328
I just realized that I'll have to get a new multimeter, mine got ... uhh how do I tell this ... yeah my old workplace stole it.
If anything blew, it must've done so internally because no MOSFET or cap is blown at all on the mainboard and I've replaced a lot of broken hardware at said old workplace (Escape Room with 1960's russian technology running in the background).
>>2852329
>better than the OEM crap
>infinitely re-replacable
My man, BMW did us dirty with the E46 and uses non DIN-standard form factors for the radio and HVAC controls. It's all fucking ugly, but if you've got something nice for me (preferably with Android auto), it should ship to germany.

>>2852564
ANYWAYS, can someone recommend me a good multimeter please. This is what I used to have, is it good enough to just get another one?
https://www.br-mannesmann.com/en/products/digital-multimeter

>>2848235
why mini jack ports have so many pins? I have two different makes of jack ports for standard 3 part jack and they both have 5 pins.

>>2852600
there's a mechanical contact that breaks when you plug a jack in
but the real question is, why are the pins in such a schizo layout?

>>2852600
Literally look at the fucking datasheet.

>>2852600

I expect for some level of mechanical robustness and footprint compatibility with the ridiculous TRRRS version.

>>2852567
It’s a bit shitty looking, but it will do the job for DC stuff like this.
So long as it updates fast enough to catch the boot-looping.

>>2852603
What's the point of this? Why would I need these pins connected, when there's no jack ergo no current? Is there any reason I'd want to have a signal there when nothing is connected?

>>2852604
Here's your fucking datasheet straight from the manufacturer
https://www.monacor.com/products/components/cables-and-plug-in-connectors/plugs-and-inline-jacks/pg-203jn/

>>2852648
Other vendors have better datasheets
https://www.cuidevices.com/product/resource/sj1-352xng.pdf

>>2852648
>What's the point of this?

once upon a time they had radios with speakers
when you plugged in headphones, the speakers would get cut off

Just how important is it to specifically use tantalum capacitors anyway? From what I understand, they can hold more charge in a smaller package, but has the technology changed a lot in recent years if I'm using an old schematic?

i need a button that lets current flow through unless it is pressed, with smooth, springy travel rather than tactile or snappy
are digikey's on-mom switches what im looking for?

>>2852757
NC

>>2852725
As far as i’m aware, tants were just ways to get amazing surface area with oxidized sintered tantalum powder giving you electrolytic-like performance in a small package without liquid electrolyte. The dielectric was the oxide, and thin hence the low voltage and probably a bunch of other shitty properties.

To on things like effects pedals and fuzz boxes where they essentially don’t know what they’re doing, it may be using these shitty tant side effects for something.

I usually try and test substitutions in a sine-wave oscillator in audio frequency range to see if I can hear any difference (e.g. audiophiles using polyester caps everywhere for no good reason).

>>2852513
Hdmi has multiple differential 100Ω impedance twisted pairs and can go at an aggregate of over 100GHz for a $1 cable. My local dollar store sells them.

This is because they were willing to take massive losses to get it into the hands of consumers in an attempt to make DRM pervasive. In the mean time, people that can use them for other than their intended purpose can benifit.

>>2852550
> no clean flux is a low grade conformal coating
Sounds like something a shitty no-clean flux mfr would say when customers complain about the residue their no-clean leaves behind… it’s a featue!

>>2852564
> non DIN-standard
Wr’re talking about finding the physical speaker wires and the 12 V line, disconnecting those and reconnecting it to a bluetooth amp and throwing it behind the oem shit.

>>2852813
>Wr’re talking about finding the physical speaker wires and the 12 V line, disconnecting those and reconnecting it to a bluetooth amp and throwing it behind the oem shit.
So a sleeper build where I replace the electronics inside the oem radio with something modern? That's already planned, but I can't get myself to do it.

>>2852642
>So long as it updates fast enough to catch the boot-looping.
That I don't know, it was pretty decent but not really high speed.
But I'm taking recommendations (as long as I can buy it off of Amazon or the local hardware store).

What's the advantage of JFET vs. MOSFET in circuits? Nowadays there's even power JFET devices available. What's the purpose?

>>2848235
Anyone know of a free program where I can test ladder logic from home?
>t. college student with currently limited access to the pic's being worked with

>>2852812
spray on flux
flux and protection of pcbs

>prototyping electronic load project
>use KUNGBER™ 0-30v, 0-10A power supply
>dies after 5 minutes @ 30w load
i've had this piece of chinkshit for 5 years and apparently this is the first time ive used it above a few watts.

>>2852827
Look at the schematic symbols.
JFETs have a junction at the gate, MOSFETs are’t even connected and have an even higher input impedance. JFETs are typically used to replace BJTs where you need a high input impedance at the gate. You can jump the shark and get JFET input op-amps as well.

