/diy/ - Do It Yourself

>>1855677
>what do you hate about it?

To begin with, it's extremely unintuitive. The interface feels clunky and outdated, and often times gets in the way of what I want to do. It also can't handle real-time changes to the circuit and on-the-fly editing while it's simulating. It makes it very difficult to experiment and "perturb & observe" my circuits.
For example, let's say I want to vary the duty cycle of a PWM input on the gate of a mosfet to observe how it changes the behavior of the circuit.

In circuit.js, I can set a slider that corresponds 1-to-1 to the duty cycle of the PWM voltage source. I can slide this slider and watch the voltage graph change with the PWM in real time.

In ltpsice, from the best I can tell, I need to add an A-device comparator with a sine and triangle voltage source on either pin. This will, at best, let me view how the circuit responds to a sinusoidally varying PWM. I can't change it arbitrarily. Then, I have to set up the actual simulation conditions. From what I can tell, it can't just display a continuous real-time graph of the circuit. I must set a specific time frame for it to run and I must re-simulate the entire circuit if I make a change. All of this has to be done through either a netlist or clunky commands, and through an interface that feels needlessly verbose, nonstandard, and outdated.

I can see where being able to use functions as input/being able to modify the circuit as text can be very powerful, but it's incredibly cumbersome for prototyping circuits.

Please help!
I have a bathroom exhaust fan/light and it has two black wires coming out, one for the fan and one for the light. I'm trying to get them on a double-pole, single throw switch, but both switches are doing the same thing.

If I have the wires connecting to their own single-pole switches, I can turn on and off the fan or light independently. But as soon as I hook both up to the double-pole switch, as lone as one of the switches is on, both the fan and light will run at the same time.

What the heck is happening? Aren't the circuits independent? I was thinking that when both circuits are connected to the same ground wire, when either of the switches is "ON", both circuits get connected. But I disconnected one of the grounds, so all I have are the neutral and hot wire for one switch, and all three wires connected for the other switch - and it still does the same thing: connecting both circuits as soon as one of the two switches is turned to on.

>>1856328
Are you sure you have a double-pole-single-throw switch? One switch that controls two devices together?

>>1856328
here's what's going on up in the ceiling. should I connect all the black hot wires together then make the neutrals independent?

>>1856358
its one of these. It has two hot wire terminals so I figured it was 'double-pole'.

>>1856360
I think the issue is that I have all the neutral wires in the same connector. I need to put all the hot wire from power together with the hot wires to the fixture, then make the two neutrals from the fixture connect to the two hot wires to each switch, then run the neutrals from each switch back to the neutral on the power.

Sounds good? I can't test it out now but I think I have the theory figured out.

Is routing one of the I2C traces under another I2C trace's pull up resistor bad practice?
SCL trace is perpendicular to SDA trace so cross talk should be minimal, right?
Any ideas on how I can avoid this? It's going to be a mess.

>>1856311
>generate square pulse
>set up pulse on/off time as function of parameters
>vary parameters in time domain simulation with the .step command
Have you tried that?

>>1856437
and I understand what you mean and what you want is not what the simulator was made for. Other products such as circuits.js and falstad are better for pedagogy and other things. But for actually designing and testing circuits I usually find LTspice the best. When I have to simulate something basic I also just go to falstad instead of going through the trouble of LTspice.

>>1856440
And your point that the interface is:
>verbose
>non-standard
I'm sorry but I simply do not agree. LTspice interface is very simple. I think people look at that shitty background, the win 98 look and just automatically get put off, but have you tried using other professional software? In comparison to LTspice Psim has terrible interface and usability.

>>1856453
And I can say this because I went on the path of hating LTspice to liking it a lot because I actually learned stuff. I remember asking myself
>lmao why do you need to select "time domain", "AC" and "DC"? I just want to "simulate"
only after learning how simulation works is that I found that is a fundamental distinction and software that "simulates" stuff just has that choice made for you, in one way or another.
While I agree that sliders would be cool, they'd not be worth anything if you coudn't edit while the simulation was running. (Falstad doesn't have logarithmic increases, you can' t choose the steps or anything. You'd need another pop-up for that and then it'd be verbose)

Is it possible to repair a T12 ballast? It is just a transformer, right? What happens to them, what kind of a damage?

I need a DC DC converter that outputs a postive and negative voltage and has a ground. Would this work, based off the diagram there's a ground on pin 3, would I just be able to hook the corresponding ground to that?

>>1856478
And the diagram in question, I'm trying to follow this, https://www.circuitbenders.co.uk/tutorials/R8power/R8power.html but can't find a power converter with a ground at a decent price or with fast shipping.

>>1856463
What are DC analyses even for?

>>1856473
Elaborate.

>>1856480
Using that as your intermediate rail will give you something that can source current but not sink it. You either a dedicated split-rail power supply, a rail-inverting charge pump, an isolated DC-DC converter, some sort of class-D-style push-pull circuit, etc. I'd personally recommend that you invert the existing V+ rail such that you keep your current ground rail as your signal ground. Better for noise reasons.

>>1856480
No, that ground symbol just means 0V. You can tie it to the ground of your house if you desire, but an LM2596 isn't going to give you a negative rail at all, just decrease the positive rail with respect to this 0V.

>>1856485
>Elaborate.
All my fluorescent lamps are flickering or won't start. New lamps made no difference. So it must be either a starter or a ballast. I couldn't find a starter so I am assuming there isn't one, and the ballast is hidden somewhere under the cover. There are a million videos on youtube on how to replace a ballast but not a single one, not a single schematic or how to tear it apart and test it. Of course I am going to buy a new one but I am just curious what this thing does other than limiting current. And what can possibly happen to it with time.

>>1856492
Ballast circuits range somewhat in technicality, some are just a transformer (or inductor?) and I think a cap and resistor, others may have more complicated active circuitry in them. Probably repairable, but if the transformer is damaged it's probably cheaper just to buy a new ballast than trying to find a replacement transformer. But certainly open it up to see, so long as it isn't potted.
Google "fluorescent ballast circuit", you should see some relevant results.

>>1856502
Yep, found a few.
This one must be the old fashioned bi-metallic starter. When the power is connected, the lamp is cold and has high resistance, so all the voltage drops on the starter and it opens and provides high voltage to the lamp until its resistance drops so the starter closes again and now the power is supplied via the current limiting choke. The only thing I don't understand is how is the high voltage generated as there is no multiplier. Perhaps those kinds of bulbs only need the mains voltage?

Are current driven indicator tubes like the IN-9 multiplex-able? With bar length being dependent on the current flowing through the tube it seems if you have a short on time it does not give the tube enough time to strike and get up to length. Using a higher peak current (like you do with LEDs to get higher brightness when multiplexing) strikes the tube at maximum length and intensity and burns out my DC-DC converter chip.

It seems like it's going to be a pain if I want to drive 10-20 of these tubes at once from a low voltage supply without being able to multiplex them. Getting 0.5A out at 170V is going to require a large transformer with thick unwieldy windings and large heatsinks to handle to 20+ amps of input current.

Additionally question, how do I prevent the tube striking halfway up as it sometimes does? I know IN-13 has an auxiliary cathode to stop that but IN-13 is also very expensive. IN-9 is expensive too but is cheaper.

>>1856485
>Using that as your
Thanks for clarifying that for me, so with something like this is the ground simply in the "middle" of the positive and negative voltage?

if I want to hand-make my own buck converter, would I use BJTs or MOSFETs?

>>1856513
No. That's still wrong. It's not the midpoint between the positive and negative. You can have asymmetrical positive and negative rails like -12V and +5V both be referenced to the same ground. In addition you can have multiple, separate, ground references in a circuit. One ground will literally be at some voltage potential relative to the other ground.

>>1856530
Yes.

>>1856485
>DC analyses
Finding the operating point of stuff
>oh I can see that pretty easily in my class AB 4 transistor circuit using time domain analysis
Yeah, try that with 150 transistors

>>1856533
please I'm so fucking confused

>>1856540
Post the 150 transistor circuit you're simulating.

Is it necessary to test the "current leakage" to figure out whether the smd electrolytic capacitors on my motherboard are bad, or is it enough to test their capacitance and internal resistance / ESR to determine whether any need to be replaced?

>>1856506
The bimetallic switch opens after some time due to heat, and this causes the voltage across the inductor to spike suddenly as its magnetic field collapses.

>>1856510
You can multiplex them with the correct kind of driver circuit if only one symbol is on at the same time. Wouldn’t be too difficult to include a decimal point on a seperate circuit. You could do it possibly simpler with normal constant voltage multiplexing and a constant-current regulator of some sort on each tube’s common cathode/anode. A TL431 would work fine, I think.

But why would you not use a few latching shift registers/drivers to avoid the flicker entirely? If power is your issue, just drive all the tubes at once but at a lower current value, should be nicer to the tubes too. There are existing purpose-built nixie drivers, so either way I’d start off looking at how those work.

>>1856530
MOSFET (or IGBT). Or even a controller with integrated transistor, like the LM2596.

If you powered an I2C sensor with a linear voltage regulator, wouldn't the output capactior together with the pull up resistors act as low pass filter?
How do you avoid this?

>>1856586
Why'd you answer my question when you clearly don't understand what kind of tube I'm using? It's not a numerical tube. It's a bar tube. Completely different. The length that they light up is directly proportional to the current flowing through them. You do not drive them like you would any other nixie tube.

I overheard someone at work say that through hole is better than SMT for hand soldering. Thoughts?

>>1856600
Are you talking about having the linear regulator output, with its capacitor, be on the same net as your I2C data or clock?

>>1856611
Oh that’s how they work. I may have looked at their datasheets in the past, but I never knew they were purely analog bar tubes. I assumed they were multiplexed like a numitron.

Might still be able to get away with latching analog circuits, i.e. sample and hold. Otherwise, just an op-amp controlled cc driver (switching or linear) with an analog switch IC before it and an array of HV FETs after it sounds like the way to go.

