/ohm/ - Electronics General: Point-to-Point Edition

[Anonymous No.2464553]

Pull down that thread!>>2457597

>I'm new to electronics. Where to get started?
It is an art/science of applying principles to requirements.
Find problem, learn principles, design and verify solution, build, test, post results, repeat.

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

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

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

>Archive of Popular Electronics magazines (1954-2003):
https://worldradiohistory.com/Popular-Electronics-Guide.htm
>Microchip Tips and Tricks PDF:
https://ww1.microchip.com/downloads/en/devicedoc/01146b.pdf
>Li+/LiPo batteries required reading:
https://www.elteconline.com/download/pdf/SAFT-RIC-LI-ION-Safety-Recommendations.pdf

>Books:
https://libgen.rs/

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

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

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

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

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

[Anonymous No.2464557]

Thread brought to you by the RTC4553 RTC, with integrated crystal oscillator, a full date-counting IC (leap-years included) with serial interface

[Anonymous No.2464572]

Trying to understand that concept:
>Power line impedance is a very important parameter on the design of power line communications modem architecture. Variations on the impedance of the power line affect the communications circuit performance.
Can you explain me what impedance means ? And what are the effects on a bad impedance situation ?

[Anonymous No.2464579]

>>2464572
impedance = resistance except it's AC
When you have two bidirectional devices communicating over the same transmission line (wired and wireless), the impedance of both need to be matched or else the signal is substantially degraded to the point it's unusable.

[Anonymous No.2464600]

>>2464579
>the signal is substantially degraded to the point it's unusable.
I see, one of the signal, emitter or receiver, can't read the message. Can it leads to escalating and unwanted problems concerning one of those 2 devices? It's a weird question, but since it's housing power, I wonder if it can leads to push a power transformer into damageable instabilities.

[Anonymous No.2464602]

>>2464600
>Can it leads to escalating and unwanted problems concerning one of those 2 devices?
I'm unsure about that, but there should be proper isolation (i.e. transformers, optoisolators, etc) between the AC line and DC circuit so if you did have a problem it wouldn't cause physical damage, only dysfunction. We're talking about ethernet over powerlines, right?

[Anonymous No.2464612]

AC fan is not spinning, just humming. I hope it is a capacitor. Can I temporarily replace it with a generic HV capacitor from an old TV? The run capacitor is dual, I think.

[Anonymous No.2464622]

>>2464612
>replace it with a generic HV capacitor from an old TV
Never tried, but imagine it would burn up because it's not electrolytic. Have you tested the fan coils resistance?

[Anonymous No.2464638]

>>2464622
I spun the fan by hand and it started. The compressor won't come on. So it is most likely the capacitor. Why would it burn out, ESR too high, can't handle the current?

[Anonymous No.2464644]

>>2464638
Maybe it overheated or caught a spike that exceeded its specs. How old is the AC unit?

[Anonymous No.2464649]

>>2464638
There's also a thermal cutoff failsafe on the compressor. If your condenser fan isn't running at high enough RPM it will cause your compressor to overheat and shut down until it cools off. You should still replace the capacitor anyway.

[Anonymous No.2464652]

>>2464644
About 10 yo or so. I meant to ask, what is so special about those electrolytic non-polarized big bulky AC motor run capacitors compared to generic film HV capacitors that are much smaller in size for the same or higher capacitance and voltage rating? I am guessing it is probably ESR/current spike handling capability. If that's the case they will overheat and explode.

>>2464649
The compressor makes that humming/vibrating sound every minute or so as if it tried to start but couldn't.

[Anonymous No.2464655]

>>2464652
Yes, they can handle normal operating surges without failing. The big AC caps are full of a liquid electrolytic and the foil + terminals are thicker. It's housed in a big heavy duty can to keep it cool and protected from the elements.

[Anonymous No.2464680]

>>2464612
>AC fan is not spinning, just humming

90% of the time, it just needs oil, and a bit of rattling to shake off film from old dried oil.

[Anonymous No.2464729]

>>2464680
How do you put oil in it? What kind of oil, canola? Turns out I have both run and start capacitors. Some models only have a dual run capacitor but no start capacitor. Why is that?

[Anonymous No.2464733]

Dropped a wrench and almost hurt this little fella. Can you see him?

[Anonymous No.2464737]

>>2464733
I toad you not to stand there.

[Anonymous No.2464741]

>>2464729
Zoom spout oil is good. There should be a port with a blue plug over it which is the drain, and yellow plugs for filling. If the fan moves easily by hand without scratching or resistance then it doesn't need oiled.

[Anonymous No.2464745]

>>2464741
And you are talking about the run capacitor not the start capacitor?
And yes, the fan moves easily, and also starts if I spin it with a stick.

[Anonymous No.2464749]

>>2464745
I'm talking about changing the oil in the condenser fan motor. BTW, what are the uF ratings on both capacitors?

[Anonymous No.2464762]

>>2464749
Ah I thought you were suggesting repairing the capacitors.
start: 216-259 uF/MFD 330V
run: 35/5 uf 370V

[Anonymous No.2464768]

>>2464762
Was hoping they were close so you could swap them to test. Anyway, you should replace both capacitors after you rule out the possibility of a failing fan motor. Does the motor look like it's been running hot? Peeling, faded paint, darkened wires?

[Anonymous No.2464783]

>>2464768
No, it looks brand new after all these years of hard work. What is also strange is that both the fan motor and the compressor motor fail to start. Perhaps that's an indication that it is the start capacitor that failed. But I don't know if a single start capacitor is responsible for starting both motors?? Yes, I've ordered both caps.

[Anonymous No.2464788]

>>2464578
I want to test a variety of different loads, so I mean ANY load, though chances are that doesn’t include 1hp induction motors or other bulky industrial loads. I guess I’ll simulate the inrush current from a mains filter cap. Maybe I should add a choke in series with mains.

[Anonymous No.2464797]

>>2464783
The start capacitor is meant to make up for sag when the whole system turns on. Without it, everything stalls and burns. The run capacitor just keeps ripple down and everything running smoothly. They work in tandem, so one failure will prevent normal operation of the whole system.

[Anonymous No.2464818]

When setting RC delay values, it seems natural to maximize the resistance and minimize the capacitance to avoid long discharge times when the circuit is turned on and off repeatedly. But is there a practical limit for how high you can go with increasing R? Say instead of 1k 100uf, I'd prefer 100k 1uf. What about 1Meg and 0.1uf? Or 10Meg and 10nf? Are there any issues with noise or whatever?

[Anonymous No.2464832]

why does digikey keep getting worse? how do we stop it?

>>2464818
you don't care about thermal noise so really the only limit is your susceptibility to leakage and loading on the output

[Anonymous No.2464835]

how does the 0.6v reference make the output of the op-amp have 0.6v output when there's no differential applies to its input terminals?

[Anonymous No.2464877]

>>2464835
because of math. do the resistor divider math, setting the op amp inputs equal, and observe the result.

[Anonymous No.2464934]

>>2464835
uhh because op amps act to set their inputs equal if feedback polarity is correct and with no input node P is HiZ and therefore gets driven to 0.6V resulting in an output of the same?

[Anonymous No.2464943]

>>2464934
it's just that i thought some of the 0.6 would reduce by the time it reaches the op-amp due to voltage drop across rf2

[Anonymous No.2464960]

>>2464943
you're literally correct but you have a fundamental misunderstanding. the voltage on the opa's + input is not 0.6V. it varies as a function of 0.6V and the input voltage.

[Anonymous No.2464969]

Does anyone happen to know a good set of plans to make a metal detector? Preferably something fairly powerful/sensitive, I want to do a bit of hobby prospecting.

[Anonymous No.2464983]

>>2464969
a little buzzed here but
>wind a copper coil with a large diameter
>wire a circuit that finds and displays its resonant frequency (could be done with an arduino if it's low enough)
>wave it around and dig where the resonance changes
i have no clue how a real detector works but that's probably how, it'd just have thresholds tied to certain deviations from its open air resonance to make it tard-friendly

[Anonymous No.2464992]

>>2464818
I wouldn’t go above 1M, if only because probing the thing with a DMM or scope starts to impact it. Also becomes prone to radiated electric field noise, so you may want to go down to 100k or even 10k if that’s an issue.

>>2464969
Watch Huygens Optics video on the topic, he shows how the dual-coil circuits act to detect phase and amplitude, so you can tell what sort of metal you’re detecting.

[Anonymous No.2465153]

What's the rundown on power stations?
Is Hobotech a paid shill?

[Anonymous No.2465157]

>>2465153
They're quite handy if you find yourself wanting to operate appliances/power gear in places with no electricity.
They're not very well suited to powering your house during a power outage.

[Anonymous No.2465211]

Absolute retard here, if a circuit says that it needs for example ±24 volts does it mean a positive and negative voltage of each 12 volts or a positive and negative voltage of that are both 24 volts?

[Anonymous No.2465219]

>>2465211

The second one: +24 and -24 for a total span of 48 volts rail to rail.

[Anonymous No.2465224]

One day I will have to give up shitting in pools.

[Anonymous No.2465283]

>>2464797
Turns out it is the run capacitor that failed.
Which makes sense since it works non-stop while his lazy buddy is only needed for a brief second and he goes back to sleep. But he still thinks he is as important. Just like some people at my work.

[Anonymous No.2465306]

Any electric engineers here? What software do you use for electric diagrams?
I've tried Eplan and Autocad electrical, both programs are unintuitive as hell and near physically painful to work with. I use IEC standards btw.

[Anonymous No.2465319]

Should I just use a 4 layer board at twice the cost or just continue the spaghetti?
It really does not like the D2 vias but it's bullshit everything is in spec

[Anonymous No.2465320]

>>2465319
>It really does not like the D2 vias
who cares what it likes? 2 layer boards are easier to bodge.

[Anonymous No.2465323]

>>2464553
look at all that electrical noise in that circuit. no wonder physicist used to 1 off mention pixies and crt demons / gremlins and say things like if you dont know how it works it could work multiple ways at the same time.

the circuit on paper in operation vs what happens in reality in operation. i think current can cross without a circuit connection 13 times on that bad boy in multiple locations

[Anonymous No.2465327]

>>2465320
Well if it just fails to print correctly then hmmm

[Anonymous No.2465380]

bass amplifier with a 3080N opamp in it, why? compressor? no schematic that I could find unfortunately

[Anonymous No.2465387]

>>2465380
Baxandall tone control?

[Anonymous No.2465443]

>>2465387
oh okay that's it
the amp is noisy so I was checking the power supply caps, not sure where else to look. next I'm going to check if the noise is on the line out or only the power amp

[Anonymous No.2465498]

I have this amp that turns on but will not produce any sound. There is a big brown mark on the outside of the case as if something exploded or caught fire on the inside. This one part in particular appears to be destroyed.

What do you think's going on here? Do transistors explode like that? Is that a transistor?

[Anonymous No.2465506]

>>2465306
If the diagram is parts-scale enough to use KiCAD, I’ll use it. Custom symbols or subsheets could make up the symbolic blocks like “VCO” or “mains filter”. Not that ideal, but it works. Never tried actual diagram or flowchart software, but I think there’s 1 or 2 FOSS examples out there.

>>2465319
Depends, how robust is your ground plane going to be, and how robust does it need to be? Also that’s an ass pinout, don’t suppose you can swap it about (e.g swapping op-amps or logic gates within an IC)? Not familiar with that connector but I guess it’s unlikely that you could.

[Anonymous No.2465507]

>>2465498
I looked up the service manual for this amp, Onkyo HT-R430, definitely it is a transistor and it does look like it caught fire. I think that implies that a larger transistor earlier in the circuit failed?

[Anonymous No.2465510]

>>2465498
The part may be fine, hard to tell from here. A brown mark is usually indicative that something was running too hot without sufficient cooling for a while, and the temperature degraded things and turned them brown. That could be a symptom of something like a transistor or resistor failing, or maybe it was caused by placing the amp somewhere with insufficient ventilation, and the resultant heat could cause damage to the high-power parts like transistors and resistors.

[Anonymous No.2465513]

>>2465510
What do you think I should go looking for to figure out what broke? I think the fuses are intact, the caps look fine.

[Anonymous No.2465521]

>>2465506
Literally cannot, it's a usb c connector. They are all wired like that.
I've pretty much done it using 2 side pcb but it's a bit of a disaster now
Once everything is accounted for I'll clean it up
Four bloody length matched differential pairs. Fucking terrible

[Anonymous No.2465563]

>>2465521
Don't route that TX/RX bus on the underside do limited crossovers only on 2nd layer then route top on 1 layer and finally pour GND on L2. Debugging ground-related issues is a nightmare.

Also. Don't put vias under the connector. They will short in hand soldering.

[Anonymous No.2465568]

>smart ass Russian shit
would there be any way to modify this circut to be able to use 8Ω speakers instead of 32Ω ones? i also made the standby current lower in LTspice by swaping out the 1meg resistor with a 5.1 meg one

>Speakers should have an impedance of at least 32 Ohms.

