/diy/ - Do It Yourself

this thread's digits brought to you in part by the MAX1574 charge-pump LED driver IC

this thread's digits brought to you in part by the MAX1574 white LED driver

Please help an Arduino fag. Why isn't pic related working?

I'm trying to use a transistor to turn the Arduino on or off. If I press the push button then the transistor is turned on. Then the Arduino should automatically continue powering the transistor via the 5V pin but that doesn't work.

>>1371679
What are the symptoms.
Does powering the arduino directly work? What transistor are you using? Sure it's not backwards?
Ditch the arduino and try powering a series LED+resistor using the transistor, make sure that works at least.
Test the voltages everywhere when the button is pressed / not pressed.

>>1371679
maybe you need to connect both ground pins to the negative of the battery? Not sure.

>>1371679
the base voltage has to be higher than the emitter voltage for an NPN transistor to pass current. no way for that to happen here
maybe try a PNP transistor that gets pulled low by a switch *and* by a digital output in your sketch. be careful that you don't overvolt the digital output port, if you have like a 5V zener you could use that in line with the I/O port and resistor to drop some voltage

>>1371696
>the base voltage has to be higher than the emitter voltage for an NPN transistor to pass current
I thought you can use a transistor to control higher voltages with a lower voltage. If that's the case then why does the base voltage have to be higher than the emitter voltage?

>>1371707
transistors are current-mode devices. you can use a smaller current to control a higher current. base-emitter current controls collector-emitter current, so you could use a PNP transistor instead (Q1, R1, SW1) and pull the base toward ground to turn on the lolduino
the MCU won't be happy with unregulated 9V on its inputs or outputs, so you have to translate the level before applying it to Q1. thus, Q2 and R2. in your sketch, turn the Dx output high first thing, and when you want to shut down, turn it low

>>1371714
Alright thanks Anon, I'll try that

>>1371679
God these diagrams are awful. How am I supposed to tell which way around the transistor is?

why is it that a capacitor doesn't make an led flash when I have it in series with one in the circuit? Am I just misunderstanding how they work?

>>1371745
to give the benefit of the doubt, i assume you're giving it a square wave from a microcontroller or something. the problem is that you're charging the capacitor one way when you put the pin high and push current through it and the diode, but then you can't pull current the other way to discharge it because the diode/led blocks it. so the capacitor just stays charged and doesn't let any more current through it.

you can solve it by putting another diode in reverse in parallel with the led. you're not really gaining yourself anything doing it this way though.

>>1371732
presumably whether or not the flat face is facing you
unless the diagram always has the flat facing you in which case it either only allows one way of placing them or the person who made the diagram program doesn't know how transistors work

>>1371746
I'm using a shitty power supply off of amazon, not sure what the waveform is, haven't been able to buy an oscilloscope yet

>>1371752
capacitors only pass AC, so since your DC power supply has almost no alternating current the capacitor won't pass anything at all, and it'll act like an open circuit. if things are sized right the led should blink only once on startup as the cap charges, though it might be imperceptible. what do you want to do?

>>1371756
it does blink, very dimly. I'm just experimenting right now, seeing what happens when I do x, y ,z, etc.

>>1371757
>it does blink, very dimly
what sort of power supply are you using there boy

>>1371758
I meant blink once when I close the circuit, if that wasn't clear
https://www.amazon.com/gp/product/B071RNT1CD/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

>>1371752
>>1371746
alright I put an LED in parallel with the capacitor, didn't change much. Only the new LED lights up, no flashing

>>1371761
just kidding I put the LED in the wrong place. They both light up, but there's still no perceptible pulsing

>>1371761
yeah if you want flashing you'll need some transistors. you won't get it just by fucking around. capacitors pass AC, they don't create it on their own.

How do I convert these k maps for the inputs of a D flip flop to two k maps for the inputs of a JK flip flop?
I've been trying for about 45 minutes and I still can't get the first one right.

>>1371774
FUCK

>>1371777
google it, this is homework

Hey /diy/, complete fucking scrub here. I just bought a cheap starter set for electronics (meter, iron, solder, etc.) and tried taking a crack at the busted AC adapter for my monitor.

Opened the sucker up, carefully discharged the big fucking cap with a rubber-gripped screwdriver because I'm not an idiot. A high-pitched squeal came from the adapter under load. Without any load it was reading the 19V it should, but when plugged into the monitor it would give around 18.5V and then drop to 0 for a while, the monitor not even being able to present an image before it power cycled. The smaller cap near the DC out jack was bulged, and I thought it would be one of those really easy "change out a cap" fixes. Electronics store was out of 680u 25V caps that would be an exact replacement, it only had up to 10V and even scrub me knows it has to be equal or higher. I opened up a dead laptop charger from a laptop I no longer even have and hey, two Nichicon-brand 680u caps are right there, 35V ones that should work just fine.Desoldered one, popped it in place for the monitor adapter. My first soldering job was admittedly a shitty cold joint with too much solder on it, but it wasn't contacting anything else and seemed physically stable so I thought I'd leave it like that for now.

Plugged it in, grabbed the multimeter. Now it's all over the fucking place voltage wise, and when I popped the lid off the perfectly fine cap I put in 10 seconds ago was bulging hard.

Any ideas what the fuck it could be? I'm not going to plug this into my monitor unless it's giving me 19V stable again, obviously, but at this point I've got nothing to lose.

If I read all this books will I be able to do any eletronic project without worries?

>>1371791
I am googling it, and I've been trying to solve it on my own for a few hours and haven't made any progress. I wouldn't be asking for help if I thought I could figure it out myself.

>>1371807
Sure you didn't put the cap in backwards?

>>1371745
maybe it's just flashing so fast you can't see it. add a reverse diode across the LED and try a bigger cap (~220uF)

>>1371809
no, you'll have to do exercises too
but you'll almost always know what you don't know and have half a chance of finding it

>>1371809
not exactly. some of the subjects you’ll see in the art of electronics need a whole other book to have a comprehensive treatment of the subject. also you need experience to build an intuitive understanding.
but those books will go a long way.

>>1371807
whining means the switching converter in the supply has decreased its frequency so much that it's entered the audible band. the transformer in the supply is literally acting like a buzzer. that can be caused by the supply either having no load or having too high of a load. in the case of no load, the supply can overshoot its target voltage and damage the caps. that should only happen in poorly designed circuits though even when components have aged. so i think something else has failed, like an output diode has shorted or something. you can check for such things but you'll only have a hope of success if you understand practical isolated switching converters (or at least flybacks) first. that's a lot of effort so if you don't want to do that then just trash the adapter.

>>1371777
I can't figure out how to write the inputs with only 3 variables down from 4 variables. If it helps anybody who could help me, C is the input that reverses the counting direction. I think my issue might be figuring out the last state of the JK flip flop

>>1371679
What the hell is it doing?

>>1371574
is bigclive gay?

>>1371910
YEs.

>>1371876

they say caps only work on AC but why is it i see the RC guys using them in the battery side when that part is DC. they do this to reduce voltage drop/ripple so what is the right answer then? does caps work on DC circuits and can you use caps every 3 meters on a 100m cable run to stop voltage drop? as i see it volts get passed from battery to C1 then charges C2 then passes to C3 etc all way to end of the 100m cable run. should no volt drop then... right?

i wish i was smarter :@

no i haven't watched that much caps tutorials sorry for sounding like an idiot.

>>1371938
forget whatever you learned about capacitors and relearn from a textbook, not some arduino or rc forum tards.

>>1371938
Calculus time:
For a capacitor, I = C*dV/dt. That is to say, if voltage across it remains constant, there is no current, and if the voltage across it changes, the current is proportional to the rate of change of voltage.

In AC, as V = A*sin(w*t), dV/dt = w*A*cos(wt), so the current is I = C*w*A*cos(w*t). Using the RMS average of these, V_RMS = A*√2, and I_RMS = C*w*A*√2. A is voltage amplitude and w is angular frequency) In other words, at higher frequencies, the same capacitor will conduct better, and increasing the capacitance has the same effect.

In DC, we need to take account of the resistance in series with the capacitor. In the simplest instance, we have the resistor from the voltage source to the node we measure the voltage of, and from there the capacitor to ground. When we switch the source on, the voltage jumps to it's maximum effectively instantly. Now the current through the capacitor is once again I = C*dV/dt, and since this same current is going though the resistor, the voltage across this resistor must be R*I = R*C*dV/dt. Hence the voltage across the capacitor is 1 - R*C*dV/dt, which is a differential equation. The solution to this is V = A - A*exp(-t/(R*C)). In "RC circuits" like relaxation oscillators, the known resistance and capacitance are used to generate a consistent delay or time period. In power supplies, the effective resistance is very low but unidirectional thanks to a diode, so the capacitor will charge up very quickly when the voltage gets boosted to it's maximum, but when the voltage before the diode drops back down again (like with a simple rectified AC power supply) the capacitor remains charged, and can only be discharged via the load. So if the load only pulls 10mA and this will barely drain the capacitor in the time it takes for another AC cycle to charge it up again, the voltage at the output of the capacitor will barely ripple.

>>1371960
Cont.
For DC power supplies we can make an approximation for their ripple voltage. Since at t = 0 the rate of change of voltage dV/dt = I/C, we can say that over the time it will be discharged, or 1/f where f is the AC frequency before the rectifier, dV/dt * t = V_ripple. Hence, V_ripple = I/(f*C), so the higher your frequency is the lower your ripple voltage, one of the reasons why switched-mode power supplies use kHz frequencies as opposed to 60Hz. Note that using a full-bridge rectifier will half that ripple voltage again, but it does cause another diode drop.

Inductors are described as V = L*dI/dt, so there will only be a voltage across it if the current within it changes. In this way, the current through the inductor in AC decreases with increasing inductance and/or frequency.

Study some network theory, it's good shit.

>>1371938
The caps used for the DC side is still kind of working on AC, the AC component of the current draw is supplied by the capacitor. Caps don't stop DC voltage drop along the cable, but can be used to reduce AC voltage drop = ripple.

Hey /diy/, I have a strange micro issue. I got a Teensy 2++ for a project and it's acting very weird. Right at the start, when I first plugged it in, it had no code preloaded even though the others had the standard blink. The flasher detected it though and I was able to flash the code myself. It went good but after a few seconds I got a "USB device drawing too much power" error and the board went dark. When I try to reconnect it, I get the same error on both PCs. dmesg shows it as "overcurrent detected with port status 0x4000". Plugging it into a powered hub doesn't help. When I measure voltage on the +5V pin, I get only 0.5V. I am completely stumped, what might be causing this?

I know how to code and wire things. What type of hardware do I need to make electronic door locks?

>>1371750
The problem is that not all BJTs have the same pinout.

>>1372010
I think for the most part all the common small-signal BJTs are (2222, 547, etc.), but it's still a pain. I hope their MOSFETs don't do the same thing because they certainly vary a lot. But formal circuit diagrams are much better.

>>1371841
AAAAAAAAAAAAHHHHHHHHHHH
GOD DAMN IT I AM A FUCKING IDIOT HOW COULD I FORGET
I know when the old blown cap was in I couldn't read the capacitance on it because it was obscured by other components. I checked, that means the white stripe would have been in the other way around. God fucking damn.

>>1371866
I thought it might have been coil whine from the bad cap, there's an inductor right next to it. It's the switching IC?

How exactly does a diode matrix work and how do you decide where to connect diodes and not? Been looking around but little info to be found

>>1372035
Looks like those are diode-based NOR gates, at least if you treat the switches like NPNs to ground, i.e. closed = logical high. Each of the 4 outputs is NOT(A+B+C) where A,B,C are any number of switches 0 to 9. What's this for?