This is electrical rather than electronic, but is there a practical limit to the length of a run of 240Vac10A cable? I want to power a bore pump, and am wondering if I can't just wire it in from the house, it's probably 400M away

>>1856675
What is the gauge of your cable? What voltage? 220V? What current?
You need to calculate the minimum wire gauge based on the voltage drop caused by the resistance at a given current. Typically you don't want more than a 3% voltage drop. I am assuming you are on 220V so let's say 210V is ok on the other end, so your max drop is 10V. Then you need to know your maximum current and the total resistance of your cable. Lets say your load draws 10A so the total cable resistance cannot exceed 1 ohm. You can use this calculator: https://chemandy.com/calculators/round-wire-resistance-calculator.htm
to check the resistance of 400m of your wire.
But it seems to me that you should be fine unless your cable is too thin or the current is silly high.

>>1856675
here is a better calculator, just plug in all the numbers and it will give the wire gauge.
https://www.southwire.com/calculator-vdrop

Can I zap my cock into instant wood with a diy electronic device?

>>1856611
I am using an op-amp+transistor (TL072+MJE340) voltage controlled current source to drive it. It works fine, that's not an issue. The issue is multiplexing it. That's what I want to know how to do, if it's possible. The tube behaves abnormally when running at a 10% duty cycle.

>>1856710
What frequency are you driving it at? Have you tried monitoring the voltage waveform as the PWM waveform turns on and off? I suspect the nixie has some amount of parallel capacitance, which may be limiting its minimum visible duty cycle at high frequencies. Though delay from the ionisation of the gas is probably a much more significant.

>>1856717
I've tried multiplexing it at a relatively low frequency of 1kHz.

>>1856750
Not sure if that's low for a nixie or not, even something as low as 200Hz would probably not be too noticeable to look at. You're just testing with PWM on a single channel, right? To isolate the multiplexing effects from the PWM effects.
Another thought, does your op-amp driver saturate on each off-cycle? The recovery speed might have some effects. Check for power supply ripple too, while you're at it.

>>1856767
There's two op-amp drivers in a cascode configuration. The bottom converts a signal voltage into the appropriate tube current for the display. The top is for gating. Yes I'm only doing PWM with one tube for now. I cannot scale this project up until I get one tube working. My power supply can't deliver enough current at high voltage to drive more than about 2-3 tubes. Turns out it's really hard to get more than 20-30mA out of a high voltage supply when you're boosting it from a low voltage source.

>>1856779
What's the booster?

>>1856547
Typicallly mosfet. Take more switch mode supplies apart, reverse engineer them, draw the schematic, look up datasheets and watch diodegonewild on youtube who does a ton of videos on the subject and is an excellent teacher. Breadboard it. Test. Analyze. Read books about transisators and retweebulate your thinking compartment to learn this stuff. There's nothing to them but you have to research dude.

>>1856793
Simple boost converter using a TL598. It's not fancy. I had intended to do a flyback converter but the E16/8/5 cores I order were way too small given I needed like 400 turns secondary and even the few dozen turns of primary were unacceptable given I needed like 20 AWG wire to handle the current.

>>1856823
Yeah, I always fall back to non-isolated converters for this very reason. The only cores I have are toroids, and while they're probably large enough for a 10-50W converter, they're such a pain to wind that I never bother and just use off-the-shelf inductors. I only bought them for low turn counts of thick wire, for mains filters and for RF.

Can anyone decipher the value of this capacitor?
I thought 10D is just tolerance, but there isn't any other value that could mean capacitance.
So I'm assuming it means 10pF ±0.5?
It's from a blown industrial 24V switching supply, the other side has a hole in it.

>>1856840
Looks like it might be a 1kV MOV instead. 10pF is too low even to be a Y cap, but it could conceivably be a 10nF Y cap. Where was it connected? On the HV side, across neutral and live, after a fuse/PTC but before the X caps? Or from live or neutral to ground?
>inb4 no ground

>>1856643
>Are you talking about having the linear regulator output, with its capacitor, be on the same net as your I2C data or clock?

Yes. The linear voltage regulator would provide the HIGH signal.
Wouldn't every I2C sensor suffer from this unless it's battery powered?

>>1856643
>I overheard someone at work say that through hole is better than SMT for hand soldering
I overheard someone at work say that through hole is better than SMT for hand soldering
yeah and it's dark at night and the sky is blue. So what?

This week in stupid ideas:
A VCA without transconductance amplifers!

Honestly way more complicated than I expected.
The multiplication is more or less implemented with logarithmic and exponential amplifiers to do exp(log(a)+b) = a*exp(b)
Since it uses diodes for the logarithmic response it's just a single quadrant multiplier, which means the audio needs to be biased so that it always remains at a positive voltage.
The complications came from the fact that the bias also gets amplified, so an amplified copy of the bias needs to be subtracted out in the end.

>>1856876
Shit, forgot the image

>>1856847
timing caps can be in the tens of pf range. i got some 47pf caps i stripped from a board. granted capacitor code doesnt really want to show anything that low its like 102 is 10 nf and 101 is 100 pf and i have never seen a code ending in zero but pf caps in the tens of pf exist for timing purposes

standard cap code doesnt apply to all ceramic caps. they have ones that are to low and have a different system for the cap code. that shit on his cap is tan on blue its not meant to be read

>>1856876
>>1856877
coulda just used a single op-amp and a single jfet, or one of those two-diode circuits that I saw on w2aew's video on making an agc

>>1856888
Jfets are all fun and games until you have to actually buy those fuckers

>>1856877
Go to bed

I'm thinking about connecting 30 - 50 temperature sensors (probably DS18B20) to Arduino i2c. Now, I'm know that it's technically possible (i2c supports up to 99 slaves) but would it practically work? Also, wire length between each sensor and Arduino is ~50 meteres (around 160 feet) and the voltage is 5V. Would reading be possible? I did find an article online saying that lowering baudrate to 9600 or lower would make it more reliable, but I'm not sure if that's possible and how (guess I'd need to tweak SCL?).

>>1856889
I've got 9 spare from buying 10 for making a mini-whip antenna. Which I haven't made yet. Anyhow they're not that expensive so long as you use them sparingly. I certainly wouldn't use half a dozen in a single circuit or recommend them for mass-production, but for prototyping and messing about they're cool.

>>1856915
that is way too much
the stray capacitance would be too high which would result in your signal looking somewhat looking low pass filtered "square" wave
however to increase distance you can to lower the bus frequency and use low value pull up resistors
You could look into CAN bus sensors

>>1856928
thanks! is there even a way to connect that many temperature sensors to one MCU then?

Can I put a resistor in line with a light bulb to permently dim it, similar to a dimmer or is there a bunch of other crap you have to do too? And if I can aprox how many ohms would I have to put in to dim it 50%? Same resistance of the bulb?

>>1856950
unfortunately, building a dimmer is significantly more complicated than that
your solution will work but all the energy that would have gone towards lighting the bulb will instead just be dissipated in the resistor, it's extremely inefficient

>>1856950
Unless it's a old filament bulb, no.

>>1856840
looks like this one

>>1857002
example

>>1856943
with CAN and I2C you just need 2 communication lines
the sensors are just hooked up to power and those 2 traces

>>1856888
I never can get the range of amplification that I want out of those jfet VCA designs.
That circuit that w2aew shows is pretty interesting though, I've never seen that one before.
He does call for germanium diodes, so that's another increasingly hard to find part that's needed.

Fuck lead free solder

How well would a cloth mask filter out sniffing the fumes of solder?

>>1857043
that's not how it works chief

>tfw nothing explained to you what a depletion/enhancement mode MOSFET was
>tfw nothing impressed the importance/significance of the difference on you, but only mentioned the terms
>tfw you just ignored the terms until now
>tfw you start building a project and realize the entire thing is fucked because you designed it around enhancement-mode MOSFETs but you used depletion-mode MOSFETs
>tfw no one told you there are actually 4 types of MOSFET instead of 2

>>1857050
An important distinction, although depletion-mode mosfets aren't all that common to be honest.
I'm surprised you manage to buy the wrong kind by mistake.

>>1856858
No, it's not going to LPF your signal. It'll look like this.

>>1856871
I disagree personally, it's much quicker for me to squirt out some solder paste and shoot a heat gun at a TSSOP than to solder each pin of a DIP individually.

>>1857060
it's because you are a richfag

>>1857068
how do we even have any poorfags in this thread? isn't anyone here either an EE or at least a middle class student in school to become one?

>>1857070
this shit's almost as expensive as my motorcycles
the only thing more expensive is radio bullshit

>>1857087
>almost as expensive as my motorcycles
dirt bikes don't count anon. i doubt i've dropped over 2k on this hobby and i can solder qfns all day.

>>1857068
SMT parts are cheaper. If you're talking about the heat gun then you can explain to me how you remove a DIP with a soldering iron

>>1857091
I've only ridden street bikes except when I took my grom into a dirt parking lot and promptly ate shit at 10 miles an hour

>>1857070
I use only salvaged shit and I'm going to be unemployed when I gradute.
T. EE brazilian student

>>1857103
any tips for the post apocalypse?

>>1857117
I plan on killing myself soon desu

>>1857037
>>1856888
After re-creating w2aew's circuit, I can say that while it works great as a AGC circuit, the linearity is pretty shit if you try to use it as a VCA.
It gets loads of harmonics if you try to feed it anything other than the signal's envelope.

is it normal to feel like a total fucking brainlet playing shenzhen io

>>1856600
>wouldn't the output capactior together with the pull up resistors act as low pass filter?
Anon, everything is a low pass filter. That's why we don't have 200ghz CPUs.
>How do you avoid this?
Design your circuits so the knee is at a higher frequency than any signals you care about. In this case though the ideal circuit has a low pass filter at 0hz (i.e. power supply provides DC with no AC component) so the low pass filter is a good thing. That's why they tell you to put the capacitor there.

>>1857227
Let's put it this way. I've heard full time electronic engineers avoid this game solely because they said it is literally their full-time job in a form of a game.

somewhat related to /ohm/

I need to make 5 of these stands that are basically small paper towel holders, but the center column needs to radiate a consistent column of light (versus the strongest point of light being just at the bottom for a flat base)

My closest idea would be to 3d print the stand with a hollow base, and just glue on a basic led strip kit that comes with the power supply. Anybody have a smarter idea? unless 6" tall led bulbs exist this seems like the easiest route

>>1857241
Including dealing with incomplete/useless datasheets.
(Looking at you PGA33X6, and the undocumented instructions)

>>1857253
I'm still a student and this game already hits way too close to home

>>1857248
You could use a fluorescent tube, although it would be somewhat harder to drive.
There’s also filament LEDs, pic related

I am looking into buying an oscilloscope because as I am doing more uC, I just can't fucking stand using a multimeter or using shoddy ass hacks anymore to measure output (like an LED).

