/diy/ - Do It Yourself

>0 Ohm resistor

cursed thread
donut post

I am wanting to repurpose a older ATX power supply to power these LED

>https://www.aliexpress.us/item/2251832761346644.html
>https://www.aliexpress.us/item/3256803142661338.html

and I was wondering if a simply SPST switch be okay to put on the green and black wire to turn the power supply on or what would I need?

>>2604940
that would be sufficient

>>2604696
A proper Field Oriented Control board will not vary voltage to change speed, it will vary the speed of its FETs switching. It shouldn't have poor speed resolution at all, I'd definitely try it as-is. The only problem I can see with it would be the back-emf being insufficient for the sensorless operation, but you may be able to amplify that or change a PGA somewhere on the ESC. A motor with a really high Kv like that is going to have a really low number of turns and hence a somewhat low series resistance, so by reducing the power output of your ESC you should end up with a lower speed at a somewhat lower efficiency.

Rewinding motors is not something I'd ever want to attempt.

>>2604700
>RPM and torque are not connected
Current is directly proportional to torque, and ultimately the amount of current going through your motor is the non-emf voltage divided by the motor's ESR. And that emf voltage will increase linearly with speed, while the non-emf voltage is just your battery voltage with the emf voltage subtracted from it. A hugely different mechanical load will change the lag angle of the rotor and cause that emf-to-rpm ratio to change I think, but for a constant mechanical load the relationship stands.

>>2604751
I'd wonder if those are meant to operate when tipping from side to side, they're also probably likely to get damaged if it tips over, and are heavier than a dedicated solution.

>>2604802
RF class D is the way to go. Also op-amps at 13MHz kinda exist, even if you can't find one within your price range you can make yourself a discrete long-tailed pair amplifier.

How should I be grounding myself (assuming I'm nowhere near a computer). What I've done is cut off a spare jug cord, heatshrinked up the live and neutral wires and used the ground wire, put a .75MR resistor on it (didn't have any 1MR) and bolted it onto a steel shelf I use to keep electronics. I'm sure it works, but is that the normal way of doing it? I have a spare plug, I'm gonna wire that up with just the ground pin instead, probably a lot safer.

Also, I read somewhere that I should be using a resistor that's rated for 400V (I live in a 240V country), is that real? Sure would be shitty if I touched a 240V wire and the shelf, the .75MR resistor immediately burnt out and I got myself a nice 240V slap

>>2605040
bro dont ground yourself. by leaving yourself floating (using good insulated footwear) you can freely grab active cables and wire without affecting shit. if for whatever reason you do need to touch something, like a pc, then just get like 10 of those resistors by one end and touch the other to earth. you should only do this when the thing youre touching is earthed. if your pc is unplugged you should just touch the gnd/earth pins of the pc power supply itself

>>2605069
>bro dont ground yourself. by leaving yourself floating (using good insulated footwear) you can freely grab active cables and wire without affecting shit.
if you're earthed through a 1 megaohm resistor (which is standard) you also won't affect anything you touch. And I do a lot more electronics than just assembling computers

>>2605040
I use the household outlet ground pins like this too, and I've seen people use them in EE labs. Technically they make special ground-only plugs for this (picrel) and those are probably what you want if you're doing this much effort

>>2605070
>earthed through a 1 megaohm resistor (which is standard) you also won't affect anything you touch
there are some tradeoffs to using a huge resistor e.g. if you rub your socks on the carpet/etc too much, it won't dissipate your body's charge fast enough and some might discharge through your project/instrument instead.

>>2605070
if you ground yourself through a 1 megaohm resistor I struggle to believe you're actually at ground potential

>>2605077
its not about being at ground potential its about dissipating energy to earth such that you dont suffer electric shock. once the energy is dissipated you will be at ground potential anyway

>>2605072
>e.g. if you rub your socks on the carpet/etc too much, it won't dissipate your body's charge fast enough and some might discharge through your project/instrument instead.
bullshit, an electrostatic charge has a tiny fraction of a coulomb of charge in it, it'll dissipate in a thousandth of a second through any resistor. Ever tried to measure the resistance of an anti-static bag? You can't their resistance is too high, but they work fine
>>2605077
as above

>>2605081
I bet you pour your used oil into a gravel pit in your front yard too

>>2605083
I pour mine in the ocean when I dispose of old batteries.

>>2605084
ok that's based actually

>>2605085
That's how you fight Greta Toonberg and George Soreass. Just dump your used oil and old batteries in the ocean. Throw bromine tablets into zoo aquariums. Grind PVC pipes into powder and dump it into municipal water supplies. Kill the whales.

>>2605091
I use latex gloves because they're biodegradable, but after reading your post....

>>2605097
I'm just doing my part to change the climate so the lizard people overheat. You don't have to thank me, Anon.

>>2605101
Lizard people? I thought it was bug people. But you're right, just a few degrees should be enough to overheat them given how hot their climate is already, plus possibly sink taiwan and japan and much of the philipines

>>2605107
If we kill the lizards then the bugs will overpopulate and eat their own babbys.

>>2605110
Ok go ahead, cook them

>>2605110
ok go ahead, cook them, this is what I used to do to bugs

>>2605124
That bug has a decent ass/tergites. Shame about her face. I will make her raise chickens and pigs.

So am I getting this right that a bunch of CMOS shift registers filled with all the same value (either all zeros or all ones, doesn't matter) will use significantly less power than a bunch of shift registers with alternating zeros and ones? Power consumption when shifting, that is.

>>2605081
Yes I think this is correct. 1MΩ is enough that static will dissipate rather easily. I think it will also prevent you feeling any shock. You'd get currents in the orders of 100s of µA, which are apparently imperceptible.

>>2605174
I suspect you're right about there being a difference, but I'd wonder if the magnitude of that difference would be significant. Is the majority of the power consumption of a CMOS IC due to charge being pumped in and out of gates? There's still the clock going back and forth (twice as often as anything else and into a fair few gates). You could maybe find an equivalent transistor diagram and see how many gate transitions happen per clock transition in both situations, I'd guess the difference would be like 100µA vs 70µA.

>>2604810
How the fuck do I count the turns inside a coil? Some Anon told me to get the wire gauge and then count the turns (>>2595916). I got the wire gauge. How tf do I count the turns inside the coil?

I have a couple of high voltage transistors that I salvaged from a CRT television and fluorescent bulbs. I’m wondering if I can use them in a situation like picrel. Will this backfeed 100 volts into my ttl buffer, or is this a standard thing transistors can do?

I want to drive neon indicators (yeah, i know it’s not resistive and I know it’s only going to light one rod)

>>2605252
picrel

>>2605257
It won't backfeed if you remain within the ratings of the BJT. Though note your circuit is in an "emitter-follower" or "common-collector" configuration, so the maximum load voltage is equal to the maximum base voltage, subtract 0.7V. You want to use the transistor in "common-emitter" configuration instead, such that the load is above instead of below the transistor. For single neons and stuff this should be fine, but if you have display elements that need to be switched at the top instead of the bottom, you'll need to figure out another solution. The easiest would be to just have a 0V rail, a negative 5V rail, and a negative 100V rail.

Don't forget the current limiting resistor bro. If the voltage isn't high enough you may need an additional inductor to give it a kick.

i need a mains frequency toroidal transformer for a linear power supply project. 120V AC input, four +30V outputs 150VA each, and two outputs +/- 15V for op amp rails. any recommendations on where to buy this, or should i just try to wind it myself... and if the latter, where can i buy mains transformer cores?

>>2605259
> emitter follower is totally wrong
Thanks, yeah, I wasn’t thinking there…I was too tired. Fixed in picrel. It’s still not obvious (to me) why it wouldn’t see the 100v across Rload… I’ll have to research that some more.

> display elements
No thank god, I think the’re probably actual vintage NE-2s.

> give it a kick
One of the other things i want to drive is an ancient solid state relay that uses a neon: from a reel to reel tape machine the size of a refrigerator.

I looked at the I/V curves for neons, and it’s this insane bullshit… in theory it should start at 90v and then you’re “supposed to” dial it back to around 60 volts once you’ve “detected” that it has “struck” successfully. The good news is the things practically run on leakage current, so I probably won’t need the TO-3 flyback transistor.

I testing some chinese capacitors bought in batch like 4 years ago on Lamazon. CapXon, Dong, etc..
I measure 556µF on Peak ESR instead of 470µF.
Is it good to solder some "overcapacity" ?

>>2605348
>chink caps
Don't use them in anything you care about.

Very basic query:

My coffee maker has a tiny LCD/Controller board that also serves as the power supply (ac cord connects to it directly). The machine looks normal up to the moment you push the button to start making coffee, at that point the LEDs on it crap out, everything on the LCD goes dark and it restarts itself. The heating element never heats up.

Things only malfunction when the machine needs to start drawing serious current, and this happens regardless of power outlet. I suspect the capacitors on the board are shot.

I can't find the board online anywhere, I'm inclined towards just finding all the same capacitors online and replacing the ones on the board.

Just wanted to hear from more experienced electro tinkerers if this is is misguided.

Pic related is the board (I checked all capacitors for residual voltage, all is safe)

>>2605381
Shit, forgot pic

>>2605383
And to be clear I'm only planning on replacing all the big capacitors, not the smc ones

>>2605325
Neon dynamic characteristics are some fun shit. You can do all sorts of wacky and fun things that may not be intuitive at first glance. Current limiting resistor solves all your problems for basic applications. Choose something that gives you the desired current based on your supply voltage and neon maintaining voltage, just make sure your supply voltage is higher than the strike voltage. The usual go to choice for a neon driving transistor is an MPSA42. Those "solid state relays" you found are probably even more slick than you imagine. The controlled element is probably a photoresistive cell and they were often used for gain control loops and such.

>>2605381
The heating element might be shorted.

Substituting a 2n3904 or a bc337 for a 2SC1740 in a video buffer will work fine, right? Just need to make sure it's biased properly? I don't really know anything about gain bandwidth product but it shouldn't matter too much, right?

>>2605396
yeah I'd suspect that or the relay

>>2605401
https://www.web-bcs.com/transistor/tc/2sc/2SC1740_sim.php

>>2605396
>>2605404
Huh
How would I go about testing those components using a multimeter? The heating element has no markings, I don't have a reference resistivity to test. For the relay I believe my countertop power supply will help.

I want to make a machine that sorts candies based on their color. I see a lot of guides already online for one but they all use arduinos. I have a raspberry pi, would I be able to use that instead?

>>2605413
>testing those components
heating element: try giving it (appropriate voltage and adequate current) power, and see if it heats up

relay: put control voltage across the gate circuit (I know it's not actually called that for relays) and listen for a click

>>2605424
>>2605404
>>2605396
thanks for the feedback, will give it a go

>>2605396
the heating element seems to heat up normally
>>2605404
the relay clicks nicely when I apply voltage

could my original intuition be correct, could it be the big capacitors?

Please, for the love of god, if this isn't a 1600 MEGA farad capacitor, Im begging you, please tell me now so that I can stop living with this kind of unlikely hope...


Seriously, what the hell IS this thing? and what is it actually for?

>>2604940
Hell, if you want something quick and dirty, just stick a paper clip into the appropriate wires.

>>2605513
they didn't have the “µ” symbol those days and used "M" to represent micro.