[Anonymous No.2465571]

>>2465568
and the simulation, because 4chan doesnt have multi file upload

[Anonymous No.2465572]

>>2465568
Add 24 Ohms of resistance in parallel with the speaker.

[Anonymous No.2465574]

>>2465572
wouldn't it still need the extra coil length cense ether or speaker would be being used as a microphone?

[Anonymous No.2465575]

>>2465513
Look for discoloured parts of the PCB.

>>2465521
0-ohm link resistors (;

>>2465572
>parallel

[Anonymous No.2465577]

>>2465574
yep, I'm dumb.
>>2465575
gimme a break

[Anonymous No.2465588]

The NE567 seems underrated

[Anonymous No.2465598]

>>2465588
>NE567

surprising: declared obsolete in 2011 yet still for sale in quantity at Digikey and Mouser.
i made my first IR remote control using one of those.

[Anonymous No.2465618]

>>2465588
Looks really neat. Maybe as neat as the CD4046.

[Anonymous No.2465638]

>>2465630
They'll have a positive tempco, so parallel is going to mean more consistent results between them, but it does probably mean needing a ≤6A 5V PSU.

[Anonymous No.2465651]

>>2465638
>30w
shit.. I am going to try to get away with using 5w and if that does not work maybe get it running on 30w with a PWM

[Anonymous No.2465657]

>>2465651
Hard to say what the actual power output is without measuring while one is turned on. If you measure the DC resistance of one chances are it will be similar to P=V^2/R, but derated by maybe up to 25% due to the positive tempco. Both require that you own one.
In general, a USB heater is going to max-out near that 10W limit per port because 10W is fuck-all anyhow and you need all the power you can get.

>30W with a PWM
Putting them in series might be fine for balancing (gets worse with more extreme tempco), that way the total power would be probably 3W or less. The tempco issue is, if one element is at a slightly higher temperature than the others, the voltage across it will increase due to the tempco, causing more power to be dissipated in it, causing its temperature to increase, and so-on. The same issue that happens to LEDs run in parallel on a constant-current source. Because thevenin and norton were both actually kirchhoff in disguise.

How much thermal output do you need anyhow?

[Anonymous No.2465670]

Should I care about isolation distance between neutral and ground? It's on a 3-pin device with no chance that live and neutral have been swapped.

[Anonymous No.2465676]

>>2465670
>Should I care

life is better when you dont.
and in this case you have no reason to care as neutral and ground are connected together at the breaker box.
it's absurd to keep things apart who are stuck together.
dont believe me? take a resistance measurement at any 120V power outlet.

[Anonymous No.2465680]

>>2465676
I know they're connected together mate. My only concern is if a short between them or other fault could cause current to flow through the ground path. But since that trips an RCD anyhow it's probably fine.

[Anonymous No.2465695]

LED bulb was flickering and really dim, opened it up to see what was going on. It's a standard rectifier, HV LED string, and linear current regulator design. In this case it has 9 LED chips, and the voltages across them were as follows:
>30
>15
>30
>120
>30
>15
>30
>30
>15
The three 15s were dimmer, having only one of the two bright squares under the phosphor glowing brightly. I suspect the death of one of the sub-chips in these LEDs increased the voltage across the linear current regulator, killing it from heat or overvoltage. Assuming all the chips are meant to be 30V, the result would be 270VDC across them and 50V across the regulator chip (230V mains). I'd expect the chip to be 400VDC rated and have thermal throttling, but maybe it sat too hot for too long before dying. Now theres basically zero volts across the linear regulator and 120V is being dropped by an LED chip. A still functional LED chip. And the resultant current is really low. I don't really understand how it can still be glowing just fine even though there's 120V across it, without either going open-circuit or short-circuit.

I have some LEDs I could try to bodge into the board to replace the existing ones, think they're 30VDC, but man I hate soldering on these aluminium boards. I'll just toss it, or maybe keep it as a flickering candle lamp.

[Anonymous No.2465703]

>>2465657
>thermal output
nothing crazy, just enough to keep a surface slightly warm, not cool. not trying to get anything hot
>LEDs run in parallel
it is better to run them in series no matter what?

[Anonymous No.2465707]

>>2465703
>nothing crazy, just enough to keep a surface slightly warm, not cool.
If you underdrive them like putting three in series would do, and don't mind ~±20% heating variability, you should be fine.
>it is better to run them in series no matter what?
Without any kind of PTC thermistor or current regulator to balance their negative tempco, yeah definitely. Never actually seen anyone use PTC thermistors to balance LED strings, though IIRC using current mirrors counts since effective hFE decreases with increasing temperature.

[Anonymous No.2465739]

>>2464553
Not much of electronics work but any ideas for an audio connector in place of Y split? I want it to be compact, 2.5mm jack would be too bulky probably. The reason is that I want to make Lightning cable for my headphones but would like to avoid remaking everything when chink Lightning DAC/cable gives out.

[Anonymous No.2465868]

>>2465306
My notebook or KiCAD, for circuit simulation is use CircuitJS.

[Anonymous No.2465869]

>>2465868
Or my breadboard if it's some simple circuit

[Anonymous No.2465874]

>>2465498
If you decide to replace it make sure you check the surrounding components to make sure they weren't damaged. Quite often transistors take out other components with them when shorting out.

[Anonymous No.2465904]

>>2464553
Which of the books in the op post should I get if I want to learn both analog and digital electronics from the ground up?

[Anonymous No.2465909]

>>2465904
Practical Electronics for Inventors is very good, I'd recommend editon three which you can find online as a pdf or even better order it paperback, teaches you a lot about analogical and digital electronics and goes indepth about a lot of interesting stuff.
For circuit diagrams, earch up "talking electronics" and get all the pdf's on his site. Pay special attention to the Transistor Amplifiers P1 and P2 as those are very intuitive books and they have no complex math or Khirchoff's laws in them. Everything is explained very nicely with lots of circuit diagrams and practical examples which you can build very easily.

[Anonymous No.2465910]

>>2465909
Is there a reason why I shouldn't just buy the newest edition on amazon?

[Anonymous No.2465922]

>>2465910
No, buy the newest edition, I just couldn't find the pdf for the 4th edition so I couldn't tell you about my personal experience with it.

[Anonymous No.2465995]

>>2465739
If you just need a mono connectors for each side, then MMCX or similar would work. If you need a 3-terminal plug for both though I’m unsure, draw diagram?

I feel that anything other than 2.5mm jacks would be too flimsy regardless.

[Anonymous No.2466014]

>look ma! no transistors!

[Anonymous No.2466087]

Only tangentially related but

>1st year Apprentice
>4th day of class
>Class consists of 2.5 hours of sitting in near silence on a laptop, doing quizzes on different types of screwdrivers, laser tools, ladder safety
>Teacher gives us a pop quiz first thing today
>It's one of those "tests" that test your reading ability, 29 questions of nonsense (draw a circle in a square, yell out your name, etc etc), question 30 says to fold your test in half and flip it over
>Pretty sure I took this in 6th grade
>Fold test in half, waiting
>After a minute or so, almost every student in here starts getting to the questions that say to say something stupid out loud
>One after the next, saying "I am halfway through the test" or "I have followed all instructions carefully"
>Teacher doing his best to hold in his laughter
>Test over
>6 out of 30 actually "passed"
I'm a retard when it comes to the trade but Jesus christ

[Anonymous No.2466124]

>>2466087
Only like 20% of Americans are literate beyond a highschool level.
60% can't follow ikea directions with big pictures.

[Anonymous No.2466159]

>>2466014
all built up, had to substatute in a 2.2μ cap

[Anonymous No.2466170]

>>2466159
odd how both electrodes are glowing, i thought that only happened with ac

[Anonymous No.2466181]

Im reading a surface mount resistor
R050 F A1703
what does this mean?

[Anonymous No.2466182]

>>2466170
the electrodes are on their side to reach the red cap and the 10k resistors on the top

[Anonymous No.2466188]

Things I forgot about breadboarding but relearned today:
>most of your issues are caused by shitty connections
>the rest of the issues are either incorrect wiring or incorrect circuit design
>the glue remnants on the ends of THT resistors IS significant and WILL mess your day up
>it can be very frustrating figuring out what's causing what error, even with a scope, because probing will move the component leads and make a lot of connection issues intermittent
>it is somewhat difficult to stop long component legs from shorting to one another
>it is rather rewarding to get a circuit working properly, and the confidence that your circuit should work first-time on a PCB is comforting
>there is no good way to turn on a BJT from an open-collector TTL comparator without somehow influencing the comparator's output voltage
>small-signal THT MOSFETs exist
>don't forget to diode-clamp high-pass capacitor outputs
Very nice, I never want to do it again.

[Anonymous No.2466192]

>>2466188
Here's the schematic I was breadboarding (mains quick-test), the bottom half in particular. The AC current signal goes through a HPF to bring it up to a 2V reference or so, then gets amplified and rectified with a single op-amp, then put through a LPF to get the average current value. Then it gets compared to a reference value, and shunted through to an SR latch made with a comparator. If the average current surpasses a potentiometer-set threshold, the latch will trip and turn off the output. Only when the load current drops again, and the button is pushed, will the load be able to turn on again. The load is turned on using a 2N7000 that activates a relay, which itself can only turn on when both reed switches are active. The reed switches are activated by magnets on the case, enforcing that the case is closed.

Also somehow I couldn't find any LM358s, so I used an LM324 for the breadboarding instead. Should be basically the same.

[Anonymous No.2466208]

>>2466188
damn dude you're supposed to trim the resistor leads. the glue should never be a problem because you should always be cutting that part off

[Anonymous No.2466209]

>>2466192

this ugly schematic style just doesnt seem to be getting any better.
- the components should be the star of the show, not the overly-long overly-bright puke-green lines.
- the number of ''connections by label'' should be reduced by 96% so people dont spend 5 minutes connecting the dots
- that full-wave rectifier is an obscenity. there's a billion examples of how to draw it right if you bing it.
see example of how sane people do it.

[Anonymous No.2466242]

>>2466209
By “connections by label” do you mean just the labels, or the power tags as well? Because I think power tags are better left as such. Actual rails on either side of the diagram don’t actually add anything of substance, and aren’t any clearer or less-clear. The signal labels I agree, but middleearth is just a virtual ground rail so it also counts as a power tag, trying to tie it in to multiple parts of the circuit with traces would make spaghetti.

I’ll swap the thing to an A3 instead of A4, that way I’ll have more room to widen the bridge rectifier and can add the current sensing line to the front of the analogue signal path.
The colour scheme is a work in progress, a partial hold-over from the original light scheme from KiCAD, think maybe I’ll swap it over to be the same as my LTspice scheme, with white components and traces, and red highlights.

Your pic is generally pretty good, a bit unnecessarily zig-zaggy, but I’m not sure how I feel about signals being rouged lower than the GND rail. Also are those power nodes labelled “Uπ”? Resistor symbols as boxes with zigs in them (but not the trimpot) feels like a strange compromise.

Thanks for the criticism though.

[Anonymous No.2466278]

>>2466209
>>2466242
Ok I think I'm happy with this. The reed switches aren't actually on the PCB though, they'll be glued inside the case, hence not connecting it originally. But since I'll be making the board by hand, having DRC alerts isn't a problem.

[Anonymous No.2466291]

>>2466278
did you trim the leads on your breadboard yet?

[Anonymous No.2466292]

>>2466291
No I'm putting the resistors back where they belong, untrimmed. I just washed them with IPA before use. I'll trim them when I put them in a PCB.

[Anonymous No.2466294]

>>2466292
>washed the resistors
I love you but you are funny

[Anonymous No.2466297]

>>2466292
when you make loops like that with the components it theoretically makes inductors (bigger the loop) in addition to getting glue in your breadboard holes

[Anonymous No.2466319]

>>2466294
Well more squirted IPA on a paper towel and twisted the legs around with it, seemed to scrape it off real well. I think maybe I’m too autistic to just trim them. I know the breadboard would look a lot nicer with the resistors flush to the board, but cramming them together becomes a lot harder, and so does clipping a scope probe to them. And like, what if I need a really long resistor to span a wider gap? On all of them at once?

>>2466297
It’s a 50Hz circuit, inductance is the least of my worries. And it’s just a test for a PCB, which will have a halfway decent ground plane.

Also the holes in this breadboard are already full of dust and probably some solder beads and at least one has a full-on solder blob dripped onto it. Maybe some RMA too since that shit gets everywhere when you’re washing it off with IPA and a toothbrush.

[Anonymous No.2466344]

>>2466319
>what if I need a really long resistor to span a wider gap
That's what jumpers are for, Anon. Order more breadboards from different suppliers and get rid of the crusty ones you have now so you don't waste more time on troubleshooting bullshit.

