/diy/ - Do It Yourself

So I have an I2C module that requires pull ups. I want to connect other I2C sensors to the same I2C line from the microcontroller. Will these pull ups interfere with the other sensors ? What do ?

>>2388749
All I2C devices require pullups. Some may have them on the modules, maybe there will even be monolithic pullups on the ICs themselves, others may require external ones. But all the pullup resistors will be on the same SDA and SCL lines, meaning they'll be in parallel. Which you generally don't want. Easiest thing to do is just not include any extra resistors and hope the existing ones won't be too large or too small when together. Personally I'd get rid of any pullup resistors on sensor boards (often I think they can be disabled by cutting a NC jumper) and just have pullups right next to the MCU.

>>2388735
Is this the coomer/electronics thread?

holy shit i could buy a resin printer and use it to expose photoreractive pcbs to make absolutely perfect high detail pcbs with no milling, or gay ass toner transfer that wants me make to kill myself.
is this 500IQ idea or what?

>>2388772
Why not buy a PCB mill and skip the chemical steps? Or a laser engraver and black spraypaint?

>>2388768
They want me to use 4.7KΩ resistors as pull ups for one sensor module. I also want to connect a 1602 I2C LCD module to the same line. Guess I will have to just add it and find out.
>>2388769
What are coomer electronics even ?
>>2388772
Some other guy in 2 threads before this wanted to do this. I dont know what happened to it. He got really deep into it.

Also unrelated question, I have a sensor connected to an arduino. I want to have some sort of ESD protection because this sensor is exposed to the 'outside world'. However the enclosure this sensor is in, is connected to ground too. I was thinking of adding a 220Ω resistor but that detoriates the performance. A capacitor might fuck up timing. Whats a cheap through hole TVS single-side diode that I can use ? Or am I being retarded ? Please help.

>>2388855
>What are coomer electronics even ?
yes

>>2388878
kek

Got a 12V adapter for this modem but they both have prongs. Can I easily connect these somehow by modifying either?

Considering opening the modem up and soldering it in place instead.

>>2388932
i think you might need to cut the male lead off, and solder a female one OR just buy a new one. You might get to use this one somewhere else.

hi, i'm looking to replace my mouse's middleclick microswitch
do you know of any microswitch model i can buy on aliexpress that could fit or do you have any ideas? i have found some square switches but they all seem too tall compared to the one in the pic

>>2388948
yeah, sounds reasonable. might just dig up another one at a thrift shop for $2. or in a nearby electronics dumpster.. thanks

Guys, i have a situation here, so my garage door has a motor with a remote, but i don't know how, sometimes when the neighbor opens theirs, mine opens too (at least i believe it's not random), luckly he gives me a call warning about it, but how can i make sure this doesn't happen? I tried looking in the manual but i didn't found anything to reprogram it to another frequency or something.

The motor is a marantec comfort 150.

>>2389003
>marantec comfort 150
https://archive.org/details/manualzilla-id-7299826/page/20/mode/2up?view=theater

>>2389003
Programming a new code is outlined in the manual, and you need to clone the original master remote to use aftermarket duplicate remotes.

>>2388878
Dude! That's disgusting. Now I have to go stop my kids from playing with mercury. Gross.

At least they can still play with lead paint chips.

>>2388971
It's a mini SMD/SMT tactile switch. Measure its dimensions including the button height.

>>2388932
Yes should be an easy hack. Just make sure the polarities match, sometimes they'll put the positive on the outside.

>>2389013
This anon is right.
Can also check the internal connections match but that's pretty standardized.

>>2389009
The original remote is dar gone, i only have aftermarket ones that have been cloned.

>>2389008
that's not it

>>2389009
It instructs on how to program new coding for the remote, but not for the head unit.

>>2389013
>>2389014
thanks for the info, i'll see what i can find

>>2389018
>>2389025
How deep do you want to go into this? You could disable the receiver on the opener entirely and replace it with a ESP32 or something to trigger the hardwired open/close button with your phone or dedicated diy remote.

>>2389025

presumably you also have a mechanical switch that lets you open the door from the inside.
you can get hundreds of diff models or remote switches that you can connect to a 120V relay.
this will let you bypass the installed tx/rx combo.
remove or short the current antenna to stop false activations.

>>2389043
Noice. Off-the-shelf and simple solution for noobs.

>>2389014
cheers, I'll try to find some schematics on the connectors

>>2389042
>>2389043
Preferably i would keep the head unit as is, simply because my parents can't be asked to use another thing.

I found on the manual how to "program" the remote, but i'm not sure if this changes the channel in rx (looks like some kind of pairing).

>>2389048
This won't work without the original remote or some hackery.

>>2389048
Oh, ask your neighbor to lend you his master remote so you can reprogram your opener/his remote, clone the code to your remote(s), then re-program the master remote back to your neighbor's opener.

>>2389050
The original remote is dead, we're talking about +20yo stuff.

>>2389052
Why would i clone the code of his remote to mine? It's an interference issue, his remote still opens his garage most of the time.

I'd thing that these stuff would have a way of changing channels, it's not clear to me if it changes anything by reprogramming it with the head unit, NOT cloning from remote to remote.

>>2389056
Nevermind. Forget it.

>>2388772
Yes, theoretically you don't even need a resin printer. Just buy a replacement LCD for it, buy an HDMI driver board for it on ebay, and put them in a box with an array of UV LEDs under it.

JUST MAKE SURE YOU BUY AN LCD WITHOUT A REFLECTIVE LAYER ON THE BOTTOM
REFLECTIVE LAYERS ARE REFLECTIVE
THEY DON'T PASS LIGHT

>>2388855
>Some other guy in 2 threads before this wanted to do this. I dont know what happened to it. He got really deep into it.
Yeah that's me.
FUCK
The metallic layer on the back of one of my LCDs could be peeled off, but I cracked the LCD in the process. I'm afraid to try on my remaining LCD, and I can't find an HDMI cable to test the current modification on. I think for my next one I'll tape it down to a piece of glass so I can't bend it. It really is tough to peel off the layer though.

I've been distracting myself programming one of my other etched boards to solder the components into my boost converter board (LM393 dual comparator as a switching IC lmao) but I really should get onto that. Especially since I've hit a brick wall with my capacitive touch keyboard project.

>>2388971
Better to buy name brand high-durability tact switches than ebay garbage. Also just cut the legs off the existing one so you don't delam the pads trying to desolder it.