>>2605516
damn. So why is it so big? What was this thing for? and what can I do with it now?

>>2605383
That's a zener-based power supply. They get hot, which could be a source of failure. R3 and ZD1 are components I'd test, same for that cap right next to ZD1. ZD2 and Q1 are also doing something, maybe a smoother rail for the MCU, but they probably aren't making oodles of heat.

Anything much on the back side of the board?

>>2605517
Probably a mains voltage capacitor, like for starting or running a big induction motor. Or maybe for power factor correction. At least that's my guess.

>>2605517
>So why is it so big
Capacitors weren't as energy-dense at the time, they had to be much bigger for any given capacity.
>What was this thing for
Probably a radio or something. Power supply smoothing.
>and what can I do with it now
Shitcan it. Vintage caps like that have almost always perished and won't do their job anymore.

>>2605454
> is it the cheap chinese caps in a hot environment for years?
Always a high probability.
That and the mcu browning out as soon as the relay kicks in…
It’s probably the drain from driving the inductive load of the relay coil initially. Capacitor should be able to smooth out the mcu supply. Go a bit bigger if need be, and over-rate temp and voltage.

>>2605540
> shitcan it
Nooooooooooo…
Restorers will hollow it out and put panasonics in there, then replace it so it looks the same as it used to. Modern caps are 10x smaller.
Don’t do this yourself, it might have PCBs in it. (In which case, it might actually work—pcbs worked well, perhaps too well). This cap is from 1957 which was prime PCB time.

>>2605555
well, I found a whole load of them, along with vacuum tubes, old resistors, the works, in this big bin at a science building they are gutting on campus. Ive collected some interesting stuff already. Is any of it worth collecting? And will anyone buy it? Or is some of it good to use actually?

>>2605561
My opinion would be to save it all and put it on the market. It’s not like they’re making tubes anymore, so finding them is getting harder and harder since they do burn out eventually.
Put “retro” “vintage” “original” “new old stock” etc. in the title.

>>2605564
Are they really not making them at all? Like, literally nobody is making them? I figured you could still get them made for special projects or something.

>>2605561
The tubes are probably still good, resistors could go either way, and the caps will have gone bad.
Tubes vary in price but the other components aren't worth much. Maybe useful to a collector or restorer, but not a lot.
>>2605570
AFAIK there's still some being made in china (often found in cheap-ass "audiophile" crap), but it's mostly just old stock people care about. There's also the korg nutube which is a VFD hybrid, again for audiophoolery.
Aside from that, maybe special transmitter or radar/military equipment but that's probably all

>>2605454
>could it be the big capacitors?

certainly.
along with 1000 other equally likely possibilities.
the odds of you actually repairing it are smol smol smol.
the odds of you getting frustrated and throwing it away are big big big.
so, save your sanity, just get rid of the whole PCB and replace with a big fat switch.
if you must have a timer, get that thing on bottom pic.

>>2605454
>>2605576

one more thing: dont throw away the PCB; frame it and hang it on your wall.
this memento of inadequacy will forever remind you not to mess with stuff beyond your ken.

>>2605576
Yeah, I have an electromechanical one, it’s going on 23 years now.
I never use the switch or let it auto shut-off; I always plug it in to turn it on and to turn it off I unplug it. No switch arcing. New outlet is $1.50… it’s going on 23 years too, and will probably last another 50 years. Used every single day of those 23 years.

>>2605257
This shouldn't be an issue assuming your BJT is an NPN and you stay within the maximum parameters of the BJT.

>>2605570
Compared to the tube heyday, no, nobody is making them. It’s not even round-off error. Go into your local independent drugstore, and ask them where their tube tester is, because you need a new RCA 845 and pick one up if the one you removed to bring in tests bad.

Like, technically, the magnetron in your microwave is a metal-can tube, and so is the neon indicator in your extension cord, but these are very very different.

>>2605257
This shouldn't be an issue assuming your BJT is an NPN (as a PNP will have the emitter tied to +100V and base to 5V, drawing very large current, not good) and you stay within the maximum parameters of the BJT. One thing to note though is the Icbo current, which is the current from collector to base when base is open. It might be a good idea to set a resistor between your base and ground if the output impedance of your buffer is too high.

>>2605570
glasstranny has a video on how to make them, it doesn't seem too hard if you're handy with a blowtorch

>>2604810
I need a 1K OHM resistor at 5watts yet I've only got 1 watt resistors could I put five together and have a 5 W resistor if they're all thousand ohm's. I want to wire them parallel.

thank you

>>2605420
>use an entire computer to use a single optical sensor and control a few servos
yes you can. but why not just grab some chink-duinos off alibay?

>>2605711
5x 1K in parallel is 200 ohm 5W resistor. so if you need 1k value, then you need to have 5x 5k resistors in parallel

>>2605717
>5x 1K in parallel is 200 ohm 5W resistor. so if you need 1k value, then you need to have 5x 5k resistors in parallel

thank you

>>2605721
If you want the bad idea version, just use the 1 watt resistor but put a big fan blowing lots of air over it, maybe even get some other metal bits in contact with it to increase its cooling area. Or submerge it in a decent sized vat of cooling oil

what's the purpose of the black wire that comes with rcbos?

>>2605521
You're looking at the back side. The other side is just LEDs, a small seven segment display and a couple buttons. The user interface.

>>2605576
>>2605591
>>2605614
Looking at the price of equivalent coffee makers these days, it makes perfect sense to try. I could replace all the through-hole components on this thing easily and for a single digit percentage of the cost of buying a new one.

I do suspect it to be a capacitor, the way the thing craps out is kinda like a brown out into a black out, so to speak. If it is the mcu then I'll have to accept it and move on, but not before i try replacing the capacitors.

The heating element is not shorted and the relay is working fine (and it's only 10% through its rated life, based on the age of the coffee maker and the datasheet).

If I need to replace it I will, but before I spend that money I'll take a few hours and spend a handful of bucks attempting simple fixes.

>>2605744
> I could replace all the through-hole components on this thing easily and for a single digit percentage of the cost of buying a new one
The hard part is the microcontroller / display driver, chances are there's proprietary code on there that you can't just download off the old one (if the old one is even functional enough to rip it from).

But the controller would only die if it somehow got killed by overvoltage spikes, but what I'm hearing sounds more like undervoltage. I'd guess the relay drains too much power to turn on, so the voltage drops and it browns out enough to stop powering the relay. Having good electrolytic caps will definitely definitely help, and may be the absolute solution if the voltage dips enough during the dips in the rectified mains. I'd try reverse engineering the power input section to see what the zener power supply section(s) require. With any luck you could just bypass one and feed the whole thing power from a 12V brick and test it that way.

That or just measure voltages while it's turned on. One hand behind the back.

>>2605740
fully insulated, switched neutral flying lead

I have an active 2.0 audio system and main speaker is noticeably quieter than passive one.
I can borrow every tool and supply item for soldering and testing, but have almost no experience nor knowledge. What should I start with and how to proceed?
What are the chances that it's some old capacitor that needs replacing and everything will be good again?

>>2605855
Swap the speaker to test the amp output. If it's still quiet, you need to pull the PCB and start troubleshooting components.

>>2605856
It could also be a shitty speaker wire, connector or cold solder joint. Volume pot is worth checking and cleaning too.

>>2605856
>>2605857
It was perfect like for like 15 years, and now for like a year main high speaker's quality and volume has 100% degraded, so I'm blaming electric components on circuit board.
Just don't know what to test and how to do it properly.

>>2605859
There's probably a small electrolytic cap in series from the mid to the tweeter which acts as a high-pass filter. Start there.

>>2605863
Any way to test it without desoldering?
Seems like there won't be esr tester either.

>>2605865
You could jumper it, just make sure the volume pot is all the way down before you power on, and don't turn it up very high. If the speaker works then it's the cap.

y did dey do it?
should i undo it?
i do has hv ! :3

>>2605872
Degauss?

>>2605872
Resistance in length?
Length for ease of repair but taped for ease of other repairs?

>>2605876
Yeah probably, it likely came out of a CRT now when I think about it.
Then I'll just undo it if I need the length, thanks.

>>2604810
Anyone know of a decent curve tracer unit or addon for an oscilloscope that's actually affordable? Looking to grade and match Ge BJTs by Iceo and hFE along with matching JFETs by Gm. I've got thousands of both sitting in manufacturers' packs that I'm thinking of piecing out to fx pedal builder types.

>>2605894
Here's one I designed/modified from existing circuits a while back. There's two circuits pictured: A complex curve tracer, and a simple V_BE matcher.

For the curve tracer:
>a bistable oscillator feeds a "triangle" wave into the base of the DUT (Q6)
>the thyristor (Q4 and Q7) and negatively biased Q3 generate a stairstep voltage, which is buffered
>i forget how the rest works lol
>the X and Y outputs are hooked to a 2-channel scope in XY-mode

A ZIF socket with at least 3 slots is used so you don't have to touch the tranny. Then I threw in fritz's simple V_BE tester, just because there was space in the ZIF I was using (you could omit this).

w2aew also has a design here, https://www.youtube.com/watch?v=ZOLLoa2fH24, but it uses a rail-to-rail op-amp

Another note:
When using "thyristor" based circuits, it's good practice to put a diode between the PNP and NPN transistors (Q4 and Q3 as shown in picrel). These circuits can have problems starting up if the PNP collector and NPN base are improperly biased, which you can see by the "V(start)=2" command I forgot to omit.

This is all demonstrated in the Korg MS-20 and Roland TB-303 schematics, which I consider to be required reading for anyone interested in discrete-component analog circuit design.

>>2605933
>>2605936
>thyristor in a linear circuit
I've never seen this before. Is there a blog or YT vid or something of someone describing how this works?
>reverse-biased BJT
Never seen this actually work in a sufficiently reliable method.

>>2606100
negatively biased, not reverse. that node at the emitter gets pulled below ground

dey rly do be lik dis

>>2606178
top two results nonetheless

>>2606179
wtf

>>2606179
> incorrect transistor symbol and random blather I didn’t read
You can ignore everything in this post after seeing how they drew the transistor incorrectly. It’s on the same level as an architect’s depiction of a chimney in picrel.

>>2605936
Yeah, on a metric european map, when you’re driving from, say Norwegia to Italy, do they specify that in millimeters?

My RTC module has a 3700000000000 pF supercap on it now that we’re standardizing in pico farads for everything now.

Now i can ignore all korg and roland schematics as any information in there is likely to be retarded and if they made anything that worked it was by luck/chance.

>>2606268
nigger are you blind?

>>2606346
Anon there's two capacitors in parallel, one is 10µ, the other is 100n = 100000p. As is somewhat common in older schematics, micro and pico farads are all that capacitors are measured in.

>>2606531
yep. I assume because decimals can rub off on schematics

I bought a Nema17 stepper motor and A4988 driver and don’t know how to set it up with my arduino

>>2606590
Sir, please unplug the device for 10 seconds and plug it back in.

>>2606583
That's no reason to use pico over nano for a cap more than 1nF.

>>2606590
Did you read the datasheets of the A4988 and of the ATmega? The A4988 datasheet should tell you what voltages it needs where, and what signals it needs.

You can also probably just download an A4988 library, but I wouldn't trust it to be well written.