>>1372037
>What's this for?
Picture was just taken from google images, not mine. I was wondering how diode matrices work, or perhaps some recommended books that explain it.

>>1372029
if you put the cap in backwards then that's that. i'm just pointing out that it's standard operation for the converter to run at low audible frequencies
during no load or short conditions, so your blown cap may only be symptomatic of another failed component. i'd definitely find and check the output diode with your meter because if that's shorted it'll both kill your cap and give you the random output voltages you say you have.

>>1372035
That particular one, where only one of the switches is closed at a time (or the input is low) is quite easy to understand. Note also that this does not work "properly" if you close more than one switch. Similarly, no switch closed might count as an invalid state as well.
All the 4 outputs are pulled high via the 4 resistors - unless they're pulled low via diodes. So, if the switch "0" is closed, all the 4 outputs are pulled low and the output state is 0000. If switch "1" is closed, then the 3 leftmost outputs are pulled low, but the remaining one is high, so the output state is 0001. And so on.
If you wanted to add 11th input/switch to it, you could implement it by first converting "10" to binary (since this is a decimal to binary converter), which is 1010. Thus, you'd need two diodes, first for the 2^0 output and the second for the 2^2 output.

If you want to do the inverse of that matrix (binary to decimal), then you have the problem that more than one of the inputs can be low. To handle that, you'll need both the input signals and their inverted versions. Then for every (interesting) input combination, you'll find the combination of the inverted and non-inverted inputs where they are all high and combine them in the similar manner as in your pic. That is, when you have the correct input combo, the corresponding decoded output is NOT pulled low.

>>1372021
kek, westfags
otoh I can't think of any SOT-23 transistors that use any but the standard pinout

>>1372005
an electrical door lock, a logic-level-controlled relay, and a microprocessor with at least one GPIO

>>1371989
sounds like the Teensy is bad, look for obvious soldering/component errors and/or contact the seller

>>1372053
Yes, seems like the chip went bad. I found that +5V has continuity to gnd but the cause is not apparent form the outside. Luckily it didn't kill any of the usb ports. But it's still weird that I was able to flash it the first time without any problems.

Camerafag here, the 556 circuit is sitting in a pile of dysfunctional wires waiting to be re-breadboarded, but that's not what I'm talking about here. I went through the definitions of F-stop and ISO and found myself a general equation for a well exposed shot:
>I = F^2/(ISO * T)
Where I is the effective relative intensity, F is the F-stop which is proportional to the inverse of aperture diameter, ISO is the sensor sensitivity, and T is exposure time. It makes sense because rearranging it 1/ISO = I/F^2*T since 1/F^2 is a unit of area. So I guess ISO has units of inverse joules, or perhaps volts/joule or volts/photon, but that's also not particularly relevant.

What I want to make is some small light sensor that gives me a reading (checking with a multimeter is fine) that I can convert into an ISO, exposure time, and F-number to tell me how to get a well exposed shot, primarily for multi-hour exposure star photography.

So I calculated this intensity value for a series of well exposed shots under different lighting conditions, where I also taped an LDR to the top of the camera and measured its resistance. For relatively normal lighting conditions (curtain up to curtain down, camera pointed at opposite wall) I got a pretty good relationship between LDR resistance and intensity. Pic related. However once I darken the room further (inside at night time with a small dimmable LED), the resistance went up past 60MΩ (the maximum my meter can read) and my camera started to tell me it was too dark for its maximum non-bulb exposure time coupled with the largest aperture size. I think I can compensate for the latter problem by cranking up the ISO, but the LDR is still a problem. Anybody else tried detecting very low light levels? I can't seem to find much documentation, besides advertisements for $100 avalanche photodiodes.

>>1372068
you can use a photodiode with an op amp gain circuit. their spectral response doesn't line up with that of the eye though so you may want to browse the options for digital light sensors and see if any of them go low enough for you. they're all surface mount so your selection is limited to the ones you can get on breakout boards.

>>1372073
Well I was figuring that since I'd mainly be using this for the stars, I'd calibrate it for that, and it would't matter if it were a visible photodiode or an infrared one (which it is). Are they better than phototransistors for low-light? I'll have a look on Digikry or Mouser to see what kind of sensitivities their sensors have. One digital sensor (OPT3001) goes down to 0.01lux as it's smallest division, which is 10 times lower than an LDR datasheet states as giving its "dark" resistance, and it uses a photodiode, so I guess that's what I'll try next. I've not had any luck with the phototransistors I bought anyway.

>>1372076
Never seen anyone using phototransistors for anything requiring sensitivity or accuracy.
0.01lx isn't particularly challenging with photodiodes, particularly if you can use a diode with large area, but as you're measuring nanoamp scale currents, some care, better than shit opamps and huge resistors (1G for example) are needed.

>>1372068
ISO film speeds are arbitrary units, but indeed not relevant.
>cranking up the ISO
>on star photography
wew lad
noise is going to wash everything right out. if anything, set the speed as low as possible, take multiple shots, and add each pixel together then normaliz in your post processing

>>1372068
>multi-hour exposure..
If you really want to photograph stars the $2.17 Shutter-Release-Cable is all you need. Maybe you never have and you never will.
https://www.google.com/search?q=photographing+stars

Some time ago I was asking /ohm/ about the feasibility of using a chinkshit SCR dimmer to control the power in a DIY welder (the three MOTs on full power were way too much.)
Thanks for the answers - turns out it actually works!

>>1372191
that's cool but where's the forklift you use to move three mots at once?

>>1372193
It has wheels, senpai, you pull it like a toy duck.

>>1372191
>full retard setting
you done good, son

>>1372053
Can you give an example of an electrical door lock?

>>1372224
Pic related
there is also a type for higher security installations with two flat magnets that attract one another, and an electromagnet that counteracts the attraction
plenty more on google images, happy searching

>>1372237
What is the "official name" for this?

>>1372253
previous pic is called something like a striker latch
Pic related is called an "electromagnetic door lock"

can someone help me figure why my flipflop isn't making the LEDs flash? They're not lighting up at all.

>>1372383
here's the schematic

So i have a water pump driven by a brushless motor I am trying to drive. im trying to control with a transistor and PWM. Works FINE when I give it the full 12 V. Moment I change it to an analogwrite, doesn't work. I'm still measuring the exact voltage drop across the motor. But I've verified this fucker works with less voltage.

I don't know much about pwm, so what could the problem be?

>>1372383
Where the w is, there is no continuity between the rails either side. Your power rails only go half the length of the board.
Depends on the board of course but these are wish? Because of the w? Pretty sure they are split. So check that first

>>1372415
If its dc brushless you know its not just a dumb motor, there's electronic circuitry inside there to sequence the coils. So its probably not happy about lower voltage.
Yes some pc fans work at lower voltage so its possible but depends. Also how much lower are you going, try hitting it full speed to get started then slow down gradually to overcome stall, getting started is harder than running. You are using a transistor or something right to drive 12v through it because I doubt a microcontroller can output enough current. Actually that's a common problem so watch out, current limit doesn't mean it will stop at the upper limit, it will try to do more but kill itself in the process. The chip that is.

>>1372419
Not much, tested it with a 9V battery and worked fine. Just using a digital pin to operate transistor. Rated for enough current and voltage. Then just have an emitter voltage of 12 V on this NPN. Voltage across the motor reads 9 V too, so I'm wondering if it's struggling to source enough current?

If I read this books and, of course, practice a lot, will I be able to consider myself an expert?

>>1372097
>1GΩ
Oh no.

>>1372113
The camera has the exposure slider on its front screen, and it only goes up to 12 units over/under exposed. With the ISO set to minimum 100, exposure time set to 30s, and F-stop set to 5.6, it will only give me an estimated exposure value if I'm not 12 units too dark. Switching to bulb mode cannot give an estimated exposure value since the exposure time is arbitrary. So by cranking up the ISO to 3200 and keeping it at 30s exposure, I can get an exposure estimate for a 960s exposure image at an ISO of 100. I'm not going to take photos at such high ISO values or at such low F-stops, but I'll be switching to them to take data points that I can convert into greater than 30s exposure times for lower ISO values and also higher F-stops.

>>1372123
I bought one already, and took it apart to make a remote and/or delayed shutter. I'll have a shutter release toggle on it also.

Do these work well?
https://www.ebay.com/itm/Solder-Desoldering-Pump-Sucker-IC-SMD-Vacuum-Sucker-Suction-Pen-Remover-Tools/281995610609?hash=item41a83f89f1:g:5OwAAOSw3v5YtP~c

>>1372416
holy fuck, I didn't even consider this and now it's working perfectly

I spent 3 cumulative hours trying to troubleshoot this fucker, I can't believe it was so simple

>>1372470
not really

Has anyone here fixed and AOC monitor before?. Im triying to fix my LE19A1333, according to schematics i should have 12V in pins 3 and 4 in the upper right conector but im getting only 0.3V, all the other sources are fine

>>1372481
Then look at the PSU board instead of the control board and make sure that the 12V supply is working. If it does, trace from one to the other. If it doesn't troubleshoot the PSU instead.

>>1372481
>>1372482
There is nothing in the 12Vpins of the PSU board, i think its get 12V from 16V and R737.
I have 16V in the 16V side but nothing in the 12V side. Maybe the resistor is kill

>>1372487
>i think its get 12V from 16V and R737
I very much doubt this. Is that diagram supposed to be the of power supply board? Because I can't see anything resembling a mains input, nor does it contain anything akin to the control board. Or does it run off a power brick?

>>1372497
Its one part of the main board, the full schematic is pretty big so the manual separates it

Any recommendation for a compact 12V solution?
Or perhaps an alternative idea one what I'm trying to do that takes less volts.

I'm trying illuminate a small picture frame. My idea was to place some LEDs at the bottom of the glass and power it with some small button cells with a tiny switch.

Catch is button cells are 3V so a pair are 6V with a nice low profile premade deal that tucks into the frame well, but won't power strip LEDS (the type you can cut that come on a reel, those are 12v). The typical string of LEDs that do run on those 6V tiny battery packs you can get from craft stores stores, the so called fairy lights, don't seem bright enough and come in a messy strand of a ton of lights.
tl;dr Is there a good way to power small LED strip with 12V to hide in a photo frame, or a better LED type that takes less power?

>>1372531
Really lazy solution would be to use two CR-P2 camera batteries (6V) in series on the back of the photo frame if nothings behind it.

Otherwise you could make a charge pump or boost converter circuit if you still wanted to use the 12V LEDs strips and run them off some AA batteries

>>1372548
10Ah Li+ power bank with 12V boost module. it'll stay lit forevar
maybe you might recover some LiPos from discarded tablets or something. attach a protection/charging board and a boost module and it'll stay lit for quite a while.
be advised, however, that the LED strips aren't very efficient. if you knew how many lights you needed, you could use a current-limited boost module to drive the LEDs in a series string directly. there are a shit-ton of ICs out there for driving strings of white LEDs thanks to LCD televisions and similar applications

>>1372548
>>1372553
Sorry I'm getting a crash coarse here. I've not really done this before and only understand some of that.

I have to make several of these and they need to look decent too. It wouldn't be many LEDs, 10 at most probably 4 at the least, 6 fits well with the current idea but I'm trying to be flexible on the project. Just lighting up some small glass frames that have etched designs on them which should capture some of the light to look neat.

I could use something other than the LED strips if anyone has an idea, I just thought those would be compact and mostly premade solution for a beginner.

All I really want to do is light up some glass, possible better way?

>>1372555
Post a picture of one of the frames if you have them handy, might give people a better idea of what you're working with

>>1372555
Also, maybe if you want something pre-made something like "2pcs DC DC Converter 5V to 12V USB to xh2.54 Step Up Boost Module" off Aliexpress

Just plug that into one of those 5V USB rechargeable batteries and you pretty much have a 12V power cell.