Bet ya can't guess what I'm doing right now, /ohm/bres.

>>2389103
Isn't that film a polarizing sheet?

>>2389276
Desoldering something you fucked up and delaminating the pad so it's even more fucked up?

>>2389280
No, polarisers are on the front and they aren't reflective.

just put in an order for my custom DAC/MP3 player.

chain is microcontroller -> PCM1796 -> OPA1652 -> TPA6120

I anticipate an SNR of around -114 dB and THD+N of around -100 dB with the volume cranked.

>>2389336
Personally I'd have attempted to get a sufficiently high impedance headphone amp so you didn't need the op-amp.
Or even have gone full class-D. Have you done the quiescent current calculations? Battery life?

I expect you're using DMA to the I2S DAC through the MCU from an SD card? Make sure there's a big bypass cap sufficiently close to that SD card IC, they can be temperamental. At least on a breadboard they are.

>an SNR of around -114 dB
Not sure I follow, are you saying there's noise from the headphones that's 114dB louder than the signal? Did you not mean +114dB SNR? Or am I just retarded and that's how all SNRs are written?

>>2389103
>Better to buy name brand high-durability tact switches than ebay garbage
as long as i don't have to spend €20 for a couple of microswitches.. do you have any brand recommendations?

>>2389377
Not for tactile switches in particular, but I know brands like Omron make good microswitches and relays so they're a good guess. The Kailh brand seen on the microswitches might be enough of a brand name. Hunt on the datasheet for a durability rating. Can probably find recommendations for brand names from other people resoldering switches inside mice on stackexchange or static forums.
The more difficult part will be sourcing the parts from ebay or amazon sellers without getting fakes. Digi-Key or LCSC or Farnell will give you the right parts, but they're not worth buying $10 worth of parts from for shipping reasons, plus customs can be a nightmare with them.

If you haven't done so already, use a Digi-Key or similar parametric search to find components with an identical footprint so you can put a name to the footprint in question and ensure that any other brands will have the right size and shape without having to look in every single datasheet first.

Are SMD electrolytic caps any good ? Do they even exist ? Or should I just stick to THT electrolytic caps and make the other passive components in my PCB SMD ? Please help.

>>2389409
Yes, no, maybe.

>>2389438
Too much information.

>>2388735
>>>/sci/14496074
>>>/sci/14502938

>>2389480

>>2389484
I'm not /sci/ OP

>>2389487
>/sci/
What are you doing hanging out in a place like that, Anon?

>>2389484
kek

>>2389491
AAAAAAAAAAA

>>2389489
looking for 3D emulator software for beginners.

>>2389495
What the fuck is 3D emulator software? Like 3DSMax?

>>2389498
Physics emulator

>>2389500
>Physics
Have you ever been to a Star Trek convention? lmao

>>2389502
no, I haven't.

>>2389516
Can you do the Sulu dance?

>>2389522

>>2389526
no (You)

>>2389531

>>2389533
>examines coin
That ain't no shekel. lmao

>>2389534
Anon came into a coomer electronics thread with a big magnet attached to a truck. I think Anon is a little bit of a silly goose.

>>2389537
Do you know suitable software, or not?

>>2389539
Yes.

>>2389535
>>>/pol/

>>2389537
lmao, >>2389491 actually works, dumbdumb.

>>2389544
No, /pol/ got taken over by masochistic trannies who like to be told they'll never be a woman. You should probably go there or /g/ (g stands for gay).

>>2389547
Star a free energy thread. Call it /FEG/.

>>2389548
you seem to be in the know.

>>2389550
Yes, I saw the homos in dresses invade in real time.

I'm the anon that asked about mountain breeze style ionizers and walton cockroft voltage multipliers quite a bit ago. Here's the result from a while back, thanks for the help /ohm/. I've since then modified it to be more like an electrostatic precipitator, but it's a bit weak for that. Either way my room smells a bit fresher thanks to the trace amounts of ozone and it also collects a bit of dust in the process.

>>2389588
>alton cockroft voltage multipliers

where are the needles, nigger?
you need very sharp points to expel the ions at a faster rate.
in fact, you have to change those needles after a time coz they become blunt as stainless steel molecules are also expelled.

>>2389603
>where are the needles
I filed and sharpened the leads of the resistors. There was a corona discharge visible in complete darkness and it was creating a breeze, a sound and the smell of ozone (if you put your nose directly up to the needles).

>>2389606
dis nigga be snortin needles n sheeit

>>2388772
Like this guy?
https://www.youtube.com/watch?v=RudStbSApdE

>>2389606
It works a lot better if you add some small bundles of carbon fibres. But nice work all the same. I get the feeling that having the positive and negative multiplier ends closer to each other will cause a stronger electric field gradient, and hence more ionisation, but I couldn't say for sure.

i'm trying to make a low power brushless motor controller (5 amps max) with onboard current sensing and i just realized i have no idea how to lay this out.

my thoughts are to do a four layer pcb with:
high voltage signal (12V, 35um 1oz copper)
high voltage ground (12V, 17um 0.5 copper)
low voltage signal (3.3V, 17um 0.5oz copper)
low voltage ground (3.3V, 35um 1oz copper)

i imagine i want separate analog (current sense output) and digital ground paths on the low voltage signal ground that are joined together in one place.

beyond this, i'm lost. i have a suggested layout for the current sense amplifiers but i'm not sure how to route everything else. ground pours on the two ground planes, stitch them together with vias? and then use vias for all ground to IC connections, and vias for routing the low voltage signals from ICs to the low voltage signal layer?

>>2389341
>Personally I'd have attempted to get a sufficiently high impedance headphone amp so you didn't need the op-amp.
I was somewhat limited in my chip selection. I was basing the board off what JLCPCB could source and surface-mount, so my choices converged to the TPA6120.

>Or even have gone full class-D.
I couldn't find any that satisfied my desire for low distortion.

>Have you done the quiescent current calculations? Battery life?
Total power draw is <1 watt (assuming my math is correct). Goal is to power it using an 18650 (I have 2 Samsung 25Rs lying around), which has about 9 watt-hours.

>I expect you're using DMA to the I2S DAC through the MCU from an SD card? Make sure there's a big bypass cap sufficiently close to that SD card IC, they can be temperamental. At least on a breadboard they are.
Yup, got 100uF tantalum and 1uF ceramic caps in parallel as bypass caps, as per manufacturer specs.