I don't really need anything too powerful and I'm 90% sure that my needs can be met with a recommended entry level scope, but do you guys have any recommendations? My only requirements is that it should be a 3 or 4 channel (I already have 3 o-scope probes from an ex-coworker), o-scope and it must be light and can record waveforms (I suppose that is pretty much standard nowadays).

>>1857254
I honestly didn’t know what that was until I took a class in digital logic.
It’s a programable logic array, with a SR latch on board, 3 inputs and one output.
Once you know that it’s actually pretty straightforward to use.
It’s a shame the game makes it hard to justify it’s use, due to cost and power consumption, because it’s actually pretty fun to design with.

I'm gathering the parts to built pic related from a previous thread. I went to the site on the pic, but I couldn't determine what voltage of capacitor to get. Should I just get the lowest voltage I can find? The LEDs that I am getting are only rated for 2v so I'm guessing I don't want a capacitor higher than that?

>>1857336
This is the circuit.

>>1857336
>>1857337
Get a capacitor with a higher voltage rating than that capacitor will ever be exposed to. For your circuit literally any capacitor is fine. 10V, 16V, 25V, 50V, 100V, etc. All will work there will be no difference between them. You never want to go lower, always higher and besides you will never manage to find a 2 or 3V rated capacitor anyway.

>>1857037
>I've never seen that one before
Actually you probably have, since I'm pretty sure it's how a ring diode mixer works.
As for linearity, I'd ditch the NPNs and what, using op-amps instead. I think you'd be able to get passable linearity just with that. Using a 2-diode op-amp peak detector circuit too.

>>1857336
Buy a reasonably cheap cap with the capacitance you need. 2V is way too low to be common, and it's unlikely that you'll be able to reuse those caps for another project. 16V, 25V, and 50V are the most common in my experience.

Can I zap my cock into instant wood with a diy electronic device?

>>1857343
it would be extremely painful

>>1857344
Post circuit plz

Is it possible to do a transfomerless mixer? Only circuit topology I've found that can do it is the Gilbert cell but I can't use the ICs because I'm using large input signal levels (>1V) and I want unity gain. I've tried discrete Gilbert cell and single balanced mixer designs with somewhat limited success, I think I'm not getting the biasing right.
I think in theory cascode mixers should be transformerless but in practice they never seem to work right. OTA based designs work but only if I significantly drop the LO frequency from what I want to use.

>>1857351
Just zap it bruh

>>1857351
Could always use a 1/10 voltage divider, then the gilbert cell, then a 10x amplifier. Is this for an audio circuit?

Making a single quadrant diode-based multiplier like I described here >>1856888, >>1857135, >>1857341 should work so long as you aim for maximum linearity with your part selection. The same sort of circuit should work for making a linear JFET design too. I think you can use an AGC circuit for negative feedback to get yourself linearity, somehow. But it might be a convoluted mirror circuit.
I can put pencil to paper and see if I can think out a circuit that will fulfil this condition, in an hour or so.

>>1857351
>>1857357
Wait, unity gain doesn't even make sense with a mixer. there's always going to be some scale factor because ___V * ___V = ____V^2, and the actual output is in V units. Should 4V * 5V = 20V?

Anyhow you could also put your signals through a log amplifier, add them, and then put them through an exponential amplifier. But that's probably pretty nonlinear too. It's also a single quadrant mixer, so you'd need proper biasing, maybe with the help of a diode clamp.

I have a basic soldering iron like pic related but Radio Shack brand (got it a few years before they went tits up). What is the max thickness wire this should be able to do? I believe it's 50w.
I've tried soldering 18ga wire with it and it just doesn't seem to get the wires hot enough, but the surrounding insulation does heat up. Also tried doing it to bike brake cables and nothing. The tip (chisel tip) also won't tin even after I sanded it down to clean it. I have thin leaded solder.

Is it the iron or did I somehow fuck up installing the tip or something?

>>1857459
You can't get solder to melt to the wires (tinning), or are you trying to solder without tinning? If it's the former and not doing 18awg, it's trash.

if electrons flow along the outside of a conductor via the skin effect, why don't they extrude wires into the shape of space filling curves to maximize surface area

>>1857464
because it only matters at certain frequencies, and not everyone operates at those frequencies. And stranded cabling does that without incurring retarded manufacturing costs. google litz wire

>>1857470
you know that phenomenon where you learn about a word and then you start seeing it everywhere? I read about litz wire in my ARRL antenna book and now your post is the 5th place I've seen it since learning about it 20 minutes ago

I guess in a way stranded wire IS space-filling so that answers that

>>1857460
This is my step by step:
I heat the iron up at max, try to tin the tip and it just beads off.
I hold the iron (chisel tip) to the underside of the wires after I have stripped, interwoven, and twisted them together. After waiting quite a while, I try to apply the solder to the top side of the wires. It's rosin-core and does smoke when it touches the iron itself, but doesn't melt to the wires when I touch to them unless I try exactly right beside the iron - difficult without having something to brace myself against to stop hand shaking.

I'm wondering if it's something I messed up to affect the heat transfer or if the iron itself is crap. Mine looks exactly like the pic but has temp numbers on the dial and is red instead of blue. It does work to solder smaller electronics components. 18ga and 16ga are what I think I'm working with. Wiring inside a tool and lamp cord mostly.

>>1857480
>This is my step by step:
>I heat the iron up at max, try to tin the tip and it just beads off.
You don't need to be at max. Your tip is probably coated in oxidation. If your top isn't taking solder, you need to clean it, ideally with a brass scrubber since it's probably caked in shit.

>I hold the iron (chisel tip) to the underside of the wires after I have stripped, interwoven, and twisted them together. After waiting quite a while, I try to apply the solder to the top side of the wires. It's rosin-core and does smoke when it touches the iron itself, but doesn't melt to the wires when I touch to them unless I try exactly right beside the iron - difficult without having something to brace myself against to stop hand shaking.
If your tip isn't tinned, specifically coated in enough solder to create a heat-bridge between the wire and the iron, the heat transfer sucks balls.

tl;dr your tip is probably fucked. It's worth getting some CCC (Yihua, Aoyue) off ebay for 25 bucks that has temp control. It will come with a new tip, make sure you tin it first.

>>1857261
afaik the ds1054z is still king for hobbyists. i have it and i like it.

>>1857070
Nope, I'm poor as shit and got lucky. When boomers see the drive and passion it opens a lot of doors. Besides, booze costs more.

>>1857459
>The tip (chisel tip) also won't tin even after I sanded it down to clean it
Modern soldering irons aren't meant to be sanded, they have a sub-mm plating on the tip that's meant to wet well to solder and doesn't take well to being scratched. So you probably fucked up and will have to replace the tip, though it's not necessarily useless as is.
If your tip has a dark colored patina on it, scrub at it with some brass/copper wool, it's indispensable for keeping a healthy tip. Scrub it while it's hot, then add a dab of solder to tin the tip, cleaning and/or adding extra flux to the tip when necessary.

What you do need is soldering flux. A nice piece of solid rosin, or otherwise some rosin paste, is absolutely necessary for all but the worst of joints. Most solder wire will include a rosin flux core, but this is sometimes insufficient, especially with hotter temperatures. Note that having a temperature too high will burn off all the flux too quickly. Setting my iron to 300-350C is what I find to be appropriate for 95% of all work. 400-420C for aluminium-backed PCBs.
I also like to add extra flux paste to the joint before I solder it, and have a small bead of solder stuck to my tip. The small bead of solder helps make proper thermal contact with the joint.

50W should be more than enough to solder some decently large components, so I don't think power is the problem.

>>1857523
>booze costs more
100% this, people who don't have hobbies but still drink are just wasting their money

>>1857070
I'm a military electrician of the nuclear variety

>>1857592
I do electronics drunk like a retard so it's a double whammy. PCB design goes much better with beer because I hate it.

>>1857484
>You don't need to be at max.
I didn't know that, I assumed the lower settings were for smaller work cause the manual it came with was pretty poor.
>Your tip is probably coated in oxidation
That's why I sanded it, but according to the other anon this was a bad idea so I'll just swap the tip to a new one I guess.
>specifically coated in enough solder to create a heat-bridge between the wire and the iron, the heat transfer sucks balls.
Nobody explained why you tinned the tip before, I assumed it helped the solder flow through some sort of pooling but this makes sense, thank you.
>It's worth getting some CCC (Yihua, Aoyue) off ebay
I don't buy chinese shit anymore.

>>1857589
>they have a sub-mm plating on the tip that's meant to wet well to solder and doesn't take well to being scratched
>brass/copper/bronze wool
Well shit. Nobody told me this. I did buy "copper wool" before but it was actually false-advertised copper-colored steel wool.
>Note that having a temperature too high will burn off all the flux too quickly.
I suppose that explains some of it too. I knew the smoke was the flux but I wasn't aware that the iron could go too high for it.

>Setting my iron to 300-350C is what I find to be appropriate for 95% of all work. 400-420C for aluminium-backed PCBs.
>572-662F and 752-788F
That also applies to loose wire-to-wire, not on boards? Most of my needs are more "electrical" than "electronical" at the moment but I'll write those down.

Thank you both for the help.

>>1857601
i'm jealous, you're lined up for a cushy sinecure at a defense contractor as a process nazi.

>>1857615
a man can dream

>>1857613
just so you're aware, heat is obviously part of what you're doing, but also speeds oxidation. Being at max heat might be necessary in some cases, but not for 18 awg. You don't want to leave teh iron on at max. And always tin it before placing it in the holster. I mean when you are not using it for longer than a few seconds, re-tin the top to keep it from oxidizing. Not just when you are "done for the day" but when you are done for a few mins, you should tin it again.
Covering it with a huge blob of solder so big it almost comes off is fine. leaving it un tinned is a one way ticket to oxidation.
Basically the tip should almost always be tinned, either because you left it that way, or are actively feeding solder on to it. You can lightly wipe off excess solder before actively using it on a wet sponge or brass brush, but even before touching it to the part you want to solder, it should be tinned again, just lightly.
And yeah, the thing which transports the heat from the iron to the part is the solder itself, so it works in your favor.

>>1857613
Applies to anything. Only turn the temp up if you're having difficulty melting the solder. If it's melting but not sticking to the tip, then it certainly isn't too cold.

