/diy/ - Do It Yourself

>>1537317
her soldering technique is doing such a fine works. how precision she is with those maximum focus yellow glasses
impressive

If R on aboard is resistor and D is diode, what is Ft?

New thread please, hard to search without title

>>1537317
If not soldering iron, what is used to repair motherboards?

Should I be worried about led poisoning when using Sn63/Pb37?
I extract fumes but I handle it with my hands

How important are sockets for ICs and transistors? I understand maintenance is easier but I don't think my 4558's will burn and I don't think my 11w soldering iron will damage anything.
I've seen people use them for 3 pin transsitors but I don't know if it's really necessary

>>1537385
It looks neater, everytime I work with perf board I awalys burn and lift some pads, make some funky solder bridges/towers or shit something up. Long ago I bought several stamped IC sockets of many pin numbers and never had an issue. But you can do without them, ICs are more though than they look.(Search for deadbug circuits). Never used transistor sockets, but I think they might had come in handy when I was making my psu and burned like 10 transistors during testing.

As we all know, this cannot work.

>>1537393
What.

And what is the black stuff

>>1537403
black magic coil

>>1537385
>How important are sockets for ICs and transistors?
Not very? Are you under the impression that they are?
I stopped using sockets for ICs once I started using in-circuit programming.

>>1537393
looks like a boost circuit, aka joule merchant

How do I calculate the efficiency of a system that has irregular shaped pulsed output?

Can someone help me here? Got this projector which won't turn on, so took it apart for fuses or something and I find this burnt resistor.
I can't find any info regarding nor the pcb nor the component, which I can sort of read "100 16M" but it's burnt out so not 100% sure.
What can I do?

>>1537439
I see an exploded capacitor?

Machine learning fag here. I'm building a small robotic arm to interface with my PC through an arduino. To test things I've bought a cheap Chinese kit arm with 4 mystery meat servos. What is the best cheap power supply for the servo shield, can I just splice the end of one of those 5v 2A wall warts?

>>1537461
Good that's a start, I thought it could be a resistance!
The + side looks like it's going to the fourth pin, second row of the connector. That connector goes straight to the PSU.
Should be replacing the cap be enough or the probably might be another fault higher up that caused the cap to blow?

>>1537473
Replacing the cap should be enough. The cap probably was old.

>>1537379
Wear latex gloves, or wash your hands after you're done soldering... It's ok to touch lead, as long as you don't end up consuming it.

>>1537485
Ok so I belive the markings say 100 16M.
From what I've collected this would be a 10pF 16V 20%
Would that be correct? (like cap C993).

>>1537432
>joule merchant
Sounds good, no more toroid winding.

>>1537501
I'd replace that C901 too while you are at it.
Where did you find 10pF at?
>100uF 16v tantalum capacitor
That's the closest I found.

>>1537512
I have read that on small caps the first tho are the significant figures, the third the multiplier.
But yeah this is a SMD..I got it wrong.
Also measured it at 7.4x3.95mm.
Does the M still mean it's at 20%?
And C901 would then be 4.7uF 25V 20% 2.77x3.75mm?

>>1537523
I am not certain as to the exact values. Surely another anon will come in here and provide his / her knowledge on the subject.
According to here though
https://www.electronics-notes.com/articles/electronic_components/capacitors/capacitor-codes-markings.php
the M does mean +/- 20%.
I hope you get it working anon.

>>1537530
Well I thank you so much! Already you have helped me really a lot!
I would be still searching for resistances now, instead I think I have found what I need:
https://eu.mouser.com/ProductDetail/AVX/F911C107MNC?qs=sGAEpiMZZMuEN2agSAc2pqyLMxN%252bdaSreOHgjUNtsmJERYGzcq4miQ%3d%3d
https://eu.mouser.com/ProductDetail/AVX/F931E475MBA?qs=%2fha2pyFadugBNWncAbXQX06zwN2R3ta7%2fKTBbfF0f7butNu1NFXIGw%3d%3d

I love milk but i hate cleaning burned milk.

So, what if i mounted this cunt on the air sucker thing above the milk cauldron, aimed at it.
Do you think it would be able to measure the raising milk surface fast enough so that arduino could smash the gas button and turn of the stove?

>>1537317

>>1537553
s/b fine, look at the datasheet for sample rate.

>>>1537597
Do you have any experience of them being used in oscilators? I tried googling ``gyrator oscilator`` and variants but have not found much. I`ll probably have use a transistor version because I don`t think there are cheap op amps that work well in the mhz range. And from what I can see I`d have to learn how to use transconductance amplifier/norton op amps and honestly they sound like vodoo to me.

>>1537530
100 does sound like it should be 10pF, but with a cap of that size I very much doubt it.

>>1537560
>those SIP components

I'm looking at powering a stereo amp and a raspberry pi off a single SMPS. Was looking at this one - Meanwell LRS-100-24. It's 100 W, 24V 5A.

It looks like it has 2x V+ and 2x V- screws, so two outputs of the same 24V voltage.

The Pi Zero only needs 5V, 2.5A. What's the easiest way to get that from 24V 5A without buring my shit up?

>>1537385
it's 2018 + 1 already, use SOICs ya scrub

>>1537646
SIP pullup/termination resistors are still legit
dem PLCC20s doe

>>1537676
adjustable buck modules from our friendly chinks, based on LM2596 or XL4015, rated at 30V input or more
you'll have plenty of power for peripherals, should you choose

>>1537676
The pi zero uses only 200mA from my experience.

>>1537678
Oh they're pullups? I guess the thing doesn't have internal pullups then. Not a bad idea in any case.

>>1537676
if this is true
>>1537679
just use a 7805

>>1537679
>200mA
run a few instances of
>while : ; do : ; done
and get back to us on that

>>1537683
old-school CPUs and periphs rarely drove their buses when not required or requested. pullups (and bus loading, and bus length, and repeater/transceivers to mitigate all the problems) were a matter for the system designer

>>1537696
a buck module from chinkland doesn't need a heatsink, insulators, screws, etc.

>>1537553

once the milk starts rising, you cant stop it anyway, so your idea is crap. get one of these. they start to rattle as soon bubbles form.

I hope this is dark enough, if not I'll be going through a fair few sheets of this stuff.

I tried washing the dust off this transparency, though it turns out that it isn't some sort of solid porous layer on one side that absorbs the ink, but a gel that does not play nicely with water. So I ruined one of my ~10 sheets, lesson learned. Still can't find a good way of telling the two sides apart (rubbing each side to see which one squeaks doesn't work) so I'm defaulting to licking a finger and dabbing it on the corner of the transparency to figure out if it has gel on it or not. And then I wasted a bunch of ink trying to get my print settings right (I have to print it not from inkscape or browser A, but browser B and it has to be set to landscape otherwise the scale will change, etc.) to get to this result. I'm getting a bit ahead of myself though, since I've yet to build myself a lightbox.

>>1537750
Oh and the scale is somehow still fucking wrong, plus some of that aforementioned gel is now peeling off my wrist like dead skin after getting sunburnt.

>>1537470
No one's answering because no one knows what you're talking about. Also we hate Arduino but that's another matter. What board are you trying to use?

>>1537754
>we hate
The royal we (pluralis maiestatis) is the use of a plural pronoun to refer to a single person claiming to represent others based on subjective mental conditions.

>>1537761
Lurk more, freshman

>>1537754
>Also we hate Arduino
yes

>>1537617
Gyrator can be used just fine in an oscillator, just replace whatever component you're expecting with the gyrated version, and it'll work as long as the voltage/currents are within capabilities.

Now you have me thinking, if you could make an oscillator out of two gyrators that are "out of phase" with each-other and are actually swapping which component they represent (rotating their imaginary vectors, so to speak).

>>1537393
That doesn't work.

>>1537393
Sorry but is there a meme I'm ruining here?

>>1537804
You actually have something auto-loading the thread to see if you are quoted don't you. Next-level shitposting.

>>1537808
? Click the little "Auto" checkmark box near the bottom of the thread

how do i measure water temperature without putting shit inside of it? something like a laser or some shit that is cheap and can worth with fartduino
the distance is about 2 meters

>>1537676

The easiest way not burn your shit up is getting a dual output 5v + 24v model.

>>1537798
Quadrature oscillator = two integrators in a loop. dx/dt=ω*y, dy/dt=-ω*x => x = a*sin(ω*t+φ), y = a*cos(ω*t+φ), a and φ determined by initial conditions. In practice, you need some form of automatic gain control otherwise the amplitude will drift.

>>1537804
No, the Joule Thief has already been arrested. Maybe you can ask your sim game to explain the working principle in some detail, like how to calculate the frequency.

Is the LoRa32u4 board the cheapest/easiest way to get an all-in-one LoRa arduino setup, or are the some other good ones? I don't need the battery capabilities as I run from mains, but I can't find anything cheaper without it.

I usually do normal nanos with external SX1278 boards and it just gets so messy with all the wires and logic level converters, so these combined solutions seems like muuuch better deal.

>>1537798
>doesn't work
or does it

>>1537754
>>1537763
>hating the most successful oshw project ever
Looks like the snobs, jocks, and cool kids took up electronics.

>>1537956
Don't lump me in with the cucks. I love arduino and use it for 99% of my projects

>>1537956
It`s just a microcontroller like most others, I dunno why some guys here frown on it. You just use the tools at hand to get the job done, tarduino or not.
>>1537967
>2020-1
>not going full analog

>>1537956
I was already cool, been programming pics and avrs about 7 years before the tarduino even existed.
Also I refuse to use their gay ass IDE and write anything called a sketch. Total kiddy babyshit.

>>1537956
>>1537999
>tools at hand to get the job done
Using it to make an LED blink is what we're against.
Many times other, more suitable devices and means are widely available for a task but the asspies opt for arduino because,
>muh arduino is so cool - look at me...

>>1537893
>like how to calculate the frequency
How do you actually calculate oscillating frequency of such circuit? You can't just use small signal model for the transistor since it goes in saturation so I'll assume you have to trow Ebers-Moll model at the problem and then what? There won't be any nice solutions to the problem because there are diodes, meaning you get transcendental equations which most likely don't have analytical solutions, which means you have to solve problem numerically. At this point you might as well use SPICE, set some cursors, read your frequency and call it a day. Am I missing something here? Is there an easy way to solve this?

>>1538030
>http://www.ganino.com/files/BSTJ/Vol38/bstj38-3-785.pdf
Small signal model, autism, frequency dependant gain.. The circuit is quite old, spice, cursors and other stuff are crutches. And it`s a non-regulated circuit, the frequency changes with load, that is why phone chargers and other stuff that use similar topology have feedback.

>>1537808
I don't see how that's relevant, also 4chan X.

>>1538030
I don't see how the particulars of the transistor matter, I think you can get it to work with a MOSFET also. All that matters is that the inductor drains the capacitor of its charge once the transistor turns on, which turns the transistor off, which boosts the voltage back up to the LED's forward voltage with the remaining energy in the inductor. In the case of an NMOS you'd just need to consider the ON and OFF modes of the transistor and ignore its gate current, only taking into account its threshold voltage, which should make the calculation of frequency pretty simple. As it requires the energy of the inductor be mostly expended before the oscillator goes on another cycle, figuring out the frequency change with respect to the load would be fairly simple.