>Not sure I follow, are you saying there's noise from the headphones that's 114dB louder than the signal? Did you not mean +114dB SNR? Or am I just retarded and that's how all SNRs are written?
yeah, i meant +114 dB. was still in the noise mindset while writing it out.

>>2389938
PCB layout is a rather complex topic.
This guy has a couple nice videos with experts on it: https://www.youtube.com/watch?v=vALt6Sd9vlY

Hello anons. Retard here. I'm trying to switch 5v with a ~2v signal using a 2n3906. The 5v will be used to drive a pin high on an arduino, so there will be fuck all in way of current. This is a speed dependent application, and analogRead() is too slow, so I need it to be hardware and I need it to be fast.

The problem is, with the 2v signal feeding into the gate of the transistor, It's not fully turning on and I'm only getting about ~2.8v through on what should be the 5v line.

What would be the best way to go about this?

>>2389982
Post circuit.
Assumption: you are using the transistor in common-collector configuration, when it should be common-emitter.

Don't forget the base and collector resistors. Good luck!

>>2389988
>Post circuit.
9000 hours in MS paint, etc. The 100ohm is there at random because it's what I had on hand that doesn't instantly fry the transistor.

>>2389991
...I also drew the transistor backwards. I assure you it's wired correctly.

>>2389938
heat dissipation is probably going to be an issue. set up everything such that high current traces are going to be on the top and bottom layers, then do your stackup.

>>2388735
Get, checkem

>>2389982
what's your Vcc? 2V will read as high if Vcc is 3.3V.

>>2389982
>>2390003
this sounds jank as fuck, but what if you cut the trace going from 5V to Vcc on the MCU and run a jumper to the 3.3V regulated output?

>>2390003
The vcc will be 5v. The 2v signal is from an optical pickup. Normally this wouldnt be a problem because i could just use analogRead, but like i said, too slow. Ive tried googling optical pickups/triggers but all i get is novelty guitar pickup stuff. So not applicable in my case.

>>2390005
I'd try that if it were a one time temporary thing, but I might be building multiples of this project if i can get it to work properly, so I'd like to keep the jank as minimal as possible.

Its for a coil on plug ignition system for some of my old shitboxes if anyone is wondering. Ive got the conversion hardware built, and the software mostly figured out. The circuitry is the major stumbling block and one of my major weaknessess here. I went with optical over mag/hall effect because of cheapness and percieved simplicity.

>>2390011
Can you bias the signal?

>>2389991

your circuit is positively retarded.
even if you flip the collector and emitter, as you said in later post.
you cant use a PNP with a 2V base signal to output 5V.
that's coz to turn the PNP off, you have to decrease the the BE junction to less than 0.6V
so if you apply 0V at the base, you get 5V across BE.
if you apply 2V at the base you get 3V across BE
neither is below 0.6 so the transistor is never off.
what you want is pic related instead.
(the resistors in pic are kinda high, should reduce by 10x)

>>2390014
I possibly could, but i think that would make the MCU more prone to interference and false triggering in a harsh RF environment like running an ignition system with 40,000 someodd volts. I'd like to get as close to a 0-5v squarewave as possible from the optical pickup.

Atleast if I understand what biasing a transistor will actually do. It would mean a more narrow squarewave, wouldnt it? Something like 3-5v instead of 0-5v, yeah?

>>2390017
May want to add a resistor here otherwise the PNP base is essentially floating. Better to tie that high when the NPN base is at GND.
You could also replace these with N and P channel FETs as appropriate.

>>2390018
I meant more biasing the sensor itself. Instead of being 0-2V, your Arduino will read, say, 2-4V, straddling the trigger.

I simulated >>2390019 with https://falstad.com/circuit/ below, to show that it works. This is the "right" way.

>>2390017
>your circuit is positively retarded.
picrel

But yes, I just mocked your circuit up on a breadboard and it works beautifully. ~.1v "off" and 4.99v on. Thank you very much, anon. You're a god send.

>>2390019
Will do. I know firsthand how god damn infuriating floating pins are, esp when you have no clue what you're doing like me.

>>2390022
you might also want to read into direct port access. digitalRead() is stilll slow as fuck.

>>2390024
I was going to use interrupts. But yeah if those still go through digitalread (which i think they do), I'll read up on it now.

>>2390027
interrupts will work. they don't go through digitalRead, they'll physically "interrupt" the processor, hence the name.

>>2389938
You dont need a stack up like that or separate planes for each domain. Just put the high current motor stuff on one side of the board and low power/voltage stuff on another.
If your low voltage circuits are between the power source and the motor driver, the current drawn from the motor will create voltage differentials in the plane the sensitive circuits are on. It likely wont damage anything but will impact accuracy, noise, etc. By moving those circuits out of the way into their own area this isnt a problem as the return currents are somewhere else. This can be done by putting a cut in the ground plane or just putting those circuits on the opposite side of the power source. More extreme is to use an isolated supply in addition to separating the sensitive stuff.
Use differential signaling for current sensing with good cmrr amps. Do that and your analog sensing can be far away while maintaining accuracy.
This is how you approach mixed signal boards in general. You dont need weird star grounds, just understand how current is going to return, avoid clashes and use differential signals.

>>2389991
Is the 5V signal referenced to 5V or GND? Is it floating? Any idea how much current it can source or what its output impedance is? Pic related are the only two good ways to do it, and if your 2V signal isn't floating you can only choose one or the other. If it's already referenced to ground you'll have to use an NPN instead.
Or use a dual inverter with complementary pairs like the other anons were saying I guess. What is that, like a switching sziklai pair or something? The doubling gain from the twin BJTs will definitely help if your signal source can't provide much help.
That or use a MOSFET.

Also 100Ω on the base is too low, use 1-10k or so instead. Consider the minimum hFE and plan accordingly.

>>2390001
>failing a get
>on /diy/
hahahaha
>>2389999
>>2390000

>>2390024
>you might also want to read into direct port access
Straying away from the walled garden of arduino is a fool's dream. First it's a PORTB.OUT, then you're buying a USBasp, then all of a sudden you're coding in vim. A rabbit hole leading only to darkness.
t. /mcg/ retard

>>2390033
>Straying away from the walled garden of arduino is a fool's dream.
i've fallen into the Arduino -> ESP32 -> ??????? trap myself.

Mountain Breeze anon here, I'm now looking to make an electrostatic precipitator which works on the same principles. For this I am going to need to generate 25-30KV steady non-pulsed DC as efficiently as possible. I see a few possibilities of generating this high voltage.