[Anonymous No.2466414]

>>2464553
1/2
>be 28yr old embedded developer but can design PCBs as well
>join company as embedded systems engineer
>company makes industrial IoT equipment and runs servers related to that equipment
>mostly write software for them, everything going well
>boss wants a few changes to be made to a product's PCB (aka next revision)
>volunteer to make the changes but boss gets an Indian company to make the changes (they designed the original board as well)
>minor changes, remove x section, add y section, replace z part with x etc, an average guy could finish them in a week or two
>create a nice document with screenshots which highlights all the changes to be made
>send the document to the indians
>next day, receive email saying they've made the changes and want us to review them
>nofuckingway.png
>turns out they only made 10% of the total changes, with many mistakes, the schematics look a mess
>have 100s emails back and forth over 3 months to get them to make all the changes
>they would made a single minor change and then emailed the schematics PDF back with "pls review sirs" for some reason and would NEVER respond to new emails on the same day
>in the process they make random mistakes like, shorting a rail to earth (not VSS, but chassis earth), permanently disable an IC by messing up pullup on enable line (ensuring board would NEVER work), connect a P-MOS backwards, etc
>decoupling is either insufficient (100n on a high current flyback SMPS) or way too much (47000u on level translator IC)
>for some reason they remap every single connection to the MCU to random pins (connected an IO to the power pins kek), nobody told them to do this
>videocall the indian design engineer with boss to talk wtf was all that about
>guy surprisingly knows his shit and doesn't seem like he'd make any of those mistakes, he assures us everything will go as planned with layout
>ohboy.png
>same story with layout, they make basic mistakes

[Anonymous No.2466415]

>>2466414
2/2
>random debugging connector besides the RF antenna section, random connector that is connected to a chip enable line for some reason, no copper keepout on screw holes so the screws would short multiple layers together, all signal traces on the internal two layers (4 layer PCB)
>again have 100s of emails back and forth over a month to get them to fix the stuff they keep breaking and the original changes that were to be made
>videocall their layout engineer with boss, surely THIS man must be off his rocker
>again, the guy is quite smart and even gives us good recommendations/options. Nothing would suggest that all the colossal errors were made by the very same man
>finally get changes made
>only took 5 months more than we expected
I don't even know what's happening anymore. Is this how the "industry" works? I don't believe it

[Anonymous No.2466446]

Why would both circuits work?
Sorry 4chin won't let me post image for some reason I really need help
https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcMBMcUHYMGZIA4UA2ATmIxAUgpABZsKBTAWjDACgA3EJmvEFBIW69wGIVSo0qSCVGgI2AMxCEa-FGtWih2emHlzIKNgHdhfMGNoWrkNgCd++bU77Y0Ufhjtm0bj1ruEqYqakHmIOF2ACbW4Ch8eFRgCfwg0QyKAIYArgA2AC4hSfGJybYhIil8VRVmJZZCfi4+EQJCPP7BZlooGqGRHq0Iqc0jbpBqrTRgTVK0s5GTUCEINGqzVIQYKOCE3RQp4EcIR5srZmsbJ+t7wbFXx7unu6y7uxnZ+UWXqbhqMx0y1a212-wof2BDhUOzuMLBwK800W5xecLsynGrghFn2C3ABhgRk4CzmkhReNk8xknhgCiAA

[Anonymous No.2466448]

>>2466446
why wouldn't they work? gate is below source in both cases.

[Anonymous No.2466452]

>>2466414
>>2466415
we had the same experience using an indian outsource firm for a large amount of braindead mechanical design work. some blessed director finally tossed out all their work after 6 months and had it redone in 3 months by the handful of in-house white people that were previously trying to assist the horde of indians.

[Anonymous No.2466453]

>>2466448
But doesn't a P channel mosfet conduct from source to drain? How could it work if its drain to source?

[Anonymous No.2466455]

>>2466453
think of it as a (very nonlinear) variable resistor controlled by the voltage between source and gate.

[Anonymous No.2466457]

>>2466452
>some blessed director finally tossed out all their work after 6 months
This doesn't happen often enough, lost cost fallacy hits them management types hard

[Anonymous No.2466460]

>>2466455
>think of it as a (very nonlinear) variable resistor controlled by the voltage between source and gate.
Thanks for responding anon
I found the circuit on the left labelled as "reverse polarity protection". I'm just trying to figure out why every circuit uses the left circuit and not the right one if both would work
Also when you refer to MOSFETs as variable resistors, do you mean that because the diodes conduct in the reverse direction or do the channels too conduct in both directions?

[Anonymous No.2466461]

>>2466460
the channels conduct in both directions. you can literally approximate D-S as a resistor.

the body diode is the reason the right circuit wouldn't work as reverse polarity protection. wire the voltage source in reverse and observe that current still flows through the diode even though the fet shuts off.

[Anonymous No.2466462]

>>2466461
This absolutely destroys my understanding of MOSFETs. Thanks anon, I needed this lesson.

>the body diode is the reason the right circuit wouldn't work as reverse polarity protection. wire the voltage source in reverse and observe that current still flows through the diode even though the fet shuts off.
Holy shit you're right!

[Anonymous No.2466467]

>>2464553
A Tesla coil has (+) voltage output.

I need (-) negative voltage output in high volts above 100,000 volts of negative voltage.

Is a Van de Graaff generator my only option?

Thank you.

[Anonymous No.2466478]

>>2466467
doesn't a tesla coil put out AC?

[Anonymous No.2466483]

>>2466478
I'm pretty sure the spark gap and capacitor make it direct current. I'm not 100% sure that's why I'm here asking for help thank you.

[Anonymous No.2466517]

>>2466483
that primary and secondary will only couple AC. there's no such thing as DC coupling via coils.

[Anonymous No.2466520]

>>2466517
>there's no such thing as DC coupling via coils.
I think there is if you turn off the coil, then turn the coil back on. That's not alternating current that's just turning it on and off quickly like a slayer exciter.

[Anonymous No.2466538]

>>2466520
You are correct. DC transformers work exactly like that.

[W Lambright !!XUvGmrfvTTp No.2466541]

>>2466538
>You are correct. DC transformers work exactly like that.

Any ideas how to make high voltage DC current?

[Anonymous No.2466545]

Why do I get the feeling that this is not going to work
This is just the port2 breakout and the ssrx/tx lines done. I havent even wired dp or dpaux or any power, or port1, or the pd controller, or the communications between the pd controller and the mux or all the passives for the pd and the mux or the fucking rom chip for the pd controller

I don't even need to do displayport but I feel like I should anyway in case I want to later try to implement it, this card will be compatible

[W Lambright !!XUvGmrfvTTp No.2466586]

>>2466483
I figured it out I can use a bridge rectifier in reverse.

[Anonymous No.2466615]

>>2466586
suck a bag of dicks

[Anonymous No.2466640]

>>2466586
dicks of bag a suck

[Anonymous No.2466641]

>>2466586
sneed's feed and seed

[Anonymous No.2466646]

>>2466641
Buck's Fishing & Camping

[Anonymous No.2466663]

>>2466181
Probably 0.050ohms, so 50miliohms. I never encountered such markings on a SMD resistor so this is my best guess, measure it with a multimeter to see if it shows you a short.

[Anonymous No.2466799]

>>2466414
>guy surprisingly knows his shit and doesn't seem like he'd make any of those mistakes
My limited experience with indian outsourcing for software dev was the exact same. How it works is they have one competent guy who handles the customer interaction and then a large team of peons with limited skills and experience. The one competent guy handles maybe a dozen different customers while the inept team is "your team". There is also some cultural thing where the peons won't do anything without the head guy spelling out exactly what to do, since thinking on their own is seen to be insulting the head guy. And the head guy won't do any work himself since that's below his class or whatever. So the trick is to get the head guy to spend his time telling the peons what to do for your project instead of all his other customers' projects.

[Anonymous No.2466862]

>>2466520
>>2466538
Switched DC is still AC. The output of any transformer will always have 0DCV bias. It’s possible that one side will have a higher peak voltage than the other due to asymmetry in the waveform, but it’s still AC.

>>2466663
Another reason why everyone should own a milliohmmeter.

[Anonymous No.2466926]

>>2465568
Replace the 100u with a bigger electrolytic capacitor

[Anonymous No.2466941]

>>2465568
A TO-92 isn’t going to drive an 8Ω speaker well at all if only for thermal reasons. A 32Ω driver drops 4 times the voltage as an 8Ω driver, meaning not only will the transistor have to supply more current for the same power output, but also it will have to drop more voltage while doing so, heating up a bunch. So either you’d need higher power transistors, or much smaller power rails.
The current-amplifier topology should mean that as long as the two drivers have the same impedance that you’ll get the right signal-to-signal amplitude with two hFEs chained like that, but that’s just an assumption. Personally I’d alter the amplifier to be more concretely biased and to not pipe DC through the speakers, if possible. A 3PDT switch/relay or even a CD4051 might make doing that easier. So might an op amp-based class-B.

>>2466926
Mate those are both power supply filter caps, they have nothing to do with the output amplitude or power capability.

[Anonymous No.2466959]

> soldering pieces of wire to the mechanical switches' pins to extend them so they pass through the 3d printed holes
> after that, soldering those wires to diodes and other pieces of wire to build a scanning matrix
Why is it so hard, anons? I already have the helping hands, but it's so frustrating when the wire springs off the pin and I get solder on my iron or the pin...

[Anonymous No.2466970]

>>2466959
>wire springs off the pin and I get solder on my iron or the pin...
sounds like pvc insulation, I recommend switching to silicone, its alot more.. limp and won't cause that kind of annoyance.

[Anonymous No.2466971]

>>2466970
That's a good tip, thanks

[Anonymous No.2467000]

>>2466959
Use your benis tweezers.

[Anonymous No.2467038]

>>2467000
Checked. Also,
> load of fugly solder joints because it takes about 20 seconds for heat to transfer on to the tiny 1,75mm pin so I just like tap the soldering wire to the iron and transfer it in the liquid state to the wires (I know, I know)
B-but i-it's a prototype so it must not be perfect, eh lads?

[Anonymous No.2467049]

>>2467038
Anon, you should be tinning the wire and pin before you solder them together. It's just as important to use good flux and solder. Chink flux is made out of slave fetuses and puppies. Chink solder is made from collapsed chink highrise scrap and mercury. It doesn't work, but it's CHEAP!!!

[W Lambright !!XUvGmrfvTTp No.2467054]

>>2466862
>0DCV bias
OK, I never heard of this, I love learning new things.

My goal is lots of DC negative HIGH voltage.

[Anonymous No.2467189]

>>2467054
How high a voltage are you after? CW multipliers or marx generators are an easy way to get to a few hundred or a couple of thousand volts, but if you’re after hundreds of thousands of volts I’d lean towards a ZVS switcher with an extreme transformer ratio, those hints generally output a constant output voltage from what I remember, so no regulation feedback necessary. In addition, for rectification you can buy silicon-stack diodes that can each support up to 20kV each for pretty cheap, which you could stack in series, but note they have 50V forward voltages so it’s not advised to pull significant currents through them. How much power or current do you need?

Tesla coils are too high-frequency to easily rectify.

[Anonymous No.2467221]

>>2467189
Dwayne is building a time traveling hurricane portal. Vote for Dwayne.

[Anonymous No.2467245]

>>2467054

[Anonymous No.2467280]

>>2466014
wat is it

[Anonymous No.2467294]

>>2467245
"SUCK A BAG OF DICKS"

[Anonymous No.2467315]

Why would you ever use a two diode center tap rectifier instead of a bridge rectifier?

[Anonymous No.2467330]

>>2467315
Only 0.7V drop instead of 1.4V drop. So less heat produced, higher output voltage, whatever. Also it would be easier to make multi-rail supplies that aren't on isolated windings. But you're basically halving the output voltage compared to putting both windings in series, which also applies to using two diodes directly off a single winding for double the peak voltage, though I think that configuration has worse ripple than the centre-tap method or a full-bridge.

[Anonymous No.2467364]

wait
SHIT
the copper is on the reverse side, not the front side
and the board i just cut out isn't symmetric

back to the hacksaw

[Anonymous No.2467377]

>>2466545
TI TPS65987? Good luck. I'd recommend more layers than 2. 6 Layer boards are pretty cheap nowadays. Does this have to pass EMI tests? Also group length doesnt matter for usb ss signals, only inter pair length matters. Unless you're using DP.

[Anonymous No.2467382]

i'm trying to understand opamps
so if i feed 1v into one input and connect the output to the second input through a voltage divider which dumps 90% of the 1 volt, it means the opamp has to output 10x more voltage to balance the rails, thus giving me 10x votlage amplification on the output, is that correct?

[Anonymous No.2467384]

>>2467382
The output would be 11v (1/11, not 1/10), but otherwise yeah that's right.

[Anonymous No.2467386]

>>2467382
> is that correct?

it's close. and that's good enough for horseshoes and for /ohm/.
google will set you right, tho, so always ask it first.