>>1537956
>thinks popularity is a measure of utility
>complains about the cool kids
fascinating
in fact, the arduino environment is in almost all cases barely good enough, at best, especially when your use cases frequently lie outside of the "physical computing" ambit

>>1538012
this anon gets it
infantilism is not cool except on >>>/d/

>>1537999
>muh 12AX7 Nixie flasher
ok kid

>>1538037
haven't heard of phone chargers using that topology, but a lot of battery-powered photo-flashes seem to use it. no feedback afaik, just a bleeder resistor chosen to balance the input current at desired voltage, if even that

So hwo do you people deal with understanding and desiging your own analog electronics using ICs?
I'm in my last year of my mechatronics and if a circuit has only 2 transistors I'm already confused.
Not that I'm going to need to know much about analog electronics in my career but I'd like to be able to understand bigger circuitry.
How did you do it?

>>1538115
Personally, I never use transistors in the linear region, op-amps only. I can't think of even a single reason why you'd want to do otherwise, excluding radio circuits. If you want to amplify an analog signal with more power than an IC can take, then you're better off using a class-D amplifier.

Take this with a grain of salt, of course.

>>1538118
I'm talking more about transistors used as switches.
But when it comes to transistor amps I quickly lose track about what's going on.

>>1538122
If you're using transistors as switches, there shouldn't be an issue. There are some simple topologies you should get familiar with, such as the totem pole, H-bridges and half-bridges, but apart from them you'll be probably using transistors in common-emitter configuration with some sort of load between the collector and +ve rail. Same applies to N-ch MOSFETS. The load will usually be some power component like an LED or heater, or perhaps a resistor for the purpose of acting as an inverter.

Transistors (usually MOSFETs) will also be used in the linear region with feedback, such as in a linear constant-current or voltage regulator, in which case you'd typically use an op-amp and a voltage divider/potentiometer in order to set the output. The extreme gain of an op-amp in this case means that the characteristics of the transistor hardly matter, and the transistor will always be driven such that the measured quantity (voltage or current) will be practically identical to that of your voltage divider or pot.

Any particular circuit you're trying to understand?

>>1538115
there are some canned topologies in which bipolar transistors are often used, e.g. single-transistor common-base, common-emitter, common-collector; multi-transistor long-tailed pair, push-pull, cascode; etc. which have been well-characterized in general (e.g. common-emitter has high input impedance and low output impedance and can voltage gain > 1) and for which there are higher-level design equations
but in general, you can approximate a bipolar transistor in its linear region as a current-controlled current sink: for a pnp, x electrons go from b to e, beta * x electrons go from c to e
if you already understand passive component networks well, check into The Art of Electronics (see OP)

>>1538122
>switches
oh, well, roughly the same idea but you can think digitally: if current is flowing through the b-e junction and the polarity's right, current will be flowing between c and e if the polarity's right
have a look at BJT datasheets, both small-signal (e.g. 2N3904) and power (e.g. 2N3055) to get a sense of the parameters in play in BJTs. some that will probably concern you in non-fault switching cases are: saturation c-e voltage, max power dissipation, Vbe-Ice curve, (sometimes) maximum reverse Vbe. but the cool kids use MOSFETs these days which are a slightly different ball game :^)
>transistor amps
those do get complicated when matters like frequency response come into play, but often they can be broken down into multiple stages of the standard topologies, plus negative feedback around the lot. if you understand Pic related, the common-collector amplifier (high input impedance, low output impedance, voltage gain ~=1, current gain > 1) you have most of the tools necessary to understand other types. when you understand the common-base amplifier (low input impedance, high output impedance, current gain ~=1, voltage gain > 1) you will also understand why not to connect loads to the emitter in switching circuits (hint: Vbe is relative)

>>1538102
The oscillator needs a loop gain of >1 and a phase shift of 180° between output and input. How can the LC low pass filter between collector and base provide that phase shift as well as an amplitude that allows start-up?

>>1538196
That assumption implies there's any kind of loop topology whatsoever. A 555 timer doesn't function on that principle, and I don't see why this should either. What about a relaxation oscillator using a relay? They're based off hysteresis, nothing to do with loop gain.

>>1538196
His circuit can't work, stop replying.

I have a circuit involving an led and motor that I measured is using at peak about 900ma. 300ma is going to the led while only 20ma is going to the motor and each are connected to a mosfet. The mosfet that is connected to the motor goes to 100f in about 10 seconds so its obvious that the remaining 600ma is getting eaten up by that mosfet and causing it to heat up right? How do I deal with that? I tried connecting the drain of the motor mosfet to the negative of the led since the led can take that current but it becomes shorted so how do i make use of this wasted current?

>>1538231
Post circuit diagram

>>1538233
Wish you didnt ask that. I've posted it here before and feel dumb asking so many questions that id rather seem like another person but i guess words wont make it clear to understand

>>1538234
I imagine that thing at the top left is an ammeter, and the one at the bottom right is a voltmeter, but what's the thing above the voltmeter supposed to be? A lamp? Or the motor? If that's what the motor is, not the inductor, then why have an inductor there? You're just asking for high ESR and transients, motors are their own inductor and should handle PWM signals just fine. As for putting a 1Ω resistor in series with the resistor, please use something a little more reasonable.

Now a great chunk of your power will be eaten up by that 1Ω resistor, otherwise since I can't see the resistance (and inductance) of your motor, how am I supposed to know how much current it should be pulling?

>>1538234
don't put loads on the emitter or source of a transistor for switching
put the load on the collector or drain because Vgs or Vbe is relative to the source terminal resp. the emitter, not to ground

>>1538234
the mos is across the battery
turn it on, it is a 'short'

>>1538241
Its a lamp but this app version I have doesnt have a motor so its just a substitute. When I attempt another more complicated circuit I'll find some proper software. The led has a max forward current of 1.8A and a pulse forward current of 2.4A. Since the power source is usb 3.0 I dont need the 1ohm resistor anymore so its been omitted. I dont have much information on this motor, all I know is it can operate from 3v to 30v and testing it I managed to get it up to 350+ma running at 9v. I put an inductor there since 5v is too low id prefer over 7v average

>>1538248
LTspice = free = not literally shit

Also don't use the 2A max of your USB power supply for current limiting, the power is going to be wasted somewhere and you don't want it to be inside your power supply.

Are you trying to boost convert the motor or PWM it?

>>1538244
Dont fully understand what you mean. The led is connected to the drain. Are you talking about the motor?

>>1538247
I see what you mean. Should the motor mosfet have its gate pin connected to pin 3 of the 555 instead? That probably would limit the v to 3.3 to the motor which is why I didnt go that route

>>1538249
>Are you trying to boost convert the motor or PWM it?
Both. I would prefer it not have pwm but pwm is more efficient for driving a motor rather than direct voltage right? Since I'm boosting the voltage with an inductor I need pwm for the mosfet gate. If I could forgo using pwm for the motor and boost the voltage using another method I would but I dont know any other means of doing so

>>1538256
You can't do both. Model your motor as an inductor with a series resistance, and don't forget to put an antiparallel diode across it. You'll want to vary its speed by changing the voltage across it if you can, no point in varying the voltage. But simply relying on duty cycle to change the voltage without feedback is a bit shit, a boost converter should use a feedback circuit of some kind to read the voltage, compare it to a potentiometer or reference voltage, and if it's larger you disable the signal to the FET. Variable duty cycle is more complicated. Without feedback you'll either be losing a lot of power or you'll be boosting it up to like 80V. To get maximum efficiency (and voltage) you need to match the duty cycle and inductance to the PWM frequency and load. More or less set the inductance that the time constant of the inductor with your load resistance is in the same ball-park to the inverse of the frequency, possibly with the duty-cycle thrown in there.

Since you're at maybe 1kHz, and perhaps the motor is 10Ω, so 10mH would make sense for an inductance. Plus or minus a factor of 10.

So do you know the (effective) inductance and resistance of your motor, assuming you're planning a circuit for a real motor in the first place? Also if the voltage across the motor is oscillating much at all you'd probably want a bigger capacitor.

>>1538264
I already have a real working circuit. I looked at the bottom of the motor and it says Type: gk26/400R so Im guessing the 400 is the resistance but I dont have the equipment to measure the inductance of the motor. Pic related has it at about 900ma at this point its running at around 90% duty cycle by my guess however I could measure the peak current at what duty cycle if needed and yes its at around 1.5khz. Its at this point where the mosfet is hot, the voltage to the motor is low and current is high whereas at around 30% duty cycle voltage to the motor is at its highest around 7v and current is 300+ma then the mosfet is fine. In actuality I just want a simple fan to turn on once in a while, I dont want to have to vary the voltage and all that nonsense the dimming i want just for the led

>>1538280
Look up the motor's datasheet for the inductance is all I can think of, that or make a relaxation oscillator with it and a 555 timer, the frequency will tell you its inductance.

If the FET is getting hot then obviously you're putting way too much current through it or somehow keeping it in the linear region.

>>1538285
Thats my main issue. 900ma is going through the circuit with the led using 300ma and the motor using 20ma meaning the rest is most likely going to the mosfet near the motor. If I could somehow transfer that wasted current to the led, motor or even limit it anything but useless heat would be great

>>1538285
or they're shitty mosfets with high Rds(on) and all the resistive joy that implies

>>1538292
do you know if the 330uH inductor is saturating? do you know if the diode is fast enough? do you know whether the MOSFET is turning on fully? (this is really easy to find out if you just look at the datasheet)

>>1538129
Thanks for the insight

>>1538292
>the rest is going to the mosfet near the motor
Basically your inductor is way too small or your frequency is way too low. The inductor is supposed to be there to store energy in, but it's filled up at maximum current/energy in far less than the ~0.5ms that the FET is on for, so the inductor just acts like a wire for all that time. Replace the capacitor with a 5n or 500p and see what happens.

>>1537736
>you cant stop it anyway
uhh yes you can, just turn off the heat you dum dum

>>1538328
If you've got one of those hefty iron elements then it will just keep pumping heat into the pot for a few minutes after turning the heating coil off.

city is a tourist spot and draws a million of drug users and homeless, they can smell copper miles away like sharks. Can`t find any junk.. I really need a microwave transformer.

>>1538333
i cook on gas. cooking with electricity is too inefficient

>>1538358
just buy a microwave a take it out

>>1538196
I've built two, both work flawlessly. If you decrease the 100Ω resistor to 50Ω the supply current increases from ~7mA to ~14mA and the LED gets brighter. It also works without load, the rectified output voltage is about 30V. With reduced gain (R=2.5K, supply current 0.3mA) the output of ~0.3Vpp resembles a sine wave. The feedback mechanism is still unclear to me.

>>1537317
>>1537321
I especially like how she's holding the heating element and not the handle, so clearly it's not even plugged in.

>>1537393
What doesn't '''work''' there is shitty schematic symbol and refdes for an inductor.
>ffs draw a coil as a coil

>>1537379
I'm at least twice your age and have been working in electronics for longer than you've been alive, most of that with tin/lead solder, and it's not done any measurable damage. The worries over lead-bearing solder are way overblown, especially for a hobbyist.

>>1537678
>it's 2018 + 1 already, use SOICs ya scrub
Surface mount is fine for production use or if you spin a PCB for your project, but for hobbyist use through-hole is still a good way to go, and using sockets on through-hole ICs not only protects your ICs but allows you to re-use them on another project later if things don't work out like you wanted them to.