>>2852918
> ladder logic
Ugh…
Anyway, it’s trivial to press F12 or Ctrl shift I on your browser and whip up some javascript one-liners to test them (assuming you know how to read/interpret them correctly). That’s what I do anyway (although I just use C instead of JS… as one would)

>>2852827
different characteristics for different uses
mosfet is off by default, jfet is on by default
I think the power jfets are mostly for RF because they're faster than MOSFETs
some other bullshit too probably

>>2852933
What do you need it o be variable for?
For 45 years I’ve used linear wall warts and regulators to test power everything. You can get them at goodwill if need be, for cheap. You can pick up good quality witching ones from Sony and Motorola as well.
Like a dell laptop supply supplies around 20 V at 2 amps. Your xbox puts out 12 V at like 10 A.
You only need 3.3, 5, 12, and ± 15 V.
Nowadays you might need 1.1V to power things like RAM chips.
If you need varying power, you hook up a potentiometer to an lm317 and put it in a box to make an adjustable output. I just measure it on my DMM, but you could spend another $1.00 and get one of those voltage panel meters from china.
Then you’ll know how to fish instead of just buying a fish every time.

>>2852937
>What do you need it o be variable for?
dicking around with power electronics. i recently used it to test an offline auxiliary power supply by injecting 30V with a 1A current limit. having a cc/cv power supply is very helpful.

>>2848235
beginner here, why isn't this oscillator circuit working? I'm getting oscillations at the inductor, but I can't get it amplified. any help is appreciated

>>2852935
>whip up some javascript one-liners to test them (assuming you know how to read/interpret them correctly). That’s what I do anyway (although I just use C instead of JS… as one would)
>>2852937
>If you need varying power, you hook up a potentiometer to an lm317 and put it in a box to make an adjustable output. I just measure it on my DMM, but you could spend another $1.00 and get one of those voltage panel meters from china.
>Then you’ll know how to fish instead of just buying a fish every time.
I guess it's boomer nonsense hour, eh?

>>2852931
>Kontakchemie
Man, fuck Conrad for closing all but one shop. There's now a fucking woolworth where my Conrad used to be :/

I feel like idiot asking this but what do you use as a ground point when testiing stuff?

Because I got a Pinecil and now I'm worried that my typical USB PSU with no ground leaks AC to the tip.

>inb4 kitchen sink
>inb4 wall radiator
>inb4 a dedicated ground stick on your backyard
>inb4 just use mains ground pin bro

>>2852978
I hope AC isn't coming out of your USB PSU.

>>2853001
https://www.eevblog.com/forum/beginners/mini-ts100-soldering-iron-and-current-leakage/
https://www.eevblog.com/forum/reviews/usb-powered-irons-(pinecil-and-similar)/
https://www.eevblog.com/forum/reviews/beware-of-the-pinepower/

Technically everything works as it should and there's no risk of getting zap'd (if you are not a sensitive component) but the folks at Pinecil or other USB iron makers should tell people about this.

Anyone here ever work as a "PCB CAD Detailer"? Is is as satisfying and enjoyable as I think it will be? Are there decent prospects for being able to make $80k to 100k a year without a degree if I do this for a few years?

>>2852725
In the case of that schematic, they don’t matter at all. You could easily use any common sort of capacitor for that, be it electrolytic, tantalum, film, or ceramic.

But in some cases a datasheet will specifically call for a tantalum capacitor, most often with linear regulators like the 7805 or 1117. This is because tantalum capacitors have non-trivial ESR, and some older linear regulators rely on this to ensure loop stability. Modern low-ESR ceramic capacitors could cause oscillation, though I’ve never seen it myself. You can use ceramics if you add a small (1Ω) series resistor, though I suspect you’d be better off taking the output off the cap, using it as an RC filter as far as the load is concerned. Modern linear regulators are often specifically rated for use with low-ESR ceramic capacitors.

>>2852757
Often datasheets will rate the hysteresis, or otherwise state the latch and unlatch force. You’ll want to find a switch with minimal difference between these force values or minimal hysteresis. Also yes, “normally closed” is what you’re after.

>>2852818
AN8009 is decent value.

>>2852827
JFETs are almost always used for analog circuits, either as high-input-impedance amplifiers, or as voltage controlled resistances. MOSFETs are much less linear and are almost always used for switching.

>>2852933
Probably the heat sink compound dried out and the main transistor overheated. Should be a quick fix.

>>2852937
Variable CC/CV is useful for load-testing. Like power LEDs, electric motors, etc. the current limit is also very handy even when being used as a constant voltage supply, since it prevents you from frying your circuit if you wired something wrong, though I’d prefer a power supply with a fast overcurrent latch in that case.