>>1372531
Well firstly I'd veer away from coin cells unless you want it to be particularly compact, but if you're not pulling much power from them there isn't really any downside. Now the only reason LED strips are 12V is because they've got 3 LEDs in series with a current-limiting resistor, so if you get individual LEDs you can just throw them behind a tiny resistor and a single coin cell. You could easily use no resistor and just put all the LEDs in parallel with the coin cell, but they may not last long without current regulation.

To see if coin cells are the way to go, take the cell's Ah rating (mAh rating/1000), multiply it by its voltage (probably 3V), multiply that by the number of coin cells you're using per frame and divide by the power that the LED array on one frame will use. The result is how many hours it will last.
>Q*V*n/P = t

>>1372556
This is one of the larger ones.

>>1372558
That is a possibility I suppose. I don't mind doing some work either it is DIY after all. I'm just trying to keep it cheap compact and fairly easy is all. And while I love DIY there isn't much point if there is something easy and already existing. Most of my work is really the etching it will light up.

>CLRW sets Z
ugh

>>1372470
Desoldering pumps are okay for removing large amounts of solder from large joints like through hole connectors.
There practically useless for cleaning solder out of a through hole so you can put a new component in. Desoldering braid is the way to go then.

I'm not sure what those chinks are selling on amazon there, it looks more like a small suction pump for holding QFPs and the like, which can be handy if you're dealing with chips that have fragile legs.

do any of these books have projects in them? the one book i have doesnt have any projects

>>1372469
Well scratch that. I've got the ISO at 3200, the F-stop at 5.6, and it's telling me that 0.1s is a good exposure at this light level, but turning down the pot any further on the LED gives me a "Subject is too dark" alert and nothing else. Putting the ISO back to 100 doesn't even give me 30s+ as my exposure but 2.5s, and it hits the "Subject is too dark" alert at the same light level regardless of ISO. Looks like I'll be stuck trying to extrapolate the sensor-voltage/camera intensity graph to very low light levels, which should be a relatively good estimate.

Thinking about it some, since the illuminance of the night sky will not change much, and since it's mostly black, I won't get much of a result by setting it up for a normal exposure. Trial and error should be more than enough for just this single illuminance, so I'll just solder the toggle switch on to the cable now and start testing. And it's raining.

I'm doing this on a couple of inputs on a mirocontroller.
But it floats like crazy. What am I doing wrong?

>>1372663
The LED acts like an open circuit as LEDs still obey the diode equation. It will only close the circuit when there is enough voltage, basically when you have a logic output of 1, at logic output of 0 you have a resistor floating in the air.

You could put a large, actual pull-down resistor in parallel with your current stuff.

>>1372668
In the datasheet it says the MCP6544 has a Push-Pull Output.
The LED is only there as indicator.

>>1372673
Hmm, sorry my mistake, I haven't read the datasheet.

Have you tested stuff on the MCU side of things?
I would test both the output of this circuit and the input of the MCU separately from each other if you haven't already.

>>1372677
I don't have a scope, so here is what I did:
>Remove MCU
>attatch pin change interrupt to pin of known good MCU
>Make a led on an output pin light up for 200ms whenever the Interrupt is called
>test on 5V led keeps off
>test on GND led keeps off
>float pin led flickers
>attatch to output of curcuit led flickers.
the in led in the curcuit behaves like it should by the way, but I guess the eye is to slow to catch it when it goes on and off quick

>>1372654
you don't need gigaohm resistors necessarily. just a dual picoamp bias op amp and some 0.1% resistors. you can make something similar to pic related, through hole, for under $5. i haven't done any math on the values but it's 1V/nA.

I just fried my arduino, probably because of a short circuit between VIN (+12V) and a digital pin (12).
What died? The board or the chip? The chip get very hot.

>>1372474
From experience I can tell you that no matter how much you think you will never make this mistake again you will absolutely waste time on this at some point, on the plus side it will be immediately obvious to you when looking at other peoples project. Yes, like a curse.

>>1372424
Try it with a capacitor, to smooth out the pwm into a constant voltage. The brushless driver is probably just pissed off at you not giving it a stable voltage.
Try it with the battery but switch it on and off a thousand times a second see if it works then lol.

>>1372642
>desoldering pump
oh the keks I am having

>>1372708
>D12
if you hook up USB and see that the voltage on VCC is still correct and the USB-serial device is still recognized, it's probably just the chip

>>1372710
this is why I love /ohm/

>>1372191
LOL Lamba core

nice.

how is polyphony achieved in synthesizers? All the basic VCO circuits I've seen and built are monophonic only.

>>1372815
I'm assuming it's done by having an individual oscillator for every note, right? Or is there a more efficient/elegant way of going about it?

>>1372815
multiple VCO/VCF/VCA sets, yes

i am completely unfamiliar with how to hook up 12v generators and electromechanical regulators to a battery. pls halp.

Does anybody here use a CNC router to carve their PCB traces and drill holes? I've been thinking of getting/making a CNC router and it's one more thing that I'd want to use it for.

I should read these books in this order?

>>1373116
From beginner to intermediate to advanced? I'd wager you only really need one of each category, but I've not read any of them. I get my electronics know-how from trial-and-error and google.

>>1373123
I downloaded the first 2 and read the summary and some pages. The first one don't seem to have projects, and the second one is only projects. Well, I'll just read and see what can be skipped.

>>1373076
Got one at work. Never used it myself, but had a couple of boards made.
Seems pretty fiddly to get to work propperly. And you still have no soldermask and no plated holes. At least for my usecase a good perfboard is more fit. That doesn't mean there is no use for it though.

>>1373228
Looks like you can buy plated inserts/rivets, and it should be trivial to make solid vias with some scraps of solid core wire, but in both cases you'd either need to leave the mask off, or put mask on after soldering them to each side. I'm not planning on using it for tiny SMD stuff anyways so if I do decide to use solder mask I figure I can do so by hand. Not like I plan on mass-producing anything anyways.

Should I connect these?

>>1373331
whoa, line and neutral produce multiple DC rails - that's awesome

>>1373332
I'm pretty sure those transitions are powersupplies

>>1373338
i really should be more careful with dickhead comments lol

>>1373331
no. a benefit of the power supplies you're using is that they're most likely isolated. it means that instead of grabbing one wire you now have to grab two in order to create a current path. your voltages aren't a risk to you, but it does make it harder to destroy your project inadvertently.

it also means you can't explode your oscilloscope if you have one.

My first ever solder joints. Any feedback?

>>1373343
>Any feedback?
When you want to take a pic to show something to people, make sure it's actually in focus.

>>1373331
If you're using a power supply from a PC (because 12V and 5V) the black (negative) conductors are already grounded.

>>1373346
Sorry, all I have is my shitty phone that can't do proper close-ups. Here are three more attempts (of the finished product), as I tried to make it clearer in a couple of different lights and angles

>>1373363
2

>>1373365
3/3

>>1373343

looks like not enough solder. you want a good looking fillet, 3/4 of those look starved.
but they look blurry like you have poor focus or a shitty ancient phone camera with the lens covered in vaseline.

also it's a good idea to reflow solder after you clip the leads. (looking at the dangling leads above) this is because clipping leads AFTER soldering them can potentially weaken or crack them, and it's also better to not have the bare (clipped) metal exposed.

some soldering specs say to bend and clip the leads before soldering. good to learn this early on.

in the other pics you posted it looks like you have some stray solder junk on the board. don't do that

some of those look okay, some look like too much solder, some don't look shiny enough (incorrect temperature)

keep at it anon

>>1373371
> looks like not enough solder. you want a good looking fillet, 3/4 of those look starved.
Yeah, I agree about that for the corner switch components. They have narrow legs and the gaps were wide, so the solder flowed through the hole.

> also it's a good idea to reflow solder after you clip the leads. (looking at the dangling leads above) this is because clipping leads AFTER soldering them can potentially weaken or crack them, and it's also better to not have the bare (clipped) metal exposed.
I've actually read about people saying that reflowing the solder like that can hide a bad joint or overheat things, while the risk of fracture is negligible. Were they full of shit, or is it a matter of opinion and personal experience?
Also, what's wrong with having the clipped lead exposed?

> in the other pics you posted it looks like you have some stray solder junk on the board. don't do that
There's a bunch of flux in some places, which I'm pretty sure is no-clean. I think I do have little bits of actual solder, though, and yeah that's my bad. Main thing I've learned here is not to hold off on clipping annoying leads (which I did because I forgot to find the wire cutters before starting).

> some of those look okay, some look like too much solder, some don't look shiny enough (incorrect temperature)
Here's a pic after clipping. This might be clearer? To me it looks like a whole bunch actually have too much solder, is that right?
>incorrect temperature
I did everything between 347-352°C (if my iron is to be believed), but I think I fucked some things again due to having unclipped leads.

Same anon learning to solder here. One led wasn't working, so I spent maybe a couple of minutes trying to desolder it. I was pretty sure I ruined it by the point I gave up, tested the board again, and lo and behold it works now. What the fuck

>>1373381
>Also, what's wrong with having the clipped lead exposed?
bare copper oxidation (is this that big a deal? maybe not, but i have seen untinned and clipped component leads look like absolute garbage just from corrosion/finger oils/moisture/whatever.)
I've also had intermittent issues due to what was possibly a microcrack, and reflowing fixed the problem.

http://www.ipc.org/committee/drafts/5-22a_d_J-STD-001F-Feb2012.pdf
>When lead cutting is performed after soldering, the solder terminations shall [N1D2D3] either be reflowed or visually inspected at 10X to ensure that the original solder connection has not been damaged (e.g., fractured) or deformed. Lead trimming after soldering that cuts into solder fillets shall [N1N2D3] be reflowed (Figure 6-2). If the solder connection is reflowed this is considered part of the soldering process and not rework. This requirement does not apply to components that are designed such that a portion of the lead is intended to be removed after soldering (e.g., break-away tie bars).

just sayin'

>>1372474
I've had this happen to me too. These cheap white, Chink boards are shit for that. There are some other Chink ones that are fine, but these really White ones are limiting in a way.

>>1373392
Alright, I'll definitely keep that in mind.

Thanks for the advice and feedback anon!

>writing a pic i2c master to control a 36ch led driver
>check scl/sda with scope
>all bytes getting sent with acks
>no leds light up
i swear to god

>>1373428
nevermind it lives.

is soldering wick a pain in the ass to use or do I just suck

>>1373446
my chinese shit is definitely a pain in the ass. there's no point even trying to use it without separate flux. it's best if you can run your flux pen directly on the length of wick you'll be using.

>>1373448
How have I never thought to put flux on the wick holy shit

>>1373452
Because it's supposed to be there to begin with?
But yeah, if you buy bottom of the barrel wick, you usually have to add flux. It can be kinda shitty even with the extra flux, if it's made of too thick wire.

>>1373441
post full porns

>>1373446
wicking and sucking each have their purposes

>>1373569
here you go. it's meant to change colors (modes) on encoder pushbutton and inc/decrement from 0 to 255 on encoder turn while sweeping the leds to provide visual feedback on the value. my algorithm didn't work though so i just overrode it and turned all the leds on at 20%. i intend to use it for volume, mouse sensitivity, and monitor brightness controls when i'm done writing the i2c code and get around to finding some usb code to steal.

https://pastebin.com/raw/D0rx09XH

>>1373577
FFC for ICSP? fascinating
cool though. maybe this is gankable http://wordpress.codewrite.co.uk/pic/2013/07/26/usb-interface-using-pic16f1455-2/

Is is better to put some external pull up resistors on I2C lines or just wire them in directly?