[Anonymous No.2467400]

>>2467382
You can calculate any op-amp transfer function really easily. If there's negative feedback (output going to inverting input, or output being inverted and sent into non-inverting input) then the two inputs are going to be forced to be the same by the output, and the inputs don't source or sink any significant current. That gives you the algebraic freedom to design any damn op-amp circuit you want. After you do such a calculation, do a check to see what the edge conditions are, e.g. at what point the output or input clips the power rails (or more accurately, the output and input voltage ranges as specified on the datasheet) to ensure that you won't be driving it someplace stupid.

As for open-loop circuits, or circuits where there's positive feedback, you only need to consider the two situations where the output is at its maximum and minimum values. From there you can work backwards by setting the inverting and non-inverting input voltages equal to see what thresholds it will change at for both conditions. Without feedback this threshold will be the same for positive-going-negative as for negative-going-positive, but with positive feedback (hysteresis) those two thresholds will be different.

With some practice doing that, you'll be able to analyse arbitrary op-amp circuits, circuits with all sorts of strange transistors and diodes and the like thrown in the feedback loop. But other things to watch out for include input offset voltage and bias current, slew rate, and gain bandwidth product. Also preventing oscillation, which can happen when there's too much delay in a negative feedback loop that otherwise has no gain limitation.

[Anonymous No.2467410]

>>2467049
It's already ordered. I guess I fucked up for not getting 60/40 solder, but whatever, the worst soldering work is already behind me. Only need to solder the knots on the rows and diodes to columns but that's much less hassle.
Also, any idea how to insulate wires so to say post hoc, i.e. after its all laid out? (Super noob here). I reckon hot glue / silicone gun is a bad idea?

[Anonymous No.2467412]

>>2467410
>any idea how to insulate wires
Wait until you know it works before you start coating everything with goo. Need to see a pic to better understand what's going on, but heat shrink is commonly used for insulating wires.
pro-tip: Hot glue is easy to remove with isopropyl alcohol.

[Anonymous No.2467437]

Should I?

[Anonymous No.2467502]

>>2467410
use kester solder and find a used old hakko iron on ebay, youll be surprised at how much easier everything becomes

[Anonymous No.2467547]

I've been invited to join the electronics team for my university's ion propulsion lab.
I need to learn power electronics fast. Do you guys have any textbooks on power electronics you would recommend?

I also need to get better at control theory. I've taken a class on linear control systems and already have Ogata's book. Do you guys know of anything else to help me get a leg up?

[Anonymous No.2467566]

>>2467547

it's pretty obvious you werent asked to join for your abilities. so forget books,just show up in tight pants and do the needful.

[Anonymous No.2467590]

>>2467566
you can just say you have no idea what I'm talking about, I'm not going to judge you

[Anonymous No.2467595]

>>2467590
okay sheldon

[Anonymous No.2467601]

>>2464612
>>2464622
>>2464638
>>2464680
>>2464729
>>2464741
>>2464745
>>2464749
>>2464762
>>2464768
>>2464783
>>2464797
>>2465283
>put some oil on it

lol.

I'm a licensed master electrician.

Replace the motor and the capacitors at the same time otherwise you will be replacing the capacitor again soon. If the motor is serviceable then replace the bearings. Putting oil on the shaft is fucking retarded and won't actually do anything at all other than make a mess.

The bearings are probably shot on the motor. Some oil might get it moving temporarily but the increased friction will just kill the new capacitor, too.

The capacitor died because the motor is failing. The motor didn't just stop because of a bad capacitor.

[Anonymous No.2467604]

>>2467437
>Should I?

Maybe. I was paid to crimp 2 wires together like that and tape them up a few days ago. Something chewed through a heat-trace and I had to terminate it again. Worked perfectly.

Alternatively, I have done this hundreds of times to temporarily close a dead end switch loop.

Fight me.

[Anonymous No.2467613]

>>2467601
>I'm a licensed master electrician.

i've dealt with you bastards a few times before.
the solution to every problem is to replace everything with new parts to pad your profits.
had to fight for months with a company that refused to replace 2 pumps which started leaking oil while under warranty.
they blamed the water! ''the fucking water hurt the poor pumps, so the warranty doesnt apply"
evil jew cock-suckers!

[Anonymous No.2467615]

>>2467601
I'm a licensed massage therapist and you're wrong.

[Anonymous No.2467617]

>>2467613
>replace everything with new parts to pad your profits.

anon, the bearings are sealed. Are you going to separate and grease the inside of the bearings? Go ahead. Other wise replace the bearings if they are serviceable, if not, replace the fan motor.

Do you think it's cheaper to buy 2 capacitors and a motor or 1 capacitor and a motor?

[Anonymous No.2467620]

>>2467617
Did you go through with the surgery yet, Dwayne?

[Anonymous No.2467624]

>>2467601
>The capacitor died because the motor is failing
Explain. Can't capacitors just fail randomly after 10-15 years? I've seen a lot of videos where people just change old capacitors and everything seems to be fine.

>The bearings are probably shot on the motor
Wouldn't that cause a lot of noise? And it spins freely if I spin it by hand. And I don't hear any unusual sounds.
Also, this capacitor is used by two motors: the fan motor and the compressor motor. I am assuming you mean the fan motor? Is it more likely to fail of the two?

[Anonymous No.2467629]

>>2467624
>everything seems to be fine
>on the video

Sure, the motor will run until the new capacitor fails prematurely.

>Wouldn't that cause a lot of noise?
No, not at all. The bearing hasn't had a critical failure, but it has increased resistance which requires increase power to move the fan, which puts increased stress on the running capacitor.

Simple as.

[Anonymous No.2467630]

>>2467613
How hard is your water?

[Anonymous No.2467633]

>>2467630
>How hard is your water?
about 6 inches but I tell girls it's 8. They can't tell the difference.

[Anonymous No.2467639]

>>2467624
>. I am assuming you mean the fan motor? Is it more likely to fail of the two?

Yes, absolutely. Compressors are generally sealed units. I've seen compressor motors run for 50+ years without an issue.

Spinning a big fan blade can put a lot of stress on the bearings, especially when dust or debris builds up on the blades throwing them slightly off balance.

[Anonymous No.2467641]

>>2467629
>puts increased stress on the running capacitor.
what stress specifically? does the stalling motor draw more current? but how would that affect the capacitor? does it charge/discharge more frequently, or that causes an increase in the ripple current through the capacitor or what?

[Anonymous No.2467642]

Hello, /ohm/ I am tearing down an older 3d printer of mine, and I want to reuse the power supply, thing is when I removed it off the frame someone mentioned mount power supply to the frame was giving it a proper grounding, I verified this by doing continuity from mains ground to one of the two bolt holes that were originally used to mount the psu to the frame, so I ran wires from two bolt holes on the psu to bare metal on the frame to keep it properly grounded.

How would I go about this if I am just wanting to repurpose the power supply to have different connectors to power devices, like xt60 connector, barrel connector, etc..

>pic related on some specs of the psu I have, this was found on the internet but mine says the same thing after personally verifying.

[Anonymous No.2467643]

>>2467641
Yeah he means more current, the effective impedance of the start/run winding decreases, meaning the current through the winding and the start/run cap.

>>2467642
Put it in a nice box with some quality high-current banana jack binding posts on the front. Then make banana to XT60 etc. pigtails. Also consider some sort of soft-start connection, since anything that runs on an XT60 off a lipo usually has its own big caps and will draw high inrush current.

[Anonymous No.2467644]

>>2467641
>what stress specifically?

oh for fucks sake, anon, figure it the fuck out.

>be 100w motor
>bearings are fucked
>drawing 120w to push this stupid fan blade
>everthing gets hot AF
>higher hear = higher resistance = higher voltage
>this stupid little capacitor can barely keep up
>MFW it fails

[Anonymous No.2467645]

>>2467643
>Put it in a nice box with some quality high-current banana jack binding posts on the front. Then make banana to XT60 etc. pigtails. Also consider some sort of soft-start connection, since anything that runs on an XT60 off a lipo usually has its own big caps and will draw high inrush current.

I am looking around for something to 3d print, but I am just concerned on properly grounding it.

[Anonymous No.2467646]

>>2467644
>figure it the fuck out.
well you are the master, and this is /ohm/.
can't you help me figure it out.

>higher resistance = higher voltage
why? i thought input voltage is constant so higher resistance means less current.

[Anonymous No.2467647]

>>2467645
>I am just concerned on properly grounding it.

grounding is a bit of a fetish for noobs.
generally speaking you DONT want your power supply to be grounded. you want it to float.
photo shows there's a ground terminal, but it's not connected to minus.
if so, it would be bad news, for example, if you connected 2 supplies in series.

[Anonymous No.2467648]

if you are into linear power supplies and need a small cheap transformer, you can buy a doorbell transformer on ebay. but there is one caveat: apparently all of them have a fuse built into the secondary winding! i was testing it and heard a gentle pop. and sure enough, i was pumping 5a through a 2a transformer and blew a fuse.

[Anonymous No.2467649]

>>2467647
Okay so if I 3d print a cover where I can mount xt60, dc battel, etc, I wouldn't have to worry about risk of shocking myself? if mains isn't grounded to the connectors?

[Anonymous No.2467650]

>>2467649
>I wouldn't have to worry about risk of shocking myself?

not unless there's some awful wiring error inside the supply.

[Anonymous No.2467651]

This anon >>2467648 fellatiates fuses.

[Anonymous No.2467652]

>>2467646
>why? i thought input voltage is constant so higher resistance means less current.

input voltage is constant. What do you think the capacitor is doing?

[Anonymous No.2467654]

>>2467652
>higher resistance means less current

higher resistance = higher voltage drop on the conductors

The capacitor is now overworked to make up the voltage for the ever increasing resistance of the overheating conductors

[Anonymous No.2467655]

>>2467648
>you can buy a doorbell transformer on ebay

Anon, you can guy doorbell transformers at any hardware store. You didn't blow a fuse. You burned out the winding. Basically the same thing, but not serviceable.

[Anonymous No.2467660]

>>2467650
thats the point im trying to make, the power supply case itself is connected to mains ground.

Here is the pic I just took, the power supply used to be secured to the 3d printer frame where I have the two wires coming off of it. but a mod I did I had to remove the power supply off the frame, so I ran wires, and secured it to bare metal some where else on the frame.

Now I am wanting to know how do I or do I need to address this if I want to use the power supply by it self.

[Anonymous No.2467663]

>>2467660
kek

i bet someone added those after getting shocked

[Anonymous No.2467667]

>>2467660
>how do I or do I need to address this

The ground connection on the terminal block right next to the L/N terminals is (should be; check with multimeter because China) connected to the case.

[Anonymous No.2467670]

>>2467663
I added the wire, but never got shocked, I just read that the power supply needs to be grounded to the 3d printer frame, so I ran wires as a lazy solution. I never got shocked or anything, I am just wanting to prevent it.

[Anonymous No.2467675]

>>2467667
Correct, continuity test verified this.

so I guess that means im okay then? Now I am wondering why people are so adamant to ground power supply to a 3d printer frame.

[Anonymous No.2467679]

>>2467675
noise?

[Anonymous No.2467680]

>>2467679
to prevent electrical infetterence

[Anonymous No.2467684]

>>2467679
I guess, maybe I'll 3d print something like a handle that acts like a insulator I guess, thanks, Anon!

[Anonymous No.2467685]

>>2467675

As long as the ground terminal is hooked up to the ground wire on the mains cable, that's all you need. Anyone going out of their way to ground the frame of the printer just doesn't know why they're doing what they're doing. The only reason you'd actually need to do so is if you have something powered directly by mains on the printer, like some heated beds.

[Anonymous No.2467686]

I saw Michelle Obama at a grocery store in Hell yesterday. I told him how cool it was to meet him in person, but I didn’t want to be a douche and bother him and ask him for photos or anything. He said, “Oh, like you’re doing now?” I was taken aback, and all I could say was “Huh?” but he kept cutting me off and going “huh? huh? huh?” and closing his hand shut in front of my face. I walked away and continued with my shopping, and I heard him chuckle as I walked off. When I came to pay for my stuff up front I saw him trying to walk out the door with like fifteen sacrificial newborn babbies in his hands without paying.

The girl at the counter was very nice about it and professional, and was like “Sir, you need to pay for those first.” At first he kept pretending to be tired and not hear her, but eventually turned back around and brought them to the counter.

When she took one of the babbies and started scanning it multiple times, he stopped her and told her to scan them each individually “to prevent any electrical infetterence,” and then turned around and winked at me. I don’t even think that’s a word. After she scanned each babby and put them in a bag and started to say the price, he kept interrupting her by yawning really loudly.

[Anonymous No.2467689]

>>2467684
np, anon. get some cheap cc/cv buck converters so you can have adjustable outputs with limited current. some have integrated meters so one less thing to add.