>>1857628
Thanks for the tip on the tip tinning.
>>1857625
Yea at some point in the past it must have oxidized because I didn't know to tin the tip the first time or two I used it, so when it wasn't melting on a wire I cranked it up to try to dump more heat into the workpiece. Or possibly when I used the "copper" wool I scratched it badly or something. I only use the thing like once every 6 months.

>>1857630
>Or possibly when I used the "copper" wool I scratched it badly or something
Probably not, I was using literal steel wool for a good 6 months and didn't notice too many issues.

>>1857340
>you will never manage to find a 2 or 3V rated capacitor anyway.
Most ultracapacitors are only rated like 2.7v, but they're pretty niche.
It's not like every project needs a 5000 farad capacitor (not, that's not a typo, I really mean 5000 farads).

>>1857672
5kF is enough that ESR becomes far more impactful than capacitance, no point in going above 1F for most uses. Only time I can see ultracaps being used practically is for long-term power storage where batteries are not desirable. But even then, these ultracaps don't have phenomenally low self-discharge currents.

>>1857341
>Actually you probably have, since I'm pretty sure it's how a ring diode mixer works.
Oh? To be honest, I never could wrap my head around how ring modulators work.
>As for linearity, I'd ditch the NPNs and what, using op-amps instead.
Sounds fair enough to me. Whenever I can't get something to work I usually just throw more gain and negative feedback at the problem.

>>1857351
Gilbert cells are usually the most precise topology for radio frequency transfomerless mixers, but cascodes are more common due to how cheap they are.
>I've tried discrete Gilbert cell [...] I think I'm not getting the biasing right.
Probably. Gilbert cells need notoriously precisely matched transistors (the IC Gilbert cells get laser trimmed).
I also have issues getting the bias right on a discrete Gilbert cell.
>I can't use the ICs because I'm using large input signal levels (>1V)
I'm with the other anon, just put a divider before and an amplifier afterwards.

>>1857360
I've designed a few mixers (on paper), where all the scale factors either cancel out or are one, so you wind up with 5V * 1V = 5V, 2V * 3V = 6V, etc.

>Anyhow you could also put your signals through a log amplifier, add them, and then put them through an exponential amplifier. But that's probably pretty nonlinear too.
I actually get pretty decent performance out of that design from what I'm seeing so far.
>It's also a single quadrant mixer
I solved that issue here >>1856877
The signal input gets 5V added to it, so any signal up to 5V peak will work, then the bias gets subtracted out in the end.
The control voltage multiplies when greater than zero, and divides when less than 0.

>topic about discrete mixer topologies
this is a good thread

>>1857682
>I'm with the other anon, just put a divider before and an amplifier afterwards.
Degrades SNR. Only reason I don't want to do it.

>>1857673
I usually see ultracaps being used in regenerative breaking on electric vehicles to buffer the power going into the batteries.

>ESR
I mean, it's certainly something to consider, but the 5000F ultracap I'm looking at is rated for 7.9kW.
Yes, that means it would take 4.6 seconds to fully charge or drain it at the maximum power rating, but 7.9kW is a lot of power to be able to supply from something the size of your fist.

>>1857688
>4.6 seconds to fully charge or drain it at the maximum power rating, but 7.9kW
>4.6 seconds
>7.9 kW hours
son, that's what we call an explosion

>>1857689
Not kW hours, just kW.
The total energy stored is 5000 * 2.7^2 = 36450 joules
Joules to watt hours is then 36450 / 3600 = 10.1 watt hours

So you could output 10 watts for about an hour, or 7.9kW in just 4.6 seconds.

>>1857688
>regenerative breaking on electric vehicles to buffer the power going into the batteries
That's actually a really good idea, and something that I may need to consider in the near future. But first I'll need to do testing without the supercaps to measure what kind of currents I see.
But for that you'd need to have a higher voltage rating than just 2V, it would need to be able to handle the maximum cell voltage of the lithium ion battery, right?

>>1857717
https://www.youtube.com/watch?v=gTt_YBzJ_Dk
Clickbai ttitle but they are real student/researchers with good projects.

>>1857782
Yeah I saw the one with the alarm clock, but from what I remember the leakage current wasn't tested, something I expect to be relatively high.

>>1857785
I'm sorry, but what do you want to do anyway? How did this talk went from this:
>>1857340
To super capacitors and regenerative braking?

I want to take out a working HDMI controller IC from a board and put it into another board. Should I use hot air gun or low melting point solder? I'm a noob at SMD soldering.

>>1856282
Component list?

is there a software or app where i can add all the components i have and it shows me projects possible with these.

>>1857823
It's called brain. If your list it means you have no creativity, knowledge or are stupid.

>>1857822

both options require skills you dont possess. recommenced watching Louis Rossmann videos on Youtube until you delude yourself into believing you can do this. e.g. https://www.youtube.com/watch?v=mr1UVPsExiE

>>1856433
think of it in 3d. is it bad to have it Xmm from i2c?

>>1857344
4u

Why are most online calculators are so bad? After checking 5 or 10 of them, I always end up writing my own. They are either written by complete retards or by students/interns and the implementation is usually so half assed, they clearly don't give a shit. Their managers probably heard from others that it is cool to have some kind of a calculator on their website, so they add one but they don't actually use it or hardly even test it.

>>1857879

pajeets get work done cheep cheep. and dont wear hiking boots in the machine room.

>>1857879
Which calculators? Also
>free web programming crap
top leek. The power electronics one are usually ok I guess.
> and dont wear hiking boots in the machine room
Indians don't even hike lmao. The good thing about going hiking is that if you encounter someone there is almost 0% chance of them not being arabic, chinese or pajeet

I always see power supplies with +/-15V and +/-20V, but the only cheap +/- rail supplies I can find are PC psu's
Besides buying a $200 lab supply, where do hobbiests even get >|+/-15V| supplies?

>>1857924
two $10 chinese switchmode supplies (once you've checked that they're isolated) wired in series

>>1857921
>Which calculators?
for example copper resistance calculator. half of them don't even bother to calculate the cross sectional area from the wire gauge or at least from the diameter. they want you to do the math first. And I couldn't find a single one that also accounts for temperature. And lots of them don't bother to convert between different units, so you are supposed to input 10km in mm or 10 uA in amperes and other bullshit.

>>1857970
Dude I really know what you mean but do you really need a online calculator for a linear formula with 1 variable and 3 parameters? just use excel lmao

>>1857924
Any $50 lab bench power supply can do +/- rails. The digital readout or the analog needle will only show a positive voltage but if you want a negative you just use the red (+) terminal as the ground reference and the black (-) terminal as the VSS rail. Now you have a negative voltage output. To get both positive and negative you need a dual channel supply.

Anyone ever powered a circuit by USB off a phone's battery? I'm making a pocket sized headphone amp and I want to use USB on-the-go or to make the phone the master

>>1857685
>>1857351
I though you anons might find this interesting, found it a few days ago while trying to learn more about SSB radio design.
http://www.radiohamtech.com/mixer%20design.pdf

>>1857351
you can get some pretty shitty mixing with a diode only. But no way for 1V, usually way more than that.

>>1858030
Less*

How would I go about creating a safe source of 1MHz ~10W sine wave? I'm thinking this will be much cheaper DIY than buying it premade. It's for striking a cold plasma.

>>1858026
yes

>>1858054
You could try generating the 1MHz wave at a low voltage and then use a power amplifier to get your 10 watts.
This paper describes a circuit to do exactly that:
https://www.wseas.org/multimedia/journals/circuits/2015/a825801-481.pdf

Essentially I have a microcontroller and I want to detect if its power source is from usb or an external source. How would I go about this. The microcontroller operates at 3.3v but both usb and my source are 5v. I tried using a zener diode (1n4729) which I have on hand but it only dropped the voltage to 4v.

>>1858072
put a resistor divider on the USB power to ground, and feed the middle to an ADC pin. You probably need a divider as many ADCs only work over limited range.

I suspect you mean something else by detect.

IF your board is open source, look at the schematic. There's a voltage regulator between the 5V input and power to the MCU. If it's not open source, trace the 5V line into a voltage regulator.

>>1858069
>UoA
>all names are indian
checks out

So I got an old analog AC voltmeter which is completely uncalibrated. I used a variable transformer and saw that it takes around 5 V to get to the max reading (240 V on the printed scale). I measured the current and it takes about .3 A to get to that. The principle of operation is that an electromagnet creates a magnetic field that influences the needle. Since the strength of the magnetic field is proportional to the current, this one must stay constant regardless of applied voltage. So getting that meter to show 240 V requires a resistor in series and the total power is 72 W (240 * 0.3) so that resistor is going to be dissipating a shit ton. Shouldn't this voltmeter generally require a lot less power to operate to have less influence on the actual load?

My thought was to actually add an iron core to the magnet so that it requires less current for max deflection or alternatively use a different electromagnet which is stronger. Would that make sense?

>>1858121
Then the system would never calibrate back to original. Find the actual issue and fix that.

>>1858121
Shouldn’t it use more than one winding? 300mA sounds fine for a current range, but drawing that much from any voltage source is a bad way of measuring.

>>1858121
>I measured the current and it takes about .3 A to get to that

that's way off base. typical small meters, even very old ones, have full-scale currents of 50uA - 250uA. larger units can go a milliamp. but 300mA is off the scale, so to speak.

I know this isn't really electronics per se but I think its reasonable to ask here. How do I read the following chart? It is a wire ampacity chart. For example looking at the 14AWG row. Is that saying "A 14 AWG copper wire in a raceway reaches 75 degrees celcius when 20A of current is passed through it"?

Not entirely sure what raceway means, I'm guessing its just some channel the wires are in as opposed to having them exposed to "free air"

>>1858357
I think so, though it could also be that "it gets 75°C/75K hotter than its surroundings at 20A". But considering the raceway one gives a lower current value for the same temperature, the reasonable assumption would be that since raceway applications can be hotter and/or have worse internal ventilation, it would be the current required for the wire itself to reach 75°C absolute.
Unless the Torbarn guys install wiring into F1 cars or something.

But they're being a bit careful in my opinion. The XT60 connector commonly used with lithium ion batteries is rated to 60A and has a maximum wire size of 12AWG. In fact, I'm going to be using up to 50A through an XT60 with 12AWG in the near future myself.