>>1373595
not clear what you mean but there's never a case where you won't need those resistors, so you'd might as well integrate them into any board you're doing. don't use the weak pullups in your uc unless it has an unusually high pullup current (~250uA at the very least)

>>1373599
Thanks man. Is this just a more general rule of good practice with all comms stuff?

E.g. If I use a UART, should I be doing the same thing? (I mean, pull or pull down, depending on the protocol)

What about with GPIO ports? Say if I connect one uC to another and have a GPIO port signal "busy", should I also be adding pull ups and pull downs? I don't really know "good practice" to do with hardware at all.

>>1373627
the advice applies specifically to I2C where the open-collector wired-OR action is integral to the protocol and pullups are part of the spec
>If I use a UART, should I be doing the same thing
not unless you plan on the UART port being electrically disconnected during use, but even then, the internal pullups should be enough
>Say if I connect one uC to another and have a GPIO port signal "busy",
unless your BUSY signal has open-collector outputs, you won't need to pull them up, and even then, the internal pullups should be enough

>>1373577
choice of driver & led vs e.g. ws2811 was because?
what software for layout? how did you position leds?

>>1373637
packaging was a constraint. the board is sized to a knob that i found (1.75"/45mm) so if I was to use multiple 8ch or smaller drivers I wasn't confident that I would be able to fit them on a board. i've had trouble with qfns so I chose the only qfp on the market that does 36 channels.

i used altium to route, with excel using trig formulas to compute the angles and positions.

ravioli is still good surprisingly. i've been top damage on t10 games all day.

>>1371574
How do I determine how much power this light needs?

>>1373707
This is the only writing on the whole light.

>>1373707
It vaguely looks like there's 4 leds in series and 18 pairs in parallel. i don't think those are COBs so you can assume a 2-3v drop. you can easily read the value off those surface mount resistors (eg 3300 = 330 ohms + (0) zeroes added on. so if those resistors really were 330 ohms, you could start with a safe current of 20mA giving you a supply voltage of 0.02*330+4*3=~18V. that's how i would do it with no other info. if you have a power supply of any known voltage and a multimeter you can calculate the forward voltage of the leds.

>>1373707
>>1373708

Give it about 1.5 amps at 37 volts to start off

>>1373710
You are correct about the number of LEDs.The resistors say 1800 on them. The only spare power supply I can find right now is 12V 3A. My multimeter is MIA but I am looking for it.

>>1373718
37 volts sounds like way too much for a small aquarium LED.

How often should I tin my iron? After every X joints? Every X seconds/minutes? Some other metric?
What about when I put my iron down for a few minutes while fiddling with components - tin before putting down, tin after (once the "few minutes" have passed and I pick it back ip), or doesn't matter?

>>1373719
i'd suggest you give your 12v supply a shot. (12-4*2)/180=22mA at the absolute worst fathomable case. realistically it'll be way less because 2vf is low. so you can see how dim it is at 12v and decide if that's suitable enough for your needs. if not then you can hunt down that multimeter.

my garage is lit by 1 shitty light bulb. its a 20x20 garage, so pretty small but still the lighting sucks. i was thinking about using one of these bulb to socket adapters and running some led shop lights. the shop lights take 48 watts each and i wanted to run 4, but i suspect 2 would suffice. i dont know the rating of the bulb socket, but id suspect its just a standard 65 watt one. how dangerous is it to overload one of these sockets by so much? should i just remove the socket and try to wire in an outlet instead? its a brand spankin new house so id rather not cut any new holes if possible.

>>1373644
QFNs don't suck. QFN die paddles suck, at least if you mean to hand-solder them.

>>1373803
>remove the socket and try to wire in an outlet instead
this

>>1373803
my dad this exactly this. it looks pretty bad desu senpai. hopefully the box underneath the lamp mount isnt round, though. if it is, throw the fixture back up and use that ugly ass adapter.

>>1373803
also, replace the bulb with an LED one and you have nothing to worry about, amp-draw wise.

Anyone know of a good small button or switch for projects?

Everything I'm finding that is tiny is momentary and I need something for turning things on and off.

>>1373829
A PCB-mount toggle switch? A computer on/off button is another way to go.

>>1373873
Just for some LEDs but it has to be tiny.

>>1373879
Then how about a DIP switch?

>>1373879
the latching mechanism takes up quite a bit of space and volume.
Pic related can be found on aliexpress
alternative: if you can afford space for a couple of SOT23 packages and a touch pad, consider a TTP223 and a MOSFET

>>1373882
>>1373888
Found some little slide switches that may work. Just took forever because I wasn't sure what to use as a search term.

Is it alright to switch out a LM393NG for a LM393ADR, the ADR performs better right?

I don't understand why my Hartley oscillator won't oscillate.

>>1374058
Don't need a base resistor? R7 is doing nothing there since the gate is already at 0V.

>>1374058
because you trolled it

>>1374058
It is hard to understand what you were trying to do.
There is no signal connection between these three parts.

>>1374078
2 and 3 are the inductors. I simply made the inductors out of op-amps instead of actual inductors.

https://en.wikipedia.org/wiki/Gyrator#Application:_a_simulated_inductor

>>1374081
That's nice, but look at what you have done:
C3 is short.
B-C of Q1 is short.
"inductors" are short as well.

Try it like this.

>>1374084
Still not oscillating but I have a -500mV DC offset now so... that's something.

>>1374085
Play with values.

>>1374105

>>1374105
>>1374106
It turns out after dealing with that short the reason why I wasn't getting a signal is because I forgot to change the initial conditions for the sim from "determine automagically" to "set to zero." It started oscillating no problem with the initial values afterward. I was hoping it'd be a bit more sinusoidal though.

>>1374109
forgot pic

>>1374109
>"determine automagically"

Cringiest term ever invented. If you are in electronics or computers and you think that's a cool term, I'd rethink my career goals. Maybe designer coffee or something like that.

>>1374111
I do like it actually. I think you're just being an autist about a term that was obviously meant to be a bit jokey. Of course there's no magic under the hood but to the layman it would seem like it.

>>1374106
Remove C4

Well fugg. Gotta abandon the gyrator Hartley osc. Once side of the gyrator is always grounded and with two of them the center tap is always at ground. Doesn't work for electron coupled oscillators where the center tap needs to be at the cathode voltage. Could probably do it with four op-amps (2 makes 1 inductor floating from terminal to terminal) but fuck that. It's a pain to make coils for low frequency oscillators but I probably have some mu-metal cores somewhere so I can minimize the number of windings. Well thanks anyway /diy/, we tried.

>>1374189
nonsense

How do I choose a hydroponics "atomiser"?
Sheer power? Geometry of the receptacle?

>>1374192
Here's an easy to use diagram explaining why you're wrong.

>>1373728
Nobody? Is tinning the iron a meme?

If I buy a pi cam from aliexpress what is the likelihood of me getting shafted?
Like what am I trading off here, picture quality?

Ive been using this solder for years now. It leaves a lot of flux and its hard to get off, isopropyl just wont touch it.

What brand/type you guys recommend is the best?

>>1374270
Check feedbacks to see opinions.

>what am I trading off here
I can only think of dead pixels, but its unlikely.

Im in urgent need for a decent bright LED solution for outdoor installation, harsh wind and rain
im looking at 100-200W Light bulb replacement strength, directional to at least the sites and color RED
Any ideas?
also, is there any LED hoose brighter than
>>https://www.amazon.de/Neon-Lichtschlauch-L%C3%A4nge-au%C3%9Fen-Gartenpirat%C2%AE/dp/B0763M7LRH/
with 100 LED per meter?
pls help out a dumb frogposter

>>1373363
shine a flashlight on the board and take a photo through the magnifying glass of your 'helping hands'

>>1373803
there are led 'corncob' bulbs with an ordinary e27 edison screw, meant for streetlight retrofits. 60 watts of led light in a 20x20 room would be fookin' bright.

Sorry for the noob question, I'm trying to learn a bit more about the esp8266, what is the difference (if any) between these two?

https://www.aliexpress.com/item/ESP8266-CH340G-CH340-G-NodeMcu-V3-Lua-Wireless-WIFI-Module-Connector-Development-Board-Based-ESP-12E/32800966224.html
https://www.aliexpress.com/item/NodeMCU-WIFI-module-integration-of-ESP8266-extra-memory-32M-flash-USB-serial-CH340G/32739832131.html

>>1374228
Well I tin the iron while soldering, but that's just how I solder.

>>1374209
First you trolled your circuit and now you're trolling yourself by ignoring the fact that you have two independent and _grounded_ inductors that are _not_ a single L but part of a network.

A PI network (4-pole, not 2-pole) can be either CLC ('Colpitts') or LCL ('Hartley'). Now you trolled away the important capacitor, C3 in your earlier circuits, so as not to see the network.

For equal L's (your case) the required loop gain only needs to cover the (small) losses. Yours is insanely large and the result resembles a blocking oscillator.

Yes, it oscillates. What does that tell you? I'm in the timezone, good night.

Could hotsnot cause problems with 230VAC lines?
I have to fix some wires but i cant avoid hotsnot getting on the pcb pads with mains.

I'm an electronics brainlet.
Apparently you can clip a signal by placing two zener diodes in series, but from this image it appears to apply to sine waves with two outputs.
I want to put them on the audio out of... whatever, to see what it sounds like. I don't think I'm supposed to bridge it from positive to earth, and the only other way I can think to wire it up is like in the pic I'll post in the next post.

how retarded am I

>>1374558

>>1371679
electricity flows from the negative terminal to the positive. probably not the problem.

let's say the he push-button has the following terminal layout:
12
34

1&3 and 2&4 will connect when the push button is depressed. you have 3 & 2 wired. meaning nothing will pass.

idk what the pin output for the transistor is soooooo .

>>1374424
the antenna on the second one doesn't particularly impress me

>>1374228
when it looks dry, and before I put it in the holder

>>1374558
across
beware that zeners have power ratings. you probably don't want to do this with a speaker amplifier of any significant size. neither the amp nor the zeners will have a good time of it
some fuzz pedals use two regular diodes in parallel at approximately line-level, as in Pic related. the graph shows a rough sketch of the transfer function but the knees will actually be sharper. add the resistor crudely drawn-in to protect the signal source, 100 ohms is usually plenty

does this stupid website work yet? Wtf is hirohito doing

https://www.bunniestudios.com/blog/?p=5259
PAPER & TAPE DETERMINED SUPERIOR PROTOTYPING MATERIALS
SOLDERLESS BREADBOARDS & THT ON SUICIDE WATCH

>>1374721
no

Hello fellow /ohm/ies. I am a noob who is planning to do a DIY portable speaker for bicycle riding at night. Initially I planned to use some chinese components from Aliexpress, Ebay etc. but after searching through local auctions and flea markets I came with a better (imho) solution. 2x20W full range car speakers (supposedly from bmw e39), TDA7297-based diy amplifier, 12v battery pack with booster and some bluetooth module (or a used Sony Walkman e053) should do the trick, in my opinion. However, the main problem for me is the enclosure. I thought about using plywood but it is way too heavy (though the sound quality is nice).
Again, I am an almost complete noob in audio-related projects so my question is: how much would I suck if I used 4 mm thick PVC pipe with foam layer as an enclosure? I could do ported design for each speaker and use a divider inside. My goal is quality, not excessive car bass. Also, any ideas about budget waterproofing?
Pics related.

>>1374767
Here are additional pics

>>1374767
Here are additional pics2

>>1374767
Here are additional pics3

>>1374767
Here are additional pics4

>>1374771
>>1374772
Ports will be properly closed and vented outside, of course, this design is crude, just to measure the dimensions and such.

is the esp32 destined to become the arduino of the 21st century?