>>1537385
If you're hand-wiring your project on perfboard then I'd recommend using IC sockets if you can afford them, you'll protect your ICs that way and also if there's a possibility of an IC getting damaged or needing to be re-used somewhere else then it's easy to just pop out of a socket.

>>1537676
>I'm looking at powering a stereo amp and a raspberry pi off a single SMPS.
Frankly, I wouldn't do that unless it's a dual-output supply, +24 and +5. The reason why is digital switching noise from the RPi might make it's way through the stereo amp and out your speakers.

>>1538115
Get this book:
https://www.amazon.com/Art-Electronics-Paul-Horowitz/dp/0521809266/ref=sr_1_3?s=books&ie=UTF8&qid=1547739081&sr=1-3&keywords=art+of+electronics
If you can't learn analog electronics from that then you should find a different line of work.

>>1538196
It looks like this is what's supposed to happen:
>capacitor charges through resistor
>meanwhile LED is supplied through resistor+inductor
>capacitor charges up enough to turn on transistor, shorting out LED load
>capacitor discharges through the inductor
>transistor turns off, LED turns back on
>cycle repeats until power is removed
however since the transistor has a dynamic range and not just on/off, I'd also think it would reach a point of stasis where the transistor is partially on, so this really won't work.

how do i make an 8 bit program counter
stacking t flip flops seems like the least retarded way but i havent discovered a way to make them because i cant into pulses
is there another way to increase a number by 1

>>1538404
she likes the way it hurts

women can be hardcore too you pleb

>>1538102
>I don't see how the particulars of the transistor matter,
Frequency dependant current gain
>In the case of an NMOS you'd just need to consider the ON and OFF modes of the transistor and ignore its gate current, only taking into account its threshold voltage, which should make the calculation of frequency pretty simple.
"no". Because of the FET capacitor you really have to take into account the nitty gritty of transistor behavior.
>blocking oscilators
>simple
To explain how they work maybe, not to describe with math.
>>1538415
This, and the dynamic range makes some input voltages to have a dampened oscilation. If it`s either too high or too low it`s just dies.

Like my pcb design, /ohm/?

It's a dual h-bridge with two DRV8838 and a few pins so that it can function as a power hub for servos

I'll have the prototypes done and assembled by january 21 but it cost 450 chinese yuan for 5 copies so it's a little expensive, but for 1000 units it's like usd 0.80 each. There are malls in China you can just go in and find people that will do this kind of job for you.

>>1537750
Use tracing paper and paint thinner vapor to blacken the black areas, not that transparency

>>1538425
>let's make red look like orange and blue look like purple
>why don't we just make the resistor white so we can see the goddamn colors?
>you're fired

>>1538394
In my view the mystery circuit is a variant of the traditional Colpitts oscillator where one capacitor is 'parasitic'. C2 is the capacitor at the base and C1 is the effective output capacitance of the transistor.

Since C2 >> C1, the high output impedance is transformed down to the low input impedance of the transistor in common emitter configuration to uphold the loop gain. The phase difference between the two ends of a resonant LC circuit is always 180°.

This is probably the simplest possible garden light circuit not invented in CN.
Kudos to >>1537393 for the original content!

>>1538415
>>meanwhile LED is supplied through resistor+inductor
Isn't LED off, not conducting any current, since the knee voltage is less than Vbat? Here is my take on the theory of the operation:
>capacitor charges through resistor
>once it reaches ~0.7V transistors turns on
>since there is a high di/dt change on the conductor a positive voltage induces on it
>left terminal of the inductor is held at 0.7V due to BE diode, since change is positive only way a voltage can induce is for collector voltage to drop
>because Vc is negative charge from capacitor flows from the capacitor causing voltage on the capacitor to drop
>voltage on the capacitor goes below 0.7V required for BE diode to be turned on
>transistor turns off
>di/dt change induces a negative voltage on the inductor, meaning collector goes higher than base
>LED turns on and "uses up" the induced voltage
>cycle repeats

>>1538309
The through hole diodes near the motor are spares I took from a microwave panel so I dont know if theyre fast enough but I guess for 1.5khz they should be fine. How would I know if the mosfet is turning on fully or the inductor is saturating? These are the data sheets

mosfet: https://www.vishay.com/docs/91291/91291.pdf
inductor: https://www.yuden.co.jp/cms/wp-content/uploads/sites/3/2017/03/170327_Automotive-SMD-Power-Inductors_ut_HP.pdf

>>1538313
If I replace the 1uf capacitor with a smaller value wouldnt the voltage not current decrease? Thats what happens in the circuit sim though I dont know if that will reflect the same in reality

how do i make a t flip flop guys
nothing is working

>>1538524
Make a t flip flop with what? Potato chips?

>>1538526
BJTs resistors diodes caps etc
preferably not too complex
ive tried some things on google but none werk
using 2n2222a if it helps

>>1538418
can you into opposite edges? use a master-slave architecture like the pros do
also make the adder and the register separate. otherwise you'll have a hell of a time trying to implement jumps/branches

>>1538429
bretty gud design. it might not matter for this circuit, but you might want to consider the effect of those little green inductors in placement of your decoupling caps (C2 and maybe C4)
those malls in china sound lovely

>>1538443
at that rate how are you gonna tell carbon film from metal film from porno film
>and dude went to work for the mafia and made cement resistors out of his local enemies

>>1538511
first, those MOSFETs are rated only for a Vgs of 10V. they barely turn on at 4V and aren't even guaranteed to turn on below that. you need something with a lower Vgs(on) for this application
next, that 555 isn't going to output more than two diode drops below Vcc (3.7V) without help. sometimes you can get around this if you add a pullup resistor to Vcc (10k should be enough) which might be enough to turn the FETs on fully... might, since, you've got two different FETs in your simulator, and (if I understand the symbol right) the one on the inductor appears to be in its linear region, not that it is necessarily anything like your real FETs anyway
next, 1uF isn't nearly even close to enough to store an amp for a full millisecond. you need more like 1000uF, maybe more
finally, those mystery diodes you pulled from a microwave... who knows what their current carrying capacity is like, let alone reverse recovery time or reverse leakage. being 0805s it's my wild guess they're rated for 0.5A max
always good that you're reading and paying attention, but you should get some more of the fundamentals under your belt so you can smell-test your designs before building them, and know how to diagnose them. switching power supplies are like wizards: subtle, and quick to anger

>>1538532
what is the pic?
>can you into opposite edges? use a master-slave architecture like the pros do
what

>>1538528
Pic related should work. Resistor and capacitor values are not too critical so you can replace them with what you have on hand (of course it needs to be reasonably close). Main thing to watch for here is current in the base. It must be large enough to put transistor in the saturation.

>>1538547
pic was not for you
this is a pretty good textbook. you should learn at least this much about digital logic fundamentals before you try to design a CPU
http://www.gatestudymaterial.com/study-material/digital%20ciruits/text%20books/Fundamentals%20of%20Logic%20Design%206th%20Charles%20H.Roth.pdf

>>>/sci/10304847

Well, /ohm/?

>>1538560
Yes it does create a magnetic field. Hole is just an absence of an electron. In order for hole to travel through a semiconductor it must displace bound electrons and since bound electrons have move magnetic field is created.

>>1538447
I made a mystery circuit too and it happily works with two LEDs in series. I can hear it in the radio on around 550kHz. The frequency changes when you touch the thing which may support the small parasitic capacitance theorem.

>>1538514
I meant the 50nF timing cap, not the filter cap.

>>1538404
Or how she's apparently soldering the battery.

>>1537560
This reminds me I need to pick up a proper barrel battery to restore an old 486 board somewhat accurately for a class I'm teaching.

Let me preface this by saying I know I'm fucking dumb with electronics.

I am trying to charge my phone with steam. Ok so what I really mean is I am building a small steam engine, and I want to attach a small DC motor backwards on the output shaft of the steam engine to act as a generator. From there I figured I could use a voltage divider to drop the voltage to 1.5V. After the divider, I figure I probably need a small capacitor to smooth out the power from the small generator. From there, a set of small diodes to ensure the battery I am charging doesn't instead power my motor. On the other side of the diodes, I hook up the rechargeable AA battery, which is in one of those cheap phone charger units.

Would this work, or am I even more retarded than I assume I am? Pic related kind of outlines my plan.

>>1538718
Looks reasonable to me, but the lower diode should not be there. There's room for improvement.

>>1538718
Why not just take the output and run it through a 5V regulator and then hook that up to your phone?

>>1538740
>the lower diode should not be there.
this - it won't work with the bottom diode there.

>>1538718
lmao use one of these instead:
https://www.aliexpress.com/item/meat/32919452954.html
It's a boost converter that will increase the voltage to 5V regardless of the input voltage, it's pretty efficient too. If the voltage output is higher than 5V (it probably won't be) you'll need a different kind of converter (boost converter).

>>1538532
>first, those MOSFETs are rated only for a Vgs of 10V. they barely turn on at 4V
I read Vgs(th) is what turns on the mosfet. Its rated for min 2v to max of 4v.
>being 0805s it's my wild guess they're rated for 0.5A max
if youre talking about the smd diodes near the pot those are 4148 diodes the datahseet says Current - Average Rectified (Io)

150mA
but i assumed since its going to my pot not the led or motor it should be fine and ive seen some schematics for high power leds that also use 4148 diodes. Ive had an issue trying to pick the right mosfet thats why i bought another model just in case but vgs of 10v, does vgs(th) mean the difference between the gate voltage and gate source for it to turn on? should i buy another mosfet that has a vgs of <5v?
>>1538618
Yeah I could try that. You might have some point about the frequency, I saw some other person conplaining about the same mosfet im using overheating while using 200hz the solution was to use something over 500 which I am but its still worth a try to get it to about 10k. I noticed that adding another filter cap parallel to the 1uf cap that is 10uf significantly decreased the heat acceleration of the mosfet

>>1538783
>Vgs(th) is what turns on the mosfet
Yes, to 250µA at Vgs=Vds. A square-law characteristic has no real on/off threshold.

>>1538785
So should I buy a new mosfet that has a vgs that is less than 5v? My mosfet let is the same and it isnt overheating, what issue would i be dealing with if i use a 10v vgs mosfet that is driven by a 5v source?