>>2605933
>that swastika in the oscillator
/pol/ approved

Herro ANON: I built up this blinkie and want to power it using two AAA alkaline batteries - adding up to 3 volt max. That's the minimum supply voltage unfortunately.

>How can I have this circuit work with under 3 volt? Right now I'm using BC548 trannies.

https://www.elektronik-kompendium.de/sites/praxis/bausatz_led-wechselblinker.htm

What does a switch rating mean when it says (like it does on the two switches in pic):
16(2)A/250VAC
?

>>2606782
NIGGER CHAD

>>2606789
16=NIGGER
(2)=CHAD
OR THE OTHER WAY ROUND?

>>2606779
Maybe add seperate transistors to turn the LEDs on. Also a CMOS 555 will use a lot less current and operate down to 2V.

>>2606782
16A normally open, 2A normally closed. Probably. Look into different NC vs NO current ratings, it's something to do with the speed of the actuator.

>>2606797
These two switches are both 2 lead, NO, push-button momentary. So nothing to do with NO and NC

>>2606801
It may just be writing for when the third (NC) terminal is installed. Maybe it's AC vs DC amps, but I doubt it.

first project and i just realized i am planning to make a breakout board for the breakout board using a perfboard (soldering female headers to perf board to plug in the esp32 devboard, and connecting those to male headers on the perf board to connect to different components). i am doing this to sort of make it easier to shuffle pins around and remove the esp if i need to program it.
is this an acceptable way to do things?
also can i connect things using jumper wires as kind of a semi-permanent solution? are they reliable or will they come loose with time? (this is a mobile project, not stationary, so vibration is a factor). because i cannot figure out how to make solder paths on the perf board so that they don't cross.

>>2606797
doesn't matter which LEDs I use. The problem is with the transistor, I think. But my knowledge is too limited.
Which transistor would be suitable for such low voltages? An NE555 is inpractical because of size restraints. (wanna replace the blinkie in a playmobil police car which was struck by a battery leakage)

>>2606811
jumper wires built civilization

>>2606823
and look where it has gotten us

>>2606824

>>2606829
>COVID pandemic, 2020, colorized

Psssst hey kid, you should know about the Xram, it's the inductive counterpart to the Marx generator. It lets you do pulsed power to do fun stuff like crush coins, make coil guns, and make big ZAPS all without capacitors. Just a bunch of copper in a funny shape and some semiconductor switches. It's more energy dense than capacitors too.
https://www.isl.eu/documents/flyers/EN/isl_XRAM_EN_nm.pdf

>>2606910
I love how the only application listed is "railguns and the like".
>It's more energy dense than capacitors too.
Specific energy too or just energy density?

more of a science question but humor me...

if I place two strong air-gapped magnets that aren't moving there is a magnetic field between them right? So would there be some kind of static electrical field directed in a perpendicular direction to the magnetic field?

>>2606934

where do you kids come from such that you kinda know a little bit about a topic but what you do know is wrong. anyway, the perpendicular field stuff comes into play when you have a time variant field, and the energy is alternating between the electric field and the magnetic field. this is electromagnetic radiation, like radio, xrays, light, radar, etc.

>>2606934
>So would there be some kind of static electrical field directed in a perpendicular direction to the magnetic field?
No. If you observe pic related only temporal changes in B field will create E field.

Professional retard here. I was looking at a tutorial online for making a delay on circuit (this image).

So my understanding is that the voltage builds up against the capacitor until it reaches a critical value that overcomes the Zener diode. That then allows current to flow across the transistor, turning the light on. The delay comes from the time it takes for the capacitor to charge.

My question is: what does the second (the lower one in the image) resistor do? Why is this necessary? When I use online circuit simulators, it works without it and just hits the Zener voltage faster.

>>2606955
Bottom resistor discharges capacitor back to 0V. If you didn't have it capacitor would remain charged to some voltage after you remove the battery and that would change the delay time when you decide to turn on the circuit again. Depending on how leaky diodes and transistors are capacitor would remain charged for hours or even days. With that resistor you ensure that you always start at 0V.

>>2606960
Got it, I see now. Thanks!

>>2606955
Reconstruct this in actual simulator and play with numbers. You will understand it much better.

>>2605383
>>2605744
it's me again, coffee-maker anon

So yeah, I replaced the capacitors. The machine doesn't brown out anymore, things stay perfectly lit up.
The problem is, the relay doesn't trip, because it's not getting proper voltage to trigger (needs around 25 V, which is the voltage at which the region of the board where the mcu is seems to run at, at least part of it.

At this point I'm assuming that the mcu is shot and that there's nothing left to do. Was a cheap fix to attempt, you win some you lose some.

Out of curiosity:
I desoldered the relay and checked the voltage across the holes where the coil pins would go.
When the "make coffee" button is not pressed, the voltage across is 20 V.
When it is pressed, it goes down to 18V.
This seems like really odd behavior. I'd imagine a fried mcu to just die or something, what on earth could lead it to have all other functions running perfectly fine, but the one thing where it should convert a button press into "change the 0 V here to 25 V for 30 mins" became "keep it at a permanent 20 V, when the button is pushed drop it to 18 V lol".
It's just such a silly way for it to fail. What gives?

damn they don't make it easy to buy American
Saw an ad for USA made variacs, go to see how much they are and the lowest end model is $500
Doesn't make sense when chink ones are less than $100 and a used American one is less than $200.

>>2607053
The next president should just start building shit with all the stolen money meant for dem programz n sheeit. Steel mills, chip fabs, pharmaceutical production, space lasers, national internet (segregated & replace TCP/IP with something new), nuclear energy, coal, oil, natural gas, pipelines, subterranean freight shipping, border walls, a new virus that only kills Han chinks, deletion of all social media, and a moon base with robuts to kill anybody else who tries to make a moon base.

>>2606949
>>2606934
>>2606945
>retards lecturing a retard

>>2606779
Use germanium transistors

>>2607053
Should be to make something that’s like a variac, but better, more flexible, and safer (isolated)

Two 60v 500w amplifiers, and you can feed it whatever waveform you want from a siggen.

Amplifiers don’t need to be audio grade, PA grade is more than adequate.

>>2607065
I'd love to hear your explanation if it will help a person out

>>2606818
What’s the forward voltage of the LED? Because you won’t be able to drive it below that, plus the C-E saturation voltage of the BJT.

>>2606910
If I put your arm in there, would you die?

>>2607045
Check the voltages going into the transistor that activated the relay.

>>2607097
How do linear amps deal with phase shift power factor? Do they just burn the reactive power? Not like current will flow back into the rails, even with diodes there.

Class D amps are what I’d use though, and they’re probably fine for reactive loads. There’s little difference between a class-D amp and a synchronous buck converter, so you might as well design your DC variable SMPS to have a waveform input as well.

Finished a design for a big club project, took one last look in 3D view. Realized the layout of the electromechanical components look like very deliberate black cock & balls.

>>2607045
>button is not pressed, the voltage across is 20 V.
>When it is pressed, it goes down to 18V.
>What gives?

those are very likely phantom voltages across an open coil.
put a 1M resistor across probes and you'll prob see 0V.
(BTW, you should be on DC, not AC, in case you dont know)

>>2607188
>look like very deliberate black cock & balls.

gay people see imaginary cocks everywhere.

>>2607239

looks to me like a fairly ordinary iceberg.
if you see something lewd, i got news for you.

Anyone have experience having PCB's made? Pic related requires a ribbon cable hdmi and a left 90 micro usb to connect. Im trying to minimize the gap to the left and the space above it.

I basically want something like a graphics card that just slots in and makes the female ends more accessible for standard cables. The actual electrical diagrams to make the pcb connection probably isn't too bad. The shape of the PCB doesnt seem standard though and might not be possible or cost effective. I pay roughly $5 for the cables. Any ideas or is wanting to pursue this a dead end?

>>2607256
am retard

>>2607258
replying to myself again, but it seems like this pcb shape would probably be the best option. My question now is what would something like this cost and is it something you can have spun up on pcbway?

>>2607265
just draw as many as you can into a standard board size and have them perforate between them

>>2607186
> burn reactive power?
Yes, the output stage has to burn it. Sine waves at mains frequencies would derate the amp power by 50% for an inductive load.

Now think back to all the tests you’ve seen people do on amps with resistive loads. I’ve seen a lot!

From Yamaha:
With an inductive load, the phase between voltage and current changes with frequency. So there are frequencies where the current is out of phase with the voltage, asking the power amplifier to deliver the maximum current to the loudspeaker, while the voltage delivered to the loudspeaker is small. This means that the voltage difference inside the amplifier is much larger, compared to a pure resistive load. A high internal voltage difference multiplied by maximum current equals high power, so the power amplifier has to work much harder. This requires a lot from the output stage (the final power transistors and their cooling) and the power supply, especially with high power loudspeakers with low impedances

>>2607256
>>2607258
I'd desolder the sockets with hot air, then make a ribbon cable PCB that gets soldered directly atop the pads for the missing connectors. That way you can actually make it just sit atop the existing board, without going over the edge at all. Make sure the traces are length-matched. If this is for re-routing the connectors to the front-plate, two suggestions. Firstly, ensure that the connectors are panel-mount, as to not rely on the board for structural support, and secondly use full size USB B as a robust and sensible receptacle in place of the shitty USB micro B socket.

>>2607289
Wonder if any power amps use PFC capacitors? Wouldn't be matched at all frequencies, but they could be sized to reduce the bulk of the power wastage (probably where X_L = R => X_C).

>>2607045

coffee-maker anon again
I fixed it!

>>2607197
good suggestion, I soldered a 1M resistor across the two holes
then I started following traces and probing around with the multimeter with the board powered on to try and figure out what was up

This anon's >>2607186 suggestion to look at the transistor nudged me in the correct direction
The problem was not the voltages going into the transistor, it was what the transistor was doing (or not) with them.

I found the pin on the MCU that changed state based on whether the board was in "coffee-making" mode or just standby.
On standby this pin was 0 V. When I pressed the button to start "making coffee" it would go up to 5 V.
Then the trace coming out of that pin hit a resistor before hitting one of the pins on the transistor. After the resistor, the voltage hitting the transistor gate was 3,5 V.
I figured out that the transistor was basically dead. It was somehow locked onto an "always on" state, and voltage on the gate wouldn't do much at all.
The transistor was this tiny sot23 SMD component on the board. I realized I had some mosfets lying around, looked up their datasheet and realized they would work fine as a substitute for the voltages in question (plus a TO220 MOSFET would be much beefier and handle the warm environment much better anyway).
I soldered the MOSFET onto the board using wires (the exposed pads on the old sot23 transistor were very crowded, I used the gate pad but the drain and source I soldered to different positions on the same trace that were easier to manage). Pic related.
The legs are all separated using thin shrink tubes, and at the end for good measure I applied a wide shrink tube to the whole MOSFET, like a sock.

I re-installed the relay, and everything was working perfectly. Coffee maker has been reassembled, it's back to the kitchen counter and making coffee again as if nothing had happened.

>>2607453
>the madman actually did it
Good job, Anon. You're in the sekret club now.

>>2607453
Nice solution.
Yeah, i dont see a flyback diode near the transistor so maybe that killed it eventually. Your new power mosfet will probably last a lot longer so you can will your coffee maker to your great grand children.