>>1374842
I thought rasPi already took that role?

>>1374767
It's not me this time I swear.

>>1374767

i cannot help you with the enclosure

however with the amp , i suggest you go with a D class tpa3116, tda7297 being AB is arguably better sound but eats way more power and gets very hot, riding outside you wont notice shit except your power going down. Using drone lipo battery's and boost it to 12v-24v for better weight reduction.

amp needs no heat sink will be a lot more efficient. might be easier mounting a 10 dollar Bluetooth speaker to your bike, but then ruining your diy project so dont do that. Anyways pretty good effort in salvaging parts to make it happen.

pic related. about 5 bux. you can also find ones with built in bluetooth.

>>1374973
neat! little 386 variant? :D

Ive got an ebike battery that wont work. Ive checked the individual banks, and most are good but one is 2.7v, but the battery as a whole will put out something like 48v.

The battery charger never indicates its putting out high current (charge light normally goes red when this happens). Ive left it on charge for days with no change in voltages, so I don't think it is doing any balancing.

Cant get anything more than 3.7v on the output side/external connector that goes to the bike.

I think the BMS is flakey, but I have no idea how to diagnose further.

Should I just ditch the BMS and wire it directly up?

>>1375014
>Ive checked the individual banks
By bank you mean ~4 cells in parallel? Your BMS will have measured that 2.7V with sense lines and decided that it shouldn't try to charge it. Either that or the BMS failed to balance the cells causing that bank to droop to 2.7V, possibly both. Replace (or rejuvenate) those cells anyways but keep an eye on their voltage as you put them through a charge-discharge cycle or two, and stop using them if they drop below their min (probably 3.2V). In that case it's definitely the BMS's fault, assuming you already checked for shorts/dodgy connections within the case.

>>1375024
>By bank you mean ~4 cells in parallel?
yes.

Now you say that about the low voltage cells, It may be that the bms has been trying to balance it but the batteries are just too far gone.

I don't have any experience with these spot welded strips, could i just cut it out and solder a new battery/strip in its place?

>>1374561
no
download ltspice, build some circuits, and convince yourself this is not the way to do it

i posted this question in the stupid questiond thread but it seems dead rn and this seems like a better thread for it

>this question might be retarded, I think ac circuits would be diy. electrolytic capacitors need to be in correct polarity to work and capacitors don't work with dc. i thought ac didn't have polarity because it switches in polarity?

>>1375031
Think we got the same question earlier in the thread. The idea that "capacitors don't work with dc" is plain wrong, they just work differently. Just check out google image results for capacitors in DC, particularly with a resistor too. Now a capacitor will resist change in voltage (calculus here >>1371960), and since the voltage in an AC circuit is always changing, it acts like a resistor, more or less. But any DC load (amplifier, microcontroller, logic gate) save maybe a heater will not have a constant power draw, the current they draw will fluctuate depending on what they're doing. A battery or any virtually other DC power supply can only provide so much current before its voltage begins to droop due to internal resistance, so people put large electrolytic capacitors across the outputs of these power supplies that resist this momentary change in voltage that would happen if they weren't there. Capacitors store energy when they're charged up by DC, and can in some cases be used as energy storage devices themselves, particularly for either low-current applications or high current applications.

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capchg.html
And the map on:
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/capcon.html
Are great resources for this sort of thing. But go ahead and download a free circuit simulator and experiment.

How difficult would it be to devise a prototype displaying pic related using 2 brush motors?


Or Would it be possible/easier with one motor, which could have so much power pushing it to the required speed and then alternate its direction with whatever required power?

I imaginr it'd be more difficult using one motor
Also open to ideas, I'm not used to doing this sort of stuff

>>1375043
Motors aren't precise, so you using torque is proportional to current probably won't be sufficient, but you can try. Instead I'd use some hall-effect feedback to program the desired motion somehow, since you'll need some way to keep the speed constant anyways. You can solve the equation easily with angular kinematic equations, by the way, though since sphere radius isn't specified you kind of assume they're point masses.

Any point to this endeavour?

>>1375045
It's for a class project, calculating the angular velocity and all isnt an issue, mainly the building of it

And so a hall effect sensor is what you mean? Controlling a motor? That does seem easier or did you mean without a motor?

Sorry im a little out of my depth when it comes to electronics

>>1375049
Yes, controlling the angular frequency of the motor via a microcontroller. Using a permanent-magnet brushless motor would be easier though, probably no microcontroller required.

But if the speed doesn't have to be exactly 20rad/s you could use a brushed DC motor with some sort of current sink and adjustable voltage rails, which would be by far the easiest method. The voltage would sit at a constant until you flip the switch, activating the current sink, which runs until it hits the 0.7V limit of the transistor and thus leaves the motor spinning. I'll see if I can whip up a circuit of that.

>>1375055
wow that's a big help man thank you

setting the initial conditions seems to be the easiest part and i should be able to set that up without even taking the torque into consideration.

the final velocity will be the same in the opposite direction, so the value acquired by the sensor should be the exact same, but the difficulty now is accurately setting the torque to 30 N*m because the time of transition is the most important part.


Would i be able to work out the math and figure out the theoretical time required, then have the sensor control the speed of the motor according to whatever value necessary in increments? seems like a backward, overly complicated way of rigging it for the correct answer

are motors really that inaccurate regarding torque?

>>1375059
>seems like a backward, overly complicated way of rigging it for the correct answer
Certainly, but it is correct.

>are motors really that inaccurate regarding torque?
They're probably pretty good, good enough for your usage case at least, but you have to set the speed by voltage and torque by current, which may take some tweaking and an overly complicated PSU for the voltage. Unless you have access to a good enough bench power supply that is.

Pic related is my shot at it. At t = 1ns the two PNPs switch over, deactivating the pull to +6V and activating the current sink. Replaceable with a DPDT switch I suppose. Since the speed of the motor is proportional to the voltage, the energy/momentum stored within the rotating system can be represented with a capacitor across the motor. This constant current flows until the voltage across the capacitor and motor hits the limit of the current sink, stopping at a constant voltage and therefore constant speed.

Measuring the torque/current and speed/voltage proportionality constants may prove difficult. Motor ESR and inductance will be easier. But by setting I*w^2 = C*V^2 you should be able to get a nice electronic model. Then you just pick voltages and currents that will result in a reasonable motor reversal time (5s or something)

>>1375068
This is kind of a lot to take in for someone who hasn't taken electrosystems yet, or any experience with circuit diagrams. Not sure how I'd account for it, seems pretty complex


would an arduino suffice for this? I understand the other symbols to a point, but why are there 4 voltage sources on your diagram? also, why is the capacitor across the motor necessary if the voltage through the motor is already indicative?


i'll be reading some of the recommended books for this, although there is a time constraint. appreciate all this immensely

>>1375076
>This is kind of a lot to take in for someone who hasn't taken electrosystems yet
Yeah, oops.

The two voltage sources on the left make a + and - voltage rail, so you can have the motor pulled to either a positive voltage or a negative, which is necessary for the motor to be able to go forward and in reverse. The other two are just used as switches in conjunctions with the PNP transistors, Q1 switching off when Q4 switches on. You can ignore those transistors, voltage sources, and the resistors connecting them. R6 is just a current shunt, the diodes are probably to prevent a voltage spike from the inductor, not sure if they'll work with a real PSU though, might want to replace it with a couple of appropriate zener diodes in series across the motor. The transistors Q2 and Q3 make up a current mirror along with R3: R3 powers the base of Q2 to allow I = ~Vcc/R3 of current, and since the same base voltage is at Q3, so the same current flows through Q3 to ground (assuming the transistors are identical). Since collector current is proportional to base current, the same collector current will flow through both Q2 and Q3, hence sinking I = ~Vcc/R3 through the motor, which is a constant current. When the voltage after the motor approaches 0.7V or so it will stop conducting through Q3, so the maximum voltage and minimum voltage across the motor will be of slightly different magnitudes.

>>1374984
Audio noob here. Many thanks for the info about the amp. I'll check it out, just haven't seen this variant yet so you really helped me.

>>1373456
Tbh I've never encountered wick that had enough flux on it. Adding flux always makes a massive difference.

I got a crimping tool for those horseshoe terminals and it works great

Now I need one for Dupont, JST and Molex connectors. Is there such a thing as a crimping tool that is cheap and will let me crimp all of them, from pic related down to 2mm JST?

in case anyone is interested in buying cheap tools:
>>65678362
>>65678412

be warned: many of these products don't even have any feedback.

>>1375194
fuck, I forgot to modify the links
>>>/g/65678362
>>>/g/65678412

>>1375179
I bought Pic related 20 years ago and it's bretty gud (can be made to work with quite acceptable results) for almost all open barrel terminals. it costs the better part of 100 burgers though
"Engineer" of Japan's PA-21 crimp tool is well-renowned but its range seems a bit limited to me. see their application chart on their product page
if you really want cheap *and* good, you might have to pick up two or three tools. for "Dupont" connectors I use the SN-01BM crimp tool and it works well and with little trouble, other than the occasional cocked angle if I crimp too hard. the SN-28B also works, sorta, but I almost always have to dress the finished crimp with long nose pliers so it fits into the shell. there are a few models/series of crimp tool on alibay that have interchangeable dies within the series, which might work okay IF you can find the right dies for the terminal types you want to crimp

>>1375179
you want a W shaped crimp die.
look at this pic, the lower jaw is a simple C while the top is made of two W of different high. the higher W is meant to crimp the contactor, the lower W crimps the strain relieve.
This should work for all pins of the style like your pic

>>1375201
also, this pic is the first result searching "molex crimp tool" at around 20€. It even has a dial for crimp force

>>1375200
That's very noice advice, buddy

Can that SN-28B be made to work with the SN-01BM dies?

>>1375201
>>1375204
Thank you very much I'll look for that

>>1375218
the "SN-01BM" I got (marked Plato) doesn't interchange with the "JX-1601" I got, the thickness of the slot in the pliers doesn't match the thickness of the mounting tab of the die thus the dies themselves won't be well-aligned. as with all things chinkshit, your mileage may vary
the SN-01BM I got was only like $15 shipped. it's possible you may find ali sellers that offer both a "Dupont" compatible crimp tool and also additional die sets

Sup lads. Long story short I am looking to build two of these oxygen sensor simulators based on the schematic from this site:
https://www.chippernut.com/blog/o2-simulator
I have zero electronics experience past splicing wires in my car and the electronics class i took in 10th grade like 13 years ago. That being said im trying to crash course my way into this because it really seems pretty simple.

Heres my questions/concerns:
The schematic calls for 2x 1.7v@20mA LEDs. Im buying a kit off amazon (https://www.amazon.com/gp/product/B073ZC68QG/ref=ox_sc_act_title_6?smid=A2XZVOPECQVFD3&psc=1)) that has a bunch of components and stuff for beginners, and among them are sets of different color LEDs. Ive been reading that red LEDs are typically the ones that are of this spec, but ive also seen them listed as 1.7v-2.0v... How do I know which one it is? Is it safe to have a range? How do i know what the specs are for the diodes in my kit if they arnt listed outright? Ive tried looking online for a diode of this exact spec but all I can find are ones from radioshack that cost 5 dollars a piece which is retarded. The rest dont seem to list the amperage, only voltage. Again i sorta dont know wtf im dealing with so is the amperage a concern here?

Another question: can I/should i use a car battery for testing this device? I need 12-15v power source, and since itll be wired into my car, I figured the best way to test is using a car battery. Is this a stupid idea? Is there a better/simpler way to do this?

Sorry if these questions are really dumb or something, again I know extremely little on electronics. Just want to make sure im not going to screw something up and fry my cars electronics.

pic unrelated

How do I a throat microphone?

