/diy/ - Do It Yourself

>>1861435

>>1861536
I got two consumer UPS units and 15 12V packs of NIMH cells. Since charging methods differ I probably shouldn't try to stuff a pack into the UPS and expect it to work long-term, right?

>>1861580
Necessary mention - the lead acid cells in the UPSs were very dead when I found them.

>>1861580
The charging isn't really that different, just NiMH are more picky and want some fancy maintenance schedule and charge cycles.

You can trickle NiMH like lead acid, but if you aren't doing maintenance cycles to prevent memory effects and etc. the NiMH will have significantly reduced life.

So yes, but the battery technology is not like Li chemistries where it will burst into flames if you look at it too hard.

>>1861585
Oh yes, I'm aware of the relative safety of NiMH compared to Lithium. I don't *really* have a use for a UPS at the moment or in the foreseeable future, but if it's possible I'll keep them around until a use arises like with all the other junk I found. Cool.

So im trying to build a portable light out of 4 100W LEDs.

The power source is x3, 11.1v lipos in series for the LEDs, and 1 single 11.1v to run the smaller bits such as cooling and the relay power.

>the bottom circuit looks retarded because I don't know to to add an LM2596 in there to pull 12V 0.35A for the cooling pump


Assuming that everything looks reasonable (I'm not sure at all, any advice/notes would be appreciated) I have a few questions from tinkering.
Mainly, Why does the resistance to amps not match with that standard formula

If I understand it correctly it's R = (Vsource - Vdrop ) / (I through LED)

Which means:
0.1 = (33.3 - 30) / x
x = 33

So why does the program show only 3.4A across the LED when I have a 100m resistor?

What I was originally going to do was a 1 resistor to give me
1 = (33.3 - 30) / x
x = 3.3

3.3amps, what I was going for....but with 1ohm resistors the program shows 1.8 amps across the resistor.

I don't get it?

>>1861614
You probably shouldn't be using resistors for this. Not to mention that your lipos will output a voltage from 12.6 down to 9.6 as they discharge, the forward voltage of the LEDs will also change with temperature. Probably best to use a switching constant-current supply of some sort instead.

>>1861621
I considered that, this is mainly to keep cost down and in it's current state is a prototype more than anything.

My main concern currently isn't so much of the build itself but why what im seeing here isn't aligning with the formulas I know. More just confused than anything.

I'm sure it's something very simple im overlooking but yah, just confused

>>1861614
Wait I'm a bit confused with what you want to do. Why you chose R= 100m ? You choose an operating current and then pick a resistance value, not the other way around. Also if your leds are 100W, pushing 30A through them will give you 900W, not 100W. (30A*30V).

If you want to operate them at full capacity
>P=100W
>I=100W/30
I=3.3A
To chose a R:
>R=(33.3-30)/3.3=1 Ohm

>>1861654
Also to get the correct values with the current you want you will have to set the voltage drop to 28V (or close to it). Why? Because falstad says "Foward voltage drop @ 1 A" so if you set it to 30 and are running them with 3.3A the voltage will be over 30V at the operating point you chose

>>1861654
>Why you chose R= 100m ? You choose an operating current and then pick a resistance value, not the other way around.
I explained it in the post, I was just messing around with resistance values because from what I was understanding the values I was getting was not matching with the formula

> Also if your leds are 100W, pushing 30A through them will give you 900W, not 100W. (30A*30V).
I'm well aware, I never said otherwise?

>If you want to operate them at full capacity
>P=100W
>I=100W/30
>I=3.3A
>To chose a R:
>R=(33.3-30)/3.3=1 Ohm
I literally said this in the post


The only thing I'm confused about, and I asked specifically about is
>>1861643
Why is this build not putting out the numbers the formula says it should?

In the program I have it set to 30 drop, 0.1ohm resistor, and 33.3v

The program is giving me 3.4A / 32V to the LED, that doesn't add up so I'm obviously not understanding something here but I don't know what.

>>1861677
Try connecting the diode directly to the power source with a short circuit and looking at a diode curve. It might help you understand.

>>1861680
for a given voltage drop the diode has a current in it, or for a give current flowing in it, a certain voltage appears across it.

Resistors placed on diodes are safety devices, used to limit that current.

Falstad is just using it's internal model to determine what is the current for that voltage, using the leakage current and emission coef.

>>1861686
Welp there goes the mistery

>>1861677
I don't know what program that is, nor how it works, but if you are forcing 14 A out of the supply, the rest of the circuit has no choice but to compensate.

so no matter what, you're going to get 3.4A through each R-LED combo.
Same thing with the voltage, it's going to 33V no matter what.

My guess is you have the voltage supply wrong. IT should be solving for current given the known parameters.

From V=IR, 3.4A through a 100M R is a really small voltage, so the rest of that voltages is going across the LED.

What you're showing is consistent with forcing 14A instead of letting the circuit solve for the Amps it requires to operate.

I want a constant amplitude sine wave from a VCO, any recommendations?

>>1861706
So far I have an op-amp rectifier feeding a capacitor with a discharge resistor. This amplitude signal is fed into an op-amp, with the other input being tied to a reference or trimpot. This second op-amp feeds a JFET to ground that makes half of the divider of a non-inverting amplifier, the output of which is tied to the input of the rectifier. The whole thing would then be inserted as the final block in a chain of phase shifting op amp circuits, which also have JFETs in their feedback networks.

>>1861686
>Falstad is just using it's internal model to determine what is the current for that voltage, using the leakage current and emission coef.
I guess that's what I'm not understanding, because no matter how I play with the drop and resistors I'm not getting the numbers I think I should in the program.


1ohm resistor, 33.3v, 28 drop is giving me 2.9a across the led.

Which it should be well above that given the selection as far as I understand
1 = (33.3 - 28) / x
x = 5.3

I just don't understand how it's getting these numbers from what's set up

>>1861706
>constant amplitude sine wave from a VCO
Give up and use digital synthesis already. Mostly joking. I don't know off hand the answer to your question.

>>1861715
Are you saying you're setting these diodes to have a 30V forward voltage? So you have 3V across each resistor giving you 3A, and we'll say there's a rounding error somewhere

>>1861719
DDS is probably a bad way of going because this will be for an additive synthesiser. Wait, is square wave or triangle wave additive synthesis any good? I’m thinking triangle wave, because I’ll more easily be able to change phase without a reference waveform. With varactors instead of jfets, of course.

>>1861706
you mean automatic gain control?

I have 0 experience with electronics outside some basic soldering. I bought a kit to build a simple analog guitar pedal, which includes the PCB, all components, detailed instructions, and a pre-drilled aluminum enclosure. I don't even own an iron yet or a multimeter.

How fucked am I?

>>1861727
No they are 28v forward voltage in that pic, I even typed out the formuls
>1 = (33.3 - 28) / I

The main power line is 33.3V
all 4 LEDs are set to 28v forward
and the resistors are all 1 ohm

Yet for some reason that yields 2.9A, and I don't understand why

>>1861811
Because forward voltage isn’t some constant. It might be 28V at 1A as the other anons were alluding to, but at 2.9A it’s 32V. Trying to match this to your actual LED is going to be pretty tough without a custom spice file, though some small variant of what you’ve got is probably good enough for a proof of concept.

>>1861804
Should be pretty straightforward. But even if you think your current soldering skills will be sufficient, I’d recommend some extra reading just to get a fix on the importance of flux, the kind it temperature to use, proper tip care, etc.
>inb4 was going to try to twist or tape connections in place without solder
Don’t, unless the pedal is designed to add noise in the first place.

>>1861814
>Because forward voltage isn’t some constant. It might be 28V at 1A as the other anons were alluding to, but at 2.9A it’s 32V.

I understood it would be changing with heat, but I assumed since there was no temperature value or anything of the sort displayed that wasn't being considered by the program.


I actually notice now though at the bottom right it actually displayed the "new" forward voltage which fits perfectly and the numbers all make sense now

thank you

>>1861800
That's half of it, yes.

>>1861829
See the shockley diode equation for more quantitative reading, it's the voltage-current graph of basically any standard diode (not tunnel avalanche, zener, etc.). Since qualitative is all you really need to know, remember that it's an exponential, not ohmic at all. If you graphed the effective resistance, it would decrease as a function of increasing voltage. But more importantly than effective resistance: small increase in voltage = big increase in current. Hence why current limiting is used rather than voltage limiting. Well, that and the negative tempco.

Depending on your setup, just dropping a volt or two across a passive current-limiting resistor might not be sufficient to ward off thermal runaway (thermal runaway is a bad thing), so watch out for that. Using some power darlington transistors you could make a relatively simple linear current source. Maybe with some op-amps and current sense resistors too, since that hFE alone likely isn't very trustworthy to keep in the smoke.

>>1861832
idk. Maybe try unity buffer -> rectifier -> jfet in ohmic mode as the collector resistor on a BJT amplifier

>>1861811
Unless this LED is actually like 9 LEDs in series that forward voltage is ludicrously high. They'll almost always be close to 3V

>>1861868
He said it's a power LED, probably one of those COBs. Though they're normally closer to 50V or 12V or 24V, never seen a 30V one.

>>1861879
Oh, I didn't see that. Those are multiple junctions in one package, i.e. a bunch of LEDs in series

>>1861836
Probably better of using an LM317 configured as a constant current regulator. Simpler, fewer components, easy to design to the point where it's nearly retard proof and will handle the power with appropriate heatsinking.

>>1861536
What is the story with this Picture you posted?

>Successfully muntzing the fuck out of a circuit while maintaining performance
Is there any better feeling?

>>1861886
>LM317
I was thinking that too, but IIRC they're not good up to 3A and their dropout is higher than he'd probably want.

>>1861904
If you read certian datasheets they explain how to increase the amps for Lm317 but it requires additional components like a pass transistor.

>>1861829
Not with heat, with current. As I said in my first reply study what a diode curve is.

>>1861904
I was gonna say pass transistor but another anon already beat me to it.

>>1861715
And you are using 28 in the formula but the drop is 30! Not 28V. 28V is the drop with 1A, it is larger with 30V! learn what a diode curve is. Also if you are worried about the voltages and don't want to learn how diode works, use a voltage source instead of a diode.

>>1861906
>>1861927
Oh yeah, never thought about an LM317 as a negative feedback loop to put circuits inside. Makes sense when I consider the CC driver and audio amp circuits for it. I suppose you could do that to get a beefier constant voltage supply too. Probably as many possible circuits for it as there are for the TL431.
An LM317 likes to run above 1mA IIRC, between that and the 3A there might be room for a darlington's hFE.

>>1861942
Yeah but the problem that arises next is heat sinks and I dont know if something like that would be something your okay with.

Like as I recall on other websites when talking about a mosfet thats rated for 5A doesnt mean you can run the mosfet at 5 amps because its package is only rated to something lesser than that. And why I mentioned mosfet is not to confuse anyone reading its just the LM317 uses the same package like a through hole mosfet.

But have you checked for Driver chips on digikey or if you dont mind like Amazon?

how difficult would it be to make my own home assistant that can unlock/open my door for me with a raspberry pi or what ever i need to turn voice instructions into mechanical force on my door?

>>1861970
No I'm not the LED driving anon, I'm >>1861836.

TO-220s should be more than fine at 5A, they're basically the same as a D2PAK and I see those (and smaller) used for 50A motor drivers. The package itself just defines the exterior, most importantly the width of those pins (0.8mm or w/e), which can certainly handle high current. If the effective resistance from pin to pin is something like 10mΩ, then you can put dozens of amps through, so long as the bond wires and silicon can handle it.

>>1861614
your circuit shows 32v on the + of each LED and 0v on the -
(33 - 32) / 0.1 = 10A
working backwards
3.4 * 0.1 = 0.34V on the resistor

i don't know what your software is but something is screwy with it. battery internal resistance/current limit? simulated wire resistance? rounding error? no idea.

>>1861999
It's just that if you use that kind of output on falstad it rounds the values. If he hovered the mouse over the resistor it'd show the precise values.

>>1862016
Also
>year of our lord 2020 and people still blame software (even though spice is older than walking fowards) instead of asking themselves what they are doing wrong

>>1861890
it's an lm7805 5 volt linear regulator

>>1861890
I'm not a big electronics guy but i think it's some kind of people exterminator.