>>1538785
I did a little more reading and it seems im not fully turning my mosfet on because the vgs is too high for my small input voltage which leads to overheating. I'll buy a logic level mosfet as that is much more appropriate for my circuit and see if the heating issue goes away

>>1538783
>some schematics for high power leds that also use 4148 diodes
if they used them in the current path, it was purely a miracle that they worked and did not self-destruct
>does vgs(th) mean the difference between the gate voltage and gate source for it to turn on?
yes, barely
>should i buy another mosfet that has a vgs of <5v
yes, and in particular look for an Rds(on) rating at a gate voltage <5V
then buy a few extra for your stock. IRLR024 would make a fine choice for this project, Pic related

>>1538794
P = I^2*R. the LED probably isn't drawing enough current through the MOSFET's resistance to cause significant heating

>>1538800
Cool thanks. I'll buy a couple of those. I read that since my mosfet isnt fully turning on the resistance skyrockets. I'll replace both for better performance however I dont know for sure if that will fix the big issue here which is wasted current. I know that when a mosfet gets hot it draws more current but thats still insignificant to the 600ma that gets lost in my circuit

>>1538808
that, and turning the frequency up so you can be more sure you're not saturating the inductor, and making the storage cap a lot larger. run through those MC34063 equations to check the values of your main components

>>1538809
Ok thanks and one last question. Im looking at the IRL/U024 and I see the Vgs is 16v then Rds(on) with several different Vgs values. Do I have to assume that since my datasheet only has Vgs of 10v that it will not fully turn on before that? Why do these datasheets not state a minimum voltage for the mosfet to become fully on or is that too abstract

>>1538813
Vgs_MAX is different from Vgs_THRESHOLD

>>1538813
Rds(on) is a measurement of the MOSFET in saturation. an Rds(on) spec at a given Vgs implies that the MOSFET can be saturated at that Vgs and will behave as the low value resistor you want
Vgs(th) is the boundary between cutoff and linear regions. it is relevant as an upper limit for Vgs when you want to ensure the MOSFET is off

>>1538808
A side note to the conversation - mosfets have a current requirement for on-off behavior. For tens of kilohertz it's usually not much, but it can grow unexpectedly depending on the type of transistor you're using and the frequency it's at. If you're using an un-buffered circuit, or worse, an op-amp with limited bandwidth, you'd have issues with turning on and off in time. An example is say the gate has 10 nanocoulombs of charge. If you wanted to switch at 1kHz, you'd only need to supply and drain at 1mA, but if you're trying for 20-50kHz, you might run into op-amp bandwidth issues or current supplying/draining, depending on how you're doing it. Some things are active-high for example, and when they're "off" they're actually floating - no draining of the charge. Not terribly common these days but it's something to keep in mind. For high-power mosfets at 1 MHz, some need more than an amp, and all that heat gets lost in the mosfet, and you run into stupid issues because the gate becomes a capacitative load at high frequencies. Only really noticeable for those frequencies and a high-voltage+current mosfet, probably.

>>1538867
Er, 10nC*1kHz, = 10uA/s. Not a lot for a small mosfet, but some are 50-100nC, and as you starting climb those kilohertz or push the switching rate...

Some dude here said he never saw any battery chargers with a joule-thief like thing. There are, big clive took two apart in a video. (I think it`s called ``non-delicated teardown of powersupply`` or something)

>>1537317
Not sure if CRT work is right for this thread but anyways. I have this old 60s Sony portable CRT. When you turn it on all you get is a white line across the center.
Any first thoughts? if its capacitors I could do that, dont have a schematic atm though. and dont worry it hasnt been plugged in for about 3 months

>>1538868
That's an underestimate. If you're using a MOSFET as a switch, you want the rise/fall times to be a small fraction of the overall cycle. Current = charge / time, where time is the rise/fall time, not the total period.

>>1538550
please tell me i'm retarded

>>1538718
>am I even more retarded
Yes. Your generator would have to output at a higher power than the battery is capable of outputting. For example, if you want to charge an eneloop AA battery with this arrangement, your system as to be delivering the 1.5V at at least 500mA. Good luck with your little gizmo lmao.

>>1538995
This also applies to this thing, by the way.

>>1538775

Notice how input voltage range is 1V to 5V, and output is 5.1V-5.2V with output current of 1-1.5A?

5.1*1==1.5*i
>i=3.4A.

Is your piss poor generator capably of outputing your 1.5V at 3.4A? Because otherwise you are shit out of luck.

>Its literally as simple as P=IE.

>>1538818
>>1538823
Yeah see I get it now. I didnt understand why exactly my mosfet wasnt turning on all the way because i was under the impression that since the vin is 5v and the vgs(th) is 2-4v my mosfet should be fine but im not working with vin im working with the output of the ic which is vin - 1.7v so its not being fully turned on now it makes more sense

>>1538928
True, I was considering a bare minimum of on/off, not even minimizing the shoot-through/non-ideal transition. "Napkin" calculation, transistor should spend less than 10% of their time transitioning on and off right? Would need 10-50x current capacity to ensure that.

>>1538912
The vertical deflection isn't as strong as it should be. (CRT is a central dot, and the deflection circuits make it zoom across the scene, besides the electron-beam circuits changing the intensity).

Simplest explanation would be the capacitors are probably shot, since everything else is probably resistors, inductors, antennae, etc for something that old.

P.s. VHF hasn't been legal to use for almost a decade... you might have a hard time finding a transmitter.

>>1539077
With a logic-level mosfet you can do about 1-10A+ depending on which you find. I bought the FDP6035AL years ago, and at 3.3V, which is about 5V at the output of a 555 timer, it's specified for over 5 amps. Newer mosfets are better.

>>1539078
Thanks for the info. yeah I know I couldnt actually pick up OTA signals anymore unless I wired it into a DT converter box. I've just had this sitting on my shelf for 4 years as a display and was curious to spend some time to get it working again. $15 in caps wouldnt bother me too much as I got the original thing for free from a family friend

>>1538718
>From there I figured I could use a voltage divider to drop the voltage to 1.5V
Everything was making sense until you got to this part

>some random lc oscilator works at 20Mhz.
>opamp square generator+integrator to make ramp doesn`t
This is going to be one of those days

>>1538912
Seconding the capacitors. I'd also like to know if it has deflection coils or plates (some of the smaller CRTs use electrostatic deflection). As for getting a signal into it, you can probably buy an RF modulator to turn a VGA signal into a compatible signal for the thing, if not you could likely hack past the demodulator and directly inject the video signal into the CRT. Since it's old you can probably find a circuit diagram on it online, if not it would probably only take a few hours to reverse engineer.

>>1539000
>implying it won't work if you try to pull 0.1A from it
That would be a fucking awful circuit if that was correct, the two different currents are simply different maximum currents as a function of different input voltages. Have you ever built a boost converter?

>>1539077
I think your FET was turning on properly, it's pretty fucking hard to get a FET to turn on partially (i.e. enough to produce a bunch of heat) without negative feedback of some kind. I think the only reason it was producing a bunch of heat is because there a very small resistance in series with it: the saturated inductor.

>>1539103
Check the gain-bandwidth product of your op-amp, often they're made with a maximum around 1MHz.

>>1539144
>gbw
I know, I`m tring to make it oscilate at audio frequencies, it`s just that the LC is much more ocmplex. It`s probably due to one being soldered and one being in a bread board.

>>1539148
I don't find any issue with audio frequencies on a solderless breadboard. The standard topology for an op-amp oscillator has the output always at Vcc or Vee, so it might be better to use a comparator instead. Plus the normal "take everything apart and put it back together on a different part of the breadboard" routine, since those breadboard connections can get bent into not making contact.

I ruined a through hole/pad and cant figure out what I need to do to fix it
>drill out broken bits
>put in new eyelet
>???
>fixed
the eyelets just seem to magically stay in and that doesnt make sense to me. In a few examples they seem to just roll themselves magically with no tools on the other side of the board

>>1539227

instead of putting the part in all the way, let it stay 5mm above the pcb, then solder it from both sides of the pcb. if it's a chip, use something like a wire wrap socket coz it has very long legs.

Is there a way to trigger on and off a mosfet using only one NO monetary switch?

>>1539144
>I think the only reason it was producing a bunch of heat is because there a very small resistance in series with it: the saturated inductor

That could be it too. But still, I read that if youre using a mosfet as a switch its better to have the voltage go over vgs(th) however im powering the gate with only 3.3v. I'm also adding another 555 into the circuit since I bought 2 just in case. I'll have one for the led and the other for the motor so the pot doesnt affect the motor speed

Hi guys, sorry for asking spoonfeeding but I want to make a supercap powered battery welder (chargeable with 12V), and I need to control its charge and discharge with MOSFETs.

I want to know if:
- Will 4x IRF540s in parallel (Rdson = 40mOhm) make the cut for the discharge circuit?
- How would it affect if I don't have any MOSFET driver ICs such as MCP1407?
- Do you have any circuit scheme that lets me control charge and discharge of my cap bank?

>>1539103
Check the slew rate of your opamp

>>1539234
Look into pullup and pulldown resistors.

>>1539236
>MOSFETs in parallel
If you're too close (75%?) to the maximum current on each FET, I wouldn't, because silicon switches are susceptible to thermal runaway. However, having 4 seperate supercaps each with their own MOSFET could work fine.
FET driving can be as simple as a 2-3 transistor totem-pole, though you can be sure that a dedicated IC will do so quickly and efficiently and will probably not be your weakest link. As far as PCB space and part cost goes, the MCP1407 may well be the best option.

I managed to add the second 555 ic into the circuit without any mistakes thankfully. The output is running at 13.6khz and the voltage difference is incredible, i've got it above 11v and this is with usb 2.0. The guy that said the inductor was being saturated must have been right, the higher frequency is causing it to collapse at a faster rate, outputing a higher voltage and the mosfet is no longer heating up. Now the question is should I bother buying the logic level mosfets?

>>1539282
gud werk. tried it with load yet?
two points in favor of buying: they would give a touch more efficiency and they'll be good to have around for other projects

>>1539282
Your FETs are obviously working just fine, good to see that you've got a good result. Note that your output voltage will be highly dependant on the effective motor resistance, so if you ever change the load you'll have to retune the frequency and possibly also the duty-cycle. If you replaced your FET and filter cap with higher voltage ones (IRF840 for example) you could keep turning up the frequency and you'd probably easily get up to a few hundred volts.

>>1537679
it's going to have a DAC hat on it too

>>1539236
>battery welder
A spot welder like this?

/ohm/ - Electronics General
>>1539340
no RULES, no restrictions, unlimited random noise.
add your babble, continue the tradition.
what about a third ohm thread?

>>1539361
That thread has no reason to exist.

>>1539078
>>1539144
Well none of the caps look busted but that doesn't really mean anything.

>>1539399
2/2

>>1539400
And a rather messy solder job

does anyone have an image of household items fuse replacements
i think it included paperclip, melted cheese, a nail, a bullet etc with the amount of amps they can take

>>1539403
Oh god that PCB looks horrifying

>>1539426
Yes, la vache qui rit is good for 16A.

>>1539439
what do you even call these in english?

>>1539444
Alu foil.

This post brought to you by the Rayovac "High Energy" 9V carbon-zinc pile.

>>1539282
>the inductor must've been saturated
You didn't even calculate your charge rate? You didn't optimize your mosfet on time for the inductance? I get this is not an engineering thread but reeeeee

>>1539399
>none of the caps have breakable lids
well shit. You'd have to check under them for downsplosions, but first, I'd check the caps closest to the CRT. If you're lucky, there could be CRT tuning resistors. Usually it's just colors and intensity, but there might be ones for the vertical and horizontal deflection. You could go as far as checking the voltages on an oscilloscope, the vertical signal would only be at 60Hz, though the horizontal would be around 15-16 KHz superimposed, if it's all one coil. Easily within the range of an oscilloscope.

If you left the CRT running for a while, a failing/failed cap would get hot.

>>1539403
Absolutely disgusting. It might just be the board got goop on it, and the capacitance of some circuit is bent, or charge is leaking. Grab a toothbrush and alcohol, scrub a dub dub.

>>1539426
For a plain fuse? Tin foil was a classic and should easily be calculable with sheet thickness and width. Cheese likely wouldn't work because after burning, the carbon is still conductive.

An interesting fuse, a thermal fuse attached to mosfets, biased with a rectifier at the max amps you want. Above the rating they would rapidly heat. Could even just use a crowbar circuit, so if the voltage above the mosfets rises above say 5V (because above 15A. For a mosfet switch, you'd want schottky diodes to prevent the body diodes from shorting out (and they'd get hot, .3-.5V @ 15A = 7.5W max dissipation).