>>1375230
That SN-01BM is also sold with a lot of dies so I'll go for it, thank you buddy

>>1375259

- dont sweat a difference of 0.3V on the LEDs. it'll work fine with any red LED or even a yellow or green. there will be such a tiny current thru D2, it wont even light up. D1 will run off whatever current the 555 uses nominally. it could be around 10mA for an old bipolar 555 or much less for a newer CMOS 555. so dont worry about current ratings for the LEDs, they can handle it.

- dont use a car battery for testing. use a tiny 12Vdc transformer, rated around 0.1 to 0.2A. this way, if you made a mistake, you wont send 100A thru your circuit. once it tests out fine, then you can connect it to a car battery.

>>1375259
the correct way would be to put 20mA through them and read the voltage drop across them. that's a bit involved, though. in this circuit, the voltage drop isn't so critical so you can assume a red LED would be close enough. you could substitute greens or yellows with minimal effect on circuit operation, but to substitute blues you might need to make R3 more like 91k if the ECU is going to be picky about post-cat O2 input voltage range
current (amperage is a term mostly only used by welders) is only a concern because the voltage across (Vf) will vary slightly with current through (If), Pic related. usually the rated voltage is specified at the rated current, which for those little guys is customarily 20mA
>can I/should i use a car battery for testing this device
sure, if you don't have another source of 13-15V handy. a wall wart that outputs 13-15Vdc would work best, if you happen to have it. the advantage of using the smaller supply is that the battery offers enough energy to melt your project into glowing slag if you happen to make a wiring or assembly error, or got a bad component in the kit. please use an inline fuse on the +V wire, close to the supply end of the wire, or use a fused lighter/accessory plug and cord for testing

>>1375282
excellent chioce, cheers

Just bought a crimper for jst/dupont stuff. Is there a go to place to get crimps/connector housings? The prices on Amazon are ridiculous and digikey wasn't much better. I'd rather not pay $.80ea for 1x1 dupont housings.

>>1375306
aliexpress, of course

Are these equivalent?

>>1375310
Well I had thought of that as well as eBay Chinese sellers. I was hoping I wouldn't have to wait 20-40+ days for stuff to arrive.

>>1375350
Yes, but technically you should have an OR gate where the select line comes in.

If I connect a negative voltage point to ground, can I generate free energy?
If yes, please nobody steal this idea before I can copyright it

What's the best way for me to run a RasPi off of an 18650, have it be USB chargeable, and have it able to be used during charging?

>>1375541
Plug it into a USB power bank. Barely any of the circuitry is reductant anyway, just find a form factor that works. Else, get yourself the USB charging and boost modules for like $1 each on ebay/Ali, they should let you charge and discharge at the same time, but check how many cells they're made for since they could push too much current through. And don't solder onto the contacts of an 18650, either spot weld them, use cells that have solder tabs on them, or put them in a battery holder.

>>1375545
Oh and you can buy empty single-cell power banks like:
https://www.aliexpress.com/item/Bush-did-911/32579622369.html
Which is a cheap option if you already have a cell, though they are only 1A max.

>>1375531
there's actually a lot of ways to generate free energy. we just hide them from the plebs.

hint: if you put a large enough current through an inductor and then cut it off fast enough you can create an infinite voltage (= infinite power).

>>1375546
>>1375541
This is the typical socket configuration used to tell the customer that the device has no feed-through function i.e. the output shuts down during charging and you cannot use it as a UPS. This doesn't mean that all others can do though. Unfortunately the charge+supply feature is rarely specified.

>>1371574
Hey swoles,

What're your thoughts about soldering up a hundred or so of these small piezo plates, and stacking them all on top of each other (with a layer of electrical insulation between them) and pushing a voltage through all of them at once.

Will it result in it extending and contracting 5-10mm, or will nothing happen? Thoughts?

>>1375586
>a hundred or so
Two or three are enough to think it through.
How do they deform?
How would you arrange them?

>>1375610
>How do they deform?

Applying a voltage results in the expanding longitudinally.

>How would you arrange them?
Literally like I said - stack a hundred or so directly on top of each other.


I'm asking if people have experience with these plates and what their theories are, not for a thought experiment.

>>1371574
Alright lads. Anyone know a decent ESC I can use for a drone that's documented more thoroughly than "push the joystick to max to go fast lol". Finding an ESC with a decent datasheet is a nightmare.

Or alternatively, is there anything else I can use to switch a low-power PWM signal to a high power 3 phase PWM signal?

>>1375586
piezeoelectric stacks have hundreds of piezo plates and only move a few microns.
so no, you probably won't see anything.

>>1375621
You can buy that shit. It's expensive. But you can atleast look at the datasheets and find out how many layers of what dickness result in what displacement.
Just search fot Piezostack
http://www.pi-usa.us/tutorial/4_44.html

Roughly, how much of a pain would it be to have a 220V 30A line installed in my lab? It's already going to be a huge pain because of red tape with the school but barring that, what's the deal?

>>1375661

easy peasy. it's more or less what every home has in N. Amerikkka to run the stove.

>>1375661
Very easy, you can find anything you need online. As far as the amperage goes, 30A is a lot. Might be a little expensive. Definitely look online for someone who has done the same/similar thing.
>>1375678
That is NOT what every north american has. That is European(mainly UK) mains power. US is 120v nominal

>>1375686
>>1375678
my bad, you said stove, youre right. Stoves use split phase 240 and a center tapped neutral.

>>1375686
M8 just connect 2x 120V outputs together, 240V.

You guys are such pussies lol

>>1375678
>>1375686
Doesn't matter how easy it is for me, I can't do it. This is at a university so facilities needs to do it, the drawings of the room need to be redone to indicate an outlet, tons of red tape. It's for an ultracentrifuge (motor spins at >41k rpm, 200k x g) which also has a vacuum and temperature control device, with the following requirements as per the manual

Instrument rating: 220 to 240 VAC, 50 Hz, 20 A 200 to 240 VAC, 50/60 Hz, 20 A

Power line range: 180 to 264 VAC, 60 or 50 Hz (single-phase), 30 A

Electrical supply: Class I

It draws 20A but needs a 30A line to handle surges when it's turned on and whatnot. I'm sure this would be simple if I could do it myself or just call an electrician but I can't.

>>1375695
https://www.amazon.com/LiteFuze-LT-2000-Voltage-Converter-Transformer/dp/B008GQTXS0/ref=sr_1_1?ie=UTF8&qid=1524763669&sr=8-1&keywords=120v%2Bto%2B240v%2Bconverter%2Bindustrial&th=1

>>1375638
>Or alternatively, is there anything else I can use to switch a low-power PWM signal to a high power 3 phase PWM signal?

Not very fond of drones, but there are integrated BLCD/3ph motor drivers. Using one however, means you should provide with the pwm signal wit a mcu or such to make the motor spin.

>>1375694
This works if he has two phases

Hello ohm. My boss got some light sockets to install into into a sandblasting cabinet. Notice anything strange here?

So I know mixing copper and aluminum wire is pretty bad news, but is it long term only? I have a project I am working on for tomorrow, and just realized some of the wires are aluminum. Can I wire nut it just for tomorrow and then disassemble it and fix it proper later? Only dealing with a Max of 4-5 amps mixed to a central power cable.

>>1375971
The external screw terminals? Also it says porcelain but it looks like plastic.

>>1375974
>>1375971
Yes exactly right. How the heck are you supposed to install this safely when the terminals are both on the exterior?

>>1375977
Well at least it's not for a water-jet cabinet.

hey /ohm/
I just picked up one

hey /ohm/
I just picked up one of these power supplies. any idea on how to remove the voltage spikes that happen on power on / power off?

Quick question about Zener Diodes.

I am trying to protect the I/O pins of my MCU, these have a voltage range of -0.6V -> 4.1V. Normally the I/O pins see either GND or 3.5V.

Is using a VESD03-02V-G-08 sufficient to protect against >4.1V transients? This Zener has a Min Breakdown Voltage of 4V (1mA) and a Reverse Standoff voltage of 3V (20uA). Here is the full datasheet: https://www.vishay.com/docs/83367/vesdxx02v.pdf

>>1375987
what do the spikes look like on a scope?

>>1375992
Low voltage zeners are a bit shit, have a look at their V/I curve and determine if the current that flows at 3.5V doesn't matter. And actually measure that voltage to make sure it is 3.5V.

>>1376002
Yea would if I could, the only useful data-points I have are (x;y) 4V;1mA & 3V;20uA

>>1376004
Well that's probably not good enough, unless you put a dropper resistor somewhere such that 1mA is enough to get the voltage down significantly.

>>1375992
Like the other anon said, low voltage zeners generally suck and that particular one is optimized for ESD protection instead of general use. It is also rated for just 3V, so your 3.5V is bit iffy.
It might still be ok for very short transients. If your overvoltages last relatively long, consider basic diode clamps (schottky to VCC and GND). Even a plain old series resistor (with MCU's internal protection diodes) might be sufficient in some cases.

>>1375091
Do you have a discord or something I can message for any further questions?


I think I figured a simpler way to make the project, simply have a motor supply constant torque in one direction up to a certain speed and then stop. Whatever time it takes to reach that speed can merry be doubled so a single direction would suffice.


You mentioned a brushless motor being easier, I could just calculate the Kt (torque constant) and apply the current accordingly but how exactly does voltage factor into this? The torque supplied doesnt have to perfectly match my calculations


Seems like all the brushless motors i find for sale have ridiculous RPM, when I only need 190 or so


I would also downsize the project, 30 N*m is way too much torque and makes the whole thing more difficult,

>>1376090
>a discord
I'm ignoring like 3 responsibilities by not logging into my discord.

>simply have a motor supply constant torque in one direction up to a certain speed and then stop
Yes, but what would you do to insure constant torque? Isn't that what you wanted to figure out how to do in the first place?

>a brushless motor being easier
Only if you were going the cheaty "digital speed control feedback" route, as they're synchronous motors (they spin at a multiple of the frequency you send to their coils), so they're easier to control the speed of than a brushed motor, which requires an encoder of some kind. But brushless motors can be driven at any damn speed you want to, though ones with few poles will be pretty jerky at low speeds. I don't suppose you've got a stepper lying about do you?

Now as far as power goes, with those4 weights and the specifications given, it will be pulling an average of 300W for 2.72s, probably double that at either end. Now this isn't unfeasible, but it will certainly require quite some electronics, namely ones that can handle a few dozen amps, so stepping down the scale of the project is a good idea.

I'm trying to replace the keyboard on a screen printer because the current one is old and a lot of the keys don't work.
The keyboard on it has a 5 pin DIN connector and I'm trying to replace it with a keyboard with a PS/2 connector.
I've cut the PS/2 connector off the new keyboard and soldered a 5 pin DIN onto it.
I've definitely wired it correctly as the PCBs in both keyboards have the cable connections labelled as VCC, GND, Data and Clock.
For some reason, whenever I attach the new keyboard, I get a keyboard error / no keyboard found from the BIOS.
The new keyboard is definetely still working as I've now soldered a 2nd cable into it with PS/2 to check it on another PC.
What the hell am I doing wrong? Is there some protocol issue?

>>1376092
> I don't suppose you've got a stepper lying about do you?

No, as far as materials already on hand I only have an arduino uno, small breadboards, and a box of random arduino parts/sensors such as a potentiometer.

I have a week to do this and factoring in my full time job and classes, along with lack of parts, I think simplifying it further would help alot.

You seem to have an engineer/physicist background, i guess the point of the project is to demonstrate linear impulse and momentum. I don't think maintaining constant torque is feasible like you said

All I really need to do is make sure the motor achieves 20 rad/s in 1.36 seconds, giving the proper average torque right?