>>1862062
it's a firecracker

>>1862090
>>1862135
It's not nice insulting Arduino approved power supplies

I have a faulty speaker amplifier that I'm trying to fix.
Should I make a new thread or can I post in here?

I've followed some threads online for this Pioneer SA-6300 blowing fuses.
I disconnected the protection circuitry and fuses stopped blowing up.
Went on to test the output transistors.
I followed a video linked in a thread but the green ones read correctly using positive on B to the other two, but the black ones works in reverse fashion.
Figured this was because left and right channel. And so I believe they are still working.

That's where I'm stuck at right now.

>>1862267
The black ones are probably PNP devices while the greens are NPN. Or the oposite

>>1862282
Ah, then it makes sense. But that also means they are working and the fault is elsewhere?

If so I should probably make a thread at some vintage audio repair forum.

>>1862286
I can't tell man, no idea of what you are dealing or the kind of circuit. my guess they are two push-pull stages (one for each channel).

>>1862267
Can you see any obvious shorts?
Do any parts feel warmer than you would expect?

>>1862267
>I disconnected the protection circuitry and fuses stopped blowing up.
Why don't you bypass it and see if it works?

What's the exact fault, no sound? shitty sound? only noise? Popping? works for a few seconds/mins then dies?
You're focusing on components without giving any issue of the problem which makes it impossible to help diagnose.
It's a lot easier if you explain the problem and what you've done to investigate / troubleshoot then ask if some random component is working correctly without any info

>>1862267

in case you dont know, class B amplifiers are set to run JUST BELOW the point where the transistors are both turned on, and cause a short circuit, blowing the fuses. a DC quiescent current is adjusted at the factory so that the transistors are turned on a bit, but still in the safe zone. in the schematic, it looks like this value is 30mA coming out of the emitter of Q18. of course, over time, components change and this value may now be too high. you might wanna check it.

>>1861816
thanks anon. i will do extra reading to make sure i don't fuck up soldering.

>>1862330
Pretty sure a totem-pole can't fail short like you're alluding to. The base will always be at 0.7V above the emitter when the transistor is on, so it's impossible for both transistors to be turned on at once. With a class AB amp on the other hand, it's probably quite possible.

>>1862351

class AB is called class B by those in the know coz it's a stupid distinction to make.

>>1862351

class AB is called class B by those in the know coz it's a stupid distinction to make, worthy only of egghead academics. also, it's a pot, not a potentiometer, like pretentious assholes call them.

>>1862365
you posted a picture of a strict class-B, what the fuck did you expect?

is there a better way to store these?

>>1862372
>what the fuck did you expect?

more latitude, less autism.

>>1862372
>what the fuck did you expect?

more latitude, less autism.

>>1862386

no better way, but needs more compartments.

>>1862351
Yeah, totally possible on class AB if you do a shit job with the bias servo/rubber diode/Vbe multipler.
Here's an amp that never draws less than 30 watts.

>>1862386
i gave up and dumped all of my jumper wires in a drawer long ago

>>1862405
where can I find an explanation on how to calculate the resistor values for the bias part of this circuit?
Most textbooks fall short when tackling emiter followers and push-pull outputs

Is it possible to reuse this touchscreen?

>>1862456
The 220Ω resistors are fairly arbitrary to be honest, the exact value doesn't matter as long as the Vbe multiplier is supplied with enough current to work.

The Vbe multipler itself is fairly simple, you can basically treat it like a zener diode where the forward voltage drop is set by R1 and R2, hence the name rubber diode.
The voltage drop is V = (1 + R1/R2) * Vbe
Since push-pull outputs have a 2*Vbe gap at the zero crossing when both transistors are off, you want the bias to make sure that the two transistor's bases are always 2*Vbe apart to minimize the gap.
Following V = (1 + R1/R2) * Vbe, to get that bias you just need to set R1 equal to R2. V = (1 + R1/R2)*Vbe then becomes V = (1 + R/R)*Vbe = 2*Vbe
(Actually you want slightly less than 2*Vbe to avoid the aforementioned issues mentioned in this thread)
What you choose for R depends on what output impedance you require, which is a bit more complicated to calculate (see https://paginas.fe.up.pt/~fff/eBook/MDA/Mult_Vbe.html))
But in most cases you can just get away with using a value near 1kΩ.

I hope that made sense

>>1862470
Maybe? Getting a good connection with those carbon traces will be difficult
Based on the number of lines, I'm also guessing that it's a capacitive touchscreen rather than resistive.

>>1862477
I appreciate the input. I'm pretty new to electronics so I'm never sure how usable items are.
fyi the screen came from a ~10 year old toy that required a stylus to work - I'm not sure if that points towards it being a capacitive or resistive screen.

>>1862482
almost certainly capacitive

>>1862489
Not him, but by what arguments? IIRC simple resistive touch screens use only four wires, two for each axis, is that your reason? I was thinking that some might use more than 2 axis for "rollover" protection.

This might be a bit off-topic, but is Gray code solely a human construction, or is it of a more fundamental nature? I'm wondering if there are any cases where it crops up more easily than standard binary. Like particular ADC or DAC topologies that it works with, or even something more fundamental. I know quadrature encoding is sort of Gray code, and I suspect it works the same if you made a rotary encoder or something with more bits, but that only happens because you lay out the contacts in that particular manner.

I suspect it's possible to make a resistor ladder DAC that runs on Gray code, which could be useful. If the resistor arrangement happens to be even half as nice as the R2R layout I'd be happy, but I suspect otherwise. I don't think Gray code applies to successive approximation or ∆∑ ADCs at all.

>>1862519
what do you mean, fundamental? Like does it arise in nature somewhere other than humans?
Do you understand why grey code is useful?
Look at it's symmetry. The symmetry is what's "more fundamental". Grey codes pop out of a symmetrical system, and rotation is symmetrical.
Grey encoding is a tool to represent symmetry. It is not, in and of itself, fundamental.

>>1862557
>Grey encoding is a tool to represent symmetry
You should go delete the Wikipedia article and re-write it 100%. As it stands now it looks like a bunch of nonsense if you are correct.

>>1862564
I think you're mocking me but I legitimately can't tell

>>1862557
>what do you mean, fundamental? Like does it arise in nature somewhere other than humans?
Yes.
I don't understand the symmetry behind Gray code. Could you elaborate? The closest thing I can think of is in this pic on the right, but even then it's barely better than binary.

also sorry to be a grammar nazi, but the grey/gray spelling reform doesn't apply to a dude's name

>>1862564
>>1862568
Oops, thought that was aimed at me, gotta read slower.

Wow, guess you can make an ADC that intrinsically runs on Gray code. Never knew about vacuum tube ADCs before.

>>1862571
Re you original question: I'm not a mathematician, but to me Gray code is simply a forced construct to do one thing: represent each successive state by changing exactly one bit so that transitions between one state and another don't have the issues that can arise when multiple bits are changing.

If there is anything "natural" about that need I can't come up with an example. Natural systems that need extreme precision and repeatability, like DNA replication, don't seem to be an application for something purely digital like Gray code.

Also, to my sensitivity there is nothing elegant about Gray code other than the fact that it works perfectly. Natural systems tend to have a wonderfully beautiful elegance. Look at the natural logarithm and its base for example, and the myriad of ways that it appears in nature as well as math and engineering.

>>1862573
Well it wouldn't have to come out of nature, even a simple or elegant mathematical or geometrical description would suffice. My reference point being "significantly simpler or more elegant than binary". The "moving from vertex to vertex on an N-dimensional cube" probably doesn't count because A: it doesn't specify direction, and B: it feels like a tautology.

>>1862571
> also sorry to be a grammar nazi, but the grey/gray spelling reform doesn't apply to a dude's name
I couldn't give a single shit about this.

Look at the pattern. When you add a bit, how does the pattern grow?

It's symmetrical around the MSB you are encoding.

>>1862293
>>1862310
I'm not sure why I didn't try to test it after disconnecting the protection circuitry, maybe I'm retarded.
I'll solder the stuff back on and test with some other less important speakers.


>>1862330
What part is that exactly of the protection circuitry?

Is it reasonable to up a boost converter to convert 10A 12V DC to 10mA 10kV DC?
What problems arise from trying to boost to such a high voltage?
Does anyone have any good guides or info on high voltage DC power supplies?

>>1862583
Yeah but binary has somewhat similar symmetries. Definitely not as good, but remember these symmetries in Gray code are there because it was literally designed to be "Binary Reflected Gray Code", it's a product of how the code was generated/postulated in the first place. Hence that reflective symmetry is of a completely artificial nature, not a more fundamental one. There are other forms of Gray code that don't have that symmetry, like the "Balanced Gray Code", but they're not as easy to generate. The fact that the entire G(N) code makes up the first half of G(N+1) makes the BRGC easier to generate and more universal, the symmetry is just a product of this design. Many of the other Gray codes have imperfect or impossible implementations for particular N values, such as balanced or Beckett Gray codes.

>>1862588
What do you mean? It's a dumb idea all around, because you'd need transistor that can stand 10kv, and the duty cycle for that will be something silly as 99.999%.
>Does anyone have any good guides or info on high voltage DC power supplies?
what do you want to do?

>>1862586
>What part is that exactly of the protection circuitry?
That isn't. That is the last amplifier stage. What we assumed you showed us with the 4 transistor picture.

>>1862588
At 10kV you should be using a flyback converter with a transformer, not a boost converter with an inductor. It's going to be real tough to find a cheap switching MOSFET that can handle 10kV, but ones that can handle 600V are a dollar a dozen. Personally I bought a bag of IRF840s, they're not bad at all. Not as logic level as you may want, however. The winding ratio of the transformer should be somewhat more extreme than the FET's max voltage divided by the voltage you want to get to (say, 600V/10kV = 0.06, so make the ratio 0.045 or so; 1/22). On the high voltage side you'll just need a high voltage diode, probably a silicon stack diode. I've got a 10-pack of 20kV silicon stack diodes (forward voltage of 50V lmao) that I planned to use for a 200kV booster. Feedback is somewhat of an issue though, even at 10kV trying to put a resistive divider for feedback is going to require a rather high value resistor (i.e. more than 10MΩ), not to mention the isolation distance. So instead of using voltage feedback off the secondary, I'd just use the primary coil for feedback, just add a diode and a small filter cap. The voltage spikes on the primary should be exactly proportional to the voltage spikes on the secondary. Add a feedback winding if you'd prefer a lower level to deal with. Then just add a TL431 or a comparator or whatever alongside some resistors on the feedback rail. I assume you're familiar with SMPS design basics? Be it constant duty-cycle bang-bang controlling or something more sophisticated. Probably be able to use an off-the-shelf switching controller, if you don't want to mess about with op-amps and the like for a variable duty-cycle.

>>1862596
High voltage converters are usually isolated for two reasons:
>you get "free" gain with the winding ration
>you get isolation from the high voltage side+usually a ground
And your control signals can be in the reasonable range (e.g, 0-80%) and not something like 99.995%-99%

>>1862605
Plus high-ratio boost converters are naturally somewhat inefficient.

>>1862599
From a schematic I looked at, that image resembled that board the most.

I soldered transistors back on and both channels actually work, with the protection circuitry board disconnected.

I'll see if it's possible to make threads on that forum I found.

Whatever happened to Frys? Are they still alive?

>>1862596
>>1862603
Thanks fellas this has helped me. I'm not really savvy with electronics but I want to build an atmospheric needle plasma and articles on the topic suggests you need a few kV, either DC or low frequency AC. apparently a matching circuit isnt needed if you stick to low freq AC or DC pulsed.
essentially its for a surface treatment tool for additive manufacturing. if you can produce a light low temp plasma you can improve adhesion between dissimilar materials e.g. elastomer on PP/PE

>>1862737
I had a friend working on plasma and "wet-ability" of materials. Just salvage a neon tube transformer or something similar. Aliexpress also has cheap HV modules. You don't need high power for plasma.

>>1862743
Also search for "voltage multipliers". You can build ones that run off mains with just regular capacitors and 400v diodes. I have already made one that goes to 3.5kV or more (I coudn't really measure). You can use a negative and a positive one for an extra "doubling".

>>1862571
>it's barely better than binary.
it sure looks way easier to manufacture a mask for gray code, if you were stamping it out of sheet metal.