There are those cheap high frequency high voltage transformers rated for 20KV but I don't know shit on how to drive them and I've tried simulating them in kicad but it craps out at higher frequencies. I know that higher frequency = higher output voltage, but how do I design/tune the circuit such that it outputs steady 5KV (which I then can multiply)? I'm guessing I'm going to have to make a diy high voltage meter and make a tuneable high frequency circuit and then just figure it out through trial and error? How would you guys attempt doing this?
Here's one of them, of course no specs or anything:
https://www.aliexpress.com/item/1005003076138849.html

There are also those high frequency neon transformers that output 5KV 30mA. I could buy one of them and rectify and multiply that with high voltage rated diodes and caps.
Here's one of them:
https://www.aliexpress.com/item/4001090859869.html

Also, and idea how much power these high frequency transformers would draw? I'm hoping it would be as efficient as one of those mountain breeze units which simply multiply mains.

>>2390123
If I were designing a HV PSU, I'd include feedback as a part of the oscillator/switcher. In a sensible method that uses a zener/reference diode and voltage divider and optocoupler. That way it's not tuning based, so it won't shift as the caps derate.

Don't necessarily need a voltage multiplier, there are 20kV diodes out there. 50V forward voltage, which shouldn't really matter for you. The caps get kinda expensive though, same for the windings, so I think you are best off with one unless you have like 2 primary windings...

On that note, consider rectifying a slayer exciter! IDK how easy it is to add feedback to them (opto pulling down the transistor base/gate would be my bet), but they're piss-simple HV circuits. Not like it needs to be high power. Could even just modify one of those $1 ebay kits.

>>2390124
>Don't necessarily need a voltage multiplier, there are 20kV diodes out there
Electrostatic precipitators need at minimum 24KV to function well, which is why I'd rather go with a few multiplier stages than have my diodes and caps rated for 30KV. Diodes rated at 30KV start getting really expensive compared to 10-20KV.
I've read that you can put multiple lower value diodes in series but reverse leakage could be a problem and my guess is they would fail early.
>On that note, consider rectifying a slayer exciter!
Ah, that seems pretty awesome and not even too hard to diy entirely. It should be able to generate 5KV which would be more than enough with a few multiplier stages. I've read about the armstrong oscillator before and it seems to be similar.

Thanks I'll have a lot to read up on.

>>2389588
How did you get the copper traces in the PCB visible? Did you tell the manufacturer that you want it that way or did you make your own PCB?

>>2390174
Looks diy to me. The substrate that the copper is adhered to is translucent. The PCB in the pic was single-sided and lacks a solder mask (the green paint on circuit boards) and silkscreen (white text labels).

>>2390174
I made them with the cheapest plain copper PCBs I could get on Aliexpress. In hindsight I could have tinned the traces, but I don't think it matters.

What is the easiest way to add a off the shelf battery management system to an arduino ? I just want to have a 5V supply from the battery. Is it a good idea to run the output through linear regulators ? I made a circuit with an Arduino running from a wall power supply. But now my professor wants it to run from a battery.

>>2390198
USB battery bank.

>>2390198
this >>2390199 is the easiest way.

for something a bit more legit, a battery shield like https://www.amazon.com/energyShield-Basic-Rechargeable-Battery-Arduino/dp/B06VVBRR7H works.

another way, if the circuit is breadboarded, is using a 9V clip with leads, feeding Vin with 9V. something like https://www.amazon.com/DIKAVS-Breadboard-friendly-2-1mm-Barrel-Jack/dp/B074LK7G86/ with https://www.amazon.com/5pack-Battery-2-1mm-Arduino-Corpco/dp/B01AXIEDX8/ would work. The onboard LDOs will take care of the voltage.

>Is it a good idea to run the output through linear regulators ?
No need, the onboard regulator on most Arduino boards should handle everything. Having multiple regulators can tank efficiency and cause undervoltage. The only important thing to remember is battery management systems should you be using rechargeable batteries. A USB battery bank and the above shield include those. A bare LiPo doesn't. Alkaline cells don't need BMS circuitry.

>>2390199
>USB battery bank.

you get way better results if you google ''power bank'' instead

I want to build a cheap pulse oximetry sensor. The idea is to measure how much red and IR light is reflected from skin using LEDs and photodiodes. I'd prefer to have the LED and photodiode pairs be well matched, so on the same chip, but I don't know what such a sensor would be called. Does that exist?

Dedicated pulse oximeter ICs are at least $6 on digikey, so I want to see if I can homebrew one with simple optoelectronics, amplifiers, and DSP. The goal is to get it cheap enough that it could be integrated into a PineTime ($25 open source smartwatch).

Two questions:
First, how do identify an existing led if there's no marking on it? Right now I'm basically stuck at it being 1x1.6mm and shining "not too bright" clear white. I would have posted a picture, but so up close it looks all terrible on my phone. Would measuring it's voltage/amperage get me probably close enough to something similar or am i destined to just try some different LEDs till I find something that's close enough?
Second, does some kind of ic exist in the sense of a smart distributor that starts gate1 with signal 01, gate2 on signal 11 and so on, like a case switch in programming without slamming down a full stm32 and if yes, how are they commonly called?

>>2390328
>smart distributor
Analog switch or mux/demux?

>>2390322
>google
When I looked up that word on Altavista it was defined as a really large number. Try yahooing Excite!.

>>2390327
emitter and detector is the terminology
2 LEDs of different wavelengths and a photodiode.

>>2390341
>on the same chip
That's exactly what I'm trying to avoid

>>2390334
>demuxer
Yeah, honestly that should work just fine for me as well. Maybe a bit of over thinking from my side. Thanks!

>>2390344
Put the LEDs and photodiode into an external finger hinge thingamajig that plugs into your watch's USB port.

>>2390180
I don't think it would matter to much if you tinned them since they're already in the PCB and unlikely to oxidize, dies AliExpress coincidentally sell the PCB you needed?

Hey, /rcg/ here. You guys might have better knowledge with what I'm trying to do. Trying to send a signal from a standard drone receiver through 100ft of ethernet cable. A second cable is meant to send video signal back up to the top. Is this possible or do I need some thicker gauge or something?

>>2390421

Cheers

what components are these? resistors?
what about the big ones?

in any case what (high-level) purpose would the components generally serve, if that is decideable without a wiring diagram?

the pcb-wire coming in from the left is an abouts 60v~ live wire (no idea why it is 60v and not 230-240v).