I'll be having kids soon, for that reason I've decided to switch over to using lead-free solder.
(I know it's not as good but I simply do not want a potential source of lead in the house while raising kids in it, lead is terrible for kids)

My trusty old soldering iron really craps the bed when it's time to do lead-free.
It's an old JBC, no adjustable temp, no extra tips, just a super basic hot pointy stick with a stand and a sponge.

I have no idea where to even start looking.
I know I want something with a temperature adjustment dial, and ideally something that can take different tips (maybe even small tips for precision soldering?)

On one hand I don't want to break the bank, on the other I want to buy something that I will never need to replace.
Suggestions?


Bonus: I could also use some feedback regarding lead-free solder. Are there different alloys out there, and if so, which one solders the best?

>>2605717
>>2605711
if all you have is 1K ohm 1W resistors, then:
if you do 2x 1k ohm 1W in series, and 2x 1k ohm 1W in parallel, you would have a 4W 1k ohm setup

Does anyone have a job where they spend a majority of their time doing PCB layout? What were your qualifications to get the job? Does it pay well? Is it enjoyable?

>>2607552
> lead free
Probably not worth worrying about. Antimony is a toxic heavy metal as well. You’ll do everything right, and then you’ll find out your baby bottles were made of BPA after-the-fact.

> soldering iron
Yeah it’s a hot stick. Find an old light dimmer to tone it down if it’s too hot. I still use my father’s old ungar with the cork finger grip.

I’m sure someone will try and sell you a cordless, internet-enabled hot stick with blockchain integration and AI as a subscription service, but it won’t be as good or last nearly as long.

>>2607552
remember that you can always use a cheap transformer from a dimmer with the old unit to get a different temp
There are pretty nice lead free aloys, and some people even swear that they are far beter than the lead ones. But i am unsure if they'll ever be safe

>>2607552
Ive spent 10 years soldering with lead-free solder on a daily basis. I dont have any issue with it.

I'd recommend a temp controlled high-wattage iron. You can get a used Weller for probably under $100, or a new Hakko for maybe $100.

>>2607474
The component marked "HEAT2" Looks like a diode, OP should measure it to check it still works, but I'm betting it does. I think the tiny transistor just died from heat exposure/production.

>>2607552
Bismuth-tin-based solder alloy is great shit. Flows like leaded and melts slightly lower even.

But I would recommend getting a temperature controlled soldering station with at least 50W of power. If just for tough ground planes and soldering heavy gauge wires. Be that a good old Hakko or Weller, or an SMPS-based chinkstation like the T12 or T3A. The TS101 or Pinecil are nice portable temp-controlled irons if you're in the market for that.

>>2607568
not my job now but a previous one i did a lot of pcb work. qualifications was the fact i was studying EE beyond that, nothing.
>pay well
nope
>enjoyable
meh
more profressional job environments might pay better and might be more interesting but in my opinion its pretty boring work. especially if youre just the person doing pcb. if youre doing the whole design from idea to product its more fulfilling

zoomer at the store tried to declare to me that there is no such thing as usb mini
I laughed and asked him if he never had a PS3 controller

>>2607552
I bought a hakko 888d like ten years ago for $70 when it was on sale. With inflation that would $140 today.
I've never changed the tip, never changed the temp from the default 750 degrees, never burnt a component, never had any problems at all. Works great for leaded or unleaded. It's such a nice tool that it played a large part in me buying Japanese stuff whenever possible, even in unrelated hobbies. They make just plain make things right.

>>2607716
what pay were you getting? I had a job once where I was doing a lot of CAD work and getting $14/hr. at $30/hr now (not doing CAD) and consider that to be good pay.

>>2607552
This is wise. My kids can't keep their hands out of their mouths and then they run into the lab amd touch shit and auuugh

>>2607718
>PS3
pff, what year is this, the 80's?

>>2607718
Still have and play my ps3.
All the usb mini connectors on those controllers still work, as well as all the other stuff with that mini connector.
Then i realized that the whole “usb mini was a bad connector” was bullshit promulgated to buy all new cables again.

where can I find a list of common basic digital circuit building blocks like flip-flops, MUXes, shift registers, etc, if I just want to know what's out there?

t. retard

>>2607752
Why is your lab not locked? Kids don't need to be everywhere.

sup /ohm/, not sure if this is you or /mcg/, here goes.

I have a CZ-101 keyboard that has gone full retard since the last time I turned it on (weeks ago). All the front panel LEDs are intermittently flickering and the thing is no longer booting up as it should. It uses an NEC 7811G-120 CPU and an MB83256 mask ROM with a custom LSI toe generator.

Per the service manual all LEDs are driven by four LB1272 LED drivers, data is shifted in from the CPU via four 74HC374 flip flop arrays. While I do not have a logic analyser on hand I can see that the signal to each of the clock pins is "tripping up" in time with the flickering of the LEDs. Clocks and reset lines across the board are as they should be, the keyboard matrix sections are sending their data but nothing is sending to the tone chip. I have an ugly feeling the CPU or ROM has shat the bed.

I'm pretty sure this thing is now a door stop, though I'm considering a new old stock 7811G just for shiggles. $20 on fleabay, a pretty cheap hail mary. I may also be looking at a corrupted ROM, in which case this thing is deffo a boat anchor, I have no idea where I would get a copy of the firmware. I don't know how prone to failure Mask ROM is though, as compared to an EPROM for example.

Does anyone have a broad idea or can point me to resources on how I could further test the ROM or CPU? I have been considering trying to dump the ROM (too see if it actually works for starters and to hopefully get a copy of the firmware) but I'm not even sure where to begin on further testing.

>>2607813
No, i don’t lock the lab, and i also leave my gun cabinet unlocked with lead shot and other ammo lying around. I can’t give up leaded ammo though.
I only hope one day I can get d.u. ammo so I can go totally lead free.

Saved a $10k VSD today.

>>2607854
vsd?

How hard would it be to make a bunch of these? Or maybe there's an easier way to do it.
It's basically an led attached to an rj45 connector. You put them in all I
a patch panel then you put a battery pack on the rj45 jack in the wall somewhere then you can see which led/rj45 connector lights up. Seems a hell of a lot faster then toning out a messy patch panel.
I inherited the messy patch panel btw so don't scold me about it please.

>>2607812
The following list articles:
>https://en.wikipedia.org/wiki/List_of_7400-series_integrated_circuits
>https://en.wikipedia.org/wiki/List_of_4000-series_integrated_circuits
The following LM series is mostly analogue:
>https://en.wikipedia.org/wiki/List_of_LM-series_integrated_circuits

Also popular without fitting into the above lists:
>NE555 timer
>L298 motor driver
>MAX232 TTL to RS232 transceiver (or 1488 / 1489 if you're edgy)
>ULN2003 digital transistor array
>TL494 PWM controller
>27C series EEPROM

>>2607827
Digital shit doesn't go wrong often, I'd expect mask ROM to last centuries. I'd check for dodgy caps or linear regulators or a PSU causing brownout conditions. If the voltage is getting low, then sharp clock edges could intermittently push the CPU down into unspecified territory. A scope might be needed for diagnosing such a voltage level problem.
Order a nanoDLA or other cheap FX2-based logic analyser anyhow though. Reading the ROM is a cool idea, you can do that (and logic analysing) with an arduino but buying a dedicated reader is probably worth it to avoid the hassle of wiring up like 24 different i/o pins.

>>2607868
Can't you just put an LED in a modular jack with a crimping tool? Then the battery on the other side needs the current limiting resistor, but that's easy enough.

>>2607874
> voltage sags
Everything seems ok when I run the test points in the SM through my Rigol but it was a cursory glance, I might recap it for the hell of it, it's only a handful of components.

> buying a dedicated reader is probably worth it to avoid the hassle of wiring up like 24 different i/o pins.
Ya don't say. I'm using https://forum.arduino.cc/t/reading-obsolete-roms/295879 and getting what appears to be garbage from the ROM but I might not have the code tuned correctly for this chip, that's a work in progress. Good to know the ROM is somewhat reliable at least. Never thought to use the mega for logic analysis though (I'm a bigger analogue guy than I am digital). Will have a google around on it.

On that topic, I have had https://core-electronics.com.au/la104-logic-analyzer-1.html on my wish list for a little bit but never really had a reason to pull the trigger. Seem legit? I'm assuming the only important consideration is the sample rate (100MHz for this unit), which should well cover this 12MHz CPU, correct?

>>2607893
>getting what appears to be garbage from the ROM
Lower the baud rate and shorten the wires. Clock speed and baud rate are connected at the benis.

>>2607893
>what appears to be garbage from the ROM

binary code is indistinguishable from garbage until you run it through the proper disassembler.

>>2607861
variable speed drive, basically a controlled (usually) rectifier back to back with an inverter so you get AC in and AC out but can control the frequency.

I repair wind turbine electronics and they're used a lot these days for allowing the turbines to have a variable speed.

They're used a lot in industry to control the speed of big asynchronous motors.

They're handy enough to fix cause 99% percent of the time the issue is failure of a power semiconductor so you just swap it out or something stupid like that guy posted like a bad solder joint.
That's if there even is an issue with them.
The field engineers that work on the turbines are retards and regularly just keep taking parts out of the turbines and saying the part is faulty until they finally get to the actual faulty part.

Then some retard in logistics finds all these parts marked faulty in the warehouse and ships them to the company I work for to be "fixed".
We clean them up a bit, ship them back and charge them a few grand for their troubles.

>>2608086
ah
I'm used to people calling them VFDs
I work on industrial controls so I've seen a lot of them. Pic related is an old Toshiba that blew up on us.

>>2608159
What did you tell the boss?

Hey, I've got an asrock h61m-dgs mobo which freezes a few seconds after being turned on, it's an old board I got off the trash so I'm not expecting much from this, but I could swear the thing seemed alright when I tested it before (basically laid it over my desk on a silicone mat and connected a thumbdrive with Hiren's Boot, fancooler over the processor, etc), however it stops working after securing the cooler and mounting it on a case.
What could it be? I've no idea where to look; checked a few voltages and components I saw on BoardViewer but it all seemed alright, CPU pins aren't bent either.

>>2608162
funny story, my boss was in the room when it blew up working on the other drive
They were out doing work during a shutdown and it blew when the backup generator came up

>>2608159
I've never seen multiple types of caps used on the snubber boards before.
Them old units that use aluminium caps for the DC link scare me. I never even heard of dielectric absorption until I discharged one of those 450V caps completely to 0V, went home came back the next morning and the fucker was back at 50V, 4200uF too so could have given me quite a zap.

What makes it worse is that if the unit needs a DC link above 450V, they connect the caps in series with a resistor in parallel to keep the voltage even so you can't charge them with a small current cause of the resistors.

I was bringing one up to full charge one day and one of the retards that work with me hadn't done a hi-pot test on the IGBTs and one failed at about 1100V and the entire capacitr bank (36 4200uF caps) discharged through it.
I was quietly working in kicad on my PC like 2 metres away (There's reinforced glass between) when it happened and it scared the living shit out of me.

Everything was coated in white dust from the ceiling tiles jumping

Hello. I've been trying to repair My Gameboy Advance SP and failing. I've cleaned it with isopropyl alcohol and games still either don't have the Nintendo logo or boot into a white screen.