Hi, I need a DIP socket for a LM386 amplifier. What's the difference between those two kinds? Are there any practical difference?

>>1861536
reposting from chink shit general on /g/

I am a semiconductor faggot. Any people here ordering generic or special I2C and SPI devices regularly?
I'm looking especially for F-RAM storage that is cheap.
https://www.aliexpress.com/item/32822241108.html

They manipulate my search, sometimes items like these won't even show if I search for the exact part number. I fucking hate Aliexpress sometimes. They control what they want you to buy. There were not even any limitations on this part for my country or payment method.

So hard to find semiconductors that have customer images proving it is genuine.
Most shit stores grind the part numbers off and sell you fakes.

Also any fram projects done here before?

hello /diy/,

I'm trying to repair a pro-user DRC4310 reverse car camera system that doesnt turn on at all (no screen no lights). Everything looks fine and the correct voltages seem present at all expected locations.
My question is about two components that are unknown to me; They both seem bad (assuming that thay are transistors) and I cannot find the datasheet for them. Does anyone here know what this part does? The part has "X74V 1A" written on it (see attached picture).

Thanks in advance.

>>1862800
the right one hurts a lot more when you step on it

>>1862800
forget those, use these. You can route resistors, jumpers and small caps under these

>>1862743
That's a great suggestion! Thank you very much I will probably buy one of these.

>>1862743
That sounds like a very intriguing investigation by your friend.

This button module for my racing sim rig wasnt turning on or doing anything from the factory, seems like its missing 3 components, presumably resistors given the "CR" prefix. Anyone know where i can find those "102" components?

>>1862960
They're just 1kΩ resistors, but those unpopulated resistors probably aren't just for the button. Beep it out with a multimeter and see what goes where.

>>1862830
Looks like a voltage regulator.

Is there anyway to reset a mcu through software. Like replicate the hardware reset pin. Im uploading firmware to the arduino micro. But it doesnt use the arduino environment. So the only way for me to program it currently is to press the reset button each time. Ive tried the watchdog method and the resetfunc method but I cannot get them to work. they stop the current program, but I still cannot upload a new hex file.

>>1862966
What programmer are you using? Usually the programmer will have an output that toggles the reset pin, at least that's what my USBasp does.

>>1862966
>>1862973

Right, responder. As for the questioner, you need to study AVRs a lot so that you can understand that there is no way for an AVR that is being programmed to at the same time have watchdog timers running or to do anything other than lay there like a bitch and say "PROGRAM ME BABY". In kinder terms, an AVR does not execute any firmware while new firmware is being written. (How is this not obvious)

So, questioner, how exactly are you uploading the firmware? USBTiny and USBASB and every other programming device has a line connected to your reset pin, and that's how serial downloading always works. Whatever you are using, either you are using it wrong (your cable is wrong), or whoever wrote it needs to go back to designing floral arrangements.

>>1862992
>>1862966

And if I haven't insulted you enough already, this sort of question might get gentler replies in the official tarduino general >>1844608

>>1862965
I'm the one that asked the question about the unknown part;
It seems a voltage regulator indeed, anon. When it's measured, it gives the same-ish values as the other (known) voltage regulators that are present.
But if the part is a voltage regulator, then there are two broken voltage regulators with exactly the same serial number and seemingly different functionalities.
It would be very coincidental (hence the question about what the part does), but it could certainly be possible.
I'll try and look for other possible defects as well, but this could very well be the problem.
Thanks for your take on this.

>>1863033
Check the capacitors at the output of the regulators. Tantalum caps usually fail short, so check those first.

>>1862992
I think you missed the point of my post.
>>1862973
Im not in front of my computer so I dont quite remember what pops up, but Im doing it through the console. Just plugging in the usb and command into the console works if you hit the reset button.

>>1861536
probably get laughed out of the thread but i want to add an outlet to a wall in my kitchen to run a fridge. i heard that i can simply run a cable from the light switch (on the other side of the wall) to the new outlet. what i dont get is how i get/connect the cable on the other side. do i just run the wall chaser up until im on the level of the switch and then drill trhough or what? or would it be better to go from a junction box.

Is it allowed to have about 2 inches of unsheathed wire outside a Jbox?

>>1863042
So you’re programming it with the native USB port on the dev board? I assume it’s using some sort of firmware to allow that, similar to how a digispark works but less hacky. I can only guess that in there in both cases the program is actually running but is able to write progmem while executing it, which really does sound hacky. As opposed to actually running the MCU in programming mode, though this only applies if the 32U4 doesn’t have facilities for direct USB programming. I’d recommend against this even if it does work, only with a dedicated programmer can you do things like setting the fuses for a different clock rate, or different brownout threshold, or even locking it to prevent reprogramming or reading progmrm, etc.

I assume you’re using AVRdude?

When I was in school and it came to transistors they would always say use 0.7Vbe. But what confused me is that data sheets would list it as something different but at a fixed current. Later on as I got into working with electronics I understood what the diode symbol on the multimeter was to which I then figure wouldnt it be more accurate to use the reading from the Forward drop voltage found with a mutlimeter?

Also on the subject of transistors what are some good resources on ACTUALLY using them such as accurately understanding what the data sheet says and how you bais?

When ever it came to actually trying to create a fixed voltage divider bias I gave up trying to figure out what vavules of fixed resistors and went with a pot instead

Has anyone realized that in the forest mims books especially the ones that cam with the radio shack kit have the basic start to a flash analogue to digital converter?

Im taking about that light meter circuit that uses those 4 package comparator opamp ICs. All you need to do in that is replace the leds with a logic circuit input and there you go. If I remember right you had to make a priority encoder logic circuit.

>>1863112
you can put all the unprotected wire you want outside the jbox baby all you gotta do have asbestos in your walls

>>1862572
That's using a beam deflection tube, it's basically an analog look-up table.
The input voltage sweeps the beam left to right, while a ramp generator sweeps the beam vertically to output the binary bits.
That could just as easily be normal binary as it is Gray code, so I wouldn't exactly call it intrinsic.

>>1862456

see figs 19-21 here: https://www.nutsvolts.com/magazine/article/bipolar_transistor_cookbook_part_2

>>1863209
At my school we only used Vbe=0.7v for convenience on simple problems.
For more complex problems we would actually have to break out all of the equations to find the q point, and keep in mind the Thevenin-eqivalent resistance of the biasing voltage divider.
Most importantly though, learn how to read the load-line graphs on the datasheets so you know where to set the q-point.

Is there any good way to perform modular division, or division with a remainder on analog hardware?

>>1863218
Yeah, not a great flash ADC, but a flash ADC nonetheless.
With 4 comparators that's 2 bits of resolution.
I don't think you could scale it much beyond 4-bits of resolution though, since that would require 16 comparators, and a 16 input priority encoder.

>>1862800

left socket is good enough. right one is cadillac model you can plug into thousands of times.

>>1863051

you wanna cut a hole in wall for the power socket so it's within the same pair of studs that the switch is between. before you do that, check that you can fish a wire downwards from the switch so it goes down close to the floor. if you cant, there's probably a horizontal bridge stud blocking your way, and you're fucked.
so, cut hole for socket, drive fish downwards, and have someone grab it at the hole. attach extension wires to fish, then pull out thru the hole.

one important thing before you do this: check that you have live and neutral at the switch. if you only have one romex cable running into the switch box (instead of two cables) then you have no neutral, so you cant use that as a power source.

>>1863218
>the basic start to a flash analogue to digital converter

this was a known technique for fast A/D before Mims was even born.

i'm trying to use an op-amp (LF353) as a comparator with a potentiometer. the voltage divider is sending 2.55V to the inverting input. it mostly works okay. the LED goes on between around 2.5-5V, as intended. below 2.5V the LED shuts off. however, between 0V and 0.9V the LED turns on again. do I need a pull down resistor somewhere or something?

>>1863246
Yeah I knew that it was hardly Gray-code specific as soon as I saw those cutouts in the mask, just I never thought about something along the lines of a TV tube / CRT that uses an electron beam as an ADC or lookup table. There's no real semiconductor equivalent to that, the closest being nonlinear optoelectronics with lasers. With tetrodes and pentodes and shit like this you really could make an equivalent circuit using far fewer vacuum tubes than transistors, particularly analog circuits. Funny that even then the thermionic solution is still bulkier and more expensive.
There was an interesting period of time where the advantages of things like CRTs and TV tubes were still better than LCDs and CCDs, so we had this fusion of tubes and semiconductors. I bet earlier than that there were some even more interesting hybrid designs out there, due to more types of tubes being superior to the semiconductor equivalent at the time. That period of time technically isn't over yet, as we still use scintillator tubes and magnetrons and geiger tubes and such, but those are all pretty specialised. I wonder when we'll start getting solid-state microwave ovens? Ones that can tune their wavelength to heat target specific bonds or to have a articular penetration depth for larger items or something. Those parts would be a real goldmine for radio hams.

>>1863254
Almost certainly not.

>>1863262
On the other hand, I'd find the production of a successive approximation register (SAR) quite interesting. That's the sort of thing Ben Eater should cover.

>>1863308
Check the datasheet, pretty sure the LF353 is a JFET op-amp and hence wants a large supply voltage and doesn't work very well down to the rails. When you feed it an input voltage too close to either rail you'll notice unpredictable behaviour, be this nonlinearity or even output inversion. An LM6132 is a rail-to-rail op-amp that will run just fine down at V+ - V- = 5V. Also consider a dedicated comparator.

>>1863308

here is some gibberish you can read. and it says essentially this: use a comparator when you wanna compare, and an op-amp when you wanna op some amps. many op-amps stop op-amping when the inputs are too close to the rails.

"INPUT COMMON MODE VOLTAGE RANGE
At the input, the CM range useful for VIN also has two rail-imposed limits, one high or close to
+VS, and one low, or close to –VS. Going high, it can range from an upper CM limit of +VS –
VCM(HI) as a positive maximum. For example, again using the +VS = 5 V example case, if VCM(HI)
is 1 V, the upper VIN limit or positive CM maximum is +VS – VCM(HI), or 4 V."

>>1863319
>>1863320
Thanks dude (or dudes)

>>1863308
>>1863321
>LF353
>input bias current
>typical: 50pA
Holy shit!
>supply voltage
>minimum: ±5V
Yep, you're outa spec, buddy.
Also the output voltage swing stays within 1.5V of each rail, and the input voltage swing goes no lower than 3V above the negative rail. Not the op-amp for 5V circuits, that's for sure.

>>1863267
Thanks

Please, help.

My 75W heating pad (pic related), which warms my chronically cold feet in bed, just broke after only 1 month of use. I know 75W is overkill, but I only ever use it at a fraction of its max power by controlling it via an outlet dimmer.

I want to make a similar DIY pad to fill in the role while my broke factory-made pad goes through the warranty replacement process.

Now, I'm having a hard time understanding how two differently resistive elements, like a nichrome wire and a resistor, dissipate energy individually when in series, and how the resistance of one affects the power dissipation of the other in this configuration. Since I'm a mentally handicapped brainlet, I need your help to spoon feed me.

So here are the approximate specs I need:
Pad dimensions: 100cm x 40cm
Heating element: nichrome wire
Power: 40 W
Input voltage: 220V AC
Resistance: 1200 Ω
Current: 183 mA

How do I electrically do this the most efficient way, without having to use resistors that will warm themselves up to a million degrees celsius in order to achieve 1,2 kΩ? I want most, ideally all, of those 40 W being produced by the nichrome wire, but I understand that might not be possible with a reasonably sized length and might require resistors to achieve 1200 Ω.

I need a reasonable nichrome wire length, given the pad size, and I need to know its diameter. Should you suggest I use resistors in the circuit, please make them only warm and not hot, for safety. We can also add a tightly sized current sensitive fuse and a temperature sensitive fuse somewhere later. I will also try to use a fireproof material as the enclosure for the mat and maybe I'll also seal the nichrome wire with heat-shrink tube.

I have a cheap Hikari multimeter, which will do the job this time.

Looking forward to learn some shit.
I await your precious help.

Thanks, anon!

>>1863319
>There was an interesting period of time where the advantages of things like CRTs and TV tubes were still better than LCDs and CCDs
I'd say that LCDs only managed to beat CRTs in terms of maximum resolution and price.
CRTs and plasma screens still have superior color production, only rivaled by OLED screens.