A current-detector and triac would also work, you'd need a sample-and-hold so say if it's above the threshold, it cuts out power until the sample-and-hold is reset.

>>1539403
>I have junk, what do?
Take it to the recycler.
No wait, first recover the museum-grade Germanium transistors and donate them to some stomptronics aficionado.

>>1539426
One individual strand of stranded wire (I forget what gauge the cable is, but maybe 24-26) is good for about 1A.

>>1539470
>>1539428
>>1539469
Well never mind, I was lifting one of the pcbs and the corner of it snapped it half like wet cardboard. i might be able to just resolder the severed connections but damn what a weak board. otherwise its going back on the shelf as as display piece

Is this bullshit? >>>/sci/10309736

>>1539479
is that dependent on length?

>>1539484
>biode
Sounds like bullshit.

>>1539484

uh...
"When transistors overfill and release power, they put out enough energy to satisfy another terminal on one transistor of the same size two times, more than is necessary for a signal, and often resistor pairs or other types of power recycling methods are used to compensate for that."

sounds like bullshit
So it's a transistor that has two outputs each with 1/2 voltage of the input.
Sounds absolutely useless but then again the same was said about lasers and actually pretty much anything ever.

>>1539403
old school as fuck. they don't make em like that anymore, and for damn good reason.
>>1539480
too bad. if it's so weak it might not be worth fixing.
that TV is adorable though, you might be able to drive the CRT with something else. give this project a try maybe:
http://tubetime.us/index.php/2014/05/31/electrostatic-crt-driver-design/

>>1539529
This is where it cracked, and there's a transistor under the end tab for some reason, probably using the metal connector as a heatsink. Maybe its savable maybe not but from researching other people who have worked on these little tvs they are a pain in the ass

>>1539530
forgot pic

>>1539530
My experience is not with TVs, but I have worked on radios with similar construction.
the dirt for sure is doing you no favors, it can get conductive over time.
Germanium transistors, if it's got those, go bad over time.
You already saw how the boards can break. If you want to fix it, you (usually) can't just solder across the gap, you must brace with solid copper wire.
If you don't have a schematic, these things are very hard to figure out because everything is packed in so close, and the layout of the traces is so haphazard.

>>1539588
well fuck, looks like i made it worse by opening it up. though it wasnt really working before anyways.. thanks though. i seem to have a knack for trying to fix something only to make it worse lol

>>1539480
>like wet cardboard
>implying it wasn't

>>1539426
looking for this?

>>1539657
>implying it wasn't
FR2, ladies and gentlemen!

>>1539286
I did but sadly a new day and a new problem. I noticed adding my wood blades onto the motor it was spinning a little too slowly. I measured the output of the 2nd 555 and it was only a few mv so I took 99% isopropyl and cleaned the board twice and nothing changed. I suspected the timing cap of the 555 failed since I couldnt get a reading from it so I replaced it with a smd cap which did something, I'm getting a little over 2v from the output but its not triggering the mosfet. The output is supposed to be 3.7v I'm not sure what happened, I didnt change the circuit at all from yesterdays night

>>1539630
>i seem to have a knack for trying to fix something only to make it worse lol
keep at it. that's how you learn.

Could you measure ionising radiation flux with a current transformer?

>>1539736
try it and see
post pics

>>1539741
The problem is a CT steps the current down by a large factor so you'd need a very high signal:noise ratio in your amplification and reception circuitry. Plus if you're trying to detect a single electron or alpha particle by its magnetic field alone you'd probably be better off with a Hall sensor. I think either one would be have to be made with superconductors though. Electrostatic detection with avalanching seems to be unbeatable for now. Excluding photomultipliers.

Could you decode a brain fart with a quantum blockchain or would that require a jargon peddler?

>>1539924
you can do anything with blockchain. stop asking.

>>1539657
based image

>>1537553
something that could be used far better with software and a webcam. electronics is rust. the future is programmable matter

>soldering iron just stops working
>sunday
pls

>>1539978
fix it

>>1539978
>soldering iron just stops working
Tool is not stong enough for iron, stay with soldering copper.
>sunday
The seventh day is a day of solemn rest.

>>1540018
I don`t know how, I think the nichrome wire just accidentaly itself. It`s not a soldering station, just a 10$ iron. I was making a ozone generator, 2morow I`ll post pics.
>>1540081
>The seventh day is a day of solemn rest.
hmmmmmmmmmmmmmmmmmmmmmmmmm. maybe you are right

>>1539333
I took your idea and designed this. However I don't know how can I replace the P mosfet in the charging path with an N mosfet.

>>1537432
/pol/ pls go

>>1537317
>tfw got started clicking on random Jewtoob vids
>see /ohm/‘s favorite BigClive
>2nd random video in and I think he already explained why my dumb light builds weren’t getting powered correctly from those USB power bank boards
The two outer pins on the USB are the 5V power pins, and then the two inner pins are data pins. If plugged into a computer, the data pins will talk and decide on a charging level like 350mA or whatever the device can output. When plugged into a a wall charger where the phone or device only wants power, the data pins get shorted together and it will send the full 1000mA that the charger is capable of putting out.

So I guess when I tapped into only the two 5V pins, the board will output a standard low current. I think I need to tap into the two data pins so when I turn the light on, the board will want to put out a lot more current. Maybe that is why it was flashing and going into some protection mode even though I was drawing 700mA from a little board that is supposed to put out 1A.

Yay or nay?

Am I learning anything?

>>1540092
I will put three of these bad bois in series, I have no idea of the voltage because it sparks when I put the voltmeter. I thas a very high output impedance unfortunaly. Tomorrow I`ll buy a shitload of .1uF or 1uF caps and try to make a more powerfull one.

>>1540121
photo is shit because I have to hold the camera still for 2 secs in a pitch black room, I busted a fuse and have no replacement in my bench so I`m using another outlet..

>>1540093
I would keep it simple and use a small 12V 20W halogen lamp along with a relay if need be. Nominal R is 7.2Ω, cold about 0.5Ω. The PTC property limits the initial current and the decreasing resistance charges the cap faster than the fixed 8.2Ω. If you want it even faster use another lamp.
The circuit I posted is from https://www.instructables.com/id/Tutorial-Spot-Welder-for-Lithium-ion-18650-Battery/

>>1540119
>the data pins get shorted together and it will send the full 1000mA that the charger is capable of putting out.
The charger will always supply full voltage and current it is capable of.
Adjusting the data pins tells the device being charged how much to ask for.
Unless the device connected has smart circuitry to look at the data pins, the state of the data pins will mean nothing.
Some devices look for certain voltage each of the data pins to know how hard to load the charger.
These chargers put resistive dividers across each data pin (from the + and - pins) to say how much it's capable of.

tldr: Unless you have a smart device connected to the USB connector, the data pins mean nothing.

>>1540133
Thanks, I didn't think on the idea of using relays for the charging part, and it actually sounds like a really good idea, I won't need to rapidly switch it on and off, and won't need as much current as the discharge circuit so I can do it without much trouble.

Anyone have a layout to a cell phone jammer?

>>1540157
metal box

>>1540157
i don't know rf but i looked into this out of curiosity. as i understand it you need to increase the ambient noise in the cellular band until the phone can no longer discern the signal from a tower. google says:
>A signal of -60dBm is nearly perfect, and -112dBm is call-dropping bad.
so i figure you have a noise floor in the 110dBm area (+-10ish?) and you'd need to add 50-60dBm of noise. math says that means you need 10^(x/10)*1mW = 100 or 1000W of transmitter power to do that. i know you don't need that much power because jammers exist and don't draw a kW, so does someone know where i went wrong?

>>1540191
actually these things are huge so maybe those numbers werent wrong

>>1540157
Couldn't you literally just google one?

>>1540191
I found it anyway. Should be good to about 100ft. Changing C1 and L1 will change the frequency it jams.

Advise me on an El cheapo ~$50-100 multimeter, please.

What is this electronic component? There is no continuity between its sides and I'm wondering if it has something to do with why my electric kettle isn't working anymore. The finger is for size reference.

>>1540224
Could be a 100Ω 10% resistor.

>>1540196
>I found it anyway
but 'it' is not what you think it is.

>>1540196
So it`s just an oscilator?
>>1540224
seems like a chunky resistor.

>>1540224

It's a resitor, probably 1W, and, if I'm seeing that color code right, valued at 1kΩ. Or 100Ω, I can't tell if that's a really faded orange or a particularly red brown, though I'm inclined to say "orange", when comparing it to the definitely-brown first stripe.

Should measure close to one of those values, either way. No continuity at all would indicate that it's fried.

>>1540233
>>1540238
Thank you. Will try to replace and see if the kettle starts working.
Btw how the hell does a resistor fail? Did it get too hot and the wire inside broke or something?

>>1540224
Kettles have a thermal-cutout to protect against nuclear meltdown.
pic looks like a fusible resistor, not enough current handling for the main heating element.
chinkshit way to fuse it

>>1540245
Interesting. Now that I think about it, the resistor really shouldn't prevent the kettle from heating up, as it is between the two indicator LEDs, and not affecting the actual heating element. Oh and I checked the heating element and I'm not getting continuity between its sides either.
Would it be fair assume that the heating element somehow failed and then subsequently the resistor did too, because all the current went to it, instead of the heating element? If so, then my last question would be this - how the hell does something as primitive as a heating element fail? It's just a big piece of metal, isn't it?

>>1540191
Jamming isn't done by desensitizing the receiver, it's done by preventing synchronisation by injecting a wrong signal at the right time. This is expensive. The broadband amateur approach could be a carefully tuned Jacob's ladder..

>>1540259
stresses induced by thermal growth will gradually widen any microscopic cracks/defects in the element until it breaks. metals also get weaker at temperature. you'd think they could design a heating element that doesn't get stressed due to expansion, but maybe they want it to break.

>>1540259
>I checked the heating element and I'm not getting continuity between its sides either.
measuring between the white and red wires at the quick connect terminals should be in tens of ohms.
2200W @ ~240V is about 26ish ohms and would be less when the element is cold.

>>1540273
Yeah well I get no reading there at all.

>>1538429 here, the boards have already been assembled. Will post pics later.

Is this BBC Micro:bit any good or is it just for teaching african refugees how to program python?

>>1540301
What MCU?

>>1540309
NXP KL26Z
It also has a magnetometer and an accelerometer and runs MicroPython

>>1540311
Doesn't look too bad I suppose, maybe a bit too IOT for me though.

>>1540312
It's made literally for children

>>1540301
Looks cute

Can I omit this zener diode? In simulation it doesn't seem to do anything at all

>>1540337
How about you post the ENTIRE SCHEMATIC instead of just that small piece of it so we know what the whole damn circuit is supposed to do?

>>1540344
The entire schematic is that picture and what it is supposed to do is in the filename

This is my reference. It says the zener is to protect against "overvoltage", but the circuit will take 9v max and the mosfet is for up to 16v, so I think the zener isn't needed

>>1540363
>up to 16V
if that's the Vgs(max) rating, and you pinkie-promise you will never feed the circuit more than 9V, you can lose the zener

>>1540244
>Btw how the hell does a resistor fail? Did it get too hot and the wire inside broke or something?
yes
>There is no continuity between its sides
maybe they're in orbit

>>1540135
some power banks are smart devices, though still very much in the brainlet category

>>1540095
this goy doesn't into organized crime

>>1540135
This was the video
https://youtu.be/jHRrnepF8aU

>>1540213
Uni-T, Extech, Klein, Southwire, Innova...there’s a ton of brands in that price range that are all about the same thing, probably many of them made in the same factory. Look on Amazon and see what has the features you want and has good reviews.