>>2390444
Take a photo with a light brighter than a candle and wipe the vaseline off of your lens. Sheesh.
>two 100k resistor
>diode
>optocoupler
guess based on Braille

>>2390439
Cables are frequency rated. Cat5 is rated for up to 100MHz, so if you're transmitting beyond that frequency you'll need to use cable that has a higher rating. Look for "sweep tested" spec.

>>2390141
>Diodes rated at 30KV start getting really expensive compared to 10-20KV.
The number of required diodes and caps won't change, just change them from being in a CW multiplier format to being in series.

The hardest part with the feedback will be the large resistors. Power dissipated in the feedback resistors will be V^2/R, and that V^2 is such a large term you'll want literal giga-ohms.
That or see if you can add feedback on the primary side of the coil with a comparator I guess. I think the voltage spikes on the secondary will be directly proportional to smaller spikes on the primary.

>>2390327
Pretty sure you just need to match them against each other after buying them separately. If they're on the same die then trying to stop light from going directly from one to the other would be difficult.

>>2390328
Measure voltage at a reasonably dim brightness, look up how much power that package can take by finding a similar one on digikey or lcsc (power is a factor of temperature tolerance and thermal resistance) and from there figure out what kind of current you can safely push through it for a 10% higher forward voltage.

>>2390439
What sort of signal? If it's analog video or something else high bandwidth, you'll need to care about impedance matching. Good luck hahahahaha

>>2390444
Optocouplers are often as a part of some (high voltage tolerant) isolated power system as a part of the feedback circuit. This tells the switching controller that the isolated output voltage is at a certain level and hence whether to increase or decrease PWM.

I need advice. I want to hide a sensor in a plastic model and have it trigger whenever I touch the model itself, but I rather not do any additional modding on the model itself, so I rather not drill a hole to insert a press button or have the model itself be on top of a press button. I thought about using a photo resistor sensor and stuff it inside the model, but turns out even though the plastic is not thick (1-2mm), it is still dark enough where it isn't working as well as I have hope.

>>2390489
Reed switch or Hall effect sensor? You'd need a magnet to trigger it but minimal modding.

>>2390489
or a tilt sensor which is basically a ball bearing that bridges contacts as it moves.

It annoys the piss out of me that they aren't going to install C50 next. Probably union workers.

>>2390489
Using a phototransistor and a big resistor could make the light sensing part work, though even with a high-pass filter, it would trigger on any fast enough ambient light change. Might be tunable, but I'd lean towards the tilt sensor.

That or a capacitive touch sensor. Putting some copper tape on the inside should be detectable with a mutual capacitance sensor. Pic related.

>>2390489
A capacitive sensor sounds like what you want.

>>2390495
>>2390496
Thanks. I'll look into it to see if I can find a hall/tilt/reed sensor that can actually fit inside my model first.

>>2390506
>>2390515
I have never considered a capacitive sensor, and never considered it can be easily done with some copper tape either. I'll look up that option as well. Thanks.

>>2390520
go with capacitive sensors. easy to implement with most MCUs (Arduino included) or moderately easy to implement with OTS components, and can be entirely hidden within the model.

>>2390477
>>2390462
thankss

>>2390477
>analog video
Yes, analog video on one of cable tethers. Does this require a thicker gauge or something?

>>2390474
The receiver is sending PWM signals to the controller so that should be only a few kHz. Can this be impeded in a long CAT5 cable?

>>2390537
Impedance matching is a complicated topic, which I advise you read up on wikipedia or wherever. Importantly, trying to connect a a signal source with one impedance to a transmission line with a different impedance will result in significant signal loss as you get reflections. The easiest method would be to use coaxial cable of the same impedance as the electronics instead, but it's definitely possible to use an impedance matching network or transformer to get it right for the ethernet cable. And then back again on the other end, presumably. You'll want to start by figuring out the impedance of both sides. Your video feed should have a datasheet you can read for that (chances are it's 50Ω or 75Ω), and your ethernet cable should also have such a rating to read somewhere.
The ham radio general might be more helpful.

>>2390424
>dies AliExpress coincidentally sell the PCB you needed?
Not sure what you mean. Do you mean the size or the layout? No to both, I designed the PCBs in KiCad to fit on 7cm x 10cm copper PCBs and then cut them to size and sanded the corners with a rotary tool like a dremel and then transferred the PCB mask with PCB heat transfer paper and an iron (one for clothes) onto the plain copper PCB and etched away the unneeded copper with ferric chloride. Then I drilled the holes using PCB drill bits for my rotary tool and soldered in the components according to the layout.

>>2390553
Anon thinks the traces are embedded inside the substrate.

>>2390553
Here's what the PCB and schematic looks like in KiCad. KiCad is pretty awesome.

>>2390337
>Altavista
did I accidentally enter a time machine and go back to 2000?

>>2390560
No, you went back to 1995. Everyone is wearing JNCOs and sticking ecstasy pills up their buttholes. Welcome to Hot Topic.
>Narvarna whining in the background

>>2390562
I don't want to have to live through the hell of having a few short years of a relatively peaceful and functioning society before it's ripped away again forever, anon. That sounds like the worst punishment you could possibly bestow on someone.

>>2390564
You think that's bad? Wait until you have to re-live 1996-present (AGAIN) because you ran out of plutonium. lmao

>>2390564
Possessing foreknowledge of coming events and trying to warn people, then watching them just smirk in disbelief and roll their eyes would be difficult to cope with. I'd probably sudoku with fentanyl.

>>2390545
Yikes, that's a lot more than what I was ready for. Thanks anon, will check up /ham/

>>2390558
>traces on same side as tht components
>on a single-sided board
Don't make that mistake again, it worked out fine since all the components were just 2-leaded and so could be flipped, but it's not a good habit for when you need ICs. I made a script that removes all copper layers of pads that are on the same layer as the THT component itself, helps me avoid the pain of realising your etched board won't work without plated through-holes.

>>2390577
Understanding won't be trivial, but once you get the theory you should know enough to buy (or wind) some cheap impedance matching transformers. Laying out those components may not be trivial, because impedance is somewhat a function of distance from the signal line to the ground/return, with any luck you can buy matching transformers that are plug and play and don't need any custom manhattan boarding or what have you.