>>2852938
A 25 W incandescent light bulb will limit a 30 V power supply to about .8A when it’s cold, and less as it heats up. You could do it with a 40 W bulb, too, with a higher initial amperage. Feel free to characterize the bulb with your DMM and a short circuit as the resistance of the bulb will go up.
> power electronics
Doesn’t usually have to be that precise.

>>2853051
Cmon, at least recommend an LM317 constant current circuit, or one of those LM2596 modules. An incandescent lamp is insufficient for testing a power LED.

>>2853016
Its part of my job but i also do everything else besides production firmware (research, design, prototypes, BOM).
It is satisfying to develop something and release it as a commercial product.
Without an EE degree its doubtful. If its exclusively CAD then an associates degree can be enough. Larger companies that can have design specialization likely have an HR bot that filters you.
Having a good portfolio of professional level designs is a huge asset once your application makes it to engineering managers.
So build that portfolio in parallel with getting a degree.

>>2852827
In analog jfets have similar input resistance to mosfets but can have much lower noise, input capacitance and higher transconductance. They also self bias as a current source since they are depletion mode.
I've only seen the power jfets used to cascode power mosfets or used as the switch in input protection.

>>2852267
Try a J113

>>2852725
Tants are ass signal caps. small electrolytics at the time were also ass but are significantly better now.
With the resistances just get an electrolytic with decently low leakage current.

>>2853130
Remember we just want to test it, and for a 1 A at 30 V rated LED, it will last a lot longer if you under-drive it. It should still light even at lower currents, as long as the voltage beats the diode’s turn-on voltage.
The incandescent bulb is “pre computed” in this case to drive it at just under it’s rated current, you don’t have to mess with your DMM measuring the current on the lm317/2n3055 pass transistor circuit.
Using resistors is absolutely fine as current limiters with regulated output voltage. In the case of an incandescent bulb, it’s just a convenient and cheap power resistor. It’s “fault” is it has a massive PTC so it will pass less current as the filament gets hot.
This can also be seen as an intrinsically safe way to do it.

>>2853163
>Try a J113
I've also heard recommendations for J201 and BF256B, to be honest I might order all three and experiment on a breadboard.
I built something else today in order to build up to the circuit using the JFETs which is going to be a nightmare to translate to stripboard, I'm sure

>>2852978

>>2848235
why can't kicad just add support for multiple pcbs in one project?

Where do I solder?

As in literally where in the house do I set up things to solder? I don't have amazing ventilation or a garage, so where do I do it?

I understand you need a fume extractor when you solder, which I didn't even know at first. The beginner tips didn't tell me that. How much would I roughly need to spend on a decent one that actually protects me from toxic fumes instead of just saying it does?

>>2853353
>Where

sounds like you've been brainwashed
you dont need any special place or special safety equipment
as a exorbitant luxury, you may set up a small fan to ensure smoke stays away from your face

>>2853353
If you really need a fume extractor you can use an inline fan and flexible ducting to exhaust out of your window. Are you soldering for extended periods of time? Solder particles don't go airborne until the alloy is like 2000 degrees, and flux fumes are comparable to cigarette smoke- the exception being synthetic liquid flux which is bad to inhale.

>>2853289
Make each PCB into its own project.

how in gods fuck do I check if my solder is leaded or not??? this garbage has NO label on it

Broke off the tip of the audio connector for my headphones. The problem is the wire is some proprietary meme cable (picrel), so I can't just go to a local store and grab a replacement; I have to pay out the ass for a new one + delivery. Would just chopping off the broken connector and welding a new one in its place work just as well or does it need some magical audiowire welding skills for the end result to work well?

>>2853377
Nvm I can just get a regular 2.5mm cable + 2.5 to 3.5mm adapter.

>>2853359
I thought flux fumes are equally bad no matter whether they come from solder or separate flux.

>>2853357
>>2853359
I thought the fumes from solder were pretty toxic, especially flux? I mean, I use the cheap ass shit so I thought it would be better to be safe.
The few times I have soldered the smell has been overpowering, so I figure maybe I need to do it in a different location that's more ventilated.
Then I saw something about fume extractors. I don't solder for long, but being new it just takes me time.

My flux is in sort of a hard gel form, it's shit. I've tried heating it up with the iron itself before but it doesn't really help.
As for the inline fan and ducting, I just don't really have the space to set up for that. I can solder pretty much next to a window though, but what do I get to extract the fumes and point it in the direction of the window? An extractor fan, or something cheaper still? It's hard to tell what's a good extractor since on Amazon it's all just blatantly fake reviews

>>2853376
It's leaded.

>>2853184
>stripboard
idk why people do this to themselves. must be masochists

How should I protect the "+" op-amp input in my led bar circuit? Feels wrong to just leave it like that, I have a feeling it will get fried by static electricity almost instantly.