There’s one called a Vici VC87 that looks like a straight knockoff of a Fluke and that’s really popular. And then there’s this list, it’s kinda overwhelming and I have rarely seen most of the brands in US stores but as I said before, many of them are likely rebrands of others anyway. I ended up getting brands sold by Home Depot and Lowe’s so I had the option of returning them if I fucked them up within a week but probably paid a couple bucks more than if I ordered a generic brand online.

http://www.eevblog.com/forum/testgear/multimeter-spreadsheet/

>>1540379
Vgs max is -8, so I'll change for one that can take -20, but it's actually meant to be powered by 4 AA batteries (6v) so it shouldn't kill the mosfet

>>1540401
nah, keep that MOSFET. it has a nice low Rds(on) rating at a nice low Vgs. just replace the zener with a 150k resistor. when power is applied correctly, Vgs will be -3.6V, which is a good solid on-voltage for that part

Goddamn it why are dip32 slim sockets such a pain in the ass to source

>>1540379
>some power banks are smart devices
Only in that if no current is being drawn for x period of time they turn off the booster.

Here's the h-bridge I posted above in Eagle form

>>1540431
Fine looking, but didn't you not cross out your name on the Eagle form?

>>1540196

that kinda circuit is fine in the FM range, not at 20x that frequency that cell phones use.

>>1540423
just use two rows of machined female header pins

scored this guy for $25 a few days back, how'd I do?

>>1540531
What for?

>>1540531
That's a cute chink chang scope you've got there.

I want to measure the rate of steam production from a sealed vessel to the outside, and I figure using pressure sensors is a good way to go. But the pressure sensors I'm finding have too high a range for my purposes (the semi-ubiquitous BMP180 goes up to 40,000Pa and has an accuracy of ±12Pa). So does anyone have any pressure sensor recommendations? The difference probably isn't going to be more than 100Pa (0.0145Psi) but may be as low as 0.1Pa (0.0000145Psi). Would perhaps a spring-loaded valve and some method to measuring the cap's displacement work better? I want to avoid moving parts, or parts that could suffer from water ingress.

Hot wire anemometer with some method to compensate for the temperature changing?

>>1540433
No, the name is on the board, I deleted it just for posting here to keep my anonimity

Also, the first h-bridge makes the motor spin both ways just fine, the second one makes it spin only one way, I don't know why yet

>>1540560
maybe take a closer look at those little buggers, and maybe surround them in flux and put some hot air on them

>>1540561
Didn't need to do that, it already works, woohoo!
Now on to ordering 1000 of them and designing the injection molded enclosure

>>1540572
>ordering
Do not forget the 2000 motors.

>>1540583
I have like 150 of them, the boards will be assembled in lots of 200

>>1540537
Oh my god, even better: the manufacturer is "Hung Chang".

>>1540541
I think you could use a temperature sensor. For a given reaction, the only variables are change in water volume, temperature, and pressure. With some chemistry you could determine how much pressure there should be for n molecules boiling within v volume at temp T. PV=nRT, something like that. It'd probably be easier just to manually calibrate it and see if there's a predictable function you can use. E.g. get a pressure gauge, get an accurate thermistor that can handle being boiled, and take 50-100 values and see if you can find a decent equation for it, and if not, interpolate the results with more reference points.

>>1540650
That's what mom and pops companies are called in China

Hello. I`m trying to make some tasty ozone. I have this 8 stage CW generator, I just made and it make some hearty zaps and a small corona. I`m using 10nF caps because they were cheap here. My question is:
Aside from going autistic and making a inverter, how would I increase the output of this generator? I want a big manly corona. Are ccfl inverters = ballasts? Can I use electronic ballasts to generate high-frequency mains-ish voltage?

>>1540673
>inb4 insulate it
Already encased in comfy heatshrink. Some magic light came out on lower stages but was contained with hot glue

>>1540673
>increasing the output
1) Reduce the resistance on the end. It's supposed to be for protecting you, and will get in the way of ozone production.

2) Each stage just retains voltage/charge, so maybe more capacitance will work? The output currents are hilariously low for ions/ozone, so I'm thinking internal resistance isn't going to matter much.

3) Higher frequency, more output nodes may reduce the wear-and-tear on each output, and since the voltage falls after each cycle from discharge, you could probably directly multiply the output by increasing frequency.

Normally at the highest voltages you'd worry about it penetrating insulation (hence why the ladder is so good, it has large spaces). Maybe you could reduce the multiplication, and instead add inductances like a tesla coil. A transfer could boost the voltage substantially, as long as you don't burn out a switching mechanism (a field collapse would ramp up the voltage like crazy, see flyback transformers). You could try a joule-thief design, but remove the LED and have a smaller inductor.

On that note, a tesla coil/transformer itself would probably be far more effective. Most high-voltage ionizers that operate off batteries/DC use a potted transformer+circuit for flyback..

>>1540684
I know. I had a tv flyback I used for this stuff but it accidentally itself. I was the guy commenting about joule thiefs and self oscilators couple threads ago because I was too lazy to solder 40 caps and diodes. Managed to get 600 V dc out of 1.25V DC. but could not do anything better with home-spun xformers.
>measuring with a cheap multimeter I got 3uA before the LCD turned black and the buzzer set off for some reason, that was with 10M ohm.
>with 1M ohm I can feel a bit of pain and my heart feels weird sometimes


>increase frequency
Yeah I know that, but how? Right now it`s running out of 220v 60hz from my outlet, I don`t have a inverter, maybe an H-bridge? I`m afraid of getting ded if I thread those paths

>>1540712
Also I`m afraid of putting lower output resistor because I have no idea of the output voltage (It`s too high to measure). And a telas coil is too big and honestly just hand turning all those coils give me back pain just by thinking about it.

>>1540712
Why didn't you try higher voltages with the transformers? Sounds like 6V would've gotten you over 5kV, but then you're risking arcing.

You could increase the frequency with a properly rated transistor/mosfet before the first stage. I know some LED-compatible dimmers should do PWM chopping, but mine doesn't. A triac wouldn't increases your frequency.

>>1540718
The output voltage can't be sustained for long with 10nF caps. Effectively at the end of the circuit you have an n*wall voltage 10nF cap charged up (minus a few .4-.5 volts for each diode). Dry body has around 500kOhm resistance, and the caps would spark before contact.. As long as you have insulation which protects each wrap against leaking, it shouldn't take too many turns (you're not working with magnificent current). PVC wire insulation is around 600V, so even 20-30 turns on a PVC pipes slid inside each other could work. Primary on the inside tube, because you need all the field contained with the second. If you did incremental tubing sizes, you could have tube->wire wrap->tube->wire wrap to prevent multiple layers shorting at the edges.

>>1540560
Start hitting stuff with a multimeter. H bridges aren't hard to figure out

>>1540755
>Why didn't you try higher voltages with the transformers?
because the oscilator circuits I was using stopped oscilating above 2v and peaked at arround 1.2v.
Also I was just making a joule thief with a third winding. Joule thief windings had 5 turns each, third winding had nearly 100. It also wasn`t linear at all in relation to turn number, probably because of the mistery ferrite toroids I found in the garbage.

>>1540684
each stage is made of capacitors in series, they store less energy with every stage, elevating voltage at the same time.

>>1540898
Each stage increases the voltage by peak voltage/cycle. The capacitance is "reduced" each time the next capacitor is charged up, but since the voltage across each capacitor rises linearly with the number of capacitors, the total capacitance is the same as each capacitance. Another way of looking at it is F = Q/V. Each cycle you're increase the charge by one capacitor worth, but you're also increasing the voltage by one capacitor worth. F is constant.

>>1540755
>You could increase the frequency with a properly rated transistor/mosfet before the first stage
what do you mean? I`m sorry but I can`t figure how this would work.

Can we just take a sec to appreciate Art of Electronics 2nd edition? Quite a bit of the older circuits didn't make it, plus the whole good v. bad circuits section was one of my faves.

That being said, the layout of 3rd is definitely an improvement

>>1538912
Might be as simple as cleaning/replacing the vertical size potentiometer. Whats the model number OP?

>>1540650
I suspect that would be far too dependant on external temperature, pressure, etc. Plus the inside of the vessel won't necessarily be at the boiling temperature of water since there would be a significant amount of ordinary air mixed in with the steam. But it's definitely something to think about.

>>1540900
>>1540898
>>1540786
managed to improve it a little by working on insulation and also made a corona thing with some VERY thin microscope thing glass and a needle. The needle glows with a faint purple. No idea of how much ozone it makes, probably not a lot but hey, it works.

>>1540906
Nevermind, you need proper +/- for the ladder to work. You could use a rectifier, and then a proper push/pull circuit. It's annoying and takes a bunch of parts.

>>1541056
Normal air temperature/pressure would be easy to have. mix of air and steam inside the chamber won't alter the temperature or pressure properties much. You really didn't describe what the fuck you're doing, so any suggestion is kinda worthless.

>>1541114
It's literally a cooking pot, with some way to measure the rate of steam leaving the pot so you can control the rate of boiling.

>>1541115
If you don't clamp it shut the pressure is going to be like the same as the room. Why not try humidity sensors above the stove, if they can withstand the heat?

>>1541121
The point is that there will be a small gap for the steam to get out, and since the steam has a finite viscosity there will be a small pressure drop between the inside and the outside as the hole in the lid basically acts like a resistor. I did the calculation with the Darcy–Weisbach (with a 4mm wide and 4mm deep hole) equation and it gave me a pressure drop of 84Pa if the boiling rate is 1L/5minutes (a very fast boil) and 0.55Pa for a boiling rate of 1L/12hours (probably a slow boil). This hole size would probably be too small for a rapid boil, so making it even tighter to get higher pressures is out of the question. Humidity sensors anywhere would probably lead to too much hysteresis.

A simple fan in a hole on the lid to measure the speed of the escaping steam would work fine, but having moving parts in the presence of liquid (especially something that needs cleaning) sounds like a bad idea. A pressure sensor can simply be gated behind a rubber membrane that will couple the pressure change without exposing the electronics to ingress.

I just now thought of this, but a rotameter like this image could be electronically encoded fairly easily if the floating piece itself were ferromagnetic, and a coil were placed at one end of the sensor, such that an increase or decrease in airflow would present itself as a change in inductance. The ball could easily be removed for cleaning, and would be easier than the rubber membrane to clean. The only problem is I doubt it would work in the case of an induction hob, but feeding some sort of control signal back into one of those beasts is probably beyond me anyhow. The idea is to build this (and a temperature sensor) into an electric frying pan so you can set the boil rate and forget about it. The thing would probably be controlled by an SCR or FET on full-wave rectified unfiltered mains, since that sounds easier than trying to control a TRIAC with an op-amp and what.

>>1540537
Thanks, it cost me $8. The guy who had it was basically giving it away, but I felt bad not at least giving him something for it.