>>1858355
Thanks, I wasn't sure what kinda values I should've expected, but what you say makes sense. It's not a small meter actually, it has a diameter of about 15 cm, but still.

The needle itself is wiggling around nicely when the whole thing is shaken, so I guess there's no issue with the hinge. So it must boil down to an issue with the magnet. I might just try and wind one myself, a drill and a bit of wire should be easy enough. Maybe see if I can decrease the distance to the needle.

If someone has an other idea, let me know.

>>1858466
The point is that the magnetic field of the coil torques against that of a permanent magnet, with a torsion spring as a return force, no? The spring is probably fine as it's sensitive enough to wiggle when you shake the meter, but perhaps that permanent magnet is off-axis or somehow demagnetised. Apart from that, I'd check for shorts in the coil. I think partial shorting is the only thing that can cause inductance to decrease, so if you figure out the number of windings via observation you could calculate the theoretical inductance and see if it gives you a proper value in the µA. Might want to measure the inductance and ESR initially in its fault condition though, for an apples-to-apples conversion. Technically, parallel self-capacitance of the windings might be a good way to classify how many of the windings are intact, but that's just a guess, and not worth attempting unless it's easier than measuring inductance.

I know how synchronous machines operate, but how the fuck do microwave turntable units work? From what I can see the coil just send fields in the axial direction. How the fuck does it spins?
http://rmdavidson.blogspot.com/2014/09/tyc-synchronous-ac-motor-teardown.html

>>1858519

See these tabs? Note that they are not the same size, not that I can explain why. But they are what make the magnetic field radial.

Beyond that you need to re-animate Nikola Tesla and beat his ass until he explains how AC motors work.

>>1858480
I opened it up again and what I previously thought was some gunk is actually the insulating material of the coil, so I'm pretty certain that this is the cause of this. Will wire this up from scratch and it will hopefully be good to go again. Cheers

>>1858523
So it's a shitty shaded synchronous motor? the tabs bend the field lines radially and from the rotor perspective it has radial-ish field? Cool. Thanks mate. I saw the tabs but thought their purpose was mechanical only.

>>1858525
money shot from the front for shits and giggles

>>1858525

never seen a meter with wire that crazy thick. it's usually thousands of turns of stuff half the size of a strand of hair. so, the 300mA full-scale no longer sounds outlandish.

>>1858666

Doesn't look like a factory winding either. No telling what someone modified it to measure.

>>1858357
whats your goal?

Why is it bad to put inductive loads at the emitter again? And does that apply to LC tanks?

>>1858403
they are rated for continuous duty in the code.
good luck with pushing 50A through 12 AWG 24h a day. it will melt eventually. for a 5 min flight that may be ok.

>>1857857
im not quite sure what you mean but i came up with an improved solution

>>1858671
yeah that's what I figured out now as well. I unwrapped it and it had 192 windings on it. It was part of some power supply somebody had custom made. The thing itself was pretty weird already, it had a transformer built in with several different windings (220v to 300, 6.3 and 2, or around that area) so it seemed to me like a transformer built for something like a tube circuit. Since I am a noob, I had some fun, wrapped a new wire around it which was around equally thick and came to 112 windings. Sure enough to reach max deflection it now takes around 0.6 A. Thus, to get down to 1 mA I would thus need around 60k turns, definitely not gonna do that one myself. I'll see if I can find a cheap replacement otherwise it goes in the bin.

Btw. I never realized what exactly was directly inside the solenoid and only saw it when taking it off completely. There indeed is a small plate in there which immediately reacts to my screwdriver. So that seems to be alright.

>>1858763
Trivial, get a coil winder if you are into radio or making your own transformers. Ali has one for ~$60 with a counter and it amazed me how much easier it made winding inductors and transformators.

https://youtu.be/OSJ61jrCV4I

how difficult would it be to mod an smd cap into a radial cap using copper wires?

>>1858723
The wire is silicone, but with any luck it will only be doing 50A for a minute at a time, when it’s climbing steeply. The first iteration should only pull that much current in the most extreme of circumstances, but my motor driver is specced to handle 50A cont. and I may try a motor with different kv. But if I do so I’ll try to pump water around the stator for cooling, so I may well run the water pipe alongside the power wires. Maybe they make water-jacketed wiring.

>>1858830
Pretty easy, but while soldering the leads to the PCB you'd likely reflow the joints between the cap and leads. So instead I’d use a breakout board of some kind, something you first solder to the PCB and then solder the cap to.

Course, if you’re just doing it for breadboarding then you don't need to worry about that, but you’ll run into metal fatigue of those solder joints pretty quickly.

>project calls for left
>I already have right
Could someone help me figure out what I need to add to the right?
Looks like 2 resistors, a capacitor, and an LED but I'm not sure what they do.
I don't need the LED

>>1858937

>>1858952
>>1858937
The LED is probably not needed, and neighter is the pull-up resistor on the LED (102).
The 103 (1k resistor) looks like a pull-up on the dataline. If your MCU has internal pull-ups, they will work fine.
The capacitor looks like it's power to ground so it's probably a by-pass cap, and probably a good idea, but skip-it if you don't have one and see if it works.
I assume this is for fucking around / prototyping?

>>1858962
>and neighter is the pull-up resistor on the LED (102).
i mean current limiting. I assume the bottom trace goes to VCC and not ground, btw. If you can verify that with a meter i think you're fine.

>>1858962
Just fucking around, could I get a ms paint schematic?
I could probably muster the parts.

>>1858937
no need for the LED or its corresponding R. according to some argie website that board costs 471 pesos and uses a VS1838 receiver. per the datasheet the other R/C are just a lowpass filter on VCC, the ~100nF and 100 ohms in this picture. realistically you only need the capacitor, and you can probably live even without that. but is it really living if you're in the hell known as latin america?

>>1858968
hat tip top kek

>>1856282
ok I got a stupid question about breadboards.

I've never made anything over 1amp so it's new to me and I'm basically making pic related.

3 Lithium batteries in series, going to have much more than 1amp which from what I understand is what breadboards are rated for.


When people say 1amp max on boards, they are talking about the connections between the holes on the boards themselves right? Like just using the holes as a support structure to actually join the pins on the other side with wire is fine?

>>1858937

you can make a direct replacement for the 3 pins but you gotta figure out the pinout. as you can see, there's no standard among these parts, so hope you have a part number. if you google what i did, it leads to a youtube video for testing one of these units (whick looks like yours) but it's one of those awful videos that wastes 5 minutes and explains almost nothing. maybe you can stand watching it more than the 2 minutes i did and get something from it other than contempt for dumb people.

>>1859020

if you run thick wires from pin to pin, you can get dozens of amps.

>>1859020
>When people say 1amp max on boards, they are talking about the connections between the holes on the boards themselves right?
yes. however, for most circuits voltage is constant and current is just whatever your devices need, not the maximum the batteries can provide. the correct answer for your application depends on what you're powering from the batteries, and thus how much current it's going to draw. if you're powering ESCs through a breadboard it's not going to work. that being said breadboards can often handle a lot more abuse than you'd think. if you're not afraid of a little smoke inhalation then you can do whatever the fuck you want with them.

>>1859024
>>1859028
So when people are talking about 1amp max, they are talking about these connections right? The fucking board isn't going to explode on me or some shit if I run 30amps?

I mean really I'm not running the current though the boards themselves now that I think about it, they are basically just holders for the plugs since the pins fit right into the holes, and I'll have thick gauge wire on the other side linking them

>>1858937
Watch yourself, the frequency that those things are tuned to differs by model. You might need to detect a 44kHz signal but only have a 32kHz signal. I think they use straight OOK, or maybe it’s Manchester. If you’re choosing your own emitter LED, then no problem. Just be sure to read the relevant datasheets. If you need to measure the frequency of an IR signal, an arduino measuring the time between pulses on a photodiode/phototransistor (or just a negative biased LED) should work just fine, assuming you don’t have a scope or logic analyser.

>>1859032
that's protoboard. it can handle a lot more current than a breadboard provided your solder joints don't suck. it can't handle 30A continuous though. if you're really just using it as a backbone and the current is in no way flowing through any of the copper traces then you're fine. be sure so choose the right wire gauge.

>>1859020
I’m also planning on making something like this, though just to put two lipos in series, I’d set some solid copper wire in the traces to beef them up. Not gonna use a breadboard because I’m not retarded.

>>1859039
oh, I'm fucking retarded then I've been calling those breadboards for years.

What's a breadboard then?

>>1859043
the terminology is inconsistent. pic related is a "solderless breadboard", now commonly just a breadboard. it has metal contacts internally that you can plug a component lead or a solid wire into. the resultant contact resistance limits the current to a few amps or so. i've put ~8A through one without issue but even that's questionable depending on your application.

>>1859047
ok I'll always called those test boards.

Well thankfully I have no friends so I never embarrassed myself

>>1859052
I've always*

If a switch is rated for 20 amps and 240v, can it run say 40amps at 120v?

or does it strictly need to stay under the rated amps regardless of the volts?

>>1859074
at the end of the day, it's power that causes shit to burn, not amperes

>>1859082
A separate but connected question would be, how would one wire a toggle switch that doesn't have the full amperage pass through the switch?

If that's a thing

>>1859082
uh...at the end of that switches day it's the amps flowing through the contacts which matters. Yes, flowing amps implies a voltage which then implies a power, but the thing that's actually transferring energy around would be the amps.

>>1859074
A switch rated for 240v means don't exceed 240V and any amps. This is possibly meaningless, but the reason is that, at higher voltages, you can have arcs across parts of the switch, which can damage the switch or whatever is connected to it.
Everything has a breakdown voltage after which they conduct due to ionization. I don't think that's the 240V rating on the switch, but it's certainly why things have voltage ratings.

Exceeding current ratings is always a bad idea. You can cause massive voltage spikes when moving metal contacts under the flow of high current, regardless of actually melting a part The switch may not mange the voltage spike.

It's probably fine but it really depends how expensive the thing connected to it is, and if you care about the risk of a fire at all, minimal though it may be.

>>1859114
That's what I figured, but I was also thinking in watts so I was conflicted, thanks for clearing that up.