[Anonymous No.2467692]

>>2467655
>You burned out the winding
Lol. You don't know what you are talking about. Straight burning out the winding would require an insanely high current. And in practice you wouldn't be able to push that much current through a small transformer even by short circuiting it because the primary has high resistance. And if you short circuit a beefy transformer, you will just trip the circuit breaker. No way. Plus I know for a fact that it has a fuse. I removed it. ffs
>not serviceable.
lol. it is blasphemy to post this on a DIY board.

[Anonymous No.2467693]

>>2467685
Sorry anon I want to verify if I am retarded or not, I ran a continuity earlier between V- and main's ground and they don't seem to be connected, I guess pic related is would be fine or would I be king of the retards?

[Anonymous No.2467694]

>>2467693
Keep them separate.

[Anonymous No.2467707]

>>2467684
>I'll 3d print something like a handle that acts like a insulator

there's nothing inherently wrong with having a grounded case.
but if the negative DC terminal is also connected to ground, that can cause problems (*) in some cases.
also, if you sit the power supply on top of another instrument which is metal, that creates an undesirable connection thru the cases which can cause the same problems(*)

(*) an example of such a problem:
you sat your power supply on top of your scope, the cases touch, you try to measure a negative DC voltage from the supply using scope probes, that would short the + of the supply thru the ground connection of the scope.

[Anonymous No.2467709]

>>2467654
Jesus no, the conductor resistance doesn't increase significantly at all. The issues is when a motor moves slowly, there's less back-emf. When it's stalled, there's none. So the only thing stopping current during a stall is the DC resistance of the motor winding (usually fuck-all) and the reactance of the capacitor. The faster the motor, the higher the back-emf, and so the less voltage gets dropped across the capacitive reactance and ESR.

>>2467693
To be honest I'd add a switch between them. So when I care about being a safety bitch I can have them connected, and it stops me getting little buzzes from capacitive coupling from the output. When I want to add power supplies in series or otherwise want them floating, I can turn the switch off and get all the range I want. So long as you remember which way the switch means.

[Anonymous No.2467714]

>>2467694
aight

>>2467707
right, thanks anon

>>2467709
Thanks anon, I may forget switch direction, so I may not add one at all.

[Anonymous No.2467716]

>>2467714
use an LED indicator with labels if you're worried

[Anonymous No.2467718]

NEW IDEA: Harmonic Lock guitar pedal
All the PLL-based guitar stuff I've seen use a phase-frequency comparator to achieve a perfect lock with no phase offset, but the CD4046 has that other XOR gate phase comparator that's got shitty locking range and will often lock onto harmonics. Well I figure that could actually sound pretty neat in a guitar pedal, you'd have like 4 CD4046 PLLs running off the input, each with a different capacitor and a potentiometer to adjust their centre frequency, so as you change a note each PLL locks to a different harmonic of that note. Then I'd have a filter on each VCO to dampen the harsh square harmonics, an AGC circuit to ensure the output amplitude of each VCO is about the same amplitude as the input, a volume pot for each, then combine them in a summing amplifier.

And even better, instead of summing it to the original signal, feed it into an open-core inductor to use as an e-bow, and feed those strange harmonics back into the guitar strings. I wonder what putting a BBD delay in series with a standard e-bow would do?

[Anonymous No.2467724]

>>2467718
Look at this guy drinking out of cups. No way.

[Anonymous No.2467758]

>>2464553
why are all thermal camera modules behind inquiry walls. just let me buy an over priced sample.

[Anonymous No.2467763]

>>2467758
Why are you a straw purchaser of military gear as an agent of Rus/Chiner/Iran/best Korea/[insert bad man here]? I think you're a national security threat. What's your social security number and date of birth?

[Anonymous No.2467767]

>>2467763
sadly i am not a russian chinese iran or north korean. i just want to experiment with it. i could flex my companies email to get it (probably for free too) but i dont feel like breaking my companies policy just to buy a camera module.

[Anonymous No.2467774]

>>2467724
huh?

>>2467758
>>2467767
Even the ones on Digikey or other distributors? You may have to buy a thermal camera product and take it apart.

[Anonymous No.2467777]

>>2467774
The ones on digikey are 3 pixels and garbage.
https://www.yangdaonline.com/sky-eye-duo-pro-3-axis-drone-gimbal-for-flir-duo-pro-r-thermal-camera/
I found a Chinese supplier
their website is very fishy I feel like if i order something from their site i am going to be put on a list then my credit card info stolen.
definitely since their drones are all advertised with a payload number.

[Anonymous No.2467782]

>>2467777
Quads of truth.

[Anonymous No.2467784]

>>2467777
>The ones on digikey are 3 pixels and garbage
Those are just PIR sensors. Actual thermal imaging camera sensors may be in a different section.

[Anonymous No.2467793]

>>2467774
Not my chair, not my problem. That's what I say.

[Anonymous No.2467825]

>>2467709
>Jesus no
>The issues is when a motor moves slowly
>The faster the motor
>the less voltage gets dropped

lol.

So, what you're saying is, when the bearing starts to fail, and the fan is moving slower, more voltage is being dropped, and the capacitor is making up the difference until it fails?

Thanks for clearing that one up, poindexter.

[Anonymous No.2467834]

>>2467825
See this? It's why motors draw more current when they're stalled. During stall, there's zero emf at all, so you've got the full 120VAC across the series capacitor and winding. During steady-state operation without any load torque at all, the back-emf theoretically gets up to 120VAC and draws no current (torque is directly proportional to current just as voltage is mostly proportional to speed). In practice this never happens since friction also increases with speed, but the running current of low-load motors can get pretty low compared to what they draw on startup. If during a partial load condition 60V is being dropped as back-emf, then the remaining 60V (± power factor) is left to be dropped across the capacitor and ESR of the winding, which is usually far lower than the capacitive reactance.

Do the calculation yourself, measure the DC resistance of a motor or find it on a datasheet, and compare that to the reactance of a ~30µF cap at 60Hz. Compare that to the uprated resistance of the copper windings at 150C or whatever it is that enamel insulation is rated to before it breaks down. No way in hell does heating up a winding cause that kind of significant change in ESR.

[Anonymous No.2467847]

>>2467652
the capacitor is a device that stores energy.
if there is no resonance (no inductance in series or in parallel), then it will charge up to the input voltage. and the discharge and charge again, like a buffer. why would it charge higher than the input voltage? are you saying wear and tear in a motor can kill a capacitor by rising the voltage across it?

[Anonymous No.2467848]

>>2467654
i don't know about a motor, but let's say you have a rectifier and a filter capacitor and a load in parallel with it. if the load resistance increases, the voltage will remain the same but the current will decrease which means less work for the capacitor as it will not be drained as quickly.

[Anonymous No.2467852]

Is this what we are talking about?

[Anonymous No.2467857]

>>2467847
Not him, but the capacitor is acting as a reactive element in order to add a phase offset to the current going into one of the windings. Kinda like a buffer, but also not, since it's in series instead of parallel. The voltage across it won't be too big of an issue, rather the increased current through it will. If you increase the torque load of a motor, that increases the current required by the motor. This higher current will drop a voltage across the ESR of the capacitor, which while reasonably low, will cause it to run hotter than designed and could lead to premature failure. Being clamped to a hot motor may also be a significant contributor to this.
Arguably running close to the rated voltage for a longer amount of time will also lead to a premature failure, but since they can get kinda warm when running normally I'm guessing it's a thermal issue.

>>2467852
Yes, though capacitor ESR is also something to consider.

[Anonymous No.2467865]

What do you need to start programming an ATTINY85? Is there any recommended literature or equipment? Won't be my first micro controller project.

[Anonymous No.2467870]

>>2467601
i dont think that necessarily always has to be the reason an old electrolytic cap would go out. couldnt it just be be cause electrolytics have a finite life span?

[Anonymous No.2467872]

>>2467777
witnessed

[Anonymous No.2467888]

>>2467865
Hardware-wise, you can use an Arduino as ISP pretty easily, but I recommend spending a few dollars to get a USBasp clone instead. Don't use V-USB programming, it's a hacky mess that uses up a good fraction of progmem. You may want to program your MCUs before they get soldered in, but generally I'd recommend putting an ICSP header on the board, and messing about with the peripherals to ensure that the ICSP programmer won't be interfered with by anything sharing its pins.

As for literature, read through the datasheet. Familiarise yourself with its registers and peripherals. Then have a look at existing projects to see how they manage to make the most out of a finite number of pins and small amounts of memories. Ben Heck does some interesting stuff with ATtiny microcontrollers, so have a look at his videos for tips. Mitxela also, but he often uses them for V-USB projects, for which I recommend using a mega8U/16U/32U series micro.

[Anonymous No.2467916]

how do i discharge caps without zaps?
i need to use some juicy caps but i can't discharge them with a screwdriver like big clive does because the zap is terrifying and loud and it gives me ptsd

[Anonymous No.2467917]

>>2467916
Resistor. Pick a value such that the time constant R*C is around 1 second. May need to use a high-power resistor if the instantaneous power would get it up above its safe operating temperature.

[Anonymous No.2467927]

>>2467916
>discharge caps without zaps?

i use a wet kleenex.
if dont have one on hand, i'll use the inside of my foreskin, which is usually moist so it's a good conductor.

[Anonymous No.2467936]

>>2467927
>if dont have one on hand, i'll use the inside of my foreskin, which is usually moist so it's a good conductor.
Fucking keke anon I laughed so hard a bunch of tonsil stones came out

[Anonymous No.2467944]

feels kinda silly to leave an etch going overnight...

[Anonymous No.2467954]

>>2467944
Nevermind, it's almost done, should be finished within an hour.

[Anonymous No.2467992]

>>2467944
>>2467954
How did it go, Anonymous?

[Anonymous No.2468045]

hello, brainlet here.
I bought an old vintage amp but the lamps are dead on them. i read somewhere that you can replace them with LEDs in serial with 3k3 ohm resistors in serial. What kind of LEDs do i need power wise? The original lamps are P/N AEL-075 at 8V/50mA.

[Anonymous No.2468064]

>>2467834
>reads a simple explanation
>autistically explains the same thing in more detail

Anon, why are you acting like such a massive retard?

You're literally the only one talking about the windings.

>higher resistance = higher voltage drop on the conductors

This statement is an indisputable fact.

>The capacitor is now overworked to make up the voltage for the ever increasing resistance of the overheating conductors

This statement is also an indisputable fact.

>No way in hell does heating up a winding cause that kind of significant change in ESR

Literally no one made this claim but here you are asserting it in every post.

Why don't you go back and read the thread before commenting again.

>The bearing hasn't had a critical failure, but it has increased resistance which requires increase power to move the fan, which puts increased stress on the running capacitor.
>Simple as.

Understanding how something works is one thing. Explaining it in a way that autists on 4channel can comprehend is another.

[Anonymous No.2468131]

>>2468045
> What kind of LEDs

typically a common white LED such as you'll find in dollar-store flashlights: typ 3V at 20mA.
the resistor determines how bright. use online LED cacalator to determine.
if it's powered from AC, you'd add a diode in series for good measure.
if you're a noob, another option is to search for an LED equivalent bulb. many replacements exist for flashlights and such.

[Anonymous No.2468140]

>>2468131
I'm guessing it's DC 8V considering thats what the original lamps use? So I guess something like this? Wonder why the site suggested 3k3 ohm resistors and what kind of leds they were using

[Anonymous No.2468141]

>>2468045
> What kind of LEDs

typically a common white LED such as you'll find in dollar-store flashlights: typ 3V at 20mA.
the resistor determines how bright. use online LED cacalator to determine.
if it's powered from AC, you'd add a diode in series for good measure.
if you're a noob, another option is to search for an LED equivalent bulb. many replacements exist for flashlights and such.
if you wanna make it ''cooler'' use blue LEDs
if yow wanna make it vintage-certified, buy exact replacement bulbs

[Anonymous No.2468142]

>>2468140
oops, meant for >>2468141

[Anonymous No.2468144]

>>2468140
130 ohms gives 40mA which is too much for a common LED. it'll die screaming.

[Anonymous No.2468146]

>>2468144

[Anonymous No.2468150]

>>2468146

see where it says ''Total power consumed by the LEDs is 120 milliwatt''
a common LED is rated for a max of 3V at 20mA, or 60 milliwatts.

[Anonymous No.2468157]

>>2468064
>>No way in hell does heating up a winding cause that kind of significant change in ESR
>Literally no one made this claim but here you are asserting it in every post.
I was responding to the original claim here:
> The capacitor is now overworked to make up the voltage for the ever increasing resistance of the overheating conductors
Is that not what “ever increasing resistance of the overheating conductors” means? If not, it’s a terrible way of describing the increased voltage drop across the ESR and cap due to decreasing back-emf. If by “resistance” you actually meant “mechanical resistance” then you’re the fucking source of the retarded confusion. So, right back at you.