>>1541133
Oh now that I think about it, that method would be somewhat difficult to design since the iron ball would be quite heavy and wouldn't be terribly suited to measuring a wide range of air flows (too large a ball means it won't measure low flow well, too small a ball and the hole itself will need to be small and won't let large flows through), plus the problem with an induction hob. So if I instead use a large (20mm diameter perhaps), hollow stainless ball in the rotameter and have two parallel plates on either side of it, I could measure a very drastic change in capacitance as the ball in the middle acts like a short between two middle plates (pic related). The ball would be much lighter and the capacitive system wouldn't be affected by magnetic fields.

To avoid hysteresis problems, this couldn't work very well with a hob with high heat capacity, so basically a gas hob or an induction hob would be the only way to go, though an electric frying pan (or other all-in-one cooking vessel) would probably work fine. A gas hob would be pretty hard to control, I wonder what valve you could use to linearly alter the gas flow, apart from a servo on a ball valve/globe valve.

>>1541154
Have you considered just drilling a hole in the lid and mounting a low-friction fan over it, to detect rotations? I have pan lids that are shit, but they're the kind with a hole built-in. 1L/12h is .023 ml/second. .023 ml/s * (.96 g / ml_liquid) *( 1 mol/ 18 g) * (22.4 L_gas / 1mol) = .027 L/s = 27ml/s = 27 cm^3/s. Another calculation using a reference, saturated steam is 1677.3 cm^3/g, so .023 * (.96) * 1677.3 = 37 cm^3 (probably the 100C is why the .hand calculation came out wrong, I used 1mol of gas at STP). Dor a square 1cm cross section of exit pipe, you'd have 37 linear cm of air movement per second, which sounds pretty damn quick. I bet you could detect that with a small fan fit for the purpose.

>>1541213
I mentioned it in >>1541133 but I'd like to avoid it if possible because the thing will easily get clogged with grease, or at the very least be a pain to clean. The problem with using a small hole is that the pressure builds up and causes the edges of the lid to start leaking significantly. 1cm is probably borderline for that, since the ~3mm holes definitely seem to have that issue. But it's also important to be able to measure very low levels of boiling, like a 1/10 or less on an electric stove.

I threw together a spreadsheet for a variable area flow meter with calculations that simply balance the gravity force of the ball with the drag force of the air going past it, pic related is a reasonable configuration. "h" is the height that the ball moves up the tube due to the influence of the flowing steam, the angle is the angle to horizontal of each side of the tube, "D" is diameter, "C drag" is the coefficient of drag, and "v" is the gas speed. Note that since the gravity force is constant, the drag force must also be constant, and so as the area of the meter changes the speed remains constant. 0.8mm minimum is a bit shit though, so I may replace the conical shape with a curved air-horn shape to get higher resolution at lower levels. Here the maximum height is 11cm, which sounds reasonable. I may or may not be designing a massive steam whistle.

>>1541240
oh whoops that's 10mm not 100mm, better recalibrate

>>1541244
The least invasive method would be a clamp-on ultrasonic steam flow meter that evaluates the travel time difference downstream - upstream.
If an immediate result is not required you could simply condensate the steam and measure the resulting amount of water, which can also serve for the calibration of other methods.

>>1540990
5-303W Sony

>>1541367
I don't need to calibrate it exactly, just have a control scheme to use it as feedback. Condensation would be too slow, but the ultrasonic method would certainly work. I'm afraid it would be more complicated and slightly more invasive than a cone on top of a pot with a ball in it, but it would certainly be easier to measure than the 0.3-3pF of capacitance between the two plates.

>>1541554
>0.3-3pF
That will be a challenge.

>>1541554
Ok, how about if instead of a ball I use a directional floater (like >>1541133) with vertical magnetisation, then use a low-pass filtered 3-axis hall-effect sensor to detect where it is? I'd need the ADC to be on the lid since I doubt long analog sensor wires will do me any good.

The other, capacitive method leaves me with only one decent method of working it: use the capacitor next to a resistor as a filter, feed it high frequency AC, and as the capacitance changes the output amplitude will change as a function of the filter's shifting corner frequency. Then I'd feed that back into an SA-602 mixer IC and get out a DC value proportional to the steam flow rate. The minimum frequency I'd be dealing with would be on the order of 1MHz, and at that point stray inductances and capacitances will become far too prevalent, so this magnetic method looks to be the way to go.

So now I just need to tune the meter shape to minimise internal pressure and minimise air speed. Problem is, for this the floater needs to have a low density to allow the hole to have a high area while ensuring the floater isn't too heavy for a low boil rate to push up in the first place. For this reason the capacitive method is superior, since the floater could be hollow instead of having to have a magnet inside it. So to progress further I'll need to check the sensitivity of a hall sensor to small magnets that are side-on. Actually using a single axis hall sensor at the very bottom corner of the meter might be sufficient to measure the magnet from the bottom, which would probably make it more responsive. I'll see if I can do this evaluation of hall sensors and magnets through datasheets alone, if not I'll need to buy one of those sensors and bugger around on an arduino.

Hello, I have a few questions about varistors. My friend has a 110VAC camping trailer heater that he applied 220VAC to, this fried a varistor on his heaters board. I want to replace the varistor (on the left), but all I can find in my scrap electronics is one with the same voltage rating, but a smaller diameter (on the right). I understand the capacitance of the varistors increases with the diameter, but I do not know how a smaller capacitance will affect the circuit.
My questions are will the lower capacitance varistor work in place of the larger one? Does the lower capacitance mean it will blow faster, or at a lower voltage?

>>1541752
It had a varistor but no fuse or PTC? It shouldn't matter what capacitance PTC you go for, the difference won't mean much. What's more important is that the maximum current increases with diameter, which should be a value that's larger than the current value of the fuse/PTC. If there is no fuse, then just go for the biggest size you reasonably can.

I don't see the point of a MOV that shorts the power rails together before burning itself out and going open-circuit, since after it's dead it will no longer be protecting the circuit. They're supposed to short the rails to prevent the voltage from getting up too high, and since they can only take this power dissipation for a fraction of a second, the fuse or PTC is supposed to trip, saving the circuit. If the MOV instead dies closed-circuit, then it's just one big fire hazard. Otherwise they're only useful for very short high-voltage spikes. You should add an appropriately-sized PTC in series with the power line after replacing the MOV.

>>1541759
Thank you for the help. The circuit does have a fuse, it blew when he applied 220VAC. So he replaced it and then plugged it in to 110VAC and it blew again. That's when he asked me to see what I could do, i'm pretty new to circuit boards and their components, but I figured this was a chance to learn a bit.
He did not say what amperage the fuse was rated for, and he removed it from the board before giving it to me. Here is a picture of the board. I suppose it would be safer to scavenge around for the proper valued varistor instead of risking using a smaller one. I think I'll replace the safety capacitor while im in there, since I found another in an old microwave.

>>1541771
Ah, that makes a lot of sense. The thing could say what current or power it's meant to take (i.e. it's a 500W heater), in which case you can assume that the fuse will be set to ~1.5 times that value or so. Obviously the original MOV was rated insufficiently for the purpose since it died, so definitely go for the biggest size you reasonably can (same pin spacing, not too expensive). My MOVs are maybe 2cm in diameter and barely cost more than the smaller ones, and I think they can take quite the beating. Suggest putting a fast-blow ceramic fuse in there.

>>1541781
>implying the purpose of a MOV is to protect against dipshits applying 2x nameplate voltage to a device
it did better than one might expect. one could argue it failed safe

>>1541781
Thanks for all your help, I hope that the blown MOV is the only culprit in this cooked board. Ive ordered a set of 14mm diameter MOVs to put in place of the 10mm one. Ill post results after I fix and test the board.
>>1541785
I still don't understand why he did it, but he seemed convinced that it was a good idea, im not sure why. The fireplace was not the only victim of his choices. He also killed among other things a brand new set of stack-able washer and dryer, though I believe the fuse saved the dryer, the washer was not so fortunate.

>>1541790
Where are you that you have 220V and 110V that easily available?

>>1541794
He apparently parked his new camper on a friends property and wanted to run power to it, so he ran 220V from a shop building straight into his camper receptacle. Thus frying a bunch of his electronics that didn't save themselves. Replacement/repair of the heater was forgotten about until the cold weather set in.

>>1541795
Ah. I guess you're the magic fixing guy now eh?

>>1541801
Pretty much. My job is aircraft electrician, which is mainly component replacement and wire splicing, but to anyone else it just means "anything electronic". I'll give a go at anything though. Worst that can happen is their shit is still broke in the end.

>>1541806
Do they run regular 12v DC systems in Cessnas n Pipers n stuff?

>>1541759
>supposed to short the rails
No, not at all. A MOV behaves more like two Z-diodes wired back to back in series.

>>1541849
hmm, zinc oxide has some nice properties https://www.youtube.com/watch?v=oAJ4XVmI4dc

>>1541806
Why aircraft and ships have 440hz networks?? I was always curious abou tthat.

>>1541903
smaller lighter transformers I guess

whats the least retarded way to decode instructions
hardcoded "if" with transistors for each instruction?
i've organized the inatructions so each component only checks 4 or less bits from the instruction bus
is there a better way using bjts resistors

>>1541908
go find this book, read it from cover to cover, and stop asking us to do your design work for you, pajeet

>>1541908
put a bit more charitably,
first you need to into digital logic
then you need to into CPU
then you can design an ISA that is actually useful
THEN you can worry about your discrete RTL autism

>>1537317
Okay, how the FUCK do I figure out how to route a PCB from a circuit diagram?
I want to etch a single layer PCB for this beeper circuit, but while it looks all neat on the diagram, in reality the pinout makes it into a shitty maze puzzle.
I could bridge the traces with wire, but I'd prefer not to. I tried auto-route in Eagle, but it generated a design that I'd be hard pressed to replicate with a sharpie.

>>1541934
practice
lots and lots of practice
hints worth considering:
>don't use autorouters. the pros usually don't, and the results you saw are why
>use a fine point sharpie, not a magnum
>feel free to go under/between ICs and other parts
>don't try to pack everything too tightly on the first pass
>don't constrain yourself to the official footprints. sometimes you can route component leads instead of traces where physical space allows, especially with offboard components like pots, power inputs, etc.
>KiCAD allows you to unlock a footprint's pads so they can be moved independently. maybe eagle does too but if it doesn't just replace the component with a couple of wire pads. putting just a mm or two between the leads of a component opens up enough space to bring a trace through

Speaking on schematics->boards, anyone have good ki-cad tutorials? Literally that's what I want to do, but the software doesn't make that easy for a novice. Figured it'd be like a sim but only with vector graphics/layout enhancements but... of course not.

>>1541848
>>1541903
Ive only ever worked on the B-52, but in training we covered systems from C130s, F15s and F16s, and If I remember correctly the DC voltage was always 28 volts. Doesn't mean they are all 28 volts, the course could have just done that to keep things simple, the B-52 DC systems do utilize 28 volts though.
As for ships and aircraft running 440hz, im afraid I don't know. The B-52 likes it at exactly 400hz, but some are a bit more picky than others. Once we had to switch out 6 different external power carts to find the one that that particular jet liked.