How would this be done? If voltage controls speed, and current controls torque, would calculating the power required work? and using any correct combination of current and voltage

I don't know if i necessarily need it to stop at that correct speed, but if i do then once it reaches that speed, the voltage remains constant? and what happens to the current? does it drop off?

Could I do this with an computer psu?

I don't suppose i could use one of these breadboards...

>>1376100
>Is there some protocol issue
I suppose so, assuming you soldered it correctly. But I'd check the datasheet of the existing old keyboard, and maybe their internals are similar enough that you can just transplant a few parts.

>>1376105
>make sure the motor achieves 20 rad/s in 1.36 seconds
Well if that's fine, just experiment with your motor and find out what voltage equates that speed with your load on it (V0), then just have a linear voltage ramp from -V0 to +V0 in 2.7s. You may want to use PWM instead, if the resultant load on your buffer is too high. There are a few ways of having a voltage ramp, such as using a 555-timer, current-source + capacitor, or op-amp.

Though it's not necessarily simpler than putting a current source across the motor itself, like the first circuit. And I think that first circuit could be simplified to 2 resistors, an LM337, a transistor, and a toggle switch plus your motor by the way. But either way you'll still need a split power supply, for which two batteries would suffice.

>>1376116
>you'll still need a split power supply
Oh this is only for going from negative speed to positive speed, you can have a single power supply for going to or from 0rad/s.

>>1376116
I think PWM would be easier with the arduino. Have any recommendations for A motor? Stepper or brushless? I imagine it'll have to be pretty small since the arduino can only put out like 30mA but I can design the prototype size around what the motor is capable of. Light materials, aluminum foil balls for the weights for precision


I guess speed measuring is also something I could solve with a hall effect sensor near the rotating balls, one with a magnet attached, measuring RPM right?

I want to at least get the speed correct before integrating the parts you mentioned for the reversal

Again, thanks a ton

>>1376126
Well for just PWMing the motor as a crude way to control the average voltage and hence its speed you'd use a brushed DC motor, which is probably easiest. For a brushless motor you'd need to just change the frequency at which you drive it and it would stay precisely at the correct speed. I'd recommend the first method because the normal brushless motor will be running at 3.2 turns per second, which I don't think would work well. Depending on how the brushless motor needs to be driven you could PWM it, but it will be more difficult.

If you just want to get from speed A to speed B without caring what kind of torque curve you're sketching then turn down the PWM on the arduino until the brushed motor spins at the right rate, and write 5 lines or so to transition from 0% to that duty cycle. Don't forget to use a mosfet and a freewheel diode.

But if you actually want to give speed feedback to the arduino then a hall-sensor near the magnets of the motor would probably be good enough, or an optical encoder (LED+phototransistor) with a piece of reflective tape on the motor shaft. Calibrating the feedback code may be less than trivial however.

The parts I was talking about weren't just for the reversal, they were a standalone constant-current supply to pull the motor at approximately a constant torque. Integrating the reversing function into your arduino model wouldn't be too hard at all, you'd just need to either make an H-bridge for your motor or use a split supply. PWM one line of the bridge to go clockwise, the other line to go anticlockwise, and if you use P-fets on the high-side you won't need a driver, just some inverted arduino outputs. Something like pic related, except instead of the two BJT inverters you'd just use two more I/O pins.

>>1376134
If the transition from 0-100% of the voltage to the top is linear, doesnt that mean that the torque would be constant anyway? Since the acceleration is the same throughout the process

>>1376134
Why do you have two bjts connecting to the gate of your fets. What purpose do they serve?

>>1376134
>For a brushless motor you'd need to just change the frequency at which you drive it and it would stay precisely at the correct speed.

Really? I thought the cheap ones used EMF feedback to set the frequency, and the good ones use an encoder or shaft sensor; BDCMs are NOT synchronous motors in the AC sense; they are synchronous only if the applied signal tracks the rotor, and the rotor position depends on inertia during acceleration, and load torque at all times. Just slapping a frequency on a BDCM does not erase this fact.

>>1376100
You sure the original keyboard is actually a PC compatible keyboard and not just something sharing the same connector?

>>1376143
Torque is roughly proportional to current, no load speed is roughly proportional to voltage.

With a constant voltage source acceleration is roughly a straight line from maximum acceleration at zero speed down to zero acceleration at the no load speed.

To approximate constant torque/acceleration you'd need to use a current source, for better results you could use a lookup table of current vs present speed to compensate for it being not quite constant.

A linear voltage ramp could be close enough too with a small inertia load.

>>1376146
The BJTs allow using a lower voltage logic input than the voltage used for the motor supply, so it'd work with for example 3.3V logic at D1/D2 to control the 12V. Without the BJTs there a 3.3V microcontroller can't go high enough voltage to turn Q1/Q3 off.

>>1376148
If you've got a predictable load it is possible to drive a BLDC entirely open loop with just a variable frequency like a synchronous motor (or even a fixed frequency if you haven't got much inertia to start up), but you have to be careful with the frequency and voltage profile to make sure the rotor can keep up, if not it's going to stall.

>>1376154
>The BJTs allow using a lower voltage logic input
This is incorrect, they're being used as inverters since P-channel MOSFETs turn on with a logical 0, so you can power both Q2 and Q3 with the same line D1.

Any ideas for converting a (stereo) DAB radio that is obsolete now to an internet radio (wifi) / bluetooth speaker on the cheap?

I'm looking at an esp32 and an I2S DAC. Planning on reusing the amp, controls (rotary encoder + numpad) and display (20x2 or something like that) if they're reasonably accessible.

Anything else I should be considering? A cheap orange pi maybe? I saw a project on memeaday using an esp8266 and a VS1011 but the VS1011 makes the whole thing too expensive, and I can't seem to find a cheap alternative. Surely it must exist since $1 mp3-players are a thing.

>>1376106
>>1376150
Never mind, I found a PS/2 connector buried inside the machine and connected a new keyboard into it

>>1376166
Read the rest of the post before jumping to "correct" someone on the internet.

>>1374113
>Of course there's no magic under the hood but to the layman it would seem like it.

so your post was for the laymen browsing /ohm/

gotcha.

>>1376154
>but you have to be careful with the frequency and voltage profile to make sure the rotor can keep up, if not it's going to stall.

correct.

so people actually do this? instead of using the normal method that simply works, where you get the rotor position from EMF or a shaft sensor? Why would you try to predict something that you can sense?

I started messing around with HDLs and FPGAs and I read that a lot of the stuff they teach about RTL-design for example in Universities doesn't really align with industry practices.

My knowledge is mostly based on uni-courses and I was wondering how I would best go about learning about industry practices without actually working in the industry.

Are there books that teach industry practices? Journals, Magazines? Any other resources?

>>1375994
here's the turn-on spike. the turn-off spike is only about 400mV above the 5V setting which is within the 1V spec they advertise. Im thinking I need a zero cross circuit but not entirely sure.

>>1373076
I got one of those tiny Chinese CNC routers specifically for making PCBs. It's a little fiddly but it works. The key is using good software to compensate for the shittiness of the machine. I think I went kicad->flatcam->chillipeppr. Flatcam can generate indexing holes for two-sided boards and chillipeppr has an autoleveling mechanism to compensate for board thickness.
It's really useful when you need a board *right now*, also for DIY smd projects. Otherwise, you don't have through plating, vias, soldermask or silkscreen, so if you can wait a week or two a fab house is the way to go.

>tfw you just need one tiny component but there's no local diy electronics store so you need a 2000 ton container ship to travel halfway around the world just to bring it to you
Is diy electronics the most environment-destructive hobby there is?

>>1375977
It might be a base for those fixtures with plastic/acrylic cover over them.

How are those cheap helping hands on AliExpress, the ones with the nice bendy arms? I need one, but I don't know if I should spring for a nice one or gamble the chink ones.

>>1376357
>just to bring it to you
>just
no, that would be competition motorsport
3/8 b8 m8

>>1376357
It's not that bad. The ship is also bringing me a one hung lao android watch

>>1376276
Well for a very simple program on a low load, a simple frequency ramp would essentially be guaranteed to get a BLDC motor up to speed.

Anyone know where I would be able to find circuit schematics for vtech toys? I want to try to make my own learning toy.

>>1376467
it's all software. there's little to no real magic in the circuitry. just write lolduino code

>>1376276
It's the standard method for initial start up of drivers that then switch to EMF sensing, since when it's stopped there's no EMF there to sense.

>>1376487
Though the models with a hall-encoder could do either.

Someone gave this to me
None of my friends will buy it
Any project ideas?

>>1376497
enjoy

>>1376497
What sort of relays are they? Are they good for mains voltages or high-current DC?

>>1376510
If the pic there is the exact model those are 10A 250VAC / 10A 30VDC.

I want to drive this pump with an arduino (https://www.aliexpress.com/item/DC-12V-19W-Submersible-Water-Pump-High-Lift-9m-500L-H-Car-Wash-Bath-Fountain-Submersibles/32838678555.html))
>12V
>1.5A at full load
I just need to turn it on for a couple of seconds at a time, maybe less, and not very often (off-time is random, can vary from seconds to minutes/hours). No need to reverse direction of the pump or anything fancy like that.

I need a mosfet, right? But how do I go about selecting one? Knowing that a digital out from the arduino is 5V I need the mosfet's threshold voltage to be lower than 5V, yes? Then looking at the graphs on the datasheet, Id with Vgs=5V and Vds=12V should be higher than 1.5A, yes? Finally, looking at Rds and knowing that Id=1.5A will tell me that the power dissipated by the mosfet will be P=Rds*Id^2, and the voltage drop over it will be V=Rds*Id. That means that if I power the pump with a 12V power supply, the pump itself will actually get a voltage equal to 12-(Rds*Id), correct?
If the above makes sense, is the IRL540 a good choice? https://www.vishay.com/docs/91300/91300.pdf

Last thing: should I put a flyback diode in parallel with the pump? Or with the mosfet? And this diode should have a breakdown voltage of at least 12V and a max current of at least 2A right?

>>1375987

nothing you can do except redesign the circuit correctly. a workaround is to install a big switch that cuts off all outputs. so you switch on the supply, it spikes, then you turn on the outputs.
or, it might help if you wire in permanent load resistors to eat the spikes.

>>1376100

should work unless you screwed up the wiring. i would have use an adapter made for that very purpose. (i bought two in 1995)
could be you missed wiring the cable shield to the metal shell of the connector.

>>1376605

you put the diode across the load (backwards of course). a 1N4001 rated at 50V 1A is more than enough.

>>1375987

nothing you can do except redesign the circuit correctly. a workaround is to install a big switch that cuts off all outputs. so you switch on the supply, it spikes, then you turn on the outputs.
or, it might help if you wire in permanent load resistors to eat the spikes.

or instead of the switch, another idea is a timed relay that kicks in a few seconds after the unit is switched on. this is used in old stereos to avoid the DC thump caused by the charging output capacitors.

>>1376100

should work unless you screwed up the wiring. i would have use an adapter made for that very purpose. (i bought two in 1995)
could be you missed wiring the cable shield to the metal shell of the connector.

>>1376605

you put the diode across the load (backwards of course). a 1N4001 rated at 50V 1A is more than enough.

>>1376181
But you ARE incorrect. 100% so. as you can see, the control signals go directly to 2 out of the 4 FETs. so, obviously the bipolars arent there for converting voltage levels, but as inverters.

>>1372663
>>1372673
>>1372679
Pls send help.