>>1859114
cont from
>>1859109

I'm in need of a 40V 30A 20A and 15A switch for DC.
Having trouble finding one that's disproportionally expensive to my build so looking for a separate solution

>>1859074
They're two seperate values, for two distinct situations.
It can handle 240V when the switch is open (not conducting) without arcing, so of course it will handle 120V. So long as it's AC, that is. DC isolation is much lower because there's no zero-crossing to break arcs. This isolation value is proportional to the distance between the contacts, maybe a millimetre or so. Double the distance, double the voltage rating. The speed at which you move the switch also matters for breaking a potential arc quickly.
But when closed (conducting) a switch rated to 20A current probably can't handle 40A current. Because current carrying capability is inversely proportional to the resistance of the contacts when closed. It's also a function of isolation distance, because arcs sustain themselves better at high currents, particularly if it's DC.

>>1859115
The watts are being taken up by the load, not by the switch. The amount of power that the switch can handle is probably just a few watts before the plastic housing starts to soften. A sustained arc will lead to quite a bit of heating, which is why it needs to be avoided.

>>1859120
Relays are probably what you want to use here. Automotive relays are frequently specced to 20A, 40A, 60A, etc. But not usually at 40V. So I'd either use a very low resistance MOSFET, or have a look around what tech the higher voltage marine systems use.

What kind of load are you switching?

>>1859074
Diferent ratings for diferent purposes.
>Current = heating in the switch, magnetic things, welding of contacts
>voltage = breakdown voltage of insulators and arcing
You should stay inside both ratings

>>1859127
Main load will be 33.3V and 29.7A DC

If that's too tall a task though, I could break it up into 9 separate channels as long as I could switch them all on/off at the same time.

>>1859138
Dude just google 40A relay idk. It's not dificult.
https://pt.aliexpress.com/item/32873374927.html

>>1859147
http://www.msrelay.com/uploads/soft/cars/MAH.pdf
>max switching voltage: 80v DC

>>1859147
>>1859148
A relay is my backup but I really want an actual toggle switch which is why I'm having trouble.

>>1859149
Why you want to pass so much current through a switch? People literally invented relays and transistors so you would not need that and use a cheap switch and a cheap switching element instead of a big chungus

>>1859153
>>1859153
>>1859149
Answering this question :>>1859109

See pic

>>1859153
>>1859154
Because I'm a fucking retard and vastly misunderstood what a relay was apparently.

Thanks I'll start reading up

>>1859155
No problem man. Always draw a circuit to have a clear idea of what is happening and we are always here to help

I believe relays can handle much higher currents than switches (of similar dimensions) because they move the contacts a lot faster.

>>1859158
It is also that due the fact of 80% of a switch being part of the human interface and doesn't deal with current at all.

>>1856840
>MOV
bro isnt this a 1kv varistor? look on google for varistor 10d 1kv, i think is this one

>>1859276
MOV= Metal oxide varistor

Please help, I'm retarded. I found an old pair of wireless headphones but the lipo battery was dead and way below the undervoltage threshold. I did want to risk starting a fire by slow charging it with a bench power supply, especially since I have other batteries laying around. The markings on the dead battery lead me nowhere so I can't find a data sheet or any specs, pic related.
I replaced the dead battery with two 3.7v 70mah batteries in parallel, and everything works well.
After sealing the headphones back up I noticed a small pcb connected to the dead battery terminals with a singular component. I did not install that board across replacement batteries I used in parallel.
My question is am I going to start a fire?

>>1859317
Pic related is the board

>>1856282
I want to make a UV disinfectant box but it needs UV light at the 200 - 300 nm wavelength. Anywhere I can buy those? Or should I just go with a basic Edison screw UV bulb.

>>1859380
You can get individual LEDs or a pre-made COB. I think component stores like digi key sell individual diodes and lamp manufacturers sell COBs and bulbs.

>>1859318
post the whole board, ideally both sides if there are components on both sides. What you posted are manufacturing ids and not helpful for googling (unlike product IDs).

> My question is am I going to start a fire?
When you charge it, you might, it's probably a charge protection board.

>>1859380
UV light is categorized into 3 bands, A, B, and C. Google UV-C LEDs.

https://www.americanairandwater.com/uv-facts/uv-types.htm

>>1859401
>post the whole board
I'll post the whole control board in a minute. I'll have to crack the headphones back open. This the reverse side of the board that was between the battery and the control board.
I used a little silicone on the pads to insulate them a bit, it's probably still kind of wet.

>>1859394
Thanks I'll look into that.
>>1859405
C is the range I'm looking for. between 200 and 280 nm.

How I read this? This is one of two caps, the other is the same on the first two lines, but the bottom is 14 20.

>Make a simple circuit for audio effects/guitar pedals
>Add voltage control for Eurorack
>The circuit complexity quadruples
I wish it was easier to make things voltage controlled.

>>1859380
A standard UVC mercury lamp is the cheapest way to go by far. There are UVC LEDs, but they're quite expensive and not even as electrically efficient. It's also possible to make UVC with electrical discharges of various sorts, but I think a UVC mercury lamp is still the way to go. Edison screw bulbs are probably the easiest ones to mount.
Such products do exist already, in particular I think you can buy them to put your toothbrush inside, but whether there exists such ones to put larger items in is unknown to me. Some UV boxes are for the purpose of curing resins and photoresists, which may or may not use UVC, depending on the resin/photoresist in question. UVC light from a mercury lamp looks like a pale blue, while UVA light from a 405nm LED looks like a vibrant purple, so you should be able to tell the difference.

But note that UVC will degrade plastics, and it's only good for surface-level stuff. If there's occlusion or multiple layers, it won't get through them, which is why I'd also consider ozone. I think you may be able to buy ozone generating ceramic plates, like the one in bigclive's recent video on a teardown of a BMW(?) AC air purifier.

>>1859552
use vactrols lmao
or H11F1s
or maybe non-isolated jfets

>>1859552
what did you do for voltage control?
i've seen some ridiculously complicated schematics based on an OTA and a shit load of other parts.

>>1859556
I know, vactrols are easy to use, but the paranoid part of me always worries about those becoming harder to source since they run afoul of RoHS due to their use of cadmium.

>H11F1
That's a $3 part, lol. For that price I can get several LM13700s and get better results anyways.

>non-isolated jfets
I never can get that to work.
It works well enough if the voltages are strictly negative or strictly positive, but I can't figure out how to get that to work for bipolar signals.

>>1859560
Usually just op-amp adder/subtractor circuits when I can, but I resort to OTA voltage-controlled resistors when I can't figure out any other way to do it.

>>1859556
>>1859571
>H11F1
why can't use just any optcoupler? why this particular part.

>>1859571
>I never can get that to work.
So long as your biasing voltage is of significantly greater magnitude compared to the signal going through them, and so long as one JFET lead is tied to a rail, my gut tells me it shouldn't be an issue. Course, this might require a divider, then the JFET circuit, then a fixed gain amp.

>>1859585
Because it uses a JFET as its coupling transistor, not a BJT or MOSFET. JFETs act like voltage controlled resistors. Current through the LED is (inversely?) proportional to resistance, without having to worry about biasing. They're not common.
But it might be possible to use an optocoupler with a photodiode (or 3-legged phototransistor's base-emitter junction IIRC) operating in the photovoltaic mode to bias a discrete JFET for cheaper. The current you'd get out of such an opto would be piddly, but possibly sufficient to bias a JFET.

Any advice removing stripped screws from electronics?

Sorry if this is off-topic.

>>1859592
By the time I have to throw a divider and an amplifier into the mix, I'd rather just go with a LM13700 and forget all that JFET nonsense, to be completely honest.

>not a [...] MOSFET. JFETs act like voltage controlled resistors
Correct me if I'm wrong, but I though all FETs had an ohmic region (also known as the linear or triode mode).
For that matter, I never was clear on why JFETs seem to be preferred in these VCA designs instead of a depletion mode MOSFET.

>>1859602
If it's just a little stripped you can try putting a rubber band between the screw and your screwdriver.
If it's really stripped you'll probably need to just drill the head off.

>>1859611
Thank you for the advice. I've already tried the rubber band tactic but it didn't work. I guess I need to go find myself a drill then.

>>1859607
Well JFETs are non-polarised so they can be used without the body diode messing up your signal. I think that's the main reason.

>>1859620
left-hand drills are meant for this very purpose, though I'd also consider an impact driver if you have one

>>1859630
Oh right, I kinda forgot about the body diode.

>>1859592
>require a divider, then the JFET circuit, then a fixed gain amp.
yikes
>>1859607
> LM13700
and a bunch of other components to go with it. so another yikes.
why are VCRs so difficult to implement in a simple way?? in 2020.
sounds like an optocoupler or even a discrete LED + photo diode is the best way.
i vaguely remember something like that in guitar pedals, basically feeding a PWM into an LED?

>>1859675
>why are VCRs so difficult to implement in a simple way??
linearity lmao
>and a bunch of other components to go with it. so another yikes.
yeah like 3-4 resistors. Faggot.

the girls are fightinngggggg

>>1859684
>lmao
whats so lmao about it, underage retard? it is a serious board.

>yeah like 3-4 resistors.
yeah 3-4. or maybe 5. and maybe a couple of transistors. and just for fucking VCR. what if you need 10. dumbfuck.

>>1859697
> just for fucking VCR
i mean just for ONE fucking VCR.

>>1859697
Those transistors are integrated on the LM13700 die, they come with the OTA for free.

>5 resistors
Wow! Almost the exact same number of resistors as you'd need for any of these other VCR circuits which have worse linearity than the OTA based approach.

>what if you need 10
Resistors are cheap. The LM13700 is fairly cheap and comes with two OTAs per chip.
Also, if you need 10 fucking VCRs for a single circuit then you've done an awful job designing that circuit.
You usually only need one or two VCRs in key places, even when converting things which weren't meant to be voltage controlled in the first place.

>>1859675
>discrete LED + photo diode
Photo diode is even more directional than a MOSFET, you have to use two diodes in series and bias them like I mentioned like 2 times in this thread before. What you want is a discrete LED + LDR, since they're much more ohmic.

Could also use a digipot I guess.

>>1859702
>You usually only need one or two VCRs in key places
Not him, but IIRC a phaser effect circuit needs one VCR for each all-pass filter, and they often have like 8. But linearity isn't too much of a concern for a phaser, being within 10% is probably fine.

And what if you need a floating VCR? Next thing you know you have a dedicated board per VCR.