Super Mario Bros 3 works fine. What part of the circuitry is that cartridge not using so it works?

>>2608171
so what brands do you usually repair?
My plant is mostly Allen Bradley, we have everything from 1305, 1336, 700, 753, etc etc
I get stuck retrofitting new drives into old 1336 buckets quite a bit. The newer drives that die we send out to companies like you. The 1336s hit the scrap bin.

>>2608174
Generally we get big Siemens or GE units.
Got a few from a place called American superconductor.
Those converters are only part of what we do though, there some older technologies from wind turbines like rotor current controllers that we do and simple power factor units with big thyristors.
We do some of the control electronics too.

A lot of the turbines in Europe were put in in the late 90s and early 2000s and aren't supported by the manufacturers anymore so the windfarm owners can either decide to invest massively in buying the more up to date shit from the manufacturers or we can provide them with alternate parts that we designed and produce for about 60 quid and pay us a few grand for it.

It's actually funny some of the shit the manufacturers do.
There's a particular unit that is used in a load of Siemens wind turbines that was manufactured by a UK company and then GE went and bought that company. Siemens and GE are obviously huge competitors and so GE basically give Siemens zero support with the units now.
Couple that with the fact that the main IGBT modules are manufactured by a company that was hacked late last year and had to close down production completely for months and it turns into a complete shitshow.
Nobody can get parts for the units and companies are scrambling to find places to fix units.
It's gotten to the stage where they're happy to send us 20 of the units and to get back like 7 that we made out of the working parts from the 20.

>>2608165
Use Linux

did you use thermal grease?

>>2608165
Check your front panel USB ports for shorts and debris. Make sure the motherboard isn't shorting against the case, and re-seat or reconnect your SSD/HDD & RAM.

>>2608165
>>2608225
and make sure you plug the heatsink fan and case fan(s) in.

>>2608223
>did you use thermal grease?
Of course. Thought it'd be overheating from something but since it booted up alright a few times I didn't pay much attention to that; though I did notice the heatsink on the chipset thing (you know the thing replacing north/southbridge stuff post 2010's) was getting rather hot even when checking the bios.

>>2608225
>Check your front panel USB ports for shorts and debris.
Yeah I cleaned it all up, gotta say it was VERY filthy, so maybe there's lint or something stuck somewhere, gotta check more thoroughly.
>Make sure the motherboard isn't shorting against the case, and re-seat or reconnect your SSD/HDD & RAM.
Mobo's properly seated on the standoffs, do I need to check for continuity on anything?
>>2608227
>and make sure you plug the heatsink fan and case fan(s) in.
Yeah it's all done, though it's a stock 775 intel cooler since I didn't have a 1155 one, board's even got holes labelled 775.

>>2608238
>I did notice the heatsink on the chipset thing (you know the thing replacing north/southbridge stuff post 2010's) was getting rather hot even when checking the bios.
>do I need to check for continuity on anything?
You have a short somewhere. Either a component that died short or a hard short to the case / nicked wire. If you clean the USB ports and they're not shorted, loosen the standoff screws and shift the motherboard a bit, then test.

>>2608238
Check all of your I/O ports while you're at it. Use a flashlight and magnifying glass.

>>2608238
is it non-conductive thermal grease? if its conductive, did you make sure to not short out components with it?

Still can't figure out how to actually bias a BJT. I was under the impression for the longest time that the base is supposed to have Vcc/2 into it but that's seemingly not the case. Am I supposed to choose Ic or Ib first? How much and why?

>>2608268
> bias
Well, assuming a small signal npn, you need to give it at least .6v on the base for it to turn on, because of the B-E diode junction. That’s the simplest “bias” scenario I think.

>>2608268
Assuming common-emitter, I think you choose Ic first, because that’s decided by your load. Then Ib is just Ic divided by hFE. Setting a base voltage is just an approximation for setting a DC base current. And it’s the base-emitter voltage you’re setting, so the total base voltage will be a that plus your emitter voltage.

>>2607802
usb micro is a bad connector. usb mini was abandoned because it's not thin enough for phones. and everyone knows phones are the most important thing in the universe

>>2608268
>bias a BJT
feed it illegitimate statistics about crime or IQ

>>2608268
>how to actually bias a BJT

this video explains the process step by step.
https://www.youtube.com/watch?v=1IYOGhS2OZc

but really, the simplest/surest way to do it is to use a collector-base resistor like in the pic.

>>2608268
1. set Vb bias voltage with divider network
2. calculate Ve = Vb - 0.7v
3. knowing Ve, set the emitter current with the emitter resistor.
4. set Vc using Rc so the voltage drop is halfway between Vcc and Vb (using V_C = Vcc - Ic*Rc)

>>2607895
>>2607899
Dropped it to 9600, still getting the same output I was getting at 115200 at least.

To be clear, by rubbish, I meant vast swarths of what appeared to be repeating nonsensical data (left). I knew it needed to be disassembled but was just suss on all the repetition. I'm still not sure if what I am looking at on the right is valid but it at least it appears to have instructions in it - https://pastebin.com/z8dWKX84

I'll spend the night seeing if I can get a logic analyser running on the mega, will order one later in the week otherwise. Cheers anons.

>>2608370
Looks like machine code to me. Hand-transcribe it, it's fun I promise.

Is it over on pad 29 bois ? Is there a way to patch this ?
Asus ME103K_MB rev 2.0 on a 10" tablet.

>>2608432
Is the fuckin thing frozen in a block of ice or is your lens covered in furry vaseline?

>>2608433
kapton tape to protect parts from hot air.

>>2608432
You need to expose the trace under the solder mask (green paint) and solder a bodge wire between the trace and the connector.

>>2608442
*blue paint

I have a BSc in Computer Science but AI is killing the industry and I do repair electronics once in a while. Is it worth doing a masters in electronic engineering in your mid 30s?

>>2608432
If it's a ground pad then you can bodge the connector to any ground on the PCB. Otherwise you'll need to do surgery.

>>2608433
>bloody thumb security located next to the lens.
here is without flux on the lens

>>2608445
This might fill in the gaps in my knowledge but I wonder if it's worth paying for. This ChatGPT shit is doing too much reputation damage to my profession. Now every retard thinks AI wrote my code.

>>2608449
Carefully remove the solder mask in the circled areas. Use a dental pick, small slotted screwdriver, 600-grit sandpaper, etc. Just go slow. Scrape/sand towards the missing pad, not away.When you see copper, flux and tin it, then tin your bodge wire and tack it in place.

>>2608451
It would make sense if you want to get into aerospace, medical or other specialized jobs. Do you have any family or friends who are in the club?

>>2608442
>>2608446
>>2608454
I see. Thanks. Will wait for the microscope's order on that one and find one scrap board to experiment.

>>2608458
Cool beans. Another tip worth mentioning is using magnet wire for the bodge. People complain about the difficulty in removing the enamel so it can be soldered, but it's preferred because you can fit it into tight tolerances and it looks nicer. Another method is using copper tape to replace the pad, but I hate the shit. Too finicky.

I broke an arm and can't solder bros. My hands are shaking, and I'm waking up with cold sweats. I don't know if I can make it without a hit of solder fumes for another week.

>>2608472
I soldered with two broken arms and a missing eye back in Khe Sanh, 1968.

>>2608475
...How?

>>2608479
Pure will.
(and robotics)

>>2608445
It’s not killing anything any more than fusion reactors have killed the energy sector.
Just put all your money into FTX crypto futures and you’ll be fine.

>>2608508
>FTX

>>2608475
I soldered at Isandlwana with no eyes and I'm quadriplegic.

>>2608517
>I AM the soldering iron

Last year I was on a kick of making led tags.

I was trying to keep costs low so I ordered all raw components, including having custom circuit boards printed. Then I would 3d print the case, which by itself took 5 hours. This of course meant I had to solder, glue, screw, and program each individual tag. Even though they sold decently, I eventually got tired because it was just so damn time consuming to do all the work to even have one completed.

are there any recommendations for how to offload as much of this as possible so I can start working on these again? I'm programming AtTinys with Arduino if that matters

>>2608551
3D print jigs for easier automation
get soldering done by JLCPCB

does a 6-pin SMD IC with the label "2501A" ring any bells to anyone? It's part of a small charging circuit for a disposable vape

>>2608249
>You have a short somewhere. Either a component that died short or a hard short to the case / nicked wire. If you clean the USB ports and they're not shorted, loosen the standoff screws and shift the motherboard a bit, then test.
Now that you mention it, the PSU has continuity from the +3v plugs to ground, is this normal? Tested the mobo's PSU in and none seems to have continuity to the case.
Mind you it's a chinaman brand PSU from 13 years ago, the other PSU I've got is connected to my desktop PC.
I've previously tested this mobo with this PSU (outside case) and it worked, so it's probably that?

>>2608253
>Check all of your I/O ports while you're at it. Use a flashlight and magnifying glass.
Cleaned USBs again, don't seem to show continuity to ground (unless it's the ground pin), the others don't seem to be clogged with anything but I didn't test continuity on VGA and the others.

>>2608257
>is it non-conductive thermal grease?
No idea, it's chinaman brand HY880.
>if its conductive, did you make sure to not short out components with it?
Applied as per usual, small spot right in the middle of the CPU.

>>2608660
>HY880
its advertised as non-conductive so should be able to rules that out as causing shorts

>>2608660
A quick look through code books and LCSC turns up nothing. I’d reverse engineer the board to gauge its input and function, then with luck you should be able to find it on LCSC directly, after narrowing down by IC type and package. Shame LCSC’s database is a mess of incomplete fields and improper classification, you’ll need some luck to succeed that path. If it’s not a chink part then you should have a better time searching on Octopart, but all-in-1 USB battery charge ICs in tiny packages are certainly the domain of the east.

Are there any good books that cover hobby/beginner tier robotics and automation?

>>2608801
Ogata's Modern Control Engineering

>>2608803
Preferably not something that requires the first couple of years of university just to start with, anon.

Can I achieve instant death by poking two needles attached to the HV side of a microwave oven transformer into my body and then switch on the primary side?
It depends on where it goes I guess, but it should stop my heart with no chance to bring me back even if a surhical team with all equipment in the world were right next to me, right?

>>2608862
Putting them under each armpit is probably surefire death. Not instant though, you'll be aware of your heart stopping and of the pain of electrical burns.
Alternatively, you could go to a 3rd world shithole and use your rudimentary electrical understanding to be a wizard. Or volunteer to be on the side of one of the dictators during the ukraine conflict, that's probably going to kill you.

>>2608862
We're all gonna be dead soon so just ride it out. Look at it this way– you got to witness the end of the world.

I'm looking for a way to make an "Ethernet shutoff switch" so I don't need to keep unplugging a wire from my patch panel over and over.

The use case is that my corporate network wifi doesn't reach far enough out into my back yard to take meetings in the garden- I have wired up an exterior rated RJ45 port but I don't want to leave it plugged into the VPN appliance all the time (Someone could gain physical connection to the network, my back yard isn't secure) and I don't want to put a padlock on the jack because then I need to fiddle with unlocking in.