>>1376637

no idea what you mean by ''floats like crazy'' -- my guess is the thing is chattering. you need to research comparators with hysteresis. maybe somehting like this: http://www.analog.com/en/analog-dialogue/articles/curing-comparator-instability-with-hysteresis.html

it's not as complicated as it sounds, just choose a feedback resistor that gives you about 10% hysteresis, and you're good to go.

>>1376640
chattering is the word I was searching for there. Thanks I'll have a read.

I don't know if there's a phrase for what I'm trying to describe, but how long did it take you all to feel comfortable with electronics, enough to know your way around building them, and could generally just go with the flow, understand what everyone was talking about, etc etc?

>>1376632
If you didn't want level conversion you wouldn't need the BJTs, can use the two drive signals connected to left side gates connected together and right side gates. Using the layout as shown just makes control harder if you have 12V controk available in as it requires having D1 always driven as !D2 and can't do braking.

>>1376682
> don't know if there's a phrase for what I'm trying to describe
fluent?

>>1376623
>>1376628
>you put the diode across the load (backwards of course). a 1N4001 rated at 50V 1A is more than enough.
Got it. And is everything else correct? I'm about to order 15 IRL540 just so I have more for future projects.
Or does /ohm/ recommend other mosfets (possibly that work both on 5V and 3.3V logic)?

>>1376682
If you get a cheap kit with an arduino and a bunch of sensors and actuators and stuff from somewhere like banggood you can pretty much start tinkering immediately.

>>1376458
>on a low load, a simple frequency ramp would essentially be guaranteed to get a BLDC motor up to speed.
>>>

I agree. Typically my applications involve a load of some sort, so I'm not familiar with these applications where a motor doesnt' do much of anything except spin.

>>1376487
>It's the standard method for initial start up of drivers that then switch to EMF sensing, since when it's stopped there's no EMF there to sense.
>>1376490

It just seems like such a clusterfuck, since the motor can start in the wrong direction.

I suppose saving 20 cents matters in some cases.

>>1376605
diode max forward current is usually specified as a steady-state value. if you're just absorbing back EMF, you're probably more interested in the peak rating
otherwise, your design methodology is spot-on. have another (You) >>1376605

>>1376730
for solderless breadboarding the IRLD series has a unique hotness: 4-pin DIP packaging. the IRLD024 is rated for 5V logic, not 3.3V, but will still pass 2-3A at 25°C at 3V Vgs, which should be plenty for prototyping, Pic related
IRL540s are good all-around switches. I recommend you also stock a P-ch complement, as sometimes low-side switching a load doesn't work. unfortunately there is no IRL9540 so you get to pore over the selector guides and find something close enough. enjoy!

>>1376845
Thanks a lot for the feedback! I actually finished placing my order a moment before your post, and I went with the IRL540. I guess I'll keep in mind to get a few p-channels too. On the other hand those IRLD024 seem quite expensive?

>>1376682
Like everything it depends how much time you invest in it. I had some electronics classes at uni and have been dabbling with small circuits, but I still don't understand a lot of what is discussed on here.
The 80-20 rule applies though (ok maybe it's like 70-30?), so if you learn the basics well that can get you already pretty far. Then it's a matter of learning with time and by doing.

>>1376861
the IRLD024 are a bit expensive, but that's less critical for prototyping, where a smol package that fits easily onto your breadboard without wearing it out can be really fucking convenient. not strictly necessary
the IRF9540 is a P-channel complement of the non-logic level IRF540. that's probably okay for switching high-side loads at 10V or higher where your low-side level shifter can be made tolerant of the higher voltages. the IRF9540 should also work well when driven by 5V CMOS logic levels and Ids requirements are only 1A or so

Once I've finished breadboarding, is it expensive/time consuming to design and make a PCB? I've never done it and don't have a 3D printer. Should I just get some veroboard and solder the parts?

>>1376898
pcbs are very cheap. elecrow sells 5pcs 4"x4" for $5+s&h. oshpark sells at 3pcs $5/in^2 but with no s&h, gold plating, and realistically 5/4 capabilities (though they state 6/6). making them is pretty annoying to be honest. i use altium and the actual process of routing is quick, easy, and fun but if you're making a simple board with ten things on it it gets real fucking tedious handling all the enterprise software overhead.

>>1376898
going from the layout to the finished board takes a while (because you sent out for it) or is mildly annoying and vaguely unpleasant and/or mildly hazardous (because you wanted that board today and therefore decided to pattern and etch it yourself from a blank)

>>1376898
>don't have a 3D printer
Can you 3D print circuits now?

>>1376806
It can't start backwards since it's a three phase drive sequence.

>>1376987
the light-projection printers are surprisingly useful for patterning sensitized pc boards

I'll be the first one to admit that I'm a beginner to electronics, but I am becoming more comfortable with it.

However, is it normal that I'm somewhat comprehensive to move up to anything higher than 9v?

>>1377034
anything 48V or below is considered low voltage and won't kill you. just don't lick it
also, a power supply with a reasonable power limit (<40W) is unlikely to create any explosion that might send pieces of component flying into your face, unless you install a polarized cap backwards or something like that

>>1371574
How stupid would it be to connect one of these wire ends (OP pic) to an alligator clip? Would it arc?

>>1377053
I just basically want to move forward to relays and automating waterpumps and what not.

>>1377019
You mean light curing resin? That's not really what I think of when I hear "3D printer". Though I guess you could 3D-print over the top of a PCB as some sort of stencil for UV exposure.

>>1377055
Arcing won't happen without particularly high current (a few dozen amps), so you'll pretty much never get it from something that isn't a welder or a car battery. Besides, what do you mean by connect one of the wire ends to an alligator clip? You mean short two voltage rails together? Or just use it as a wire in a reasonable circuit?

>>1377055
remember your flyback diodes and don't plug things in backwards. you'll be fine.

Dingus is need of help here, had a power surge during a storm last night. The ac adapter on the left stopped working, I'm hoping the cell phone booster I had it plugged into is okay.
I ran to walmart and grabbed this variable voltage ac adapter on the right. None of the connectors fit my device because standardization is for plebs. Since my old adapter is trash could I splice the end and insert the wires into the two holes you see there? Would it work? Is it safe? It would only be temporary. I just want to know my device still works before ordering the new adapter.
Yes I will use a surge protector next time.

>>1377170
>Would it work?
yes
>Is it safe?
not in a consumer sense obviously but if you're asking the question you're fine.

>>1377170

the more legitimate (but still amateur-certified) way to do this is to cut both cords, strip wires, and splice together, then tape it all up.
you need a voltmeter to check polarity, tho.

>>1370907
>>1370914
>>1370925
Someone asked me to try my component tester with an inductance. I actually managed to find one, according to its color scheme it's 100uH with a 10% tolerance. The tester measured 0.09mH and 3.7 ohms.

hey guys i need some help making a very simple circuit, ive tried teaching myself electronics but im a little dyslexic about it. I'm good with my hands though and am determined to make it work

what i want to do is make a random number generator. It will be a simple circuit and the output will be a single flashing LED. I want it to randomly generate in binary, so the light off would represent 0, and the light on will represent 1

here's how i need it to work: it has to have two operations, "calculating", where it flashes on and off rapidly, and "display", where it either shines the light on or off.

flipping a switch will make it "calculate", and flipping it back will make it stop on whatever the light was flashing at the moment the switch is flipped. That right there is the part im having the most trouble with, how can I get this circuit to recognize what the light is displaying in that instant, and to hold it? I really need a degree of accuracy here, the "random"ness factor comes from the instant i decide to switch back, with the calculating phase being too rapid for me to track and predict

i know its kind of a weird request and i feel bad for coming here with no concrete understanding, but i have this idea for kind of an interactive wall made of hundreds of these little circuits. kind of an art piece with randomization being the theme. its all in my head right now, but I would really appreciate some insight on how possible this is, i mean if I could get it working with just a few switches and resistors soldered together that'd be great, but if its too complicated for that and I need to get into arduino to host these operations, I'd rather hear it from you guys than try and assume because the answer isn't immediately clear to me.

I'll be lurking in this thread if any of you can get back to me on this. Thank you for any insight you can give me

>>1377482
I personally don't know what IC can achieve this, but I think Arduino is the easiest solution for this.

>>1377482
Also, maybe this will help you: https://sci[add -
here]hub.tw /10.1109/ICECS.2010.5724498

>>1377500
its difficult to understand as a layman but it sounds really similar as a layman, maybe all I need to do is take the individual units from this circuit since i dont need big numbers generated

>>1377218
Sounds fine, is there a problem?

>>1377482
Use a D-flip-flop for the memory latch. You can probably make one out of a half dozen transistors and resistors if you don't want to buy an IC for it. But chances are your RNG will produce a really fast oscillation anyways so you'll need to step it down with a frequency divider anyways, and frequency dividers are made of flip-flops anyways.

What are you using to generate the numbers?

>>1377482

cant think of a way to do this with just passive components, but if you can into chips, then what you wanna do is very simple. you need an oscillator that runs at high speed (fast enough that you cant anticipate a certain count when pushing the STOP switch) and a binary counter chip. you put a toggle switch between the two, so when you open the switch, the counter stops at some unknown count. when you close it, it counts fast again.

see pic attached. this particular counter (4040) can have up to 12 binary digits, but you use only as many as you like. if you want 4, use the Q0 to Q3 outputs only.

I want to play audio samples with a micro controller but I don't know.how to go about doing it. I want to be able to play different samples by pressing buttons. Anyone have some good sources where I can learn this stuff? I'm guessing I would need to have a wav file and then the waveform would need to be filtered and amplified before reaching the speaker but I don't really know how to go about doing this.

>>1377569
I forgot to mention the microcontrollers I have are arduino uno and msp430

>>1377537
>Sounds fine, is there a problem?
not really, some anons asked me to check how well it did with inductances and so I did

>>1377067
the standard sensitized resist films are sensitive to the shorter wavelengths that the UV-curing printers emit. in some experiments I did a while back, it took about 120-150s to expose the film. in theory (have not tried it) you could just cure the normal resin in place as a resist layer, wash uncured resin off in IPA, etch in a compatible etchant, and remove with a strong solvent. in theory, that should be good for 5/5 spacing

>>1377218
cool, cheers
>>1377537
I had asked earlier, was curious about precision, a bit disappointed it didn't read into fractions of a microhenry

>>1377482
so a coin flip kind of thing? two traditional means of generating real (as opposed to pseudo) randomness are Zener diodes and B-E junctions, each reverse-biased. take the (analog) noise, optionally amplify it, run it through a comparator to get binary, feed the binary to a transparent latch, and just turn the latch enable on/off with the switch

>>1377569
when I did this many moons ago, I put my samples in a serial EEPROM with a custom format, because the PIC I was using didn't have enough flash and/or CPU time for FATFS or variable audio formats or any SD card control crap
in short, when the button calling for a sample was pressed and released, I looked up the address/length in the EEPROM's "directory" page, held the flash's nCS low, sent a read command for the sample data, and just streamed bytes out as I needed them until the sample length was exhausted. then raise nCS and wait for the next button press. those bytes were loaded into the PWM generator of the timer, whose output pin was low-pass-filtered and amplified to a speaker. in my case I had one button and samples were pseudo-randomly selected so I combined the button with the nCS line to fit an 8 pin micro
grab the datasheet for the chip you'll be using, and read up on the timer (especially PWM) and the SPI port. you'll need to use a timer interrupt if you want playback frequency to stay constant. good luck

THIS WAY PL0X
>>1377579
>>1377579
>>1377579

>>1377482
>>1377500
https://en.wikipedia.org/wiki/Linear-feedback_shift_register
output can be from any stage of the shift register, clock frequency sets speed, start/stop via enable/disable clock.

>>1371679
why would you do this? it's only going to run while you have the button pressed.

>>1377569
DFPlayer