[Anonymous No.2468158]

>>2467648
>>2467692

What's the point of putting a fuse in the transformer that can't be replaced? It's obvious you've burnt the transformer. The wires don't have to melt completely. They just need to be hot enough to melt off the insulation and cause a short circuit.

[Anonymous No.2468159]

>>2468158
>What's the point of putting a fuse

ever been in a room where a transformer has burnt up?
it's a smell so bad, no white man can smell it and live.

[Anonymous No.2468204]

>>2468158
Because burning insulation makes the innermost windings go short-circuit, increasing the current drawn significantly but still not enough to trip a breaker. This is a massive fire hazard. The fuses buried in windings are usually thermal fuses.

[Anonymous No.2468282]

>>2468158
>What's the point of putting a fuse in the transformer that can't be replaced?
Stop embarrassing yourself further, anon.
Do you think I would lie about the fuse? I pulled it out, I fixed the winding, it is working.
Every single doorbell transformer sold in the US since 1997 is equipped with a thermal fuse. It can be replaced. You are probably not aware, but there are many different types of fuses. Here is the picture, educate yourself.


>They just need to be hot enough to melt off the insulation and cause a short circuit.
But you would have to be completely retarded to achieve something like that. I don't know, you'd have to short the secondary and leave it on for a few hours or something. You must be speaking from experience.

[Anonymous No.2468343]

>>2467992
good for now

[Anonymous No.2468403]

THT or SMD? Which one is superior?
I think SMD>THT because SMD parts are often cheaper and don't require drilling fucking holes in the board.

[Anonymous No.2468413]

Has anyone used LGBTs in a linear mode? I know they are mostly high voltage high current switching devices but I have a bunch of them from an old TV and the datasheet is not available except for some basic parameters so I am not sure about their SOA in DC mode or even if they are specified for continuous duty since they are 4-layer devices.

[Anonymous No.2468422]

>>2468403
Yeah in general SMT is the superior. The caveats in my mind are:
>anything particularly large or heavy, where the mechanical strain on purely surface-mount connections would be significant, the same applies to often-used board-mounted connectors that have nonzero insertion force
>power packages that are best mounted directly to a heat-sink (why the fuck are all modern fets in powerpak so8)
>parts where you want thermal seperation from the board when soldering / when in use, like heat-sensitive sensors or thermal fuses
>high-voltage situations to stop you chaining a bunch of 0603s
>cases where the pcb itself is flexing and may crack solder joints of more rigid parts
>ancient parts that only come in THT
I've forayed into making purely SMT boards since it's easier for a home-fab, but it becomes a bit obtuse. There's a reason professional boards still have a handful of THT components.

>>2468413
I've always wondered this myself, and had a quick look about a year or so ago to find anything on this, but I came up blank. Considering they're basically just a BJT coupled to a MOSFET, I'd simulate how they function as two discrete components as something like a current regulator. See if they're prone to oscillation in common-emitter or collector forms and where caps might go to mitigate that, and also see what kind of current ends up going through the MOSFET and at what gate voltage. I suspect if you're dumping 10A through the BJT, the FET current will be only 0.2A or so, which I think puts it in the SOA of a power FET regardless of gate voltage. But it's possible that the hFE of the integrated BJT is only like 5 or 10 or something, and there's probably no real way of figuring out what the SOA of the integrated FET is either. So honestly I'd lean towards making IGBTs out of MOSFETs and BJTs, or even MOSFETs and Sziklai pairs. Assuming two hFEs stacked is too much to supply from an op-amp.

[Anonymous No.2468424]

>>2464553
Alright lads I need your help.

I used to watch a YouTube channel four years ago on the basics about electronics. It was some wiz kid who had several thousand subscribers. He did a great job at explaining concepts in an easy to understand way. One day he randomly announced he was leaving for several years on a Mormon mission, but had prerecorded one hundred videos and set them to steadily release in his absence. His subscribers were worried for him, but he said in the video he would be staying away from the internet. So he went cold.

I unsubbed from this channel and have since forgotten what it was called. Now four years later I rediscover and old screenshot I took from one of the videos. If someone can help me locate the name of this electronics channel from the screenshot I would appreciate it. Thanks.
>pic related

[Anonymous No.2468462]

Looking nice. Just doing a dry fit before I solder it up.

[Anonymous No.2468476]

>>2468462
>before I solder it up

not even soldered yet?
it's so simple, i'd have finished that in an hour on perfboard using point-to-point wiring.
in economics there's the concept of opportunity cost: the time you lose making and assembling a PCB could've been so much more personally- or professionally-enriching, like say, reading Tolstoy or 'baiting to tranny porn.

[Anonymous No.2468483]

>>2468476
I don’t actually have all my parts yet, somehow the LM358s have vanished from my collection. I don’t really feel comfortable using perfboard, since it’s a PCB with mains voltage on it, and it needs to be strong enough to push-on and pull-off the spades. So FR2 wouldn’t be strong enough, and FR4 has plated rough-holes which make attaining isolation distance by removing copper impossible. Even if I could add solder mask to the bare wires. The wiring on the back would also be a complete mess, the M3 holes for standoffs would be weak due to intersecting with neighbouring holes, and it would just not be fun to make.

Etching boards is fun, especially since it’s kinda rare for me to get to dump many hours into a halfway productive endeavour. Plus it’s also a test of my etchant solution regeneration, which is an ongoing chemistry project.

[Anonymous No.2468506]

I have a crystal with a load capacitance of 8pF and an ESR of 150Ω. What capacitors and resistor do I use in picrel?
I've done some googling, and picrel already contains the values I guess I need, but I'm not sure about the resistor and it'd be nice if someone else were to check it for me as well.
Also, is there any equipment I can use to test the correct working of it?

[Anonymous No.2468553]

Why are so many of the boomers who are into electronics faggots or crossdressers etc?

[Anonymous No.2468558]

>>2468506
>I've done some googling
you should consider doing it again

[Anonymous No.2468559]

>>2468558
Any hints on what to look for? A friend sent me some STM32 application notes which I'll look into next.

[Anonymous No.2468560]

>>2468553
Sexual dementia. Ballzheimers. Forgot about the benis. Many such cases. Sad?
Also, /g/ needs to solder their headphones and dilate.

[Anonymous No.2468573]

Clothesdryer made a grinding noise. Figured the bearings were toast.
Turns out the Magnets separated from the rotor. Any one ever seen an electric motor fail like this? I've only ever had motors fail because the coils burned up.

[Anonymous No.2468582]

>>2468573
Were you overloading the dryer with wet clothes or running it like a commercial dryer 24/7? Don't forget to change the belt.

[Anonymous No.2468585]

>>2468559
1) manual of the mcu or whatever is the chip you are trying to clock
2) general info on how crystal oscilators work and why sometimes they might need resistors

[Anonymous No.2468608]

>>2468422
I've been researching various options and implementations including commercially available dummy current loads and also linear bench power supplies and they either use TO-3 BJTs or TO-264 DC-spec'ed Mosfets. Look at this SOA for MJ15024. But it is hard to search for parts based on SOA. Of course I could a gazillion of cheap parts in parallel but that's not the best option.

[Anonymous No.2468618]

>>2468045
So i have encountered an issue, 1 of tthe 3 original lights work, the spec sheet says they are 8V, 50ma, I measured the voltage output from the terminals where the broken lamps were connected and it measured at 30V. The working lamp measures 8V as it should.

Do i just say fuck it and get a 12V leds with tthe appropriate resistors or is something seriously fucked in there?

[Anonymous No.2468622]

>>2468618
Were they (original incandescents) in series with each other?

[Anonymous No.2468623]

>>2468622
yea, no sure if in series witht each other, they have seperate poles they are wired to

[Anonymous No.2468625]

>>2468623
Post the whole schematic. 30V is way too high for parallel 8V incandescents.

[Anonymous No.2468626]

>>2468625
there you go anon

[Anonymous No.2468628]

>>2468626
S1-3 is the culprit. Got any contact cleaner? Check for bad solder joints too for good measure.

[Anonymous No.2468636]

>>2468628
S1-3? Im guessing thats the mode selector switch? How the hell would I even go about troubleshooting that?

[Anonymous No.2468637]

>>2468636
The schematic shows the bulbs connected to the selector switch. If one bulb works, then it's fine up to that point. The problem is in the switch, wire wrap/solder joints, or bulb sockets.

[Anonymous No.2468640]

>>2468636
>How the hell would I even go about troubleshooting that?
Continuity test, then resistance test.

[Anonymous No.2468642]

>>2468637
so do i peel open the metal shielding on top and have a look inside? Besides, if the switch is fucked im guessing its not something i can just replace considering the age of the design.

[Anonymous No.2468645]

>>2468642
Follow the schematic and find the jumpers that you need to snip to disconnect the switch, then manually jumper each bulb/socket to test. If everything works, tear it down and find a replacement selector switch and solder the jumpers back together.

[Anonymous No.2468647]

>>2468637
Solder joints look fine to me.
>>2468645
Alternatively, could I just do a 12V led with a 1.8kohm resistor?

[Anonymous No.2468664]

how do i get into electronics?
where do i start?
any starting kit recomendations?

[Anonymous No.2468668]

>>2468664
Learn to sew.

[Anonymous No.2468669]

>>2468582
We don't even use it in summer. I guess they're just not made to last. Gotta sell new dryers somehow, right.

[Anonymous No.2468670]

>>2468647
>could I just do a 12V led with a 1.8kohm resistor?

the lamps are powered from 30V but have two power resistors in series to reduce the voltage to 8V.
that's why you measure 30V across the open lamps.
so, you can use any kind of LED you want as long as you add the appropriate resistor in series.
go back to the LED calculator, and replace 8V by 30V and it'll tell you to add 1.5K ohm in series.
you already have 352 ohms present on the PCB but that's so small compared to 1.5K that you dont have to compensate for it.

[Anonymous No.2468674]

>>2468647
Okay so the 30V reading is normal?
So if I got a 12V 10mA LED with a 1.8K ohm resistor to split the power more evenly it will be fine and the switch isnt fucked?

[Anonymous No.2468675]

>>2468669
What's the brand and approximate age? Appliances in general are getting more expensive and less durable. Samsung made exploding washing machines once. lmao

[Anonymous No.2468680]

>>2468674

that'll work.

[Anonymous No.2468683]

>>2468680
thanks anon, i'll get on it tomorrow

[Anonymous No.2468701]

>>2468664
I'm down to point you toward good resources and help you out with basics if you have Discord comet#7400

[Anonymous No.2468717]

>>2468675
AEG and about seven years. I ordered a new motor. Should be here in a couple of days. Pleasantly surprised that they sell replacement parts. At least it isn't a write off

[Anonymous No.2468718]

>>2468424
Tanner Tech

[Anonymous No.2468724]

>>2468573
Think I saw that on an EEVblog video once. On the bike generator that rubs against the wheel. You should be able to glue them back on, with the right adhesive, but chances are there’s a bit of magnetic dust about the place that won’t be too easy to remove.

>>2468608
Or SMPS into a resistor.

>>2468664
Start with a simple project, that should determine what parts and tools you need to buy.

[Anonymous No.2468851]

>>2468718
Thank you so much!

[Anonymous No.2468858]

>>2468506
The crystal's datasheet will tell you what capacitors to use. Also this is only really an issue if you're building something like a digital clock. The crystal will still oscillate even if you use capacitors off from what the manufacturer intended. The frequency will just be off by 0.0001% or so.

[Anonymous No.2468938]

In a typical functional diagram of an op amp based linear voltage regulator, vref is applied to V+ and the feedback is applied to V-. But I noticed that LM317 seems to have the opposite connections which is why the values of R1/R2 are reversed. Why is that and how does that even work? If feedback is applied to V+ then isn't it a positive feedback? I tried to simulate that in LTSpice by swapping V+ and V- but it just doesn't work at all.

[Anonymous No.2468945]

>>2468851
>>2468718
>>2468424
>Tanner Tech
comfy channel lads

[Anonymous No.2468955]

>>2468938
Where the voltage divider is doesn't really matter. By adjusting the voltage divider on an LM317 you're effectively changing the reference voltage, while adjusting the voltage divider on the op-amp regulator means changing the feedback voltage, both do the same thing more or less. If you used a PNP transistor instead of an NPN (can get arbitrarily close to Vcc with a RRIO op-amp) you'd need to flip the two inputs since high output voltage would mean low output voltage. Also don't forget a capacitor on the reference voltage. Such circuits have much better high-frequency ripple rejection than monolithic regulators, and can be improved further by powering the reference voltage off the regulated output voltage (bootstrap required).

The LM317 is a strange case since it's got an adjustment pin, but a fixed regulator with a ground pin just uses no divider at all. It's just a way of saving pins compared to using seperate pins for V_output and V_feedback.