>>2853402
>fried by static electricity almost instantly

it's an analog chip
you can hit with 1 million volts of ESD and it wont even notice

>>2853399
>must be masochists

either that or they like to have a project completed from start to finished before lunch

>>2853399
> stripboard sucks
Yep. You’re better off with perfboard and wire it up on the back.
I don’t even understand how people conflate basic principles of electronic construction to come up with such a thing.

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

>>2853402
Assuming the signal input is under 1 V p-p, put two series diodes going in each direction to ground. The diodes won’t turn on until about 1.4 volts and shunt/short-out any higher voltages on the inputs.

>>2853396
good enough for me thanks anon

>>2853409
Surely it's much more tricky to wire up all the components on the underside of perfboard, while avoiding shorts, compared to soldering them into stripboard?
Stripboard is the only reasonable way to build a one-off or prototype circuit, unless you want to design a whole custom PCB and wait for it to be shipped from China at great expense.

Did something happen regarding regulations for ISM bands around 169MHz? All front end IC I could find for that frequency are being discontinued so I'm being forced to rely on remaining stock to avoid having to place a switch+LNA+PA+filter. Feels bad man

>>2853421
Shorts and stripboard go hand-in-hand. Never had a short on perfboard because there’s no pads. I’m not cutting traces with a dremel or exacto knife either. It’s a lot easier to cut wire.
There is one thing worse than stripboard:
There’s this german guy on youtube that makes traces using these huge “solder worms” that he blobs on instead of using wires, or traces. It’s insane…
…and also a contender for looking even worse than stripboard (if you didn’t think that was possible).

Something that looks better than stripboard, but still terrible, are these PCBs that have the solderless breadboard layout…presumably so you can just transfer your breadboard project onto the PCB, solder it all in and voilà. Like stripboard, it’s not a very space-efficient way to do things, either.
And in both stripboard and breadboard layouts you *still* have to have a lot of jumpers.
Perfboard just has short and optimal jumpers…half the time you can just use the leads on the component.

>>2851895
Following on from this I have ordered 40 potentiometers for 8 USD, I can't wait to see what's wrong with them

>>2853353
>Where do I solder?
Anywhere. Only would suggest not doing it where you eat/prepare food.
>you need a fume extractor
Not really, unless you're doing it day in and day out. It's an irritant, but not particularly toxic.

>>2853416
Nice idea. I was thinking of adding e.g. 24 V zener diode to ground for some basic over voltage protection.

>>2853405
Famous last words

>>2853421
lolno
routing is the part that takes time. routing stripboard in a way that doesn't make horrible ground loops is a nightmare.
with perfboard, you can just lay a ground track or power track wherever it's needed and be done.

>>2853392
https://www.ifixit.com/Guide/How+to+Make+an+Affordable+Soldering+Fume+Extractor/170836

Just build something like this, does't really filter air but sucks it away from your bench and may even work with extra ducting to guide fumes further away.

Mine runs with 12V USB-PD decoy board, battery power ain't mandatory.

>>2853460
So, flux is made of rosin which comes from trees.
You’re getting far more shit in your lungs standing around a campfire for a few minutes.
And far more weird shit from you wife’s candles, incense, and those stupid, overpriced ultrasonic vaporizers from saje with oils containing god-knows-what.

>>2853353
I open a window, hold my breath when soldering, and take a breath to the side when I need to. That’s with rosin at least. For some of the nastier synthetic fluxes (not yellow), the fumes can be real nasty.

>>2853376
If it’s unbranded and from china, it will be leaded. The flux will be trash though. Not only is it cheaper, but the Chinese just can’t help putting lead in products sold to gweilos.

>>2853399
It’s the quickest form of prototyping for more than 10 parts, though that doesn’t include the time spent laying it out. I use KiCAD to lay them out ahead of time with a 0.1” grid. It allows for much higher component densities (especially passives for analog circuits) compared to dot-board, and requires zero shitty solder blobs or funky wires going through one hole but being soldered to another. If you’ve done things properly, you’ll have your ground and power rails running the length of the board, same for any oft-used voltage references. Which is convenient. Some circuits you’ll find are already laid out for strip-board adoption, but those are usually shitty single-rail audio circuits with electrolytic caps in the signal path directly.
If you want to just start soldering components to your board without taking the time to lay them out with a netlist or ratsnest, strip-board isn’t for you.

>>2853402
Series resistor, followed by a pair of fast diodes going to the rails. If you care about input offset voltage, you’ll make that series resistor 16.5k and then add a 100k resistor to ground.

>>2853427
>no pads
Are you using unclad punched laminate? With that stuff your parts aren’t held very solidly.

Also there’s a little twist tool for easily removing the copper around a strip-board hole. You’d still use a knife if you wanted to remove the copper between holes, but that’s not usually required.