Looking deeper it seems like all modern ethernet cables regardless of category have 100Ω impedance, so an off the shelf "ethernet over coax balun" may just work if the coax impedance is correct. If not you may be able to modify them, so if they're available and cheap I'd buy a couple to mess around with. That characteristic impedance will change if water gets inside the outer insulation though, so be sure to seal that well with JB weld or whatever.

As for the control signal being transmitted, I'd buy some RS485 transceiver modules or ICs, and see about jamming all your signals into one digital channel (presumably that's what it's like just before the RF modulator in an RF controller) to send down another twisted pair. You'll have two extra twisted pairs, across one I'd probably put a simple signal (probably no need for an RS485 transmitter) for an abort sequence that triggers it to ascend when the UAV no longer receives it.

>>2390588
>Don't make that mistake again, it worked out fine since all the components were just 2-leaded and so could be flipped, but it's not a good habit for when you need ICs.
Yeah I noticed something was off when I already etched it but then I realized it didn't matter for this circuit so I just went with it.

Does anyone know if I get an Altium trial and use it indefinitley by installing in a VM? Or can it just be torrented anywhere? Trying to learn because nobody gives a fuck about kicad for actual jobs.

i have never dealt with fridge electronics. fridge had a sizzle and a pop today, burnning smell, breaker tripped, etc. took a pic at the back. did this shit melt? wtf happened. pls halp my food is melting

>>2390647
There's a capacitor and relay on the side of the compressor. Start there.

>>2390647
?s could be bleeding/filling pipes for the refrigerant loop, doubt it's anything electrical.
It's impossible to say what caused it for now, but I guess it wasn't on a fastblow fuse or else the breaker wouldn't have tripped. I'd first look for a wiring diagram on the inside/back of the fridge, in the manual of the fridge that came with it, or an online service manual. If you can't find one, take photos of things before you disconnect them so you don't get lost.
As for the actual smoke-causing fault, it's possible that something organic just got in and shorted out the mains, or that a capacitor gave up the ghost, or that the motor windings acquiesced too much dust and overheated. An expert could tell based on the olfactory quality of the smoke. The motor start/run cap would be my first bet as they don't necessarily last beyond 10 or 20 years. If it's the motor windings you're fucked. Arguably the compressor could have gotten gummed-up or otherwise stalled the motor, causing it to overheat and shart itself, but compressors should be more bulletproof than HV electrolytic caps any day of the week.

Clean it as you go, it's good practice and should help identifying things.

>>2390647
Those two pipes are there for the filling process and maintenance. If it was leaky you'd smell it, it smells terrible.

>>2390655
This is good advice.
The pump is probably fine, check for caps and other electrical equipment. Also very possible a bug or something got in and shorted stuff out.

What does it do when you plug it back in?
Also you in the US or BR? Guessing the latter.

refrigerators would be so much more enjoyable to work with if everything wasn't fucking foamed and spot welded together. Nightmare taking one apart.

>>2390680
>If it was leaky you'd smell it, it smells terrible.
Not really, freon doesn't smell of anything. If anything, it is a bit sweet like some solvent.
>>2390647
If you have tripped breaker - compressor is probably shot for good. Or it can be just a capacitor. Actually it might be even a capacitor, which is easy to replace. Give your fridge has one that is, because some don't have caps, they have relays.
Can you replace compessor yourself? Not really. Because if you had all tools for HVAC - you wouldn't have asked (or would have asked more direct questions), but if you don't have HVAC tools - you will waste a lot of money buying them
>>2390655
>HV electrolytic caps any day of the week.
They are paper oil caps usually. In good ol days it was PCB oil, nowadays is veg oil.

>>2390655
>>2390653
>>2390680
>>2390684

i checked it out a little more with schematics. the welding gave and the refrigerant leaked, nasty shit on the bottom. god r134 smells awful. Definitely fixable, but this fridge is 18 years old. i think ima get a new one since i can afford. plus this is my first home. thanks everyone.

>>2390686
I've patched up a fridge in the past with jb weld. Ghetto as fuck solution, but it worked and its holding up so who am I to complain.

>>2390686
refrigerant loops are brazed. Fridge is fucked unless you feel like learning how to do that. Also, that smell isn't the r134 itself, it's probably the oil that came out with it.

>>2390558
update to kicad 6 buddy, you are missing out. Add some mounting holes.

>>2390690
i do not feel like it, lol. i am sure that i am paying like 1/4 of my elec bill for this old ass 18 yo one, so i'll probably save more energy on this energy efficeient one i ordered in the long run.

>>2390699
Could arguably buy a replacement compressor + coolant loop, but probably more expensive than a decent 2nd hand fridge.

>>2390686
>nasty shit on the bottom

every fridge has that, BTW, in case you never looked before.
it's the defrost water that accumulates in an evaporating pan.
it gets black and evil looking after a few years.

>>2390704
>black and evil looking
rayciss

>>2388735
Dunno where else to ask and need to run my thoughts:
Do Apple cables really need a fucking handshake before they start charging?! My colleagues keep breaking the pins inside the microUSB charging ports for the props at work and I'm sick of fixing them so I ordered lightning charging ports which don't have any "floating" pins and would be resistant to damage (replacing a cable is easier than replacing a fucking charging port).
Looking at the schematics, it tells me that pin 1 is GND and 5 is power but I only get 1.27v from both pin 7 and 8. I assume the cable is waiting for communication from either line 7 or 8 so it can figure out which way it is inserted and then send 5v to either pin 5 or 4.

Is my assumption correct? If so, here's my question: How the fuck do I insert a handshake into this shit? Do I need a special chip or can I do that with a simple resistor?

>>2390813
Why apples lightnings? Use USB C and get rid of all the micro USB cables.

>>2390892
Because they also broke some USB C ports. I have no Idea how. I've settled for C now. Thanks though Anon!

>>2390897
Are your colleagues actually gorillas with a phone addiction? Do they play Candy Crush all day? Are they Obamaphones?

>>2390901
>Are your colleagues actually gorillas with a phone addiction?
Yes
>Do they play Candy Crush all day?
Among other things
>Are they Obamaphones?
This is Germany, sir, we don't have that here.
Jokes aside, I'm the only one with an Android, everyone else has an iPhone. Ages range from 18 to 25. Well, there's this new guy who's 40 but he hasn't broken anything yet.
It also might be the customers who fiddle around with the props and think that it's a switch or some shit because it's dark and all and then my coworkers break them for good when plugging them in for the night (Did you guess that this is an escape room btw?).