[Anonymous No.2468967]

>>2468955
LM317 is strange indeed, it has no ground pin so it is floating, and I am not sure how the opamp and Vref are grounded, relative to Vout?

>If you used a PNP transistor instead of an NPN (can get arbitrarily close to Vcc with a RRIO op-amp) you'd need to flip the two inputs since high output voltage would mean low output voltage.

Here is a picture from wiki, it makes sense now. But that's how LDOs work but LM317 is not an LDO so it must have an NPN output.

>Also don't forget a capacitor on the reference voltage.
I had to mess with some external RC compensation since my simulation oscillated when I tested in a constant current mode (current sink). I had to add a large cap from the opamp output to the feedback path and also a small resistor on the feedback path. But that's simulation. I hope I can find a good internally compensated precision opamp so I won't have to deal with this nonsense. Poles and all.

[Anonymous No.2468988]

>>2468967
I just chuck a capacitor from the inverting input to the output. If there’s no resistance going into the inverting input add that too. 100pF is usually enough to dampen oscillations and not impact the circuit itself I find. That’s in real life, never had a simulation oscillate on such a simple circuit, at least not without upping the gain a bunch with a BJT pair. Common-collector or common-drain configurations are generally more stable than common-emitter or common-source configurations, so an LDO design is going to be more difficult in that respect. FETs much more so, in case you want some sort of extremely low dropout like the HT7833.

[Anonymous No.2469060]

Someone give me this, I promise I won't play garbage. Please.

[Anonymous No.2469079]

Getting somewhere now. Miraculously everything fits into the case, though it was an unexpectedly close call with that ground binding post. Glad my math checked out and the LEDs and button and pot were properly matched up to the holes.

It's still missing the hinging lid to cover the binding posts from fingers. It's the quicktest btw.

[Anonymous No.2469118]

>>2469079
Ah, didn't see this one coming. Pushing the button SW1 acts as a charge pump to shove high-current through C2, and into D12. So it's causing a big power supply dip (down to 4V from 12V), even though it's a 100nF cap compared to the PSU cap of 22µF. Even the voltage before the linear regulator at the 470µF cap is dipping, which I would never have expected. The obvious solution is bodging a 1kΩ resistor in series with SW1 or C2 and tuning the values to still get a decent signal through. D12 exists to stop the voltage at that node from exceeding the rails, so putting a resistor in series with it would interfere with that. I think I'll try replacing C2 with a 10nF just to be sure though. I'd likely put such a bodge on C2 since SW1 is a structural component.

[Anonymous No.2469143]

>>2469079
what do you call this type of spring type terminal blocks, what to search on ali?
something like this?
https://www.aliexpress.com/item/3256802624338788.html Yours look nicer.
I don't necessarily need the complete device, but I like the idea of quickly and safely plugging bare wires instead of looking for a power cord and a couple of twist-ons of the proper size every time I need to test a transformer etc.

[Anonymous No.2469147]

>>2469143
spring clip speaker terminal

[Anonymous No.2469202]

Guy with the custom keyboard here, it sort of works, but some buttons are fucked up and don't register (entire first row actually). Probably a bad solder or misoriented diode.
So I am going to play around in KiCAD and design a PCB for it. Can I make through-hole designs with the program or is it just for surface mounts?

[Anonymous No.2469204]

>>2469202
kicad has plenty of footprints for through-hole components, and you can always make your own.

[Anonymous No.2469208]

>>2469204
Cool, thank you.

[Anonymous No.2469249]

>>2468683
Update, i completely butchered the wiring for the led, heatshrink caught fire at one point, couldnt heatshrink the space between the LED and the resistor because the heatshrink was small in diameter and wouldnt go over the resistor to get it over there (non adhesive) and im ashamed.

On the other hand, started replacing some of the caps and i think i did a decent job (considering i havent soldered for over 6 years and have extremely shaky hands)!

[Anonymous No.2469329]

>>2468403
SMD all the way nigga
not only it allows for tiny pcbs but also like 80% of parts today aren't even made in tht form anymore
smd parts are also much cheaper, like resistors and shit
stock up on 0805 that is the best size to use

[Anonymous No.2469461]

>>2469329
whoops i meant to type 0603

[Anonymous No.2469490]

>>2469329
>>2469461
But 0402s fit perfectly across the leads of SOICs!

[Anonymous No.2469507]

>>2469490
>0402
terrible aesthetics. the uncanny valley between 0603 and 0201. the most cringe footprint.

[Anonymous No.2469522]

>>2469507
1206 isn't cringe. It makes people angry. You can kill a man with a sock full of 1206 inductors.

[Anonymous No.2469524]

>>2469522
agreed, 1206 is in fact the second most based footprint after 0612

[Anonymous No.2469599]

It's babies first "proper" PCB time. Could i get an opinion if it's realistically possible to fit all this together without switching to a third layer? Board outline, LEDs and JST connector(J1) are pretty much non variables due to physical restraints/short cable. The bottom pin row may be moved left/right, but not too much upwards. Though i do think it wouldn't be much help anyways. I do wanna let it the smd parts get assembled, so maybe it's just using smaller caps/resistors?
>inb4 why so many external GPIO
Technically i only need 22 out of the 23 pins, but i thought i just add it, because it won't make a difference.

[Anonymous No.2469608]

>>2469599
Poor ground plane :(

[Anonymous No.2469622]

>>2469599
How hard did you try to rearrange the components to minimise trace lengths? Are you able to change the pinout to J4? I'm guessing it might be easier to run traces on one side alone to J4 if it's positioned correctly, with U3 or whatever else is necessary being routed on the other side. You should probably add a ground plane on both sides, and those bypass caps Like C5 going to a ground plane far removed from the QFN itself is egregious.

Also add Mitxela's melting script.

[Anonymous No.2469673]

>>2469118
Wait no that's not the problem at all. I just had a couple of spades swapped so my relay's freewheel diode was going from positive to negative instead of negative to positive.

haha

[Anonymous No.2469700]

I dont really get it

[Anonymous No.2469702]

>>2469673
Aha it works great now. Except the offset voltage. For whatever reason, the Idecouple node is sitting at 2.00V and the Irect node is sitting at 2.96V with no current, even when GNDA is 2.15V. Now the Idecouple mismatch could be the 10MΩ DMM loading it down, but that definitely doesn't apply to Idecouple being a volt too high.

I'm going to try decreasing R14 a bunch (ideally to the realm of 1-10k) to see if it's amplified input offset voltage, but the R14/C7 time constant needs to be significantly lower than 50Hz so I'll have to hunt about for my non-polar electrolytic caps. I can't really make R16/R15 larger since R16 needs to be significantly smaller than R17 in order for the rectifier to sink and source current well. Not that it's an absolute requirement, it just helps keep the filtered voltage more indicative of the average current, instead of the peak current.

The other issue may be power supply ripple, which gets kinda shitty once the transistor turns the relay on. 0.33Vrms from a 22µF cap, which I guess I can use to reverse engineer the load current if I'm not tired. Should be able to fit a bigger cap in there to drop that by a factor of 5 at least. The GNDA voltage also has 0.046Vrms of ripple, so I should upgrade that 100nF to something a lot larger while I'm doing upgrades. I doubt power supply ripple influences my op-amps themselves considering their PSRR, but the initial current amplifier may be affected.

[Anonymous No.2469703]

Can anyone tell me from this
>>2469700
about U1 and U1, I dont understand why there are two different partls listed, could both be used?
R1-R3 I also dont get, what does 4k7...10k ohm mean?
Can any resistor between does values be used?
Sorry for the retarded questions

[Anonymous No.2469705]

>>2469703
From what I remember, the soviets did their own versions of the 7400-series ICs. Even if not, all it needs to be is an open-collector hex inverter with the same pinout, of which there may well be half a dozen different ones, not including logic family varieties.

4k7...10k just means the resistors can be anything in that range without making a difference to the circuit operation. I'd go 10k since that's easier. They're pull-up resistors so values are non-critical, so long as the value isn't so low that the current drawn by the inverter is too higher, or so high that the edge transitions are too slow.

[Anonymous No.2469717]

>>2469608
That's pretty much what I thought doing it.

>>2469622
>How hard did you try to rearrange the components to minimise trace lengths?
Not at all? This is hands down trying to connect everything at all. In the first iteration i had most GPIO lanes off ground plane, but then xin/xout for the oscillator were sitting on right side..
>Are you able to change the pinout to J4?
Maybe. Would have redo the plug.
>it might be easier to run traces on one side alone to J4 if it's positioned correctly, with U3 being routed on the other side.
I'll give it some thought, thanks.
>add a ground plane on both sides
Sigh. Probably..
>those bypass caps Like C5 going to a ground plane far removed from the QFN itself is egregious.
Well yeah. Again it's my second iteration, where I just try to connect everything at all, without creating a super blob on ground plane.
>add Mitxela's melting script.
Looks funky. I'll look into it after I get this sorted.

[Anonymous No.2469735]

I'm a total electronics noob. I'd like to build some of the babby's first projects mentioned in the OP but I have no fucking idea where to start.
Do I just go on the internet and search for schematics for a certain project or should I try to find kits that I could build?
I found some velleman kits I'll buy first and try to do them. Where else to find kits to build in EU?

[Anonymous No.2469738]

>>2469735

https://www.youtube.com/results?search_query=electronics+kits+for+beginners

[Anonymous No.2469759]

>>2464553
What's a good budget 100ah lithium battery brand?
Chins? Redodo?

[Anonymous No.2469778]

When making transformers, why are the actual number of turns much higher than the winding ratio. For example if the winding ratio is 1:50, the actual winding are something like 50 and 2500 turns respectively?
I've been trying to learn about Transformers and cores and shit but it's actually not very common knowledge.

[Anonymous No.2469790]

>>2469778

if you made a xformer with 50 turns and plugged it into 120V, you'd get 100A of current, smoke, fire, and a blown breaker.
more turns = less fire.

[Anonymous No.2469975]

I would like to power a PC fan (12v 100mA) from a 24v supplie. Is there any mosfet I can get that would do the job quick and easy or do I need some type of buck conversion?

[Anonymous No.2469994]

>>2469735
Electronics is a multifaceted engineering discipline. There’s design of block diagrams (e.g. the VCO connects to the VCA and the LFO), design of circuit blocks (e.g. assembling a common-collector amplifier), manually prototyping and soldering, designing the interactions with the outside world (e.g. enclosure, sensor placement, airflow), programming if relevant, and PCB design. Even component selection is probably a layer. Depending on what you want to get good at, there’s a few different ways you could practice. Copying circuits online won’t really give you component selection or inter/intra-block design at all, but it will train your soldering skills and get you learning to read schematics and understand to them to some extent. Copying other people’s projects is how I’d recommend starting off, especially nice ones on Instructables or Make or wherever that have easy to follow tutorials. Feel free to replace resistors with potentiometers to experiment what changing the values do, it will give you a better intuition as to what’s going on. And read up the theory about it all.
My first project was copying the Elektrosluch circuit, was pretty fun.

Remember, breadboards are temporary but knowledge is forever.

[Anonymous No.2469999]

>>2469778
more turns = higher inductance = higher reactance = less current or a given frequency
You’ll find really low primary turns numbers on very high frequency circuits like Tesla coils, because higher frequencies mean higher reactance for a given inductance. When it comes to switching DC into a transformer like in a flyback or forward converter, there’s an inductive ramp due to V = L*dI/dt. Switch a voltage on to an inductor or transformer primary, the current through it will linearly increase at a rate of V/L amps/second until you turn it off again. Increase L and you decrease the peak current. Decrease the time period and you decrease the peak current, which is why switching converters can use much smaller transformers with fewer windings than mains-frequency transformers.

But ultimately the average current going into the transformer is directly proportional to the average current coming out of the transformer (by the turns ratio), so if you want a kW output power you’ll need some pretty high peak currents on the input, which could well mean less turns.

[Anonymous No.2470004]

>>2469975
>PC fan (12v 100mA) from a 24v supplie

you can make/buy a PWM controller easily.
by setting it 50% duty cycle, you get an average voltage across the motor of 12V.
but why bother?
the fan takes such little current that sticking a 120-ohm 3-watt resistor in series is all you need.

[Anonymous No.2470005]

>>2469975
>>2470004

PWM pic for ref.

[Anonymous No.2470009]

>>2470004
Pretty sure it would be 25% duty cycle to make the RMS voltage 12V. But I’m. It sure if magnetic motors care about RMS or average voltage. There’s also the matter of peak voltage across the internals, I wouldn’t necessarily trust the hall sensors or any capacitors to handle 24V without giving up the magic smoke. Power resistors are kinda clumsy, little buck converters are cheap as chips so I’d use one of them to be safe.

[Anonymous No.2470019]

is Tech Connect trooning out?

[Anonymous No.2470020]

are you a mica person or a silicon person?

[Anonymous No.2470025]

>>2470020

silicone all the way.