>>2853579
>shitty single-rail audio circuits with electrolytic caps in the signal path directly
Why is that a bad thing? Obviously I know because I'm very clever, but I'm asking on behalf of other anons who might not know

>>2853604
Well single-rail designs are generally simpler for small circuits, and don’t really make as much of a difference for BJT circuits where you need to add DC biases anyhow, but they always require at least two DC-blocking caps on the in/out jacks. For multi-stage op-amp circuits especially, split-rail will be superior for noise and for simplicity reasons, even if that means you’re just buffering a voltage divider.

The problem with these DC-blocking caps is they result in a turn-on pop sound, and will add some low frequency attenuation if the capacitor isn’t large enough. The cheapest large caps are electrolytic capacitors, but they don’t sound too good. Even using electrolytic capacitors as part of an audio low-pass filter won’t sound brilliant. You only ever use these when you can’t avoid it, like on the output of single-ended single-rail amplifiers (which are a bad idea in the first place) or in passive crossover circuits. Large ceramics with their voltage dependence are probably even crustier sounding though. You want to use metallised plastic film capacitors.

For bypass caps on your op-amps or other power rails, there’s nothing wrong with electrolytic or ceramic or tantalum caps.

t. op-amp supremacist

>>2853604
>>2853633
Addendum:
If you’re making a guitar effects circuit, you probably don’t care about it being high-fidelity. Do whatever you want.

Now I’m wondering how crusty you can get modern ceramics to act, would they sound nice and gritty? Or would they just add some soft clipping? Flesh Simulator’s neon bulb distortion circuit was pretty cool looking.

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

M
U
R
ACID
T
I
C

>>2848235
Just about to wind some custom gate driver transformers to drive some 12V FETs which are switching a 3.7V circuit
Why are magnetics so cool bros? Do you do anything with magnetic components?

am i too paranoid about this?
plan is to put a 24v dc (not ground referenced) sensor inside a grain bin which is the tallest structure around. The grain bin is obviously grounded.
The sensor feeds inside a smol industrial controller.
Is there a risk for lightning to hop onto the wire and wreak the cabinet? the cabinet is usually not connected to the grit or ground at all since its plugged.
is a contactor in line with the sensor placebo? contactor has 6kv spark gap rating

>>2853736
There are also dedicated surge arresters

>>2853742
surge arresters are literal placebo for lightning strikes, they break through without upstream GDT, on the grid they slowly wear out without indication that they are used up and lightning protection (for mains appliances) is a 3 stage protection. I and doubt they would protect a low voltage circuit

>>2853636
this post reeks of audiophool BS

how the hell do you solder those pesky fake SMD packages? These are crystals but also some electrolytic capacitors have this fake SMD design. What's the point of this plastic bottom pad?

>>2853442
Oh. I do see it generally recommended >>2853392

>>2853460
I know most people have parts like this lying around to make it themselves but I honestly don't. It's why I'm just considering an extractor fan but in need of a non dogshit one. That's if I need one at all. I don't solder daily, or anything, infrequently if anything but the strong fumes are a bother since they linger for so long.
I am close to a window, a big one but it's not exactly airy. I would hope a fan would take the fumes and throw them out the window but the reviews I see for the fans typically complain about fundamental things like the fan extracting the fumes to the sides instead of behind which sounds daft

>>2853579
>hold my breath when soldering, and take a breath to the side when I need to
This kind of sounds uncomfortable, I didn't really know you had to do this either.

>>2853766
You could just use a desk fan and zip tie an activated carbon foam sheet to it for odor control.

>>2853763
>fake smd
I think you might be retarded that's just an smd package

>>2853770
>just an smd package
there are all kinds of different SMD packages but this one is "fake" SMD because it has leads
that are tucked under the bottom but you can straighten them and make it a TH. and it doesn't even matter since my question was about that plastic pad which makes it awkward to solder but you have no idea so you instead decided to add your shitty take about what SMD is even though you have no fucking idea what you are talking about.

>>2853793
you're not supposed to solder them, you're supposed to use solder paste and stick the whole board in an oven

>>2853793
Do those square leads look like they fit in a round hole? I can't tell because I'm retarded.

>>2853827
Yeah they are bent, see pic

It's another 3dpg moron trying to understand what's going on in his printer.
Wouldn't it have made more sense to swap both bed inputs, so it would be just one fat trace between them? And if not why not turn the MOSFET 180degree?

>>2853736
I'd want to use optical and/or galvanic isolation. Toslink cables are a cheap way to make very large electrical isolation gaps.

>>2853763
KiCAD has footprints with extra wide pads for hand-soldering those.

>>2853793
>leads that are tucked under the bottom but you can straighten them and make it a TH
Wait till you find out about PLCCs.

>>2853854
What's that bridge rectifier doing? The screw terminal pinouts are like that because they want to keep positives on the left and negatives on the right, though I've seen positives in the middle and negatives on the outside on occasion. In this case it really doesn't matter. MOSFET placement is probably just convention, but keeping the heat-sink on the far side might be beneficial for one reason or another.