My idea was to take the blue pin on your standard T568A run and run it (with a diode) to a mechanical toggle switch. that switch will then take that voltage and if it's "ON" send it to a bunch of transistors in parallel (1 per pin) to close them and allow the other 8 pins to connect. If the toggle switch is opened, it will cut voltage to the transistors and disconnect every pin.

Convince me this is a bad idea and I should just get a small padlock/unplug it when I'm done. There shouldn't be too much of a problem tapping on the blue pair since it's just for poe voltage, and the device I'm plugging in does not use poe.

>>2609013
>blue pin
You'd probably want to switch the white/green and/or white/orange wires.

>>2609013
I don’t know about that.
The diode itself can do the switching. If you attenuate the ethernet voltage to be under .7v the diode won’t be on and nothing will get through. To turn it on, add back a bias voltage with a switch.
The pulse transformers on (real) ethernet block the dc offset. You might need a video amplifier (it’s like an op amp but faster) to boost the signal.

Or, you could do what I did, and get an old d-link 4 port switch at good will, put it between the port you want to protect, and just toggle the power to the switch with with a switched lamp socket. The lamp even indicates if it’s on or not.

>>2609029
I was going to switch all 8 wires.

>>2609063
>switch all 8 wires.

then you'd use a relay as transistors are not good at switching in both directions.

anyway, what i'd do is use of the numerous ethernet switches i have in series with the signal.
just unplug it, and the connection dies.

>>2609013
Buy a managed ethernet switch that has this function?

>>2609063
MAC filtering is a thing ya know? Blacklist everything except your devices MAC address, turn off DHCP and configure precisely one static IP address in a weird range

Leaning Kicad by trying some easy scheme from scratch. I feel I'd better draw 2 rows of female 1x15 connectors to host the Arduino Nano than using it in the schematics. Is it the right way?

>>2609136
The symbol doesn't need to reflect the actual pin configuration of the device, usually doing that makes things less clear. Though regardless of the symbol configuration, so long as the PCB footprint is correct (pretty sure KiCAD has footprints for arduinos) it doesn't really matter. Making your own symbol is a good thing to get som practice at.

>>2609132
I don't have direct control over the meraki provided by work, if it was on my home network I'd mac filter it but I'd have to convince the sr. net engineer to do it and I don't want to bother him.
>>2609113
If the port sees any non-authed devices on it portsecurity will lock it. Could do a dumb switch like >>2609034 said but that's cheating, isn't it?

Probably a bit dumb but I want to make sure I have the right parts in mind, if anyone has any recommendations for this simple thing I'm making.

I just want to be able to power an LED bulb by turning on and off a small button using preferably a battery pack of AAA or AA as an energy source. Rather than make a mistake and buy something incorrect I just wanted to ask here.

Wiring wouldn't be too long, a yard at most.

>>2609254
> meraki
LOL, physical network switch as a service.
I saw an ad for those and I thought it was a spoof right off the bat.
They literally force you to pay for a shitty web UI when everybody else just includes it as part of selling hardware.
Choosing those…. I’m amazed you can work from home, my grandmother has more technical core competencies.

What's the best way to individually address bits in an array of flip-flops? Is a binary tree of MUXes really the best way? I feel like that's not an optimal solution.

t. retard that never studied electronics

>>2609390
>What's the best way to individually address bits in an array of flip-flops?

Meaning what? Like you want to be able to toggle individual outputs attached to a common bus?

>>2609392
I want to read and write a specific flip flop in an array of ~64 of them, using an integer value from another array of six flip flops to address that particular bit.

>>2609325
>power an LED bulb by turning on and off a small button

your local dollar store will likely have several models to choose from.
there's even a unit you can control both with a remote, or through Alexa or Google assistant.

>>2609393
>array of ~64 of them

sounds like a problem much much better suited to a software solution.
rather than a 5-gallon bucket of chips.

>>2609393
>read and write

WTF are you doing? It sounds like you're designing a tiny, 1-bit memory cell.

The first things that comes to mind is attaching them all to a bus. Having to give each flip-flop a full-on, 6-bit bus decoder seems excessive, though. It's probably better to multiplex them the same way you would with an LED array. Each flip-flop only needs a single 2-input AND gate (one input on row, one on column) to be addressed in that arrangement. Reading could be done similarly, by simply adding another AND gate that activates if the address gate and flip-flop output are high. In other words...basically the same way that real memory modules are addressed.

If your input logic is already in binary, you will need some additional decode logic to convert that into row/column. Not sure if the transistor count in that additional logic would end up exceeding the give-everything-a-bus-decoder idea or not.

I want to make a high-power variable benchtop supply using an ATX PSU's 12V rail as a base. I'll use a split-H boost-buck topology (pic related) for the continuous current at both input and output, using 60V FETs fed by high-current half-bridge gate drivers.

The thing is, I'd like to have an optional linear post-regulator for reducing switching noise. As far as I understand, there's two ways of doing this:
>1: Have one feedback loop feeding the pass-transistor base such that the output of the pass-regulator output is X volts below its (low-pass filtered) input, then use a second feedback loop to control the switching controller such that the output of the linear stage is always at the set-point
>2: Have a feedback loop feeding the switching controller to ensure the input to the pass-transistor is always X volts above the output of the pass-transistor, then a second feedback loop feeding the base of the pass-transistor such that the output of the linear stage is always at the set-point
I think 2 would be more stable, and have a more accurate output voltage. But then it's less easy to bypass the post-regulator. I can't really find any examples of this kind of dual cascaded feedback loop, just people talking about putting fixed linear regulators after switching converters.

I'd also like to know how you're supposed to feed the seperate half-bridges. If I use two seperate TL494s, they might end up fighting a little close to the 12V output area, which I don't think is a problem, but it's kinda janky. Are there dedicated ICs that can control such a PSU? All I'm finding are full-bridge controllers designed for isolated converters. Anyone have any thoughts?

Since the topology is bidirectional, I might as well also turn it into a variable load. And I want the ability to use an external analog feedback loop to control it. Might as well try to make it use the full 264W of the ATX too. God I love feature-creep.

>>2609398
Ah, I'll clarify:

What I mean is an LED bulb like this.

>>2609440
forgot pic because I am waking up

>>2609342
Meraki are just a VPN appliance. It's not something they give out to anyone who WFH (most people are getting back in through normal over the internet VPN), it's stuff they give out to people who have physical hardware requirements. I'm in It, and have some basic network understanding, but I'm not net eng. I have 3 computers and several POE devices I need to plug in for work, so a meraki hardware VPN appliance makes the most sense.
>They literally force you to pay for a shitty web UI when everybody else just includes it as part of selling hardware.
we're a cisco shop. It's a regional company that makes about 7 million dollars (revenue) a day- I'm sure they're fine.

>>2609475
If you can use an external antenna with your existing AP, get a unidirectional antenna (Yagi) and a signal booster if necessary.

>>2609136
Why does it have two RESET lines?
Is this something I’ve been missing out on? Can I get a high-end mcu with, say, 64 reset lines?

>>2609592
You need more reset lines in spaaaaaace.

>>2609441
So you want to power the bulbs with those batteries? Either:
>make a ZVS inverter that steps up to 120/240VAC
>take the bulb apart, hope it's not a HV string of LEDs, and bypass its step-down converter with your own step-up converter

If you just want to add a remote cable switch for a bulb that itself is running on mains voltage, then you don't need batteries at all, just a mains voltage rated switch and some twin-core flex that interrupts the live wire. I'd ensure there's plenty of quality insulating heat-shrink tubing around where the wire meets the switch.

do you guys recommend to use pcb coatings?
I mean for the finished soldred pcb, I am a noob, and i am worried about some powered transistors that need heatsinks.

>>2609689
Nah this has to be either remote or pressed by a button while being handled. For a prop in a play.

>>2609718
"low voltage LED candelabra bulb"
"CC/CV boost converter"
"USB power bank"

>>2609721
You can skip the CC/CV boost converter and just get a USB power bank with 12V output. Then you connect a USB cable to the power bank and solder the red/black leads to the bulb socket with a switch that makes/breaks the red lead. Make sure the power bank is powered on. lmao

>>2609726
>connect a USB cable
barrel plug

>>2609718
I still don't get it. Questions:
>1: are there size/shape limitations on the light itself
>2: are there size/shape limitations on the remote control
>3: can there be a wire between the remote control and the light
>4: can there be a wire between the light and wall power
>5: does the light need to be bright enough for illumination, or just to see the light itself?

If no for 3 and 4, the chad method is to use a fluorescent or neon "candle bulb", and activate it remotely by turning on a tesla coil under the stage.
If no for 3 and maybe 4, there's a variety of remote activated relay or remote activated mains switches out there, though the relay side's form-factor may be too restrictive for those.
If yes for 3 and no for 4, then look for an appropriately shaped battery powered candle and bodge the wires in place of its on/off switch.

>>2609749
1. Yes, standard socket for a bulb/LED bulb
2. Preferably a single button that can be pressed once to turn on the light and again to turn off the light
3. Yes, it wouldn't need to be long, definitely not longer than a yard.
4. No, no wall power method will be sufficient for the prop
5. It needs to be able to be seen beneath stage lighting

>>2609785
Just realized if I could find a remote control bulb socket for an LED bulb that'd be it.

>>2609793
Ones I can find so far seem to demand being plugged into an actual socket though. Fugg

>>2609785
>Yes, standard socket for a bulb/LED bulb
I meant more its surroundings. Like, does it need to be fitted into a candelabra? Or a hand-held mock-candle? Or is it fixed into a wall or ceiling?
>Preferably a single button that can be pressed once to turn on the light and again to turn off the light
Again, size limitations of the whole thing. Does it need to be hidden inside the hand or pocket? Concealed beneath clothing on a wrist or chest?

The simplest solution seems like it would be to find a 12V/9/6/5V light bulb that has the appearance you're after, then cram an appropriate battery inside something nearby. Activating a mains-voltage bulb without it being plugged into the mains is going to be expensive and bulky.

>>2609702
what coating are you talking about? why would transistor heatsinks factor into that? assuming youre talking about a production level design, i use a conformal coat across all sides. this saves the pcb and components from shorts due to condensation or random conductive dust

I've been working as a mechanic (and by extension, an auto electrician) for the last 8 years and feel like taking on electronics as a hobby.
I understand the basics of DC circuits and diagnosing some basic digital/analogue stuff - PWM controlled actuators, crank/cam angle sensors, throttle position sensors, solenoids, relays, etc. My main interests are playing around with microcontrollers and learning how analogue radio/audio stuff works.

What book should I read?

>>2604810
What is "asynchronous communication between two data buses"
What would this chip be used for?


https://www.ti.com/product/SN74AVC4T245/part-details/SN74AVC4T245RSVR

>>2609858
no books, learn by doing. grab anything not tarduino and get cracking. also visit /mcg/ though sometimes /g/ autists come in and derail everything with their x vs y software arguments

>>2609865
level translation between two voltages without pull up/down resistors. IE: communicating with a sim card requires data to be sent at 1.8V. most simcard interface chips or mcus are 3.3V. you chuck one of these chips inline with the data bus and till convert 3.3V down to 1.8V for tx, and 1.8V to 3.3V for rx

>>2609866
I don't know jack shit about how analogue circuits like op-amps or frequency generators work though, one of my goals is to build a modular synthesizer from scratch.