[Anonymous No.2470029]

>>2470019
no that's just the typical american body in 2022

>>2470020
hot glue

[Anonymous No.2470035]

>>2470025
>butt baby? challenge accepted

[Anonymous No.2470067]

>flicks molten solder off lead into eye
>for the 5th time this week

[Anonymous No.2470068]

>>2470067
>flicks molten solder off lead into eye
Safety first, newb

[Anonymous No.2470113]

Does anyone here have recommended AGC/compressor designs? I'm thinking either LM13700 or a diode-based single-balanced mixer like this pic. Or a vactrol but the delay on that might be too long. I'll also need to use it as a VCA to change the amplitude of other signals such that they all match, which I think all these can do, though the voltage mirrors needed by the op-amp method do take up a fair few components. Not sure about using a JFET, signal amplitude is somewhat limited with those, but I guess that also applies to an LM13700.

[Anonymous No.2470136]

I am surprised that nobody ever discusses measuring temperature of power transistors wired in parallel. Not even on diyaudio. It seems like a natural solution: instead of measuring current, attach a thermistor to every transistor and balance current based on the temperature. No need to get fancy like running a PID, just make sure the temperature of each transistor is below a certain threshold and limit power. Should work for power supplies, amplifiers, dummy loads etc. What do you think? As for attachment, high temp silicon epoxy should work well.

[Anonymous No.2470145]

>>2470136
Power transistors often have a low-value power resistor in series with them, if you made it a thermistor and bonded it to the transistor itself you'd get automatic feedback on each transistor, though no total balancing circuit.

>silicon epoxy
Is that a thing? Even if you meant silicone I think silicones and epoxies are pretty different unless you're doing obscure chemistry.

[Anonymous No.2470147]

Can i stick a bare wire into this type of spring clamp connector or it requires a ferrule?

[Anonymous No.2470148]

>>2470147
solid core if it's big enough

[Anonymous No.2470235]

Third attempt. Much less ground plane fuckery iin this one.Think i'll clean it up and then roll with it.

>>2470147
Do NOT use stranded wire, if that answers your question.

[Anonymous No.2470259]

>>2470235
Your bottom-right bypass caps are still going somewhere silly to get to the QFN’s ground. Why not just put them on the other side of the board under the IC? You’re already having pins poking through the back so it’s not like it will be flat. Putting U3 behind the QFN might make routing easier also, maybe.

Consider adding an anime girl or other decal to the front.

[Anonymous No.2470267]

>>2470259
You mean something like this?

[Anonymous No.2470279]

>>2470267
By other side I meant on the back instead of the front. Moving bypass caps to between the IC and the long connector wouldn’t do much good since there’s a trace between the IC ground and the actual ground plane it would connect to, but I guess it would be fine if you moved the trace below the cap. Also check if each cap is meant to be at a particular VDD and/or VSS pin, often I’ll see caps placed specifically next to AVDD and such. Not that it really matters if you’re not doing analogue stuff.

Also unresolved ratsnest over on the left near C15. Here appear to be two C15s.

[Anonymous No.2470300]

>>2470279
>By other side I meant on the back instead of the front.
That sounds like a cool idea, but i don't trust myself with a particular piece of hardware on both sides.
>Moving bypass caps to between the IC and the long connector wouldn’t do much good since there’s a trace between the IC ground and the actual ground plane
It seems so obvious. Hopefully i see these things quicker with more exercise.
>it would be fine if you moved the trace below the cap.
Again.
>check if each cap is meant to be at a particular VDD and/or VSS pin
Will do.
>unresolved ratsnest over on the left near C15. Here appear to be two C15s.
Indeed. I almost forgot that 1V1 rail. Only a C13 and C15 though.

[Anonymous No.2470318]

what is the best way to measure voltage drop over a resistor with an op amp? basically i have a shunt and i want tho measure the voltage drop it has with an arduino but i need to use an op amp since the voltage drop is too small

[Anonymous No.2470327]

I've salvaged some somponents from CRT TVs and I found what appears to be ferrite coils with a magnet on top. Anyone has any info on the purpose of saturating the core or whatever this thing is supposed to do?

[Anonymous No.2470331]

>>2470327
>ferrite coils with a magnet on top.

dont believe you.

[Anonymous No.2470340]

>>2470331
Photo with a Kamelåså attached to it

[Anonymous No.2470356]

>>2470327
OK, a quick search explained a lot https://www.eevblog.com/forum/beginners/why-stick-a-magnet-to-an-inductor-permanently/

[Anonymous No.2470401]

[Anonymous No.2470408]

[Anonymous No.2470409]

[Anonymous No.2470431]

>>2470235
>Do NOT use stranded wire,
>>2470148
but tinned wire should be fine, right? should be as good as solid

[Anonymous No.2470441]

>>2470318
Instrumentation amplifier.

[Anonymous No.2470451]

Not sure if this is the thread where i should ask this but i want to buy a multimeter from nearby and so far i found a Maxwell MX25306 and a Maxwell MX-25302 also some cheaper different brands. What should i look out for in them aside from the tester cable? Would be used for small electronics, would like to test my not working monitor as well with it. So it would be for general usage as far as i'm concerned.

[Anonymous No.2470462]

>>2470451
Have a look at the “EEVblog multimeter spreadsheet” to get an idea of what people normally pay for those specs, and what you may be missing out on. Generally you’ll want autoranging, and true RMS is nice to have also. Multi-displays can be really handy for displaying voltage and frequency at the same time, or maximum value and current value, or whatever. Less readily listed as a spec is the delay when you tap the leads together in continuity mode. Some meters also have a warning buzzer for when you plug the leads into the current slot, which would be pretty handy.

[Anonymous No.2470468]

>>2470451
>Maxwell MX25306 and a Maxwell MX-25302

both are way better than what you need.
both can read millivolts and microamps in both AC and DC, and that's all the basic functionality you need; the rest is gravy.
the '02 can measure 10Mhz vs 200kHz for the '06 so i'd go with that one.
neither is auto-ranging, which means it's made for adults who want measurements QUICK, not limp-wristed faggits who cant turn a knob.

[Anonymous No.2470483]

>>2470462
>>2470468
Thanks.
Don't you mean the '06 can measure 10Mhz vs 200kHz for the '02; so you would choose the '06?

[Anonymous No.2470496]

I'm trying to split a power supply for an effects circuit. My problem is that adding the LED as in pic related messes with the voltage at the middle point ground. What is the error?

[Anonymous No.2470500]

>>2470496
Below LED V_forward you get a split supply, above V_forward the LED has super low impedance. Since the parallel 1K is 100x smaller than the 100K you end up with a voltage divider like this
100k/(100k+1k)*Vin-V_forward=10.8V ish

[Anonymous No.2470502]

>>2470496
Just put the LED+resistor from V+ to V-. Same for any DC load. An unbuffered middle ground rail like that should only be used to sink AC currents that get eaten up by the capacitors. It's advised to buffer it with a spare op-amp.

[Anonymous No.2470580]

>>2470468
>want measurements QUICK
>turning knobs
pick one

[Anonymous No.2470583]

Has anyone used IRFP250s in linear mode? I found a box of them that I probably bought on ali a while ago. I wanted to test their SOA and the first one I tried did well at 30-40W levels but died quickly at 7A at 14V ~100W. So I am wondering if that's too much for them to handle or perhaps because they are fakes. Or both.

[Anonymous No.2470588]

>>2470583
IIRC what matters for linear FETs is how they’re constructed. Planar, trench, epitaxial, whatever the terminology is. Even if the datasheet doesn’t have an accurate SOA for linear operation, it may well tell you how it was constructed. Then it’s just a matter of testing that for a few hours or days. 100W sounds like a fuckton just from thermal reasons, so for an experiment I’d aim to keep it at a constant 60C or so with a thermistor bolted to it, while the transistor is bolted to the big heat-sink. That should rule out death from bulk over temperature, and just leave you with spot failures.

Test some IGBTs too.

[Anonymous No.2470595]

Thoughts on resettable fuses? Sounds too good to be true. Why use glass or ceramic fuses in multimeters then? I've burned so many of them mostly when measuring current. Sounds like a resettable fuse could be handy in this situation.

[Anonymous No.2470604]

>>2470595
OK so sounds like the obvious catch is they are only good as a protection against slight overcurrent, say 2-3 times more than the rated current. But they can't protect against a massive short since they can't withstand infinite current. Plus they may turn out to be be pretty slow compared to fast fuses.

[Anonymous No.2470624]

>>2470595
Honestly you’re better off with a solid-state overcurrent latch. Pass the current through a MOSFET and measure current with a sense resistor, which turns the FETs off if current gets too high. Add a button to turn the FETs back on again. If you do it right it can react fast enough that it stays within the FET’s SOA even for massive current spikes.

[Anonymous No.2470730]

What is the fastest oscilator you can make?

[Anonymous No.2470765]

>>2470730

this bad boy puts out around 600 terahertz.
and you can do the same just rubbing sticks together.

[Anonymous No.2470773]

>>2470624
What happens if the mosfet burns out creating a short from drain to source? Perhaps further overcurrent will do more damage and will open it, but who knows.
Then there are further issues if there are several of them connected in parallel. I guess the first line of defense is try and prevent excessive I*V which would help prevent catastrophic failures in the first place. But there is no guarantee.

>solid-state
which is another mosfet basically. which can get destroyed as well. there must be a simple fuse there somewhere. to at least prevent fire.

[Anonymous No.2470786]

>>2467400
>If there's negative feedback (output going to inverting input, or output being inverted and sent into non-inverting input) then the two inputs are going to be forced to be the same by the output

This.
I spent so long analysing op amp circuits by trying to figure out which of the configurations the op amp was in (inverting amplifier, buffer, integrator, etc).
If you just forget that shit and see that there's a pathway from the output to the inverting input then the op amp will do whatever it can to keep the inputs at the same voltage.

[Anonymous No.2470806]

>>2470765
What? Is there a circuit inside it that does that or are you talking about the light emited?
I am interested in the fastest possible circuit for any kind of oscilation. Like square wave, sine wave whatever. I count the "main" frequency. Actually I would also be interested in the fastest possible of each kind if possible (square, triangle etc)

[Anonymous No.2470817]

looking for suggestions for household material to use as a conductor in a 3D printed 18650 7S battery holder to power a speaker, I don't really want to buy nickel strip if there's something I can repurpose to make the tabs

[Anonymous No.2470894]

>>2470588
>Test some IGBTs too.
Tested one of the many I pulled out of an old TV. It did well up to about 40W but then I probably turned up the pot a little too much and it died silently as a true warrior. Last I remember the current was about 5A and voltage probably about 14-15V. So sounds like FET or LGBT, the optimal safe power dissipation should be kept below 30W.

Oh and it also unsoldered itself while dying. No sparks or anything. The collector wire just fell off. Wicked.

[Anonymous No.2470931]

>>2470730
I’ve got some RF transistors lying around, I could whip up an air-coil and use a film capacitor to make an LC oscillator with the transistors for a few hundred MHz. Can’t say the result would be very sinusoidal though, chances are it would clip a bunch, though there are simple ways around that.

>>2470773
Yeah it suffers the same issues as the resettable fuses in that it won’t like really large current spikes, but it acts a lot faster and can be adjusted. It would be well-complimented by a normal fuse.

>>2470817
Why not just have solid-core copper wire coiled into a flat surface? You can add a bit of springiness to the 3D print itself, that or add a tightening screw or a pillar for an actual spring to push a contact into the cell.

>>2470894
Mate you’re basically just testing normal thermal failure, not that specific kind of micro-scale thermal runaway you get when using them in the linear region. Also how’d you burn out a pot?

[Anonymous No.2470942]

>>2470931
>how’d you burn out a pot?

[Anonymous No.2470954]

>>2470942
Oh “turned up the pot too much” instead of “burned”. Don’t mind me.

[Anonymous No.2470956]

>>2470894
>testing normal thermal failure, not that specific kind of micro-scale thermal runaway
how would you test each of these types of failures?

>Also how’d you burn out a pot?
No, the pot controlled the gate voltage so I cranked it up and killed the FET.

[Anonymous No.2470971]

>>2470956
As I alluded to before, put a thermistor on the FET and use an op-amp to keep it no higher than 60C or so. Then use it to dissipate as much power as it will do without surpassing that temperature limit. It will probably oscillate if you just use the thermistor for feedback, so you’ll either want to put a really aggressive PID controlling network on that op-amp, or use something else to limit current or voltage drop. I’d lean towards the former just because it would keep he temperature really constant, which would be more like a worst-case scenario and the least like what a switching transistor would see.

[Anonymous No.2471226]

new thread plz. current thread has "end-of-life" status

[Anonymous No.2471246]

>>2471226
Not recommended for new designs

[Anonymous No.2471265]

>>2471226
oh shit you're right, baking now

[Anonymous No.2471267]

>>2471266
>>2471266
>>2471266
NEW THREAD