>I bet earlier than that there were some even more interesting hybrid designs
The Vidicon camera tube used in many consumer camcorders comes to mind.
Vidicon tubes were the source of the (Micro) Four-Thirds camera system still used today.
>we still use scintillator tubes and magnetrons and geiger tubes and such, but those are all pretty specialised
Don't forget Klytrotrons, Inductive Output Tubes, and Traveling Wave Tubes, all of which are commonly used in high power radio transmitters or satellites, since they still far outperform semiconductors.

>>1863319
Yeah, it's pretty neat how some of those ADC designs work.
Just about the simplest one you can make is a up/down counter being controlled by a comparator to compare the current output with the input.
Works fairly quickly as long as you signal doesn't jump up and down a whole lot.

>Analog modular division
>Almost certainly not.
Damn it. I have this design, and it works, but it depends on the comparator's output voltage to be exact.

>>1863267
>pair of studs
thanks for your answer! shoulds have mentioned that its a solid concrete wall

>>1863433
>its another episode of Anon wakes up to the smell of long pig

>>1863433
hello my brazilian friend. Please I don't understand what you are having trouble with since you just gave all the spec of your to-be heating pad.

>>1863444
Hi anon. I appreciate your humurous concern. Here in the south of Brazil, where winters can be quite cold, we have thermal sheets for our beds. They're somewhat popular, but I don't use them because they heat up the whole bed and I don't need that. I'm warm in my core but legs and feet are cold. So I have opted for a pad targeted at the profissional market, which is used in beauty clinics for reduction in body measures or whatever bullshit. They are smaller and I can position them to just heat up my legs and feet, which is perfect. These can get pretty hot and are made for 10 mins of use on max setting and 40 min on half power setting. Even on lowest setting, they are still too fucking hot to use in bed. So I use a very similar model of pic related. That way I can get the power even lower, and achieve maximum comfy.

I have researched accidents involving thermal sheets and found 0 cases. I think this is because they are made in such a way that when it fails, the circuit opens and it just stops working. We have INMETRO, which is our regulatory institute that certifies products for efficiency and safety before they can be sold. That is probably playing a role in that regard. I have a safety mindset too, and that's why I have mentioned my intention of adding more than one type of fuse in my project, and the use of flame retardant material.

>>1863459
Yes fren, I have the specs and understand them fully on a macro level. I don't know how to best translate them into a physical project. Here are the things I need help with:

- Sizing nichrome wire gauge
- Sizing nichrome wire length
- Sizing resistors (if resistors are needed)
- Making the nichrome wire use and dissipate most of the power in the circuit (40W), not resistors
- Making resistors (if needed) not get hot, and only warm

So you see, that needs some calculation and balancing, for which I do not have the sufficient skill yet due to a lack of knowledge.

Ideally, I need most of my resistance (1200 Ω) on the nichrome wire. But that could require an insane length since nichrome wires are pretty low resistance per meter. This makes me suspect that I will need resistors, but I don't know how to calculate the power dissipation for each resistive element when they are connected in series, since one resistance will affect another.

If you can help me with the items listed above, I would really appreciate, fren.
Thanks.

>>1863481
>making resistors not get hot, only warm
>power in the wire = power out of the wire (heat) - power in the wire (I*V)
The temperature of the wire depends on the surroundings and how you build the bag OR you can have a control mechanism that makes the bag stay in a preset temperature (your pad probably has one).
>wire gauge and size
Whatever do you have available, to calculate the resistance of the wires you use the formula R= r*(L/A) where r is the resistivity, L the lenght and A the cross-section. If you want to use less wire you clearly need one with a small gauge for a higher resistance.
>>1863477
>fuses and flame retardant
Those are not for personal protection, they are for network, device and fire safety.

Meu conselho é não faça isso. 90% do trabalho de importancia é na parte material/mecanica do produto. Voce vai estar colocando uma resistencia não isolada em proximidade com o corpo por periodos longos de tempo. Eu sou engenheiro eletricista e posso te dizer que isso é burrice. Voce não achou mortes em relação a heating blankets pois ninguem é burro o suficiente para fazer um em casa.

What I recommend for you is to use a isolated DC supply/batteries to feed a peltier element or a PTC one with a very low temperature. Wrapping yourself in a live wire is NOT a smart idea. You could also do the thing with the regular hot water bags or fill a bottle with water, put it in the micro wave, wrap it in a couple socks and put it near your feet while covering them.

>>1863477
You could use an infrared heat lamp pointed at your feet and achieve the same result.

>>1863503
Thanks fren.

When I said:
>making resistors not get hot, only warm
I meant an actual resistor, the electronic component made out of ceramic. I understand that 'resistor' can be anything that resists electricity and that might have caused confusion.

What I still don't understand is how two differently resistive elements, like a nichrome wire and a resistor, dissipate energy individually when in series, and how the resistance of one affects the power dissipation of the other in this configuration. I should really learn how to calculate that, but haven't yet found the info.

I guess what I actually want to know is how the schematic of these heating pads are designed. That would make things super easy. I would open my broken pad up and reverse engineer it, but I don't want to risk voiding the warranty.

>power in the wire = power out of the wire (heat) - power in the wire (I*V)
Ah yes, I understand the concept of power dissipation. So interesting. But in this case I'm much more interested in the produced power (40W) of my pad and how to achieve it on the heating element, with minimal losses anywhere else. This is because I have empirical knowledge that suggests 40W will mimic Position 1 on my broken heating pad. Then, I can proceed and regulate the final temperature with my outlet dimmer.

>R= r*(L/A)
Thanks fren. One more thing learned.

>>1863503
>Fuses
>Those are not for personal protection, they are for network, device and fire safety.
I understand and accept the technical and normative interpretation, fren, but you have to agree that if the circuit's nominal operation is at 190 mA and I add a fuse which opens at 200 mA, you can bet that fuse will protect me when shit hits the fan. The same goes for a temperature sensitive fuse. Let's say I get my lab thermometer, do a measurement in prototype phase, in loco, and estimate my heating element to operate at around 45 ºC when under my sheets. So I add a fuse which opens at 50 ºC, and so the fuse goes and PROTECC. So you see fren, these components protecc the circuit but also protecc me. I can add 10 temperature sensitive fuses all over the pad. Whichever area shorts and heats up to 50 ºC, bam, open circuit. Protecc succeeded.

>Meu conselho é não faça isso.
I understand it's a bit risky, fren, but I trust a robust circuit that I created with care, knowledge and dedication, that I understand fully. I won't do this on a whim. You can see I'm trying to understand all of it. As an engineer, you probably know many devices on the market are technically inferior to DIY counterparts. DIY ones can not only be safer and more efficient, but also cheaper. If you are a casual bigclivedotcom spectator you already know this. So many poor designs for minimal cost of production, fren, they just sell it everywhere and DIY ones are way superior.

>>1863503
>Voce vai estar colocando uma resistencia não isolada em proximidade com o corpo por periodos longos de tempo.
I will isolate the heating element fren, of course. There are some tubing materials we can buy for high temperature operation. It doesn't have to be heat-shrink tube, though for my low powered pad it will do the job. There are also some isolating tubes that can take super high temperatures. I think they are made out of fiber glass or something similar. These tubes are usually found in ovens (I have repaired a few ovens). I can use these tubes on my pad.

>use a isolated DC supply/batteries to feed a peltier element or a PTC one with a very low temperature
I see, that would be easier to make. The problem is that this will lower my cost efficiency to a such point it would just be logical to just buy another factory-made pad. Since I'm heavily interested in learning all about heating pads circuitry, making these shortcuts will not only make the project more expensive, but more importantly will severely hinder my learning curve.

>Wrapping yourself in a live wire is NOT a smart idea.
I understand, fren, but in a well made circuit make the risks are super super low. You just have to imagine all the ways the circuit can short, or the ways each component can fail, and make sure all failures lead to an open circuit, fast, way before you get hurt. That's what my factory-made heating pad did: something broke and it stopped working, open circuit. I'm unharmed.

>>1863503
>You could also do the thing with the regular hot water bags or fill a bottle with water, put it in the micro wave, wrap it in a couple socks and put it near your feet while covering them.
Unfortunately I have already tried these methods and they are way more inconvenient, offer little or no control, and perform worse. So my solution has to be electrical. The factory made one was perfect and made me achieve extreme comfy. I won't stop using it. When it comes back from warranty, it will go under my sheets once again. For now, though, a DIY one will provide me with comfy and knowledge in the meantime.

>>1863520
I doubt it, fren. I need a spread out warm surface under my sheets, like the pad I have that just broke. It's pretty perfect. I guarantee a lamp won't provide the same comfy in the same way.

>>1863551
Isolation in the sense of it having a transformer in the current path.
>>1863554
>>1863549
Fuses look instantaneous to us, but they take time to break electrical contact they are not made to protect people from electrical shocks, they are made to blow up in case of a short to stop things from burning.
>>1863549
>What I still don't understand is how two differently resistive elements, like a nichrome wire and a resistor, dissipate energy individually when in series, and how the resistance of one affects the power dissipation of the other in this configuration. I should really learn how to calculate that, but haven't yet found the info.
Isso é lei de Ohm basica. (V_sobre_o_resistor)^2/R = P_dissipado. Se voce não sabe disso voce não pode mexer com circuitos que podem te matar.
>>1863556
>I doubt it, fren. I need a spread out warm surface under my sheets, like the pad I have that just broke. It's pretty perfect. I guarantee a lamp won't provide the same comfy in the same way.
Aqui na noruega os bares e restaurantes tem lampadas termicas para os clientes na calçada. Voce fica muito quentinho mesmo que la fora esteja 0C. Uma daquelas lampadas usadas para criar galinhas seria ideal.

>>1863551
>but you have to agree that if the circuit's nominal operation is at 190 mA and I add a fuse which opens at 200 mA, you can bet that fuse will protect me when shit hits the fan
false
By the time fuse breaks you can be pretty much already dead (lethal current is 20mA, can be lower depending on environment)
If you want to protect yourself - GFCI outlet is the way to go

>>1863559
>Isolation in the sense of it having a transformer in the current path.
Ahhh, that makes a lot of sense. But of course, that's my goal. In order for my DIY 220V AC heating pad to work at 40W, I need to either lower my current, my voltage, or both. I'm still trying to learn and see what are my options, since my biggest obstacle right now is speccing a given nichrome wire length for 40W. That's why I'm here boys, to learn from you. I'm not intimidated by electrical warmers, it just needs knowledge to be executed properly.

>>1863559
>Fuses look instantaneous to us, but they take time to break electrical contact they are not made to protect people from electrical shocks, they are made to blow up in case of a short to stop things from burning.
>By the time fuse breaks you can be pretty much already dead (lethal current is 20mA, can be lower depending on environment)
Right! I've forgotten about this. Thanks. I learned about fuse response times a few years ago when I was studying a bit about surge protectors. It's really fascinating how long they can take to respond, when compared to the (near) speed of light at which electricity travels. My bad frens, still learning.

>Se voce não sabe disso voce não pode mexer com circuitos que podem te matar.
Yes fren, that's why I'm looking to dominate the knowledge before I do it. Will I choose to lower my operating voltage to 12V and use a higher current, or will my heating element operate at 220V at a very low current? That's a decision I have to make, and then proceed with my project.

>>1863569
It doesn't matter, you can get 1:1 transformer. Dude you are having trouble with basic Ohms law. You shouldn't be wraping your foot in electrified wires

>>1863562
Forgot to quote you on my previous post, sorry fren.

>>1863559
>Aqui na noruega
>living in Norway
That's a big privilege fren. Such a wonderful country. I need to visit it someday.

>os bares e restaurantes tem lampadas termicas para os clientes na calçada. Voce fica muito quentinho mesmo que la fora esteja 0C. Uma daquelas lampadas usadas para criar galinhas seria ideal.
I think I've seen that before! They are propane powered, right? I've worked in an agricultural company which made those gas warmers for chickens in poultry sheds. My problem is that they can degrade my room's air quality. I think they burn oxygen, right?

>>1863572
No. They are just a big filament lamp.