>>1859708
>what if you need a floating VCR
A properly designed op-amp circuit shouldn't require the VCR to be floating

>>1859705
wait maybe you can use a varactor instead and still get a voltage controlled all-pass filter

>>1859708
>6 resistors and a single LM13700
Wow, so hard.

Any precision screwdriver sets recommendations? Bought a cheap Baku, recommended by a friend, and one point broke within one fucking use.

>stand on ground
>die when you touch live wire
>isolate from ground
>destroy parts with ESD
Is there any environment where a worker or user has to take cautions from dangerous voltage and also care for sensitive parts? How do they manage?

>>1859730
That's not how ESD works. If you're isolated from ground and you touch an ESD sensitive part that's also isolated from ground all that happens is the ESD sensitive part is pulled up to the same voltage potential as you. If you touch the part and it's connected to ground then all your really doing is touching ground which means you die because you're also touching high voltage and you've made a current path.

>>1859730
protect your shit from ESD?

>>1859719
here is a good VCR. just a few wires.

>>1859734
>7 components, 6 of them being resistors, is equivalent to a board with hundreds of chips on it.

Even the phaser example given only needs like 8 VCRs, and a phaser is complex enough that the VCRs are actually a relatively small part of the complete circuit.

Question related to the OTA but not really to anything else, but why are we still using an OTA with an internal (effectively external) Darlington pair buffer? Why not just put two more buffer op-amps on die and internally connect them to the output of the OTAs? You can eliminate three connections on each side. Just whack the new OTA in a 12-pin DIP and you're golden (maybe 14-pin DIP with two NC pins since 12-pin is not really common). I'm sure the transistor buffers were a good idea when this chip was designed back in the 70s but nowadays it's antiquated.

New revision, call it the LM13900, I don't know.

>>1859747
yeah it seems like a waste. I can make my own damn darlington if I want, I'd rather take 3 OTA's per die

So I was thinking how I wanted to progress my hobby PCB prototyping. I'm in a 1st world country and can afford to have ying in the mainland send me cheap PCBs, but where's the fun in that.
I've bought a 3018 CNC for milling, and screenprinting materials for the soldermask + silkscreen.

>Here's my thought process:
1. Design in CAD (I like altium)
2. Test design
3. Output design
a. PCB traces layer
b. Soldermask layer
c. Silkmask layer
4. Convert layers to gcode
5. Mill out PCB board
6. Mill or dragknife cutout the soldermask layer in laminate plastic
7. Mill or dragknife cutout the silkmask layer in laminate plastic
8. Screenprint UV resin using soldermask stencil
9. Screenprint white paint using silkmask stencil

I can't see anywhere where a hobbyist is going to this level of autism other than backwater PCB houses in third world countries. I figure if I get the pipeline right, it'll be trivial to get decent quality PCBs within the hour.

Anyone have thoughts on this?

>>1859761
Oh good. I have some artwork for a stripline mm-wave LNA I've designed. Can you do RF stuff and controlled impedance traces? I'll pay the cost of shipping.

>>1859763
Sorry not a chance lmao, but I like the optimism. Will post updates on how it goes though I expect this to be a few painful months after work to get an automated pipeline I'm happy with. Key bit isn't to duplicate a PCB house but to have my 1-off boards looking much, much nicer.

>>1859747
>>1859753
Don't forget that the output of a OTA is a current, not a voltage, so you can't directly pipe the OTA output into an op-amp.
You need a load resistance to turn the current output into a voltage first, and that load resistance changes depending on the application of the OTA.

Also, some OTA circuits use the buffer as a part of the feedback path, and the circuit's actual output is before the buffer, such as the voltage-controlled resistor that's been brought up in the thread.

>>1859770
>Don't forget that the output of a OTA is a current, not a voltage
Well yeah, that's why you need a buffer. Current sources are always high impedance. You don't really need a transistor buffer specifically though. Feeding your current source into a non-inverting FET input op-amp buffer should work just as well and give you a nice low impedance voltage output. You could integrate all that on die, derive the supply from the power rails for the actual OTA and internally route the signal from the OTA to the buffer. Then you only need one pin for the output. I am also down with the other guy's idea of just putting three OTAs on chip and basically forcing you to use an external buffer as well.

>>1859782
>You don't really need a transistor buffer specifically though.
Duh
>internally route the signal from the OTA to the buffer. Then you only need one pin for the output.
>give you a nice low impedance voltage output
By definition, an OTA outputs a current as a function of the input voltages and the bias current, I_out = I_abc*(V1-V2)
If you only have a voltage output, then it's not a trans*conductance* amp, it's just a shittier version of an op-amp.

>>1859794
Yeah I know, you'd still keep the bias current pin. That doesn't go away. I made an error when I was looking at the LM13700 pin out. You'd actually be reducing the pin count with an op-amp by one pin per side. You could eliminate the buffer input pin and just route that internally. Attaching an external load resistance is unnecessary with an op-amp.

>>1859796
>bias current pin
I'd hope so, I_abc is essential to what makes OTAs so useful.
Did you mean the output current pin?

>Attaching an external load resistance is unnecessary with an op-amp.
Why do you say that? The load resistance is what determines the output current to buffer output voltage relationship, which varies between circuits.

>>1859552
Use LEDs and receivers. Makes everything easier. But yeah, I get you. Mixers and VCA are way too complicated and bad with discrete components.

>>1859607
>Correct me if I'm wrong, but I though all FETs had an ohmic region (also known as the linear or triode mode).
Yes. but for mosfets you can buy at the store that region is VERY very VERY steep (low resistance) and not very standard because they are made for power electronics.

>>1859823
>receivers
?

>>1859825
Steep doesn't imply low resistance on its own, just that the voltage required to get from comparatively low resistance to comparatively high resistance is really tiny, and not very consistent, and probably dependant on temperature. But with a matched pair of MOSFETs and an op-amp you could make a sort of resistance mirror, by feeding the gates the same voltage, using one FET for feedback, and the other as your resistance.
Also is there a name for that sort of thing? Where you approximate an actual feedback loop with two parts that should work the same? Like what those dual photodiode analogue optocouplers do. Phantom feedback is what I call it in my head.

>>1859826
Phototransistors, LDRs, photodiodes or even other LEDs with some amplification. I used a LED and LDR pair in a heat shrink tube in my pedals just to avoid dealing with OTAs, buying Jfets or an expensive IC. I even got to do a 2 stage phaser which took about 4 of those pairs and it worked well.

>>1859830
Pretty sure LDRs are the only ones that work for this. Photodiodes and phototransistors are directional and have an effective resistance that changes with voltage. Even if you could bias your signal to avoid the directionality issue, there's still the nonlinearity to worry about. LED+LDR is somewhat standard though, they're often called vactrols.

>>1859832
What do you mean with directionality? It worked fine for me. It was a single quadrant multiplier tho.

>>1859834
the directionality is the single-quadrant stuff.
what was your THD like?

>>1859829
>resistance mirror
Scratch that, it looks like MOSFETs aren't ohmic enough. Resistance changes with current even though gate voltage is the same.

>>1859835
not so good but the sound was great. LDRs are also single quadrant friend, what you said made no sense

>>1859837
>LDRs are also single quadrant
LDRs aren't polarised, so there's nothing stopping you from using them in a circuit with no voltage bias on the input signal; an input signal that swings to two quadrants. The signal feeding the LED would need biasing though, so it would be limited to two quadrants. Could chuck the LDR as a part of a standard inverting op-amp for example and it would work just fine.

>>1859838
Ah in that case yes, for some reason I was only thinking about voltage dividers

How hard would it be to make one of these?
I need a midi in sound out device that can syhnhesize a bunch of instruments, I've searched for arduino projects but I havent found anything similar, most results point me towards midi USB controllers.

/ohm/ approved yes/no?

>>1856282
Roll

>>1860090
For USB you want an MCU that can handle native USB, like a pro micro or STM32. Or just bitbang it through VUSB like mitxela does. Check out his YT channel, it might have what you're looking for.

>>1860137
Kinda. When they work they're great (I have one), but the build quality can be a bit of a shot in the dark. I've heard of people getting ones with poor solder joints on the internal PCB. If you already have another passable iron that can reflow such joints, I'd go for it.

>>1860150
>rail gun
Good luck bro

Alright, newby retard here. I'm looking for an electrical part but I don't know what it is or I can't find one.

I think it's a rotary switch however I need it to be 360 degrees with 8-12 positions (or around that, I don't want it the turn radius for a change to be too small). I suppose a rotary encoder would work too however the detents on encoders aren't usually strong enough and there are too many positions usually (normally ~20).

Again, the rotary switch needs to function 360 degrees and keep going all the way through, so you can turn it as much as you want.

The problem is I can't find one. I look up shit like "rotary switch 360 degrees", "rotary switch 360 degrees 8 position/12position" etc. etc. and I can't find shit on google or amazon. Any help would be appreciated.

>>1860310
okay, so as soon as i posted this and did some more looking i think i found a listing that matches.

https://www.amazon.com/uxcell-Split-Positions-Rotary-Switch/dp/B008MJWTN2/

The problem is, how am I suppose to know if the rotation is "unbound"? As in, when you turn 360 degrees, can you keep going around?

If someone could verify for my dumb ass it would be appreciated.

>>1859798
>The load resistance is what determines the output current to buffer output voltage relationship

It doesn't actually matter. Trust me I've done this experimentally. Using LM13700 on a breadboard configured as an amplitude modulator I've bypassed the internal Darlington buffer entirely and routed the the current output to an external TL072 configured as a non-inverting buffer. It works just fine.

>>1860310
>>1860313
What do you need it for?

>>1860316
I'm making something of a button box thingy for a military flight sim.

Basically, the rotary switch is going to control some settings for bomb release. The thing is, I'm going to map it to the sim in a way that doesn't matter the specific setting but is rather controlled by left/right motion of the switch. That means an encoder would be closer to the real function but again the encoder is too "fine" in the rotation and the detents aren't powerful enough for the tactile feel I want, so I'm just going to make a workaround with some rotary switches.

That means the rotary switch needs to be able to rotate without being stopped on reaching the last position, and needs to have the switches throughout the 360 degrees. I have some 8 position ones, however they completely stop once they reach the last detent (maybe ~270 degrees through)?

I think the one I linked is fine I just don't want to buy them and figure out that they have a detent. I don't know if that's normal for the ones like that because I'm a noob to this shit.