>>2390905
At least they give you something to do so you're not dragging ass all day.
>I'm the only one with an Android, everyone else has an iPhone
High fashion doesn't matter to old farts. Those kids are getting ripped off and they love it.
>Did you guess that this is an escape room btw?
That explains a lot. How did you get into that job? Are you independent?

>>2390913
>At least they give you something to do so you're not dragging ass all day.
Ha. Haha. Ha. ha. Anon, I'm not dragging my ass, without me, the fucking thing would collapse like a card house. My main "job" is being a University Student (B.Sc. Comp Sci) and being a fucking GameMaster is a "Minijob" (jobs up to 450€, tax-exempt). Boss saw that I used to be a system-administrator in my resume and asked me if I could fix this small shit a year ago. Thing has spiraled out beyond good or evil.
>High fashion doesn't matter to old farts. Those kids are getting ripped off and they love it.
I'm not the oldest one either Anon ... I'm 24. I've never owned an iPhone and I've been using Smartphones for 11 years now.
>That explains a lot. How did you get into that job?
My brother wanted to get a job, got into there, I needed a new job because rona had eaten into my savings, so he got me that job as well.
>Are you independent?
No, see above. I work there. As a GameMaster. I also fix shit there on the side because I'm a tech-dude. But I might go indipendent. I have something else lining up for me so I hope I can double dip, fixing his rooms and doing my other job.

>>2390924
>Ha. Haha. Ha. ha.
Sorry, Anon. Sounds stressful, but it's a great opportunity to get experience doing cool stuff without spending your own money.
>I'm 24. I've never owned an iPhone and I've been using Smartphones for 11 years now.
Happy to hear you didn't fall for the hype. I've hated Apple since the time prior to my conception.
Anyway, best of luck with your schooling and future opportunities, Anon.

>>2390932
>Sorry, Anon. Sounds stressful, but it's a great opportunity to get experience doing cool stuff without spending your own money.
It is. But I learned A LOT about /ohm/ and shit. And my boss is paying for all tools and also materials that go into repair and things I build for the room. I'm currently also building another riddle completely from scratch that the people who originally built the room fucked up.
>I've hated Apple since the time prior to my conception.
Based
>Anyway, best of luck with your schooling and future opportunities, Anon.
Thanks!

Dipshit here (this guy >>2391025).

I have a 10 to 4 BCD input that I built with diodes, but the signals are 12v and I cannot let 12v touch an arduino so I'm using an opto-coupler to keep the arduino safe(ish). But how the fuck do I actually get the signals from the opto-coupler to the arduino? I have two configurations but everyone online is only mentioning B, but I would rather want to do A because that would really clean up my code.

What do? I can post more info if needed.

>>2391140
You can use four voltage dividers instead of the opto thing. But how does it clean up your code to use A? And anyway, to use B you would use the built-in pullups instead of that mess.

>>2391151
>You can use four voltage dividers instead of the opto thing.
Probably, but I "have" to use the opto-couplers. We already had some arduinos burn up because the people who originally built this room are backwards assed savages. Their solution to stop one device literally eating arduinos after a single use was to lay another ground line directly to that device. It wasn't in their plans or anything, they came up with that out of nowhere.
>But how does it clean up your code to use A?
Long ass description:
I assume that I will have to pull the inputs of the arduino high and sense a voltage drop with it. But to understand why that would fuck up my code you need to know the following:
The goal of the game is to press buttons within a timeframe as they light up. Lets say I have 4 leds, and 4 buttons (to simplifiy, in acutal, it's 10 leds, one inside one of 10 buttons each). The leds and buttons are coded bitwise into a variable each (lets say integers). When none of the leds are on, the status of that variable would look like this: L0000 and the variable for the buttons like this B0000. If I want to set one of the leds to flip on, I use a function that flips one of the leds, let's say L0010. From then on, the arduino is polling the input until the timer is up. If the button var matches the led var (L0010 = B0010), you pass and the next round starts. If they don't match, you start over. If my assumption is correct and the inputs indeed are high, the button var would look like this: B1101 instead and I would have to run an inverter. Nothing too hard, but I would have to check for it all the time and it would cost some cycles every time I would have to convert. I know, I'm bitching on a very high level, but I'm trying to hit a moving target (I didn't even know how to hold a soldering iron on march 2nd this year).

I can compile you the entire working mechanism if you want, but it will take a while.

>>2391173
>if you want

lol. you took care of my interest in that wall of text. I'm sure someone will come along who can help. have a nice day.

>>2391173
Sounds like "Simon".
https://github.com/urish/tiny-simon/blob/master/tiny-simon.ino

>>2391140
>>2391173
You can have the optocouplers either common-emitter or common-collector, doesn't matter since the phototransistors have a floating base. But whichever way you go, you'll need pull-up/pull-down resistors respectively. Circuit A would work with pull-down resistors.

Also the arduino may have a register you can set to invert input logic levels. I know my MCU has one to invert output logic levels, not sure if it works on inputs too. Either way I bet some demorgan's laws could help you clean up the logic. Not like you'll be pressed for processor time.

Additionally, it's an option to flip the opto LED circuit so they're turned off only when the button is pressed. Though I'd definitely go common-collector before resorting to that.

t. starting another unskilled retail job tomorrow

How would an AC motor behave if the mains power is stepped sine wave rather than a true sine?

>>2391271
You won't notice any difference.

>>2391271
Probably a bit more coil whine, especially at higher harmonics. If it's an asynchronous motor it may run a bit faster or slower, may have somewhat higher or lower torque. The torque itself may be a tad stepped. Make sure that any inverter you try to power a motor with is rated for significantly inductive loads, PFC correction caps would likely be recommended to drop the required maximum VA.

Say I have regular 120v mains power.
And I want to use an isolation transformer.
Can that transformer provide meaningful surge or even lightning protection?
How far can I space the two sides apart?
Insulation?
And could the isolated side be grounded while still providing surge protection?
Would it be just it's own grounding rod or can I also tie it to the main grounding rod?