>>1863577
>>1863572
https://lista.mercadolivre.com.br/casa-moveis-decoracao/lampada-incubadora

>>1863578
They are also used in physiotherapy

>>1863578
>250W per lamp
>9h of operation a day
>30 days
>67,5 kWh/month
>R$ 47,25 a month to have my feet warm
KEK! Also, my feet is cold in the summer too, so my inverter air-conditioner will need to rev up even higher to keep up, therefore the cost of running the lamp at R$ 47,25/month will get exacerbated by the AC compressor running at higher revs. Too inefficient, fren. It's way more efficient to have a pad under my sheets while my blanket preserves the thermal energy so it can get absorbed by my feet with minimal use of power.

>>1863571
I understand you concern, fren. But look at it like this: I do not completely understand electricity, until I do. When I do, I'll design the project and fulfill my practical need.

I understand Ohm's Law, mostly but not completely. I lack the knowledge of many of its ramifications. I know how how current and voltage affects power and vice-versa. I have a basic grasp of how resistors affect circuits in different ways, but I'm not 100% familiar with it yet. There are infinite theoretical possibilities to achieve 40W, but practical real-world possibilities are finite.

Here is the catch with this situation: there are two kinds of people, the ones who take knowledge shortcuts in order to make things, and the ones who strive to become specialists before even laying a brick. I'm more on the side of the latter.

I'm getting there, fren. Don't worry.

>>1863584
>I understand Ohm's Law
No you don't. You said you don't understand how power is divided in a series circuit with resistances. You got your advice, have fun.

>>1863591
Thanks fren. Will do. Electricity is all about fun. If you can teach me and help fill my knowledge gaps, that would make me happy.

>>1863596
Use a single IR lamp on a dimmer and use black sheets or wear black socks, then read Forrest Mims.

>>1863597
I actually have a professional Taiff hair dryer I've been using for years as a kick-start of warmth on my feet and legs. Since the hair dryer uses high velocity hot air, heat exchange and warming up process is faster than anything I have ever seen. I never stopped using it even after buying the heating pad, because it's so quick to kill cold feet and legs. If I get 2 blankets over my legs and feet, I can maintain the heat without the pad. But at night, while I sleep, the second blanket usually falls partially or completely on the floor. I probably do some involuntary movements while I sleep, so that's why. Hair dryer: great, but not the perfect solution for the long run.

The reading recommendation makes me happy, fren. Will load it up on my Kindle right now.

anyone know if there are 4-6 pin single part T flip flops?

>>1863477
I just wear two pairs of socks and long-johns

>>1863652
PIC10F320 is a sot23 with an internal configurable logic cell that will do whatever you want :3

How hard would be to make a motherboard for a homemade general-purpose cpu?

>>1863734
not hard if you're messing with DIPs. if you mean a board for a modern CPU then you've got a lot of rabbit holes to go down and you'll be spending a lot of money on a many-layer board that won't work the first time.

>>1863735
Connect elements like ddr4, pci-e and this kind of things, that do basic things. Modern hardware may do everything from security up to survilliance build into controllers.

>>1863737
built*.

>>1863737
i'm generally ignorant here (i have no idea what the fuck a northbridge is or where to buy one) but to give you a sense of scale, go design a point-of-load regulator for the core that can provide 1.2V at ~50A with the necessary transient response. that's not an easy thing to get right the first time, and my impression is that it's one of the lesser issues in mobo design.

>>1863740
South and north bridges today are united or merged together

>>1863743
Meaning they've been rolled into the CPU, rather than being separate components.

>actually manage to solder 4 wires without taking 7 hours and ultimately failing like last time
>accidentally melt casing to a battery more valuable than the thing I was fixing because lol hf $1 iron and no stand
sagusa

>TFW an example design from the datasheet crashes the circuit simulator
Well fuck, I guess I'll just have to build it in real life on a breadboard and test it the hard way.

>Walking up hill under 345KV power lines
>Hissing and popping from lines getting louder as the hill gets taller
>Sapling next to me tall enough that there's barely more than an insulator's length between its top and the wire
>mfw

>>1863978
Cool. I want to see that shit, high voltage freaks me out but fascinates me. Ever since I was I kid I loved playing with flyback transformers, don't know how I'm still here.

>>1864022
Here's an old pic of a different part of the same lines. I don't have a pic of the part with the sapling under it, but the line was only 20 or so feet off the ground

>>1864025
I found another one that kind of shows it. It's up the far side of the middle hill, if you zoom all the way in you can see the lines come really close to the ground

It's the perfect place if you ever want to steal a shitload of electricity

Schematic calls for RC4558P EasyEDA has 4558 by PUPLOP for $0.04
I've never understood compatible components and I've always kinda trusted my shop clerk when he tells me "This is just like the LM317 but from some other manufacturer".
How can I find out if this 4558 by a chinese manufacturer will work like a RC4558P?

>>1864053
Click on the datasheet PDF. It's likely the same. There are so many things to doubt about chinese products but cheap ICs are not one of them..

I know how you can use a laser cavity with inhomogeneous broadening as a comb filter, producing a bunch of harmonics around the central wavelength, each splitting the cavity into a large integer number of different wavelengths. All of these harmonics add together to create incredibly short laser pulses.
Is there a way to have multiple oscillators running through the same BBD comb filter? Just an amplifier and a comb filter and maybe an AGC?

>>1864054
I checked the datasheet, but I don't really know what I'm looking for.
I usually just use them to check dimensions

How would I find the loudest possible relay that trigger at 3.3V? There's no sort by decibels option on digikey or mouser

>>1864087
you're doing it wrong

>>1864094
I mean probably, I need relays for the sole purpose of them making the click noise when it switches over. I'm an aero engineer but I've worked in a research environment for a long time and that's lead me to designing boards for our payloads alongside a working knowledge of this stuff.

I'm trying to do like a goodbye prank since I'm leaving soon, just an RPi connected to our server thats controlling like 30 relays and I'm just going to remote access from time to time to trigger them en masse. I'll hide it in the ceiling tiles.

>>1864097
FYI: The Pi only supplies 50mA of current on the 3v3 rail. Check below for an alternative. You could also use speakers to play the sound of relays switching (loudly).
https://www.amazon.com/Channel-Network-Wireless-Control-HLK-SW16/dp/B07HD959X6

>>1864102
Ahh I see, I didn't know about that on the gpio pins. I have a spool of mosfets laying around and some old pis, I could use like a wall wart for a 5v rail and then use gpio to control the mosfets and put power into the relays. With what I have now it might be cheaper than buying a 50 buck thing

>>1864097
For a clicking noise, sending a rising (or falling) edge to a speaker (or piezo especially) makes a pretty good click. If you use a speaker that isn't a piezo, put a capacitor in series to block DC, but I would highly recommend these cheap piezo discs, they're a dime a dozen. Might want to put some reasonably high voltage to them to get really loud clicks though, like 50V or more. IIRC this is the method simple geiger counters use to get their clicks. Just use one MOSFET to switch each piezo, probably no need for a half-bridge, let alone a full-bridge, and also no need for a diode like with relay coils.

If you do go with relays, noise is proportional to switching speed, which is somewhat proportional to both rated DC current capacity and rated voltage. So a 150A DC relay rated at 48V should give you a loud clack. Gonna be pricy though.

>>1864115
Let me order some piezo buzzers thanks for the suggetsion, I've got a cart going right now with like a handful of different relays so I can get a feel for the sound level. My plan is to try to utilize every gpio pin on the pi, like having a lot of these going at once and then stopping so the quantity of them might help to produce a much louder noise.

I'll start workin in kicad and get some drawings going I guess

I'm trying to build some stuff for a play. Do you know what I could do to release a lot of compressed air quickly? Like in a explosion of sorts? All those shirt guns and similar things are manually operated, what could I do to make an eletrically operated one?
>gas relay
>compressed air electronical valve
>electronical gas regulator
Either give me retarded stuff or some high end industrial things.
>just use gunpowder or some other theater explosive thing
Not in america or europe unfortunaly, those things are hard to come by

>>1864130
Something that could pierce the container would also work, there is no danger involved because it's quite small

>>1864130
How much psi are we talking? You could use a gas solenoid but they'll be in the $50+ range probably.

>>1864130
Get one of these and hook a motor or stepper or something up to the handle

https://www.grainger.com/product/1WMY1?gclid=EAIaIQobChMI8Inh-uPM6gIVxsDICh0E-w2XEAQYASABEgL1o_D_BwE&cm_mmc=PPC:+Google+PLA&ef_id=EAIaIQobChMI8Inh-uPM6gIVxsDICh0E-w2XEAQYASABEgL1o_D_BwE:G:s&s_kwcid=AL!2966!3!281698275765!!!g!472905489838!

I want to make a battery system for a raspberry PI 4. Ideally I'd be able to plug a USB-C to it, on any of its voltages to it, charge a integrated power bank, and deliver power to the pi. I'd also want an accurate reading of the power left, probably communicated to the pi via I2C. How exactly could I design something like this?

/g/ are braindead consoomers and probably never even looked inside anything so I'll ask here.
Has anyone ever seen headphones that are particularly well built internally? I buy the meme of the week $20 over ear's for the past 7 or 8 years and at best get MAX 2 years with me being a jew and hacking them back together , but I've had shit last only 3 months, with an average of maybe a year. I treat them like shit so some of the failures have been physically breaking, but most are wire releated.
If it's got a removable plug, the female plug in the unit wears out. If it's a fixed wire, the casing is problably vinyl and work hardens, eventually causing the inner wires to be exposed. Or tiny wires going from the drivers break or come loose, causing intermittent issues that eventually fail.
I really wish some offered a model with a female 3.5mm headphone jack that has the threadding that is used for those 1/4 audio adapters so could thread the cable in and minimizes wear on the port while still offering a changeable cable and make the driver wires real god damn thick so they don't come apart as easily and are easier to solder back on.

>>1864227
I have a Beyerdynamics Custom Pro set of headphones that have held up for quite a few years now.
They're built really solidly and sound great.

>>1864227
Design a keyed magnetic coupling adapter.

Check it out
Ghetto silk screen.
Would look better on copper but it will do

>>1861536
So I have a device that uses a 3.43 bi data line. The chip I want to use operates at 3.3v. Do I need a logic converter?

>>1864398
no

>>1864325
What method have you used? I've done silkscreen once by doing regular toner transfer on component side and leaving the toner as silkscreen but it's black only and not really durable. Your looks much better than regular black toner.

>>1864087
>loudest possible relay that trigger at 3.3V?

even fist-sized relays are pretty quiet. what would work well would be a vintage mechanical blinker. new, they're like $30+ but i found some at the thrift store for 50 cents. and one is enough to wake people up coz they're made specifically to be heard.

>>1864475
It's not ink, it was carved with a drag bit into other side of a single sided pcb so the lines look a bit rough, letters are barely readable, it would look much cleaner if it was made into copper since fiber glass is real shit for engraving

>>1864498
>It's not ink, it was carved
Oh, I thought it was some kind of paint. It doesn't look bad at first glance and at least it's just as durable as the PCB.
> letters are barely readable
You can always make them bigger - I've done this when doing toner transfer method and results are much better with bigger letters - especially if you have space for them.
Still it's an interesting technique that I've never seen anywhere else. Engraving bare glass fiber is not the best idea out there but now at least you have a silkscreen that won't erase itself.

Those 3D printed parts made up some kind of ghetto CNC machine or they are just used as jig for manual moving of the parts to engrave?

What do RF jfets (like the 2n5485) do that non-RF jfets (like the J112) don't?

LTspice shows that both pass RF frequencies (>10MHz) without major attenuation, so what's the hangup?

>>1864511
Looks like the output capacitance is a lot lower on the 2n5485 (2pF) vs the J112 (28pF)

>>1864227
>$20 over ear's
Buy chink shit, get chink shit. But if you're willing to modify them, I'd add globs of silastic around where the wires are soldered to the drivers for strain relief, and maybe replace all the wires with silicone insulated ones.
Earphones made with good materials are expensive, because of how large they are. You can get great IEMs with excellent build quality for much lower prices than you can get the equivalently built headphones, which is why I'm never considering buying headphones for the foreseeable future. They still have replaceable cables, and are easy to convert to use bluetooth with bluetooth cables.

Are center tapped wall wart transformers a thing that exists? I haven't been able to find any on aliexpress. Is there a keyword I should try? There are plenty of regular center tapped transformers but I'd prefer something with all the necessary bits already on.