>>1541957
Digi-Key did a bunch of shorts on KiCAD 5, can't speak to quality, but it's in bite-sized pieces so you can bail out without much investment if you like
if you'd rather sit with a bitchy Dave Jones for a couple of hours
https://www.youtube.com/watch?v=qpw9dKxL2Ho
>like a sim
that seems a matter of expectations vs. workflow
to start, you draw the schematic, which might entail creating your own symbols, then (KiCAD unique) you associate the schematic parts with pc board footprints, which might entail creating your own footprints, then you open up the board and place the parts, then you start drawing connections, moving the parts around if and as necessary. eventually you'll see the "unrouted connections" counter drop to zero, at which point you might be done, unless you want to tighten/neaten your design
at some point in the process you draw board boundaries, maybe add cutouts, mounting holes, etc. etc. etc.
once you've finished with all that, you then File->Plot your design to gerbers and send them off to your board house or File->Print them to paper for toner transfer or just to guide your sharpie work, or you can hit Alt+3 in pcbnew for the 3D renderer then export to your favorite 3D CAD/CAM tool, or just sit back and fap

r8 my PCB design

It's kind of an arduino with a built in h-bridge just like this one: >>1540431

>>1541934
Dot board layout is always sharpie-friendly.

>>1541934
Change the design rules in Eagle so that it produces thicker traces with more clearance

>>1542007
Mother of fuck, how do you get this good?
That's incredibly tidy compared to the best I could manage in Eagle.

>>1542012
I thought about setting up wider spacing and fat ass traces more befitting my shit ass sharpie technique, but where do you do that?

>>1542017
Edit > design rules
BTW you can route that by hand if you have highe than 85 iq

>>1542017
>Mother of fuck, how do you get this good?
He doesn't. He missed several components from your schematic, and left an ugly jumper wire under the IC.
One can do better than that.

>>1542071
Here are the part references, if you have troubles identifying them.

>>1542075
And a 3D view, just because I can.

>>1542071
What's with that nasty 90-degree on the trace for Vcc 1 & 8?

>>1542092
Still not an acid trap, but I could have done it better, I admit. That's an easy fix though.

>>1541934
I don't know, but I want to hijack this post to say
>FUCK the 555 and FUCK its garbage pinout
Topological shitposting in chip form. Fuck.

>>1541849
Yeah ok they don't go all full short like some sort of capacitor, but if you feed them 220V they short anything over that. You know what I mean.

>>1542077
Place your components better, that's what I'd do. In fact, that's what I'd be doing right now if KiCAD hadn't crashed after I recreated your circuit diagram and tried to assign footprints, deleting fucking everything. I was going to say that there's no real reason to use Eagle over KiCAD as they look pretty similar to me, but I guess I'm wrong.

>>1542099
I'd agree with this. Can we get a breakout board for SMD 555s except with the pinout swapped to something more conducive to not being a pain in the rectum?

>>1537317
>Horowitz and Hill, The Art of Electronics
Not buying any books made by a hymie

>>1542176
Yeah, and relativity is a jewish lie.

>>1541953
>>don't use autorouters. the pros usually don't, and the results you saw are why
>citation needed
auto route has uses, but you don't use it for everything.
>>1542006
>Ref des overlapping
minus two points. vertically orient them the same too, unless you turn the board when you are soldering.

Where can I go to buy decent power modules?
I buy plenty of cheap chinese crap, but now I need a cc/cv buck converter and I want something I can trust.
Digikey is usually my goto, but they don't have stuff like this.

>>1542308
Something wrong with alibay modules? www.pololu.com may be more reputable.

for the anon who was working on the Johnson counter from hell and needed a convenient, quick way to build arrays in KiCAD, this just saved me an hour or two:
https://kicad.mmccoo.com/2017/02/01/replicating-pcbnew-new-for-arrayed-sheets/

>>1542308
just derate it by a few tens of percent and it should be fine

>>1537317
Only the most dedicated electro-engineets have mastered the deep state of concentration where you can hold the soldering iron around the heat coil without feeling any pain.

>>1537956
The problem with arduino is that it adds another level of abstraction to the avr processors. And while it helps you get started easier you have the job of doing all the work of porting it to (meaning doing everything the arduino abstracted in the first place) avr it later if you want to deploy it to anything which isn't an arduino.

>>1542529

you make a good point: if instead of making it easy to use, they had made it crazy hard like most processors, then it would remain obscure. almost nobody would use it, and everyone would be so much happier fingering their buttholes.

Can anyone tell me if it is normal for a torque screwdriver to still adjust the torque setting regardless if it is in locked or unlocked position?

Is this normal, or is it broke and I need to return it?

>>1542625
>thinking most processors are crazy hard to use
maybe you should have stuck with your original calling that you named at the end of your edit

>>1537317
I just realised that OP didn't put /ohm in the title, and now we've got two /ohm threads up. I guess we'll migrate to it in a day or so when this one dies?
>>1539340

>>1542811
>so when this one dies?
this one has already reached the bump limit
it's dead, it just hasn't fallen into the archive yet

>>1542824
It hit the bump limit a day or two ago, we usually continue to post on here for a fair while after hitting the bump limit. A half dozen threads ago we hit 500 posts, if I remember correctly.

>>1542829
I try to avoid creating a new one until page 10. SOME other anons seem to wear their panties a couple of sizes too tight to wait that long

>>1542811
that one's made of fail

>>1542870
>that one's made of fail
It's worse than this one for sure, but at least it has /ohm in the title and it still has a fair few people on it. If you do make a new thread when we hit page 10, make sure to post the "NEW THREAD" announcement on both threads.

>>1542529
>look up register in data sheet
>change register bits in code
Not hard.

>>1542343
>>1542499
i'm not just about rating, but reliability, consistency, low drift.
I have a similar one right now, based on the XL4016, and it's all over the place. unusable. It's charging a Li-ion bank (16.8v), it'll be on duty 24 hours a day, I need something trustworthy.
Usually I find what I need on digikey, but not this time. constant current/constant voltage adjustable dc-dc converters seem harder to find than they should be.
if I can find nothing, maybe I'll just use a bench power supply.

>>1542947
For Li-Ion charging, you should be able to find a name-brand (solar?) charger since it's a common enough thing. If there's a name-brand IC on the board that would certainly help, I think TI have a fair few solar panel and lithium ion charging solutions. If you find an IC that suits your purpose, you could have a look at its datasheet and look for a recommended circuit design and see if you can find anyone making that sort of solution.

By "all over the place" what did you mean? Did you compare it to a proper li-ion charge curve? Were the balancing lines functioning properly?

>>1542949
this is kind of a non-standard setup because it's what I could cobble together using some 3Ah cells I got surplus. Each one has its own protection circuit (making this project pretty safe overall).
The battery is made of 4p4s packs, stacked in parallel. The idea is, it's modular, each pack has its own balance circuit, and I can keep adding more packs in parallel whenever I have time to build them. So, a single balance charger won't do it, I'm relying on these HY2213-based balance boards to keep the cells in balance. They're all the same, brand new, so I don't expect them to get severely unbalanced.
All I need is a cc/cv supply to trickle charge the battery. no requirement for fast charging.
In this one I have, it's the current that's all over the place. It'll deliver a constant current into a dead short, but as the resistance increases (and voltage increases), it just kinda cuts off, the voltage drops way below the set open-circuit voltage. like many reviews say, these things are inconsistent in quality.

>>1542958
Do protected cells have overvoltage protection, or just undervoltage? If it is a simple constant-current supply you want, perhaps a name-brand LED driver would be more up your alley, since they're often made for constant-current. Phillips or Kree or one of that lot may produce something like what you're after.

>>1542966
These have undervoltage, overvoltage, and overcurrent protection.
But I don't just want to charge them until they cut off (at 4.35v), partly because that's still overcharged, and partly because the highest-charged cell in the series string will cut off first and leave the others undercharged.
Inconveniently, the balancing circuits I have won't bleed down overcharged cells enough for the overvoltage protection to release, so the only way the cells will balance themselves is if the charge voltage tops out at 16.8v or slightly under. Too high and the balancing resistors will be constantly burning off power.

>>1542973
I think it should be possible to power a constant-current driver with a constant voltage source and have it not go higher than the source, but I guess it's hardly ideal. You could always design your own charger.

>>1542979
>You could always design your own charger.
that might be the route I'll have to take. I didn't plan on having to do that, I thought this would be an easy thing to find.
I actually thought I'd be able to readily buy an AC-DC converter that could output constant current. surprised I couldn't find one.

>>1542982
Troubleshooting the existing thing is also a possibility.

Just in case you ever wanted to see the inside of a 70s Eaton Viking clock radio

>>1540133
Update on the welder:
I tried to make it the simplest way I could: with an arcade pushbutton as switch (which I could just switch with my legs), four IRF540N in parallel as the power switch and a couple more components.
Result? I discovered the IRF540s had a much bigger Rds(on) than the one I need (about 40mOhm), so instead of generating a spark in the terminals, when I pushed the switch, the MOSFETs heated rapidly.
So I bought some IRFB7430s in AliExpress, which have a much bigger current capacity and lower Rds(on) of 1.0mOhm.

Gonna try with a car relay too to cope with the import anxiety.

>>1543162
it could also be that R2 is too big, discharging all those gate capacitors slowly and leaving Q1-Q4 in the linear region
you are heatsinking all those, correct?

>>1543162
Look at the copper rails, drain tabs used, pins unused. Good wiring is no less important than low Rds.

>>1543180
>it could also be that R2 is too big, discharging all those gate capacitors slowly and leaving Q1-Q4 in the linear region
you are heatsinking all those, correct?
I took care of using 12V to prevent them to stay in the linear region but I didn't care about parasitic MOSFET capacitances. Will keep an eye on this.
About heatsinking I wanted to run a concept test so I didn't care. I guess thermal runway is really something.

>>1543180
I'll gather some new materials and try to build the wiring again with some more love and better materials.

Sorry for these newb mistakes, it's my first time designing something for tens of amperes.

>>1543189
Copper rail image courtesy of
https://www.instructables.com/id/Tutorial-Spot-Welder-for-Lithium-ion-18650-Battery/
About spotwelding connection tabs on 18650s. Smaller capacitor than yours, same voltage.

>>1543162
Any reason you are using TO220s instead of the much bigger available packages, like SOT-227?

Also as an aside I find it mildly annoying that datasheets list these absurdly high current limits, as if the source pin on a TO220 would be able to carry 195A no big deal.
It turns out the test methods they use involve exotic cooling which lets them use such high numbers for the absolute maximum limits. Look at http://www.irf.com/technical-info/appnotes/an-1140.pdf to see what I mean.

>>1543202
>Any reason you are using TO220s instead of the much bigger available packages, like SOT-227
I didn't know of them. Also availability. I live in South America, delivery from any American retailer is very expensive and retailers here have a very limited supply of components.
Will check that application note out.

>>1542625
>you make a good point: if instead of making it easy to use, they had made it crazy hard like most processors, then it would remain obscure.
Microcontrollers sold by the billions before tarduino came along and there were was already a decently large communy of hobbyists. There were free compilers, free IDEs, and custom programmers. They weren't obscure if you were even mildly into technology.
Basically you talk like a fag and your shit's all retarded.

>>1543202
>source pin on a TO220
very short in the image

>>1543202
>not calculating the maximum current from R_DS_ON and the thermal data

>>1543350
did you?

>>1543370
I usually do so since the maximum current is quite dependant on the ambient temperature.

>>1538404

>>1543464
*snap*
I'll add this one to my cringe in pain collection

>>1543488
Q: What's a worse mistake than dropping your soldering iron?
A: Catching it.

>>1543609
>A: Catching it.
*catching it wrong.

>>1537322
Hard to say without a symbol. Post a picture or scan of the schematics.

>>1537322
fuse