>>2609866
>grab anything not tarduino
arduino is perfectly fine

>>2609858
A read through a book like "practical electronics for inventors" should give you the basics. From then I'd jump into a more practical look. That's both making and simulating circuits, but also researching existing circuits, reading application notes, that sort of thing. I'd recommend checking out some of the resources in the OP.

Also if you're into audio you'll love this guy:
https://www.sound-au.com

>>2609870
Thank you

anyone has resources to teach me how to work around CRT/Vaccum tubes? Like safety precautions, etc.

>>2610200
Shango066

>>2609866
> no want book.. anon want smash
Too bad magazines like elektor, eti, radio electronics, etc don’t exist anymore

>>2610246
Check this link from the OP:
https://worldradiohistory.com/Popular-Electronics-Guide.htm

It's not the same, but it's something.

>>2610200
one hand behind your back
also isolation transformer if working on old radios

$1987

>>2610272
Please take me back...

>>2610278

>>2610278
To shitty slow computers?

>>2610292
>computers that are fun and comprehensible instead of controlling you
Why do you think I'm into embedded?

>>2610292
It can run Wolfenstein 3D.

>>2610294
>instead of controlling you
Why do you think I run Linux?

ACT NOW
SUPPLIES ARE LIMITED

>>2610301
I also run Linux, but Microsoft and Oracle are major sponsors of the Linux Foundation.

>>2610312
>36 years later
>marketing is exactly the same

>>2610316
>I better hurry up and invest
lmao

>>2610318
>prolog
>lisp
The residuals of "symbolic" AI. Now everyone is into NNs.

>>2610322
I wonder how a neural network would fare on a Mac cluster.

EPSTEIN DIDN'T CPAP MACHINE POWER CORD HIMSELF

>>2610322
Nothing really new in NNs either other than hype.
It’s been used in post offices to read your chicken scrawl for decades already.

>>2610343
Now microsoft just uses the clang front end for their C++ compiler.
They can’t even compete against freeware and occasional volunteers/internet randos.
Even their C++ templates needed to be hacked into their preprocessor since they didn’t even have the ability to maintain, update and improve what they had.
What’s left? Taking others’ code and wedging it into bing and trying to steal apple and sony’s innovations.

All I need is QuickBASIC and gorilla.bas and I can ruin AI and take over the world.

Looks like the laser has some delay between turning on and off, which is a pain. There's also a significant difference between its effective intensity in long straight areas compared to tight squiggly areas, so obviously the acceleration limitations of my laser are limiting me.

I'll try just tuning the maximum speed down and tuning the laser power down, hopefully that solves my problem.

>>2604810
I'm a total noob at electronics, currently learning. I can't seem to grasp one thing;

I've been taught that the current primarily travels from the ground to vcc via electrons. And that makes sense in context of connecting the resistor with kathode of a diode. But when I measured current on the diode pins, it was actually higher on the anode. I know the diode itself eats up a little bit of current but I can't understand why the measured current is actually higher from the plus side. The same goes with resistors, I noticed.

Does this just mean my multimeter is favoring the + to - flow in it's outputs? Or should I measure with inverted probes and assume it makes sense now since electrons are negatively charged?

>>2610368
Two things:
Electron current is fucking stupid because everyone uses conventional current (i.e. what naturally flows from positive potentials to negative potentials). Unless you're talking about what happens inside semiconductor devices or electron tubes, it doesn't matter what charge moves. Even when you're using semiconductors or tubes, it basically doesn't matter since you only ever care about the current outside them. As far as your multimeter is concerned, a positive current is just a higher voltage on the red side of the current sense resistor than on the black side.

You don't measure current "on the diode pins", you measure current "through the diode pins". Current is equal in series, so there should be no difference between current on one side of a diode compared to on another. Did you mean measuring potential on either side of it? Post a damn circuit diagram.

>>2610368
Are you a physicsfag or a chemistryfag?

>>2610306
>$1000 for a fucking programmer
>not counting inflation

>>2610375
>"Just put it on the company charge card."
>buy 2, keep 1 free because "it broke".
lmao

>>2610368
>>2610372
I meant voltage but I kept writing current. Sorry.
The diagram is literally picrel. I just wanted to know why the VOLTAGE was higher on the plus side of the diode but I think you explained that well enough for me
>>2610374
I know a fair share about doping the semiconductors but I'd rather say physicsfag

>>2610386
>I'd rather say physicsfag
If you serve two masters you'll end up schizophrenic. You have to forget everything you learned in physics textbooks and go into practical logic mode.

>>2604810
not really strictly electronics related, but I'm sure you guys know.

I want to run a 120v 1100' / 335meters long wire (with a 20a breaker) to feed an electric gate, a couple of lighting poles, and maybe something else along the way. would a 12awg / 2mm stranded copper wire suffice for that distance? or should I opt for a thicker wire? aside from the gate electronics and the LED electronics, nothing that is sensitive to voltage sag will be exposed. yes, I'm trying to save every penny possible

>>2610386
Yeah you just look at the circuit as 3 different nodes of different relative voltages. Typically we treat the 0V or common or ground or whatever as the reference point for all other measurements, so in this case the voltages of the three nodes would be 5V, ~4.3V, and 0V.

>>2610397
What's the maximum total current draw? Chances are it's way lower than 20A, and its what you calculate your wire size based off.
12awg is kinda expensive though, if you really need wire that thick you'd be better off using a microwave oven transformer at either end, in order to have a 2kVAC transmission line using thinner wire. The resultant cable would need to be insulation tested appropriately though. Otherwise using a more tame 120/240V isolation transformer at either end should definitely be a consideration.

Even a solar panel or other energy harvesting methods might be on the table, considering you'd be spending in the ballpark of 200USD in wire.

>>2610401
max power with the gate (~700w) running, would be about 1,300 watts including the lighting (~600w).

I don't know much about solar panels, but sadly the whole area is covered with a bunch of thick trees, with not a single patch of clear sky in sight, so I don't know if that would work.

if it was just me walking through that path, I wouldn't mind using a microwave oven transformer, but I don't wanna end up killing someone with my butchered together DIY, so I'm gonna have to pussy out of that one.

the 120/240V isolation transformer idea seems nice. I could just get a bunch of LED street lights that play nice with ~240v, and then just convert back to 120v at the gate.

>>2610401
>ballpark of 200USD in wire.

dont listen to this guy.
you can pick up 1000 feet of thick wire almost anywhere for free.

>>2610419

>>2610415
>(~700w)
Shit that's kinda a lot. Can't you downgrade your gate motor?

> I could just get a bunch of LED street lights that play nice with ~240v
Oh there's LEDs all the way along? Yeah it should be possible to get LEDs that work fine off that, though if they're dual voltage they'll be the SMPS-based ones instead of the linear regulator ones. Arguably more shit in them to go wrong. Or order a bunch of cheap LED bulbs from UK/EU stores.

Ehm is there free online brushless DC engines calculators? Like estimation of torque/KV rating by number of windings etc. I'd like to make one from scratch with my own characteristics and I don't have money to pay someone to design it.

I will buy my first bench PSU, just how bad are the $40-60 ones?
I will mostly use it to find shorts in laptop motherboards.

>Spend more
I still have to buy a hot air station and a thermal camera so budget is low.
Siglent would be a example of a good brand based in reviews, what else is around?

>>2610614
A PSU is usually a good first project, I’ve never needed one. No idea how good any arbitrary chinese power supply is, they often change their designs based on buying large batches of components for a run.

I often just use a wall wart and a cheap dc/dc converter module with current and voltage if I need something like that, which is rare. Generally I think you only need 1.1, 3.3, 5, 12, and ±15.

Anyone recognize this famous PSU built in the 80s and still in use today?

Either that, or prepare to buy a noise filter (and an oscilloscope to test said noise filter)

>>2610614
>>2610650
I also recommend it as a first project, though I'd suggest to build one with some sort of current limit. Be that a variable supply with actual constant-current output (handy for testing NTC loads like LEDs or lasers) or just an adjustable overcurrent latch, whereby reaching a threshold will shut off your output until you push a button.

Simple designs include:
>single tap transformer or AC/DC SMPS with linear output adjustment
>manually ranged multi-tap transformer
>a modified DC-to-DC module with adjustment pots
More complex designs:
>auto-ranging multi-tap transformer using relays and a linear output stage
>custom variable DC-to-DC converter
Very complex designs:
>custom variable DC-to-DC converter with linear output stage
>custom variable direct AC-to-DC converter

Then there's a lot of freedom as to what adjustment methods and display methods you use. I'd recommend a quickly updating (at least ~2Hz) 2 decimal-place digital panel meter for something simple and cheap, and using either a multiturn potentiometer or two potentiometers for fine and coarse adjustment. Nothing wrong with analog meters, but their scale can be a bit coarse. Digital display+control systems can ensure that you can set your supply to output an exact voltage very repeatably, so if you're using an MCU and ADCs to drive a custom display, you might as well also use rotary encoders and DACs to be output references for the PSU itself. Adding coarse and fine adjustment is then just a software problem.

Say I have 20 some poles I want on an analog switch, volt/amp could be 50v 5A. Is there a good way to do this without relays? I want to use the same breakout board, stepper drivers, power supplies, etc, to drive two different types of CNC machines depending on the switch. I've found some small 10PDT switches but that's about it.

does anyone have a recommendation for a "circuit cookbook" filled with ready-to-implement good-enough circuit designs

>>2610614
I have an AliExpress bench supply
It can do the loads advertised

>>2610686
What's inherently different about all the analog projects? Just an amount of components to solder? Because obviously the beginner won't create his new schematic, he will take already existing one and just solder it.

>>2610723
If it’s just power (or anything else with mono-directional current) you’re gating, and all the outputs are allowed to be referenced to one another, then an array of MOSFETs should be fine. Demuxers or SIPO shift registers or counters to control them from fewer pins would be what I’d do if I/O is at a premium.

>>2610730
sound-au.com

>>2610752
Yeah but you can still hunt about for existing circuits with specific characteristics, like “using LM317” or “using TL431”, with a little guidance. And even a beginner could probably make crude mixings of existing circuits, like adding the over-current latch from one circuit in front of another.

Also “beginner project” implies something early in your journey of creating projects, not necessarily in learning theory or even breadboarding.

>>2610686
> current limit
I use these specialized, hermetically sealed, glass PTC resistors with backlit real-time current indicators:
Light bulbs.
For 12v, you can get a vast array of these things and they are fine with hooking them up directly to a 200A 12v source. Anything for cars, basically.

I don’t like overcomplicating things.

>>2610723
Just add a single
SPDT switch to your 10 pole rotary switch and <blam> … it’s now effectively a 20 pole.

That’s how the 12% boost switch works on >>2610650

>>2610723
Use a patch panel instead

>>2610816
Not the anon you replied to, I believe FPV drone community does the same concept, uses a light to absorb a short when they first power test their drones.

>>2610816
They require a bit too much room between your operating current and your "trip current". As in, the maximum current at which you can use it without dropping appreciable voltage is going to be a lot lower than the minimum current at which it would act to thermally protect something like an MCU connected in reverse. Probably fine for more brute-force things like discrete transistor power amplifiers, but they just don't have the finesse for precise use.