>>2853893
>What's that bridge rectifier doing?
Idiot proofing, so you don't mix up input polarity.
>The screw terminal pinouts are like that because they want to keep positives on the left and negatives on the right
Seems fairly obvious, but i thought short power lanes are better for interferences and what not.
>keeping the heat-sink on the far side might be beneficial for one reason or another.
Damn, haven't thought about it like that all.

>>2853917
>short power lanes
You can see the enormous copper pour they're using between those pins, resistance won't be a problem.

>>2853925
>resistance won't be a problem

in other words, it's futile

>>2853925
> copper pour
It’s really the opposite.
99% of PCBs are designed completely wrong—they’re still using a “point-to-point wiring” philosophy and understanding of the problem.
What you *should* be doing is designing and optimally routing/minimizing the spaces between the copper. Not the copper itself.
Here’s a picture to help right-think people into using modern 1960’s PCB technology effectively instead of in the primal (ape-like) and troglodytic way that I see most PCB designers cobble together their “design” they send to production.

>>2853956
A: you almost never need that much copper
B: traces being too close for too long will capacitively couple
C: you're best off with a ground pour that goes between signal traces to minimise cross-talk
D: high thermal mass where you don't need it makes soldering more difficult, at least without thermal reliefs like in your pic

>>2853965
> need that much copper
Need?…. Want! I want *all* my free copper from jlc and pcbw.

Got this AO3400/3401 H-bridge fabbed and soldered up by JLC. You see those caps, C3-6? Yeah those should be from gate to source, not from gate to drain. Now I have to desolder them, otherwise the circuit will kill itself from shoot-through.

Works great otherwise, it can theoretically handle like 12A of load, and it's way smaller than a relay circuit.

>>2853854
Function follows form, anon. I'd even go a step further and put both + terminals top and both - bottom, like on your PSU.

>>2853769
that simple, huh? I have a desk fan.
I don't have large enough zip ties I believe, or activated carbon foam sheets, but I image they can be picked up pretty cheap. thanks.
i thought i was inhaling all that toxic shit but seemingly it's not so bad. only thing with a fan is wouldn't it make it less stable working on small parts and having things wiggled around with the air?

>>2853956
>Let's do it like the guys who barely knew EMI and had yet to discover EMC

>>2854010
Oh, and it's still getting shoot-through when I mess about with the button pins, I'm guessing the voltage at nodes "inve" and "invr" is getting floated kinda funny and is turning both FETs on at once. Time to add a dual comparator IC.

What's more important for soldering, tip quality or station quality? I've got a cheap station I bought a while ago that works poorly (tip doesn't get heated evenly, poor heat retention, oxidizes quickly etc) and I'm wondering if it's enough to replace the tip or if I need to get a whole new soldering station

>>2854083
this is just gearfagging, buy whatever and make stuff
and tin your tips when you're done with them ffs

>>2854084
I don't need anything extreme but what I've got right now is preventing me from working properly, I'm just asking where the bottleneck is. My tip oxidizes within 10 seconds of tinning it and doesn't melt solder on one side so big blobs get stuck to it.

>>2854087
>My tip oxidizes within 10 seconds

- you have shitty tips
- your temps are way way too high
- or both

>>2853766
> know most people have parts like this lying around to make it themselves but I honestly don't.

I had a old 12V 120mm PC fan, had to get that USB-C PD decoy board (3 bucks), a pack of filters (25 bucks, probably can be bought cheaper) and cardboard box from trash. If I wanted more extraction (tube out of window) I'd get some plastic box and use it.

Are you really sure this hobby is for you?

>>2848235
i'm trying to understand FBSOA for MOSFETs used in linear applications.

RDS(on) is 0.085 ohms.
ISD (continuous drain current) is 33 amps max.
PD (maximum power dissipation) is 180 watts.

at 1 volt i would expect:
V/R = I ==> 1 volt / 0.085 ohm RDs(on) = ~12 amps
(12 amps^2)*(.085 ohms RDs(on)) = ~12 watts

this falls well within the 180 watts maximum power dissipation and 33 amps continuous drain current rating. so then why does the FBSOA chart show a maximum of 3 amps?

https://pdf.datasheetcatalog.com/datasheet/fairchild/IRFP250.pdf

>>2854170
or put another way, why are they claiming that 1V is limited to 3A by RDS(on). RDS(on) would have to be 1V / 3A = 0.33 ohms for that to be the case?

>>2854164
everyone starts somewhere anon
i don't have an old pc fan lying around.