>>1864535
A 1/4" stereo female jack on the headphone housing would be more stable. Then use a 3.5mm patch cable with a 3.5mm to 1/4" TRS inline adapter on the headphone end.

>>1864485
>mechanical blinker
Thermal flasher?

>>1864577
>Are center tapped wall wart transformers a thing that exists?
what makes a transformer a "wall-wart" transformer?
Do you mean "do wall-warts exist with center tapped transformers"?
That'd be a gigantic waste of space and expensive material if the wall-wart only outputs one power rating desu.
you might want to try "dual supply" but shoving that into a wall-wart is not marketable.
What you're looking for is a dual-supply, and if you have an idea of what voltages you want, you can search in application specific ways. E.g. +/- is typically audio, where-as +12V/+5v/+3.3V is typically for digital shit.

>>1864609
>Thermal flasher?

never heard that term, but sounds like the same thing. video here: https://www.youtube.com/watch?v=HdERp5e78eM

>>1864577
>Are center tapped wall wart transformers a thing that exists?

they do coz i have one, but it's the only one i've ever seen. one way to get sorta center tap is to use a +5, +12 transformer, which are common enough as they're used for external HDDs and external DVD writers. by using +5 as ground, you get -5V on ground, and +7V on +12. for op-amps, and such, the 2V imbalance makes zero difference.

>>1864633
I have a similar one, but for audio use so it has a proper split rail. Meant to be used with an ipod dock speaker that my flatmate left behind when he got arrested.
Nice tip on the +5V centre rail, but if the PSU is grounded then your ground is offset, which might be potentially a bad thing. Can't do that trick with a multi-rail SMPS though, as those rails can't sink current.

I'd still probably recommend buying a dedicated split rail SMPS, as opposed to a bulky old transformer.

>>1864577
Uh, hey listen, if you're looking for any tap configuration other than 12-0-12 then just wind your own or be prepared to pay someone else a lot of money to do it.

>>1864649
is that 12 rms or 12 peak?

>>1864535
For any sort of money I would be willing to spend whenever a pair broke (maybe $70 at best) I expected basically chinkshit with a slightly better marketing department behind them. I can't imagine them putting a drastically thicker wire in the driver or putting in attention to detail with strain relive where it wouldn't shake loose or break eventually for a minor increase in cost. Maybe a better cable, rubber perhaps, but I'd be skeptical. And if they lasted about as long, I'd be even angrier than usual. Cost wise I think they'd have to last like 5 years minimum to make me start to come out ahead. That's not to say I don't want something better built, but most people are dur it has metal or more bass when it comes to any sort of review, so I'm not sure I'd find someone who was autistic enough to actually tear something down to the point where they can say it's actually built better internally.
When the sony professional ones were still like $70 i'd think about those, but I just looked up a video of someone doing repair on them, and the driver wires look pretty thin as well, not that I'd expect 12 gauge wire in them ,but it doesn't look much better than the internals of these. Plus those have a wound up cord.

I'm not fond of working on consumer electronics, and given how tiny the stuff is and my less than ideal equipment and micheal j fox tier dexterity with a soldering iron. So I don't want to take something new and fuck with it since it's lifetime expectancy is likely to go from short to crib death.

>>1864653
+/-12Vrms relative to the center tap. Could also be 24-12-0. Same idea just your zero references isn't the center. You may be able to find 6-0-6 transformers as well. Asymmetrical center tapped transformers, good luck. You'll never find exactly what you want. Symmetrical center tapped you may find particular common output voltage configurations without too much hassle.

>>1864665
If it's not built with threaded inserts, a reasonable clamshell that actually goes back together without warping, and proper strain relief and wire mounts, the things that would make it easy to modify or repair, it's not worth buying in the first place.

>>1864668
Not the original question asker, but why 12Vrms? That's ~16.25V when rectified and filtered, and 32.5V is getting near the region of being too high for some ICs.

https://www.ebay.com/itm/936B-Iron-Soldering-Station-SMD-Desolder-Welder-Welding-w-Stand-Sponge-ESD-110V/361070621272
Are these any good? I've seen videos on them and they look okay, but I'm not sure if this one is not a knock off of the knock off sine it has the name in the picture, but not listed in the ad.

>>1864673
But at what cost do you actually see those features or some examples of models that have them? Using over ear is non-negotiable, I'm not going to go spend $300 on a pair of headphones because they are more serviceable and if anything that costs less suffers from the same problems as the $30 ones, I have to opt for cost unfortunately.

>>1864673
The input voltage is specified in RMS so why would the output be specified in peak?

>but why 12Vrms? That's ~16.25V when rectified and filtered, and 32.5V is getting near the region of being too high for some ICs.
You're supposed to linear regulate it down to the exact voltage you need. It would be a bad practice to just rectify and filter the voltage straight from the transformer and use that in the rest of your circuitry. Some loads you might be able to get away with that if they aren't super picky about input voltage but for any digital stuff you will use a linear regulator to get the exact voltage you need.

>>1864695
I figured for audio stuff you'd just use the filtered rails without a linreg, but I guess just relying on that PSRR isn't ideal. Definitely wasn't thinking of doing so for digital stuff, though I've seen very poorly regulated wall-warts that I'm pretty sure I've used for digital stuff before. Probably not an issue as they'd have an internal buck. I'd think that linearly dropping the 0.414*Vout would be too wasteful, especially with higher voltages, but I guess that primary voltage can vary by 10% at least and cause equal variations in the secondary. Good to know, thanks.

>>1864633
>>1864622
>>1864637
I'm needing it for some audio equipment that connects to other things like a stereo. If the ground on my equipment is really +5v and the ground on the stereo is 0v the audio cable is going to short the two together and melt down isn't it?

>>1864719
Quite possibly. But you could probably open up the PSU pretty easily and swap the ground lug to the 5V rail instead. Gonna get some weird shit with the ground-plane not being grounded, but it should be fine.

>>1864719

I have 7 transformers of that sort, diff brands, but they're all 2-prong jobbies, so have no reference to ground to screw things up.

>>1864507
>not the best idea
it beats fucking around with toner transfer and other chemical shit, like this the actual manual labor in making the silk screen takes just seconds since it only involves changing the endmill for a drag bit

also as i assumed it looks way better if done on copper, even the tiny letters look crisp and near perfect, since i mostly use 2 sided pcbs for my projects, this is great

>>1864759
Consider melting some black or white wax and rubbing it into the grooves, before polishing the top to remove the excess. Basically what Clickspring did to his skeleton clock's face.

>>1864767
That seems like a lot of work, is there any benefit?
After you solder components on most of that wax will be melted and gone anyway, plus the visibility in current state is good enough where it's easily readable for soldering
One more thing i should solve is since i don't have solder mask (i mean i have it in a tube but applying and UV curing it is too much work and i'm lazy) so i wonder if there is some common house hold chemical i can put in a spray bottle and spray the finished pcb with it as conformal coating (as in it will go on after the components were soldered)

>>1864507
>Those 3D printed parts made up some kind of ghetto CNC machine
FOrgot to answer this, no the router is made from aluminium, it's for milling wood
The plastic jig i designed and made from PLA on my 3d printer, i needed a way to hold the PCB without covering any of it at all with clamps since i am using 100% of the surface so there is no safe place to put any top clamps
The L bracked is also great for making double sided pcbs because i can just easily flip the PCB and get near perfect alignment without having to drill any aligment holes

>>1864777
>That seems like a lot of work, is there any benefit
To improve visibility and remove any burr. I probably wouldn't do it on copper, but for the back of a 1-side board. It might be possible to solder over a board with it in place, especially if you used something like black epoxy instead of something that would melt like wax.

>>1864812
Honestly that is not really worth it just to increase visibility, on the fiber glass side the visibility is perfect, it's the detail that is missing and vax won't help with that since the detailed shapes simply aren't there, the bur i could remove with simple sanding, but not really worth it, since silk is just one time assembly guide and it will work great for that

>>1864816
>the bur i could remove with simple sanding
I know, just if you did that I suspect the legibility of the lettering would drop considerably.

>>1864816
>>1864819
Oh but to clarify I wouldn't bother with this for such a one-time assembly guide. Maybe for one of those PCB anime girls that's all the rage with zoomers.

>>1864827
>PCB anime girls t
what?

also if i wanted to make a pretty pcb i would just pay chinks $2 to get 5 perfect pcbs made and shipped the only reason why i didn't is it takes a month to deliver and the postage costs $6 so it's shit if you need fast prototypes

>>1864829

>>1864830
but why? the consumer won't even see the pcb at al since it will be hidden in a case

>>1864834
IIRC it's not a product for consumers, but for enthusiasts. No real reason not to do so if there's nothing else to put on the PCB anyhow, unless you're worried about a few extra square centimetres of gold plating.

>>1864834
>since it will be hidden in a case
not if it's clear acrylic, as all cases should be

>>1864498
holy shit i take it all back, look at what swiping the board lightly on a sand paper just TWICE did

it looks fucking impeccable now, this is incredible, i am so fucking hard right now

even the letters are easily readable now

>>1864882
That looks really good anon, when are you going to start taking commissions for board designs?

>>1864897
Step one is bankrupting jlpcb by undercutting them

>>1864900
I'd actually pay a much higher rate for hte novelty of having my board manufactured by a fellow /diy/er

Can anyone help me with converting a laptop keyboard into a USB keyboard? I took the keyboard out and it had a ribbon cable with 24 connections to it.
I just don't know how to figure out what each connection means and how to use it. I know the keyboard is configured like a matrix and you have to scan it row by row. How can I figure out where to go from there?

>>1864928
Many mechanical keyboards are oriented towards some customization so most likely you could use some controller from them, see https://deskthority.net/wiki/Controller_matrix_traces

Just trace out which buttons do which actions and see if there are compatible controllers. If not, you'll need some development board, like teensy to be able to program to your specific layout

>>1864930
Thanks a lot : D

Anon with broken speaker amp from previous thread reporting back.

With the help from a guy over at a forum I found the broken part and I just replaced it, it's fully operational again.

A most amazing kind of feeling.

Is it true that amber LEDs lose alot of efficiency when heated up?
I heard they get less bright if hot.
Why is this the case specifically for these LEDs?
Also wouldn't a hot LED get more efficient since current increases when temperature rises. Thus the luminous flux would increase aswell, no?
Datasheets claim the opposite and I can't explain why.

>>1864930
Not him, but are there any recommended controller ICs for USB keyboards or joysticks? Be nice to make custom game controllers with ease.

>>1865137
Radiant power is not solely a function of current. The hotter an LED is, the more leakage current there is. This is the same leakage current that you get in reverse bias, or when the voltage is too low, it doesn't produce light. On a more microscopic scale, the hotter the PN junction is, the more electrons have the thermal energy to jump across the band-gap. This applies to all LEDs, maybe amber LEDs are worse, but I've never heard of that before. Solar panels are also less efficient when they're hot, as the photovoltaic effect is more or less the inverse of semiconductor electroluminescence.

>>1865144
>recommended controller ICs for USB keyboards or joysticks

there's 1.1 million youtube videos on making your own game controllers, all using the same parts, essentially.

>>1865160
Do they use arduino pro micros instead of dedicated ICs?

Can someone explain to me the difference between a UJT (unijunction transistor) and a JFET?
They look like they're the same thing to me.

>>1865179
>Do they use arduino pro micros instead of dedicated ICs?

there are essentially 2 varieties of builds: one programmable with a computer like the Teensy, and one with a certain hardware board specifically made for the job, like pic related.

>>1865182

red the UJT wikipedia page to learn that it's COMPLETELY diff from a FET. it does not do any kinda signal gain, it's just for switching, or creating pulses, somewhat like a 3-legged 555.

>>1865179
Looks like a bunch of them actually use arduino unos instead of pro micros. I guess they technically have native USB through the Mega8, but it's a bit shit. The only other thing I could find was some guy selling a PIC on a board with USB B socket and a bunch of headers for pots and a keyboard matrix and/or rotary encoders. For 25GBP.

I guess I'll lean towards a cheap MCU with native USB then, like the Mega8. Could probably make a fuckton of these and sell them for a lot less than 25GBP.

>>1865194
>Teensy
Sounds overkill to me, but maybe it means less latency.
>pic related
Isn't that just a board out of an existing product? Are boards like that sold separately on alibay or wherever?