>>1860327
Figure out they have a detent stopping you from continuing from the "last" to the "first" switch, I mean****

>>1860327
You should search for a 'continuous rotation rotary switch'.
I most just see results for 6-position switches though.
I would recommend just using a rotary encoder and then adding your own detents to be honest.
All the detents are is little divots and a spring loaded ball bearing pressing into the divots.
You can also try modding a switch to remove the rotation stopper, it's usually just a little tab sticking up from the switch's contact deck.

>>1860327
I'd almost certainly use a rotary controller and make it less sensitive in software. You can do that with a 40 cent microcontroller and ten lines of code.

>>1860314
I call bullshit unless you’re just trying to describe circuits 2 or 3 from pic related in the most opaque way possible.

You can’t pipe the current output directly into an op-amp as shown in circuit 1 since op-amp inputs are high impedance. Due to the high impedance, the OTA’s output will shoot towards the power supply voltage in an attempt to push I_out into the op-amp’s input.

>>1860393
Pic

>>1860310
So long as you’re not limiting yourself to buying from chinky sources, you can buy rotary encoders specified by number of positions per rotation (or degrees per position) and also probably specified by detent strength. Search on digikey or wherever. With any luck, both values are included in the part name so you may just be able to search for them on alibay or amazon.

I connected a cheapo flashlight led that I really needed to a 9v battery and it smoked. Dead. So if I was to connect a 1k resistor to it would it limit it 1k or subtract that from 9v? Don't anything about electronics. Dumb me.

>>1860424
If the LED is simply an LED, it'll have a voltage drop of about 2V when it's operating normally. So 9V-2V=7V would be across the resistor. Gives you 7mA. Probably low if you're trying to give off a lot of light

>>1860434
More like 3V.

For the anon first messing about with electronics, here are some simple equations for you: V=I*R, voltage = current * resistance. A resistor has a constant resistance, but an LED is closer to a constant voltage drop; a large increase in current I will result in a relatively much smaller change in voltage. For the sake of mathematics we just simplify this to a constant voltage drop, 3V for blue or white, 2.5V for green, 2V for red. If you’re familiar with E=hf and the unit of electron-volts, 700nm red light is more or less 2eV, and since each electron emits a single photon, this should make sense.
Secondly, P=V*I, power = voltage * current. Increase the power rating of a part and it will pop. Read the datasheet for this. The most common resistors are made to handle 1/4W, while small LED packages (3mm and 5mm) are closer to 50mW, I think.

Is there any better way to perform modulo division with analog electronics?
This circuit works, but I don't think it would be feasible in real life since it relies on the voltage comparator's output level being symmetric and very precise.
I want to use this for CV processing and quantization, so a difference of a couple dozen millivolts would be a problem.

I love these fuckin things. 15kw? No problem.

>>1860536
Love that pretty blue glow too.
Don't stare too long though since that mercury vapor can put out some pretty intense UV-C.

>>1860327

you can get rotary thumb switches of either push-up-and-push-down-keys type, or rotate-wheel-with-your-thumb type. most are 10-postion, but any reasonable number of positions should be find-able. you can get BCD (4-bit binary) outputs or 1-of-10 outputs.

>>1860563
One will have to sell his kidneys to buy those. Just use an OLED instead

If I get really good at electrical how likely am I to be able to get a job? Or is the market over-saturated? How does it pay?

Do you ever use EE calculators that accept inches? How would you even enter inches in a field, as a decimal fraction? I haven't seen any calculators that would accept 5 1/8, how would even enter that? But entering 5.125 instead looks confusing. How is typically done? Or EE calculators are all metric and EE engineers in the US understand the metric system?

>>1860698
EE calculators are overwhelmingly metric. Occasionally you run into calculators for electricians that accept inches. If I have to use imperial for whatever reason I'll convert the fraction to a decimal
>5.125 instead looks confusing
It's better than 5 1/8
>EE engineers in the US understand the metric system
If they're not a boomer, yes

>EE engineers

>>1860542
Well aware. Pump enough through these and some x-rays are generated. Very cool tubes but fucking dangerous.

out of curiosity, is there such thing as an analog circuit that generates the laplace transform of an input? Fourier transform?

>why yes, I AM an Electrical EE Engineer how could you tell

Electronic brainlet here. How do I improve my understanding of electronics so that I don't have to google everything? I have some projects I want to work on but I am severely lacking in ability to carry them out. I want to understand how to create low frequency AC electronics (kHz range) without blowing myself and my equipment up. Nothing high power, maybe 10-50W.

>>1860784
Vocoders use a bank of 8 to 32 bandpass filters to analyze a signal, which can be viewed as a crude Fourier transform.

All-pass filters can be used to shift a signal +/-90º, which approximates a Hilbert transform.

>>1860799
oh fuck I completely forgot about vocoders, that's right
I always thought of them as a digital thing, I'll have to look into analog vocoders

>>1860794
Learn your basic building blocks, and keep a small collection of them written down for reference.

I've collected a handful of designs for current mirrors, current sources, differential amplifiers, etc.
Whenever I design something I just start slapping those building blocks together like Lego.
Once I get the circuit working I go back and try to figure out what can be simplified.

>>1860784
https://sound-au.com/project136.htm
>>1860787
the depression, unemployment and autism maybe
>electrical EE

>>1860721
Makes sense. However I periodically run across B(H) magnetization curves with some weird units for H, such as oersteds or Amp-turns per inch. Some of them are from old books but some modern time steel companies publish their graphs in weird units as well. Oh and B in kilolines per inch^2 is my favorite.

>>1860849
I had to use some of those non-SI units but on the only context where they are useful : studying motor design and functioning. Most SI stuff for magnetics is very out of scale for practical stuff.

I have 3x 12v items in here, and it looks like a cluster fuck. Can I get a 5A transformer in here and split the output to power the 2A router, 2A modem and .2A fan?

I'd use the splitter shown here: https://www.ledsupply.com/images/products/secondary/lo-x-21fm-x-1.jpg . Probably put in a second fan with the extra connection.

Maybe throw in some in-line 2am fuses, but unsure if necessary.

>>1860889
>Most SI stuff for magnetics is very out of scale for practical stuff
You know the unit of teslas? that's really in scale. The higher end Nd magnets commonly used in BLDCs have a magnetic field in the realm of 1T. Even if the unit is 15E-6 in SI, there's nothing wrong with adding an SI prefix before it like we already do with mA and µA, or µF and nF and pF. The whole "out of scale" argument is completely bunk.

>>1860889
>>1861096
yeah, i asked because i use all kinds of units in transformer calculations. and all of them can be "in scale" depending on the scale. but i sometimes have to convert to T and amp-t/m for the sake of sticking to the same units.

>>1861096
Doing math in non-base units is ass though.

Should I got for the Elegoo Arduino R3 Super Starter Kit ($30) or the Mega 2650 Most Complete for $60?

>>1861138
which is why i am adding conversion options to my tranny calc so no matter which units different magnetization curves use, i could just type the values as is and internally convert to SI and output results in SI as well.

>>1860980
>it looks like a cluster fuck.

sure, coz your box is too small. tear down the wall and put in a bigger box that'll hold some network attached storage, a firewall appliance, a MOCA adapter, a cable box, and a USB stick with your emergency porn stash.

but if you must, it's certainly possible to share a single wall wart, using one of those splitters (assuming all the power jacks are the same size), or just cutting and splicing wires. no fuses coz that's just for faggits.

also, consider velcro instead of tie-wraps. if you get the brand-name stuff, the adhesive will stick forever.

What's the most user friendly circuit building program for a retard just getting started?

>>1861256

this, plus multimeter, hands, and a brain. you get real results in real time, real measurements, and real world faults. a software emulator will gladly let you put 1 million amps thru a transistor and not think anything is wrong with that.

>>1861256
I know that autodesk has tinkercad, which is free.

>>1861264
How is simulink off mathcad?

>>1859154
What program is this?

>>1861266
Tinkercad's all I know about

>>1861270
http://www.falstad.com/circuit/circuitjs.html

>>1861263
>a software emulator will gladly let you put 1 million amps thru a transistor and not think anything is wrong with that.
>Just 1 million amps
Lol, I put an exa-amp through a JFET on falstad once by accident before the simulator crashed.

>>1861277
Thanks that's exactly what I was wanting in a program, just something to play around with.


Which I now realize was in the OP I just didn't see what it was with how the OP was worded my bad

i'm getting really tired of retarded aliexpress shiping costs

>one 1x1mm chip costs $1 in postage,
>2 1x1mm chips cost $4 in postage

>>1861252
I think I'm going to put all that other shit in another room. The previous owner had his modem in here, so I just followed suit.

I used velcro initially but the adhesive slicked off after a few months due to heat.

i want to accurately control a bldc motor (robotics application). at the moment im planning on using low kv motor, probably gm7008l from aliexpress.
somewhere down the line, im planning on controlling 6 or so bldc motors from single computer.
i found two popular controllers: odrive and vesc.
anyone has experience working with these? vesc seems to have easier to work with pc tools, but odrive is about 30% cheaper.

>>1861243
the cheaper kit. electronics is a self-driven hobby. the kit is just to see if you're into it and get basic foundational knowledge. after that you'll have to come up with your own projects, and you'll have to order new parts for those projects no matter how many widgets you've accumulated.

I built a 1A 5V buck converter so I can charge my phone while I'm riding my motorbike
It works as intended and I can charge my powerbank with it but I'm scared it will ruin my phone
looking at the charigng circuitry it's hard to tell how it works as the circuitry is under the can

I also tried to charge an old MP3 player but it didnt work

Is it safe?

>>1861385
yah I hate that shit, I just do multiple orders.

>>1861385
fuck, I thought they automatically did combined shipping pricing
I've been overpaying out the ass

>>1861536
>>1861536
>>1861536
>>1861536
>>1861536
>>1861536

>>1861529
only some of the shops do, other shops only do it partially where they combine 5x 2$ shipping into say $4

>>1861435
What do you mean by accurately? Also why BLDC for robotics applications?

>>1861513
shouldn't be any issues
Devices should have internal charging circuitry and as long as voltage is within limits it will simply stop charging once fully charged
For MP3 player - some devices might only begin charging if only approved chargers or communications are established. Apple does this shit and it's easy to implement this with usb type-c