>>2854181
>i don't have an old pc fan lying around.
..and you also missed my point

>>2854184
no i get it
it was a diy solution i'm not keen on diy-ing
but it's more that if I have to buy all the parts anyway, then it works out about the same as just buying an extraction fan

i need to drop a 5v rail down to around 3v to keep a BJT's Vce in its rated spec. its an emitter follower. i could use a resistor to drop the voltage (with a bypass capacitor since its a follower), but i thought dropping it using three diodes might be safer, since then there would be no possible way for the collector to hit 5v no matter what the base is doing.
1) is this logic sound?
2) whats the best diode flavor to use for this purpose?

>>2854196
>3v to keep a BJT's Vce in its rated spec

small BJTs typ have a Vce of 25-30V
maybe you missed a zero somewhere
in the very unlikely event that 3V is real, diodes will kinda work but will add a bit of non-linearity

>>2854210
behold, the most delicate BJT known to man.
>diodes will kinda work but will add a bit of non-linearity
certainly less non-linearity than a resistor, no? at any rate, if the collector voltage is actually dropping an appreciable amount then the input voltage is definitely "out-of-spec" for my particular circuit (or, now that i think about it, the collector current is likely over-maximum), so i dont really care about performance in this scenario, i just dont want it breaking. it might not break at 4v but it doesnt perform correctly above it so i dont see any reason not to use a diode drop for safety.

>>2854216

4V is with base open, but your base isnt open
so 13V is the max it can handle under normal operation

>>2854220
>so 13V is the max

just noticed that's with base short-circuited
so still dunno what the max is under normal operation, prob 30V like a regular transistor

>>2854220
>>2854221
my understanding of open-base vs shorted-base ratings is that the differentiating factor is whether or not the collector-base leakage current is following back into the emitter (which gets multiplied and thus causing runaway): if the base-emitter is shorted then a very small portion of this current flows back into the emitter, but if the base is open then almost all of the current flows back into the emitter.
in my circuit pic, the impedance looking out of the base is 51 (or more), and looking into the base is 180 times hfe (or less), so the Vce breakdown should be slightly less than 13v, but not over it. in the event the input impedance of the next stage drops dramatically (such that the impedance looking into the base << 51), OR if the output impedance of the preceding stage rises dramatically (albeit easily avoidable, pic rel), the Vce breakdown would also drop. so yes, Vce would normally be around 13v, but when putting in safety features you have to think about worst-case scenario fuck-ups.

>>2854087
Eating tips is a common problem with lead-free solder, from what I've heard. Especially the off label chink alloys.
If youre gonna use lead free, you might have to shop around, maybe check eevblog, for an alloy that isn't caustic to tips and buy legit

>>2854083
Tip quality. If your station is just a power-controlled unit instead of being temperature-controlled there's a bit of room for improvement on that front, same for if there's really poor tip selection, but generally tips make all the difference. There's good variety of 936/900 series tips, in T12 tips, and in JBC tips like the T245 and T215.

As the other anons say, lead-free solder can be really bad for tips that aren't designed for it, in that case I'd get different tips if you can't avoid using lead-free alloys.

>>2854170
I think it's just a pessimistic worst-case Rds ON. Though maybe they drew the graph wrong, because it's 0.33Ω at the far left, but 1.67Ω at the DC inflection point. All other graphs I've seen have been a consistent resistance.
Also for linear applications, see these documents:
https://ntrs.nasa.gov/api/citations/20100014777/downloads/20100014777.pdf
https://www.nexperia.com/applications/interactive-app-notes/IAN50006_Power_MOSFETs_in_linear_mode

i want a split my board into two separate boards: a power board and an everything-else board. JST connectors appear to be popular in hobbyist circles. should i be worried about the hot voltages momentarily connecting before the ground line does?

>>2854379
>hot voltages momentarily connecting before the ground line

if the positive connects first, the entire circuit attains the same charge as the positive rail
but since there's no ground, there's no loop, no current, no potential difference
so nothing can happen until the ground connects as well
meaning, there's nothing to worry about
except zombies, spiders, bathing, dentists, the color yellow, and belly buttons

>>2854379
Arguably, but you shouldn't be plugging a connector that's already live into a board anyhow, aside from a proper external connector like an IEC.

>>2854443
oh yea, forgot about that. ill add a push-button to the power board.

How come I2C is being used for external modules like these? I always thought it was intended solely for communication between ICs on the same board and would be adversely affected by transmission through wires, connectors etc

>>2854464
I2C runs into problems in the ~1m region, not in the ~5cm region.

I understand I can't connect a US 120v GFCI or surge protector directly into French 240v, but would these work dowstream after a voltage converter?

>>2854611
>dowstream after a voltage converter?

if that's a transformer, then definitely no
if it's electronic, probably not
you'd need a second GFCI after converter

>>2848235
Hello everyone! I am making a chlorate cell with a temperature control. My original idea was to use DS18B20 for that and pair it with arduino nano. Do you think this is a good idea? Is this thermocouple enough to sustain conditions in the cell? If not, what can I use instead? Thanks in advance.