A CC linear regulator or latch is a relatively simple thing to add to a circuit, especially if you don't care about it dropping a volt or two such that you can get away with a large value sense resistor and no sense amplifier. Or a fucking LM317.

Feature creep is a good learning experience.

Making a PCB to route high gain audio, are thicker tracks better or more probable to pickup interference?

>>2604810
i want a bipolar power supply, -30V to +30V 3A maximum with constant current and constant voltage control. schematics like pic related will work but only one of the outputs will be regulated.

can i make separate +30V flyback and -30V flyback converters, join the floating outputs together to make a virtual ground, and have a regulated (albeit non-tracking) -30V to +30V? or will the control loops fight each other?

I am contemplating either fixing hundreds of footprints or kms. Why the fuck are there ZERO standards. Every fucking footprint downloaded from places like Snapeda or Ultralibrarian etc has a slightly different style. Some are outlined, some are not. Some have this stupid green background (Mech-15 I guess) and also purple (another mech), sometimes one of these, sometimes neither, the overlay outlines have arbitrary width, standard shapes (like diodes) have varying proportions etc. Seriously try downloading a couple basic parts like SMD caps and see for yourself. But the real question I have, what's the point of these mechanical layers if they are not printed and it is only the silkscreen that matters?

Specifically, what's the point of using a mechanical layer for a component's courtyard if there is the overlay layer for that which is the one that actually matters??

>>2611016
Just create your own footprints. Why would you trust some pajeet-made footprint that probably has other issues as well? Does it even match recommendations in the datasheet? SMD parts are mostly standard anyways so default CAD library should have footprints for 90% of components.
>>2611017
In the end only layers in the final gerber output matter. Other layers are used by CAD software to perform DRC so consult manual on what exactly it does with those layers.

>>2610924
Wider traces increases capacitance but reduces inductance and resistance. The capacitance is between the traces and ground plane as well as between the trace and your environment. So yes, wider traces could pick up more interference, so try to keep it thin unless you need to keep resistance and inductance low (or need some kind of impedance matching). You could also keep your wide trace further away from your smaller traces which will reduce crosstalk, but the wider the trace the more space is needed between two lines. You can also reduce noise from the environment by using two grounded plane layers and have your signal run between those two planes, but whether that is necessary or not will depend on your design specifications and what your'e trying to accomplish. .

>>2611018
Yeah, but I am just trying to understand, for example, some SMD resistor footprints do not even have soldermask, and some like this one has both soldermask and Top Paste. Here, I moved the pad of the way, what is the purpose of this top paste polygon? And then a masked pad is placed on top of it.

>>2611021
Just chuck it in the bin and stop overthinking about it. Follow IPC-7351 if you want to larp as an engineer or just copy from standard library or the datasheet. There is no point in banging your head against the wall because some pajeet made footprint in a certain way. At minimum you need
>pad - this is on copper layer
>solder mask - this is cutout in that green plastic looking stuff on top layer to expose copper, depending on your software this could already be automatically included with pad
>solder paste - this is used to create stencil for paste application, again software could automatically add this with pad
Optionally
>silkscreen - print on top of solder mask
>courtyard - for DRC
Z ordering (what comes on top of something else) doesn't matter because PCB layer stackup is fixed.

>>2611063
>standard library
Is there a standard library for Altium (without an online access)?

Also, do you know why some pretty standard footprints do not have solder masks? I have some SOIC-8 SOP-8 footprints and all of them but one have a solder mask around the pads. Also tons of basic SMD resistor/capacitor footprints from SnapEDA/Ultralibrarian do not have solder masked pads either. I understand it could be pajeet and all but clearly there's lots of parts like that so I was wondering if there is a good reason for that.


>no point in banging your head against the wall
I have a problem with that. I always get stuck and can't proceed and will bang my head for weeks and procrastinate over simple things but I can;t help it. And with these footprints I am completely paralyzed since there are hundreds of them and they are not neat enough and I want to make them neat and that makes me anxious and I cannot make any progress so I am just researching the alternatives. Maybe I will just create my own footprints from scratch.

Also why are some pads rounded and some are square? That drives me absolutely fucking nuts. Same kind of style, SOIC or SOP or TSSOP, etc. but some are rounded and some are not. WTF?

Wait, I have to PAY for IPC-7351B?? What the actual fuck? Can't find it on torrents.

Found this doc:
>IPC-7351B Naming Convention for Standard SMT Land Patterns

Fuck, this is another can of worms. Now I need to rename all my footprints and make sure the symbols get updated too. And it is not easy to set all the proper names, it is so confusing.

Wow, look at a typical "standard" KiCad footprint. No soldermask at all. But it uses like 3 or 4 different layers. Why so much useless garbage.

>>2611094
>some pretty standard footprints do not have solder masks?
This is most likely because CAD will automatically generate them from pad. If you want to adjust mask or paste then you can add additional polygons for that.
>Also why are some pads rounded and some are square?
This is due to soldering because it can happen that solder won't flow into sharp corners, but you shouldn't worry about this because it's manufacturing autism for commercial scale production.

If you're worrying too much about visual design and things being nice you're in the wrong business. Engineering is all about embracing approximate ugly solutions, only art and science care about being beautiful, because real world is inherently ugly. You need to learn to embrace wabi-sabi and taoist philosophy of learning to not force yourself upon the world because that will only bring you unhappiness.

I'd like to build a switch mode power supply using only components I have laying around. This means toroid cores and magnetic wires. Can a suitable transformer be built using toroid cores or do we need air gapped cores for all kind of SMPS? Any efficient topologies with toroid cores?

>>2611094
>I always get stuck and can't proceed and will bang my head for weeks and procrastinate over simple things but I can;t help it.
Kek, i'm also like this. That being said, take your meds, now!

>>2611122
KiCAD has soldermask with all its footprints. Just it's hidden behind the copper layer by default. Swap to viewing F.Mask and you'll see a purple layer. It's automatically added with a pad when designing a footprint in the editor, but you can remove it if you want a pad with masked copper on it.
>Why so much useless garbage
The grey box on the outside is for DRC, basically no reason not to include it since it speeds up your workflow in ensuring no components are overlapping.
The cyan stuff is the silkscreen, labelling which part is R6 is pretty handy when hand-soldering so you can ensure you get the right part in the right place and in the right orientation. Again, no reason not to include this.
The other grey lines on the inside is the "F.Fab" layer, which is purely used for in-software reference. If you don't have any silkscreen layer, or the board is too cramped to label each IC, you can still use this to see the orientation and name of a given part by referencing the KiCAD file. Unlike the silkscreen, F.Fab and B.Fab carry both reference designator (Q2) and part name (BC557) information. They're not only useful when assembling your PCB, but also when designing it in the first place. Can't imagine shoving about a bunch of objects with no label on them.
Also SMD components come with a paste layer that's hidden behind the copper.
And there's the associated 3D model files.

If your pirate version of the latest industry meme cloud software doesn't let you have footprints, you should look for either an extension that will give you a bunch, or download a pack of them all made to the same standard. I know Digi-Key had a bunch for KiCAD, not sure if there's similar things for other software.

>>2611139
Air-gapped cores are only used for flyback converters, which are an isolated version of a boost converter. Forward converters, the isolated version of a buck converter, don't have this restriction. Think you can use a series inductor instead of having an air gap.

>>2610817
Common rotary switches might not be able to handle 5A. Mine are only rated at 300mA and they're 26mm wide chonkers, though that rating is at 120VAC.

>>2610961
Why flyback instead of using non-isolated topologies? It makes finding appropriate existing magnetics a lot more difficult, or if you're winding your own a lot more work. You could use a buck and Cuk converter to get above and below the 0V rail from a 30V source. Or anything with two seperate switching controllers really. That makes them both able to track properly, which is a good thing considering you could fry something if the side with feedback gets an unexpectedly stronger load than the side without feedback.

More importantly though, why would you want a bipolar supply for anything other than audio? If it's for audio, why would you want variable instead of just having 5 or so different bipolar rails? Having current limitation is still a decent idea, but since it's mainly for protection as opposed to driving constant-current bipolar loads (these don't exist) you could either have an overcurrent detection latch that just turns off an output relay, or maybe hook up linear pass transistors instead of linear regulators and have a CC circuit connected to them. With some sort of thermal protection for them, like strapping them to thermistors. If you use LM317s or equivalent as pass transistors with a dedicated CC feedback loop, then you can use their built-in thermal regulation.

Generally I'd recommend making your low-noise bipolar audio supply be a seperate project from your high-power CC/CV load testing / battery charging power supply. The former using an iron core mains transformer, the latter using a SMPS of some sort. A cheap way is to buy a CC/CV DC-DC converter module (e.g. LM2596) and just run it off an AC-DC brick.

>>2611142
Thanks.
OK, the silkscreen/overlay + courtyard make sense and the Altium libraries I use have them too. I didn't think there was a reason to include anything else. F.Fab is kind of odd, but I get it why it might be needed, it's just there's too much text on it. It may be similar to Altium's Mechanical layer 13 (pink) which is used as yet another outline +3D is attached to it.

>>2611166
ive just read through all your bullshit. fuck you are the most autistic sperg ive ever seen on this website. if you hate the way other people have done things, just fucking do it yourself. its not like theres a tool within every fucking pcb design software that lets you do this. god i hate larping faggots like you
>muh standards
fuck off the only thing that matters is if it works or not

>>2611142
>KiCAD has soldermask with all its footprints. Just it's hidden behind the copper layer by default. Swap to viewing F.Mask and you'll see a purple layer. It's automatically added with a pad when designing a footprint in the editor, but you can remove it if you want a pad with masked copper on it.

Why is it completely hidden by the copper pads? Shouldn't it extend around them?

>>2611189
Well the copper pads are slightly oversize in general, so I guess it doesn't matter.

Which are more electrostatically sensitive out of transistors and mosfets or are they pretty similar?

>>2611293
mosfets are notoriously pop sensitive. why? do you keep blowing up your trannies or something?

I shorted a cap with my multimeter probes because I was getting impatient waiting it discharge. It sparked and now the voltage doesn't go down at all. Is it over?

>>2611299
>the voltage doesn't go down at all.

your cap has been transmuted into an infinite power storage system.
you'll be a trillionaire as soon as you submit the patent.

>>2611295
I had this notion that mosfets were more sensitive but couldn't remember where I got that idea, just looking for confirmation. So they are. Where it doesn't matter I'll use transistors then

>>2611317
>mosfets were more sensitive

history lesson: static only became a huge deal when the industry came out with these satanic CMOS (FET) chips.
TTL and analog are essentially immune to such trivial threats.

Any tips for routing a one-sided PCB? What are your thoughts on jumpers? Should i use them?

>>2611299
what kind of pussy ass cap (fr fr) can't handle getting its legs shorted

>>2611325
>your thoughts on jumpers?

your time, your choice.
you can spend 20 secs soldering one, or thirty minutes trying to eliminate one.

>>2611328
Are there any disadvantage of using jumpers than traces?

>>2611330
yeah, if it triggers your OCD and makes you want to an hero for being suboptimal

>>2611330
>jumpers vs traces

one is a bit of copper carrying electricity from one place to another.
the other is a bit of copper carrying electricity from one place to another.
so, huge difference.