>>1865194
I think I'm finding something similar with the search term "zero delay arcade usb". Not expensive, but it could be cheaper. But I don't think it has analog inputs.

I wonder if I could bit-bang it just fine with an attiny13 or something?

>>1865198
>Are boards like that sold separately on alibay or wherever?

of course. https://www.newegg.com/insider/so-you-want-to-build-an-arcade-stick/

>>1865197
That's not really the impression I got from the wiki page.
First off, the wiki doesn't give any equations to model the UJT with, so it doesn't really explain them that well.
Second, both UJTs and JFETs have almost identical internal structures.
Third, based on wikipedia's description of how a UJT works, I seem to be able to get similar behavior out of a JFET (red is input voltage, yellow is current across the resistor).

>>1865264
>current across the resistor
There shouldn't be any current in the resistor at all

>>1865275
There will be in the AC case, anon.

>>1865284
There should be no net DC current I mean. If I can read that graph properly, it's all below the x axis.

>>1865292
The input waveform has a 5v bias to it, so it never goes negative.
Instead it swings from just below to just above the point where the jfet starts to conduct.

>>1865305
>The input waveform has a 5v bias
Then why bother having +5V on the JFET's source? Kinda misleading. Falstad's implied ground reference is also a bit confusing.

>>1865317
It was just a quick and dirty test.
I had the source connected to a voltage source so I could test a wide variety of biases quickly.

Op's pic is all wrong. It needs a fat polarised cap of 16mf in paralele.

My BBDs arrived, should I make a comb filter, or something else?

i recently patched up my radio, wasn't working for probably a year and a half, just didn't have time to look at it. it's fixed now and sounds better than it did before, but I have moved since the last time I used it and now I am a lot farther from any broadcasting areas. I want to diy an antenna and am finding out this is a total rabbit hole. Any anons here diy antenna? or have specs on one that performs pretty well? currently just have a beater dipole made out of speaker wire but I want to be able to have better signal on a few of the channels I already get

>>1865393
What frequencies are you aiming for? I've heard good things about the pa0rdt mini-whip active antenna, but as a capacitive antenna it needs to be mounted far from walls or conductors.

Also check out the ham radio thread.

Is there any simple rules for adding caps /resistors in a circuit im making?

I dont have a EE degree, i just wanna make simple stuff without having to worry about frying main components.

>>1865409
There aren't simple rules, besides the addition of bypass and PSU filter caps, and the selection of shunt regulator/current shunt/current limiting resistors. For the most part, resistors and capacitors are parts that decide the particulars of how a circuit block works, and you'll construct a functioning circuit out of these blocks. This is especially true for analog circuits, where the selection of resistors and capacitors will determine the gain and frequency response of circuit blocks like amplifiers or filters. If you're doing analog circuits, then you'd better learn how op-amps and/or transistors work, because you can't easily bodge together circuit blocks without that knowledge.
For solely digital circuits on the other hand, knowing simple things like pullup and pulldown resistors is probably all you need to know.

What sort of circuits are you planning on making? There are circuit simulator software packages out there which might save you a few components, if you understand their limitations.

Just adding an accessory to my car.

power is coming from a fuse piggyback on 20a fuse slot, going to a switch (from auto zone), then to a 12v water pump. ground out to chassis. Pretty standard for a car accessory, pretty much all work within the confines of 12v of power source. Switch should be able to handle it as that is exactly what those brands are built for. just a little unsure.

>>1865409
>are there rules to adding letters to my words and words to my sentences? I'm just trying to write something that makes sense
yes

>>1865409
I studied EE for 12 years before i worked up the courage to even add a simple switch between my battery and a light bulb

>200% harmonic distortion
I don't understand, it looks pretty sinusoidal to me.

bought 3 pairs of gx16 9-pin connectors and 1 pair of gx12 6-pins, they seem to be panel-mount types for the males (on the left) but i need this for wire-to-wire connection so i need a housing for them like the female connectors (right) but i can't find any online
i know they exist but where do i get them?

>>1865471
Are you sure n004 is the net you want to measure? I suggest adding net labels on nets you're measuring, helps avoiding errors

>>1865471
>looks pretty sinusoidal to me
Shouldn't, with schematic as is some clipping on peak values should occur with possibly some distortions around zero cross

>additional U.S. sanctions on Huawei prevents Huawei from using EDA made by U.S. firms such as Cadence or Synopsys
>it also prevents companies using U.S.-made EDA from selling designs to Huawei, so even ImgTec (UK-based, China-owned) can't sell core designs to Huawei now

https://technode.com/2020/05/20/export-ban-ii-huaweis-harsher-higher-stakes-sequel/

How long before open-source semiconductor design tooling becomes insanely good with money and developer hours pouring in from mysterious sources?

>>1865623
The design isn't the problem, the thing is semiconductors are one of the only few things impossible to diy.

>>1865623
Stallman was right

>>1865627
so I CAN fuck my sister then?

>>1865631
Its the states rights to decide whether you can or not

>>1865623
UK also is purging the chinks from the 5G network in bongstan

>>1865396
thanks, didn't know there was a ham thread on the board

i got this wire from chinks
https://www.aliexpress.com/item/32954226474.html
and it has polyurethane coating which means it should be possible to solder it right? i set my iron to 400C and tried sodlering it but i am having no luck i just can seem to burn through the coating

>>1861753
i would've thought that additive synthesis would be best for sine wave since its easier to control the harmonics and wont sound shite quickly

but i don't know i ent tried additive that much in a while anyways

>>1865739
you solder metals, not plastics, so yes, you'd have to remove the coating first.
polyurethane coating has nothing to do with soldering other than getting in the way

>>1865752
this coating should be burnable by a soldering iron, but even if i set mine to 400C is still only works sometimes

>>1865754
Get a solder pot.
https://www.youtube.com/watch?v=M02FbfUvg2I

Hi, I'm building a DIY Ebow.
Does anyone have any idea what the 47nF capacitor and the 10 Ohm resistor after pin 5 are for?

>>1865811
RC filter

>>1865816
Thanks

>>1865538
Pretty sure, adding a label causes no difference.

>>1865573
No zero-crossing nasties, the fft looks just fine, with a tiny bump at the 3rd harmonic. It's surprising how good the tops of the waves look, just a tiny tiny bit of clipping.

But now I see that the debug screen thinks my fundamental frequency is 1kHz, with appropriate multiples of that as harmonics, instead of the 1.56kHz that I've actually got.

>>1865811
Read the datasheet of the LM386, it might say something about it.
Also what are BUZ1 and BUZ2 meant to be? IIRC the input should be a pickup while the output should be an inductor.

Also post your project when you're done, I'm interested in Ebows.

>>1865816
a 2MHz RC filter?

>>1865851
Pretty sure with a 10Ω resistor it has to be a snubber, not a filter. The 10Ω is just rudimentary current-limiting and to prevent the cap resonating with the load, while the 47nF cap suppresses any spikes. The snubber circuit only does something if the output load is inductive, which I suspect is the case. But I can't think of a situation where you'll end up with high frequency transients you want to block.

>>1865851
not relative to the output, no, but those two components in that configuration are filtering. the rest of the network needs to be taken into account.
Remember that's feeding back into the inverting input, and the output is across BUZ2.

>>1861536
What's the difference between a motor run capacitor and some rando voltage and capacitance-matching cap from Digikey? Is it resistance?

>>1866024
uhh...voltage.

>>1865884
Is a snubber not a type of filter where you come from?

>>1866034
>Voltage and capacitance matching cap
>The difference is the voltage
Genius.

>>1866024
Current carrying capacity, basically yes it's the ESR.

>>1866037
Well, as opposed to an RC filter. Compared to an RC filter like was originally suggested, a snubber uses the characteristic impedance of the surrounding components to filter out transients. Same with filter caps on a PSU, or bypass caps across an IC.
Shoulda said "it has to be a snubber, not an RC filter". But now I'm wondering if it's not a snubber but some way to improve the frequency response of the output coil. >>1865885 doesn't quite make sense to me, as the only way a cap to ground on an output would be feedback is if there's some sort of current-based internal feedback inside the amp IC.

Checking on stackexchange, I get a few answers:
>The .05uF and 10 ohm resistor on the output act as a phase-shifter, to cancel out the internal phase shifting caused by the LM386's own design. It is common to see this on many IC amplifiers.
>It bypasses the speaker at high frequencies, reducing the effect of its inductance.
>This is the "Zobel network". It is for stability against oscillation; it provides a load for the amplifier at high frequencies (which the speaker doesn't, being an inductor). This feature is very commonly seen in the designs of audio power amplifier output stages.
The latter seems to make the most sense. Interesting. I guess it technically is current-based internal feedback, as the amplifier behaviour is load dependant.

>>1866053
>But now I'm wondering if it's not a snubber but some way to improve the frequency response of the output coil.

datasheet says you put that in to prevent the amp oscillating under certain conditions. simple as that: RTFM.

>>1866044
> some rando voltage
eh, this reads two different ways

>>1866099
Wait, it really says that? I couldn't find anything saying that. Sounds like the Zobel Network I found anyway though.

>>1861536
Roll for challenge

>>1865851
>>1865884
>>1865885
>>1866037
>>1866053
>>1866099
Oh thanks, I was wondering why the circuit still worked even though I had forgotten to put in this part. Makes a little more sense now.

>>1865834
>Also what are BUZ1 and BUZ2 meant to be? IIRC the input should be a pickup while the output should be an inductor.
That's right in the video https://www.youtube.com/watch?v=1iccGqmEC-M
The guy uses modified magnetic buzzers (transducer) with different coil resistance to act as a pair of pickup and driver. I had to open them both, remove the diaphragm and glue the magnet that was on it to the core.

The prototype is done, it's actually my first electronics project. I borrowed the breadboard and the wires from a friend.
I tried it on my bass, I guess it kind of works? The G string sustains fine although the coil have to be exactly aligned with the string and nearly touching it for it to work. Unsurprisingly the other strings are probably too heavy and the ebow doesn't work. Right now I'm wondering if there are any ways to get a stronger magnetic field out of the driver.

>>1866195
>magnetic buzzer speakers with diaphragm removed and a magnet permanently fixed to the coil
So they're basically both pickups. I'd consider temporarily replacing your output pickup with a higher-impedance speaker load of some sort, so you can hear what the input pickup is picking up. If you can hear it on all the strings, then it should be good on that front. As for giving the output driver more oomph, I'd first try with some standard radial-lead inductors. Lower impedance = higher current = stronger magnetic field, so you'll want to try a few values to see what the strongest magnetic field you can get without overheating the IC. I don't think you need to worry about distortion. You could try increasing the gain on the amplifier, assuming it isn't already clipping. If the gain is already maxxed out, I'd replace the amplifier with a comparator and a MOSFET driving an inductor. I'm pretty sure feeding it a square wave won't be an issue.

I think an inductor will be just as effective (if not more so) than a dedicated pickup (inductor + magnet) but I'm not 100% sure, so do some testing.

>>1866152
>Wait, it really says that?

Zobel network is correct.

Not sure if /sqtddtot/ or not, but my refrigerator is getting too cold all of a sudden, as cold as the freezer side, so my milk is solid right now.
I wanted to check if the temperature sensor died and if that isn't it, maybe a valve of sorts that controls the flow from the compressor? Not sure what else could fail though, as the compressor and the freezer side are doing just fine.
So before I take it apart, what else could be the problem? And when I do have to take it apart, what could other potential troublemakers be?

>>1866335
Failure in the door, either leaking or not closing properly. Makes it turn more often
Source:
>mine has that problem

Is there a proper way to disconnect these types of terminals or is using a screwdriver to untighten the female's lips and then retighten them with pliers the correct way to do it?

Can someone explain to me why you would add traces between two components on a PCB when they are connected with a "ground plane" (not sure if that's the right term here, there's no layer inbetween but the whole area outside the traces is ground).
As you can see here on the right all 3 of these pads extend outwards to the grounded area so why is there there need for a trace between them?

>>1866416
you pull

>>1866461
Well that's what I thought you had to do but the male connectors with the holes seem to get hooked up by the female connectors and it looks like you can't disconnect them just by pulling. I think this is done so that it can't disconnect itself over time.