>>2854663
I would put the temperature probe in a glass tube rather than expose it directly to the electrolysis solution

>>2853376
update: they said its unleaded
good enough

>>2854676
Just treat it like it's leaded regarding safety and you should be fine

>>2854684
>should be

Unusual combination. A PNP transistor and an N-channel FET in a 6-pin package.

>>2854663
Pretty sure stainless won't resist oxidising chlorine species. Glass-encapsulated like the other anon suggested is likely your best option, though there may be metals or passivated metal coatings that you could apply to a conventional stainless probe. You may also want to consider a way of measuring pH or conductivity independently from the power source.

What do people actually do with chlorate cells anyhow? Also what anode are you using?

>>2854699
Isn't that what an IGBT is made from?

>get an idea of something I want to code
>open IDE and start coding immediately
>get idea for electronic circuit
>have to place an order at Digikey, pay 8 USD shipping, and wait a week for it to arrive
How do you anons deal with these time delays, have several projects in progress at once? Keep a huge amount of components in stock on the off chance you might need them?

>>2854752
You have to already own all the parts. I also order PCBs for general purpose use, like a TL494 breakout board, or an IR2184 board with slots for MOSFETs of my choice. Picking an array of more general-purpose parts means I can prototype something thoroughly, maybe even deploy a one-off project with these PCBs just soldered together. I 3D print my enclosures, so it's no trouble fitting 5 PCBs into a single box with perfectly matching mounting holes.

>>2854752
Order multiple MCUs and respective breakout boards, regulators, passives, connectors/headers, and programmer(s).

>>2854690
Idk man I never soldered. I wanna get into it but I got a few other projects I gotta tackle first.

>>2854752
Proper planning. By the time it arrives I'm already deep in some kind of diagram or went through everything in my mind 20 times.

>>2854010
>>2854071
Ok so the client needs it working 100% reliable, and there's no time to order new boards, so I designed a daughter-board to sit atop it with an LM393 acting as a schmitt trigger circuit to ensure the gate voltage is never at some funny intermediate point for longer than it takes for the FET gates to charge. With debouncing too.

To that end, I hand-painted a PCB using nail polish as my etch resist. Just tested it on its own, and it works well. It's going to stack atop the existing board with 4 wires between them, and a bit of liquid electrical tape to ease the strain on the thin wires soldered to SMD pads. I'll stress-test it soon by connecting it to the driver board and messing with it to ensure there's no shoot-through error.

Can anyone figure out why this circuit would kill the comparator? It's just meant to be a normal schmitt trigger. The other comparator in the package is configured the same way. This is the 2nd time I've had one die. The caps are 22nF not 10nF, and there's also a 100nF cap across its rails. The supply voltage is 14V.

>>2854743
Not quite, the PNP and FET are independent devices in one package. What would be the use case for that.

>>2854699
>>2854743
>>2854849
I'm confused. Who would need this?

>>2854835
Well, whatever the problem was, I swapped the LM393 (that came from aliexpress so you never know) for a TS3702 push-pull comparator from LCSC. Pullup resistor R10 be damned. Max voltage is 16V so it's kinda sketchy, but eh it should be fine for a test at least. 555 timers would probably be more robust ways of getting hysteresis, though for my next circuit I'll be using actual half-bridge driver ICs with shoot-through prevention.

>>2854850
Well I was thinking they'd just get the user to configure it as an IGBT if they wanted, but there's also the possibility the transistor is being used as some sort of tempco matching thing. Maybe the temperature coefficient of the BJT is equal and opposite to that of the MOSFET, and so is used as part of a feedback loop with an op-amp for linear MOSFET use. Or linear PNP use I guess. It could also be used for measuring the temperature of the MOSFET, though a normal diode would typically suffice for that. BJTs are better than diodes for lin/log conversion, but I doubt this is the case here.

Maybe there's some obscure usage that these two components are often found next to each other in (besides an IGBT), like how the FS8205A dual MOSFET package is used for BMS circuits.

>>2854849
>>2854850
Yeah the datasheet says the PNP has a higher current rating than the MOSFET, and it's designed to work as a "load switch" or "power switch", I'm guessing it's meant to be used as an IGBT. Can't imagine why anyone would use one instead of a MOSFET, but maybe it's more reliable when considering static or something.

>>2854321
very cool links, anon.

>>2854752
> Keep a huge amount of components
Yeah. You see that interview with marie condo after she had kids? She said minimalism is bullshit.
A huge amount of components doesn’t take up a lot of space though.
Always order extra for the next project.
A microcontroller is something that’s in the middle ground.

>>2848235
Real novice here.
I have ancient 6.3V 1500uf electrolytic cap.
what happens if i parallel 400V 2.2uf capacitor from a led bulb on it?
its on a single phase buck converter output.