>>1866422
In this cases it looks like just some errors were left

Most likely case is that ground traces were laid out given that rather random approach, then copper pour specified with settings, maybe by default, that did not pour over same net objects

>>1866335
There's a defrost heater for the evaporator coil.

Is there anything necessarily wrong with this design? Replace CLK with raspi GPIO pins to get the idea of what I'm doing. The right side are relays.

According to the circuit simulator it functions how I want it to (cycling the relays), but I'm not sure if there's additional nuances to it.

>>1866526
Connect a diode from across each coil. As a rule of thumb, never cut off the current to a inductor abruptly without some other path for it to take.
To be clear: Diode from coil ground to the other coil terminal

>>1866529
Thanks, I added that in. I come from an aero background but I've picked up tidbits of ee here and, would that be a flyback diode?

>>1866532
Yes that would be it. Mosfets have what is called a body diode but unless it's some semiconductor fuckery in a high power converter module I think everyone adds flywheel diodes.

I'm feeling fucking retarded here.
Like ok, I've connected normal pn cables to different terminals before.
But now I've got like mini cables, you know, like cutting an ethernet cable, and I've got to connect those to 5 terminals, do you really just cut the end and connect to the terminal like normally for each one, or is there some kind of adapter or method you use?

>>1866200
>I'd first try with some standard radial-lead inductors
Something like this? https://www.mouser.fr/ProductDetail/Bourns/RLB0914-102KL?qs=SUrgTEU5cSoXW47y1y0lkw%3D%3D
Would you happen to have an idea of the order of magnitude of the inductance needed? Can I just swap them around until I find one that works without altering the rest of the circuit?

>If the gain is already maxxed out, I'd replace the amplifier with a comparator and a MOSFET driving an inductor. I'm pretty sure feeding it a square wave won't be an issue.
Yeah with the 10µF capacitor between pin 1 and 8 the amp is functionning at max gain (200). Would you mind explaining what replacing the LM386 with this combination will do?

>I think an inductor will be just as effective (if not more so) than a dedicated pickup (inductor + magnet)
Does the magnet make a big difference?

>>1866540
Yeah that sort of thing. I ordered a 20-pack of 220µHs off aliexpress a while ago, haven't needed any more since then. Found a reel of 1000 4.7µH SMD inductors in a junk shop though.
>Would you happen to have an idea of the order of magnitude of the inductance needed?
No, but the datasheet says the minimum load impedance is 4Ω, so I'd go for that. If your maximum frequency is 500Hz (IDK, it's a bass), then that would be an inductance of 1mH, at which point the ESR would almost certainly be above 4Ω anyhow. Looks like going somewhere around 330µH would be good for this.
>Would you mind explaining what replacing the LM386 with this combination will do?
A comparator has an effectively infinite gain, so feeding it a sine wave will result in a maximum amplitude (i.e. from 0V to the 9V rail) square wave. Using a power MOSFET as an output stage won't be linear, but it will let you drive higher power loads than just 700mW. Though there are other issues with driving an inductive load with a single MOSFET, as compared to driving it with a half-bridge. Not sure if the good-old antiparallel diode will give you a usable output signal or not.
>Does the magnet make a big difference?
The magnet produces a permanent magnetic field that interacts with nearby magnetic objects. When a metal string is plucked in the vicinity, the moving string will move the permanent magnetic field along with it, distorting the entire magnetic field, even the magnetic field within the nearby coil. A changing magnetic field in a coil produces a changing voltage, which is how pickups work, but I've no clue if it will have any impact the other way around.

>>1866416
>>1866461
>you pull

no you dont, you wiggle it laterally while pulling. the wiggle prevents breakage.

>>1866526
Connect the coil between the drain and +5V. If it's connected to the source, Vg=Vgs+Vs, i.e. the gate voltage needs to be at least Vgs[thresh] *plus* the voltage required by the coil. And 5V isn't going to be enough.

If you look at existing circuits which drive relays (or any other load where you have access to both sides), it's always done that way. In cases where you have to connect to the source (e.g. because you don't have access to the other side of the load), you need to use a "high-side driver" which can shift the gate voltage above the drain (supply) voltage.

>>1866538

what a dumb question. you wanna connect wires to terminals but dont tell us what kind of terminals.
(it's obviously not ethernet coz your pic comes from an internet video that shows how it's done.)

>>1866799
He doesn’t need 50 amps, just use an NFET+PFET instead.

>>1866799
>>1866821
Wait, I misread this. Yes I agree that you should always use a switching MOSFET in a common-source configuration (read: source of transistor should be at a voltage rail). High-side drivers are interesting.

Also check that your MOSFETs have a Vgs_threshold less than 5V (or 3.3V, not familar with raspis). If you can't find a MOSFET that switches on at your voltage level, you may need to use an NPN transistor to turn the MOSFET on with a pullup to a higher rail, or just use an NPN on the relay instead of the MOSFET. The relay won't be using that much current anyhow, no particular need to use a FET in this case. Still need the diode whatever you do though.

>>1866532
>>1866537
Why is it called a flyback diode? Flyback converters, from what I understand, work just like a boost converter but with the output magnetically coupled through a secondary winding of the inductor, and boost converters can't have flyback diodes because it removes their voltage spike. Or do they work like buck converters instead?
Sounds like a misnomer to me, that's why I prefer the term "freewheel diode".

>>1866876
I don't know why. In portuguese we call them "diodos de roda livre","grampeadores", "diodo de comutação" and a bunch of other names. IIRC it has something to do with Old TVs and their electron gun

>>1866801
Well yeah I am dumb.
First electronics project that is not arduino shenanigans.

Pic related, screw terminal.

>>1866946

red pin terminals go down to 22 gauge, but you got 24, so maybe you can solder them. technically , these are solderless terminals but they accept solder readily. i've done it before.

>>1866946

red pin terminals go down to 22 gauge, but you got 24, so maybe you can solder them. technically , these are solderless terminals but they accept solder readily. i've done it before.

trick is to insert wire from the back, then push down on it with soldering iron tip at the front, and dump a lot of solder, which will be sucked in to the tunnel of love.

>>1866955
Thanks man.
>trick is to insert wire from the back, then push down on it with soldering iron tip at the front, and dump a lot of solder, which will be sucked in to the tunnel of love.
Great tip, I'll do some practice before I try on the actual wire.

There was an experiment forest mims had in one of his books involving a LED in reversed bias driving a 555 timer oscillator. The whole point was to demonstrate some type of capacitor action with the LED in response to light.

Has anyone seen this or even have the circuit for this?

>>1866997
Would just be a photodiode feeding that one FM/PWM input on the 555, no?

>>1866355
How would I go about finding that other than feeling around? Because I've been feeling all around and can't really feel a difference in temp somewhere

>>1866488
Should I check that if there isn't a leak?

>>1867026
Okay so I've turned it off and on again and now it returns to normal temperatures (or it's dead), the airway that goes to the refrigerator side stays closed now. Though it wont last for long as the last attempts would return to freezing in about a day..

What is the black contact panel material called? Does JLPCB or kicad even support it? Will it work if it's just exposed copper or hasl?

So i tried ordering this riveting tool for making vias.
Can't wait to try it out. Costs 7 yuros so not that bad.
Although i am going to use it with the cheap chink rivets since this store sells 25 rivets for 2.5 yuros, what a fucking joke, that is 2.5yuros worth of rivets to make one board basically

>>1867074
carbon print contacts

>>1865811

>>1865816
>>1865851
>>1865884
>>1865885
>>1866195

Its called a Boucherot cell. Common for the outputs of amplifiers to prevent oscillation or other poor performance due to capacitive loading.

>>1866526
As the switch is between the +5V supply and the coil, wouldn't it be better to use P-channel FETs?

Couldn't this limitation be overcome by replacing R5 with two parallel resistors, each having opposite-polarity 1N4148 diodes?

>>1866422
That looks like a mistake, but when mixing analog and digital components you usually want to break up the ground plain to keep digital noise on it from effecting the analog. These two plains are connected together by a trace somewhere.

>>1867155
Is there some paint or any way to apply this over exposed copper?
Can exposed copper work? I've seen aftermarket Gameboy clones just use exposed traces as pads

>>1867026
>>1867035
If the evaporator coil is getting cold then the compressor is working (and start/run cap), there is refrigerant present, and the door switches are working. Check your thermostat relays for the heater coil and the heater coil resistance (~30 ohms).

>>1867096
Oh you're buying a legit via punch? Tell us how it goes. I've been interested in those and bought a few sizes of via rivets, but the punching tools themselves are kinda expensive for what's just a piece of metal with a specifically-sized tip, and IIRC you need a different punch for each size of via. I tried 0.8mm vias first, but they were too small for some components.

>>1867186
Oh neat, that's exactly what it is.

>>1867268
>using an entire inverter IC just so you can control two LEDs with a single GPIO
Why not just tie one LED to ground and the other to Vcc with a resistor either on each LED or on the GPIO? Picrel. Setting the output to high-z would only turn the LEDs off if their combined forward voltages was higher than the voltage supply, if not you'd just blow the LEDs up. If using red or orange or yellow LEDs on a 5V supply, put two in series I guess. But either way it sure beats having to get an inverter IC. No clue why they're bothering to use a double-LED package as if they're attempting to save space.

>>1867330
Exposed copper is probably ok, though I'd personally go for hard gold. Might be able to find that shit online in a hobby-size syringe, but you'd need to dig deep. Maybe search on forums dedicated to modifying/repairing controllers and the like that have these PCB buttons. Where are you going to get silicone dome button pieces from anyhow? Salvage them? Or are you remaking that PCB as a repair?

>>1866876
I think just from how it sounds, when you cut off power there's that sudden negative voltage (or something) and so the flyback diode prevents the back voltage from "flying back at you" or something.

>>1867330
>Can exposed copper work?

copper wont work for very long. it'll oxidize and stop working after some years. gold-plated will last longer, and carbon much longer.

>>1867388
I thought the "flyback" was referring to the HV flyback converter in a CRT television circuit and how the ramp generator needs to fly back to the start after each line, but now that I think of it the ramp generator doesn't need to be HV since it's just powering the deflection coils; the HV generation is for the electron gun.

>>1867348
>Oh you're buying a legit via punch
yeah, i tried to order cheap punches from chinks before but they just torn broke the vias
i am not sure if this one will work better but 7 euro is pocket change so i want to try it
maybe it won't work too because i am using cheap chink gold colored rivets instead of the copper rivets, but copper rivets cost 45 euros for 1000 pieces while the exactly same ones from brass or whatever that is from chinks cost like 4 bucks per 1000. And i am not willing to buy the expensive rivets so if it doesn't work i will just go back to using resistor legs to connect vias.

I ordered the 0.9mm punch because for logic connections this is more then big enough, and 95% of my vias are low current, and if i need something high current i can just do connect the via with a wire instead

>>1867492
I have no idea, I'm just making assumptions here

>>1867651
CRT beam deflection is close enough. Memes develop a life of their own.

shieeeeet i thought kicad was smarter than this, i ran DRC and kicad was like
>everything is fine and dandy pal, no issues whatsoever, go ahead an send this shit in to chinks for manufacture, *thumbs up*
well shit is not fine!
this ground fill ain't connected to anything, in fact a bunch of ground fills aren't connected together
if i had this made, i would be in for a bunch of manual jumper wire soldering to connect all the ground planes
also i posted the screenshot of this board here several times and not one of you dicks pointed it out. i hate all of you

>>1867819
time to install altium

>>1867823
>altium
Tell me altium dev, how many working particle colliders were built in altium?

>>1867846
i wonder how many board revisions CERN had to do because of discontinuous fills

>>1867850
big brain boys at cern don't make such trivial mistakes, that must be why there is no ground fill check in drm

i want to control a bldc motor actuator (RMD-L-7025 on ali).
the integrated controller has vcc and gnd pins, but no breaking resistor pins (or any braking resistor inside).
can i drive this actuator with a power supply? if i understand correctly, bldc generate large currents while braking. usually this is beneficial, since power is pushed back to battery. as far as i know, most power supplies dont sink current and might even burn if current is sinked to it.
is there any way to drive this motor with power supply? do i need a capacitor (super large one) to sink the current to?

>>1867819
I swear kicad should have noticed that, it's noticed discontinuous fills on my boards