>>2391383
>that transformer provide meaningful surge or even lightning protection
It will block transients to some extent, but you'd likely be better off with a dedicated mains filter. Common mode suppression choke, normal mode chokes, X2 and Y caps, all that. And a MOV and fuse/polyfuse/PTC thermistor. Maybe even a mechanical breaker if you want, or one of those surge protector outlet things. But I'm unfamiliar with lightning and surge protection so take that with some salt.
>How far can I space the two sides apart?
Depends on the geometry of the transformer. It's somewhat common to have [] or toroidal core transformers where the primary and secondary are on opposite sides from one another, which could get you maybe a cm or two of isolation, but EI cores are more common and they only have a mm or two of isolation, assuming the windings aren't stacked. I wouldn't worry much.
>insulation
Transformer wire is enamelled, though it's definitely common for there to be paper or mica or plastic insulators between the primary and secondary windings. Can probably expect a few thousand volts of isolation across a good insulator even if the windings are stacked, though that doesn't speak for creepage through any gaps or around the edge. Laminated iron core transformers are also electrically conductive, so I guess I'd lean towards a toroidal core.
>grounding
There's reasons for and against that. I think it's generally safer without coupling it to the other mains in the event of a surge. Maybe hook up an RCD to its neutral to have a ground lug on the outlet without actually grounding it, should be safer that way.
All grounding rods are equal, if lightning pushes your ground potential about it's going to manifest itself on all ground rods nearby.

>>2391399
>dedicated mains filter
not him, but I'm having difficulty remembering these from school..
These aren't Hi-Q 50Hz Bandpass notch filters, are?

>>2391399
Could an isolation transformer be used with a pool pump?

>>2391410
>These aren't Hi-Q 50Hz Bandpass notch filters, are?
No, they're designed to prevent high frequency noise getting from the appliance in question out into the mains to interfere with other appliances, or from coming from other appliances to interfere with it. Just a few chokes and caps, nothing tuned.

>>2391423
Yes, so long as the transformer is rated for the VA. If the VA are significantly higher than the W rating, you may want to add a capacitor for power-factor reasons. Not a bad idea to use an isolation transformer considering the possibility of moisture ingress, though personally I'd rely on an RCD instead. Less chance of electrocution that way, assuming you dangle a ground wire into the pool for it to trigger off.

>>2391451
>nothing tuned.
ok, that makes sense.
LP, I assume.

>>2389850
shieet that looks pretty tempting. it's better than milling for traces, but milling has several huge advantages such as you can also make super precies tht holes and slots, not to mention alignment holes for easy making of double sided pcbs
fuck man i wish chinks had 2 day shipping for like $5, the only reason i even bother with making pcbs at home is i don't want to wait 3 weeks every time i make a new pcb design

>>2391264
Thanks Anon! How do I calculate the value for the pulldown resistors?
>t. starting another unskilled retail job tomorrow
Fuuuuuuuuuuuuuuuuuuuuck! I know the feel bro, I feel with you. TAKE MY POWWWEEEEEEER.

>>2391668
>How do I calculate the value for the pulldown resistors?

make them 10x lower than the input resistance of the thing you're gonna connect to.
if you have no idea what that is, 10K-100K will work 99% of the time.

>>2391673
Thanks, got it!

What's the cheapest CNC I can make 2 layer PCBs on? I feel like I waste so much time revising my designs before sending them off to China and having to wait a week to see if I messed up...
Can they do holes?
Can they do vias?

>>2391856
It's not worth it, trust me. Unless you want the shittiest oversized tracks, you're better soldering breadboards.

If you do go that route, make sure you get a decent spindle since anything underpowered will break your V bits.

>>2391866
That >route<
Kek

>>2390503
what makes you think those 0603 caps are all the same value
what if c49 is 100n and c50 is 10u? I'd think they have loaded up the cart with all resistor values the same from one tape, since there are no markings on them.

I'll be running some electronics off a 7805 voltage regulator inside a relatively noisy environment. lots of RF around it and (potential) inductive voltage spikes in the 10-18v supply itself. What sort of filtering should I be looking at to make sure my circuits are atleast somewhat protected? a small filtering cap?

>>2391856
You're probably better off going the etching route. There's tons of ways to add the resist layer to a board. Toner transfer, lasering, hand painting, etc.

For a fun robotic arm project, do you think it would be better to attempt to build the arm from scratch? Or buy one of those kits on amazon?

>>2391856
>2 layer
it'll be a pain in the ass to mill. i'd recommend etching, then drilling and pressing in and soldering rivets like https://www.amazon.com/Quality-Elements-Copper-Double-sided-circuit/dp/B015CV377O/ for vias.

>>2392067
you can get great specs if you print and design your own. it'll take a bit of work to do your own design, though. faster alternatives are open-source arms and the Amazon kits.

>>2391926
depends on the specs of your components. if they're tolerant of these spikes, who gives a shit. otherwise just add a shit ton of filtering caps and maybe ESD protection diodes to protect against surges.

RF shouldn't be a problem unless you're using RF/high frequency components, in which case careful PCB design is required. if you're just annoyed about RF noise on an analog line, you can use a low-pass filter to get rid of most of the static.

>>2391856
any chink wooden router can mill pcbs, double sided are pretty easy as well, you just have to mill two alignment holes

>>2392096
>if you're just annoyed about RF noise on an analog line, you can use a low-pass filter to get rid

RF doesnt care about filters: it hippity-hops over them like hare over grass.
only a metal cage can stop its flittery flight.

tried soldering balled chip today
hated it
never again

>>2392102
Sounds like you own one yourself, what model would you recommend?
>>2392094
I fell out of love with quemical processes when I took up screen printing but I could give it a shot, do you think I'd get better results than a cheap mill?

>>2391926
You mention lasering, this sounds precise and not something I can fuck up.
Can you show me an example?

>>2392124
just get any of the cheap chinkshit ones
or if you want to be fancy buy wegster

>>2391856
Holes will not be plated, which put a significant restriction on board design. Vias can be made by filling and hammering wire or dedicated rivets into small holes. If you look about on youtube you may find a plug-and-play one but it's not going to be particularly cheap. Bonus cost for a springloaded bit to carve away soldermask.
Compare that to photoexposure through LCD lighbox (>>2389103) and laser ablation etching setups, see what stands out at you. If you can buy a 2nd hand broken MSLA printer and don't mind etching then fucking go for it.

Mills can get good results, I've seen people do QFPs with them if they're well tuned, but that relies on the carriage being sufficiently rigid and the V-bit being sufficiently sharp. Don't think you're going to get that rigidity out of the box with a desktop CNC mill, but the really small ones might be good. Bed height calibration is a must last I remember.

>>2392126
https://youtu.be/RuSg7-hMaQg
He's only doing 1-sided though, and he's using a laser powerful enough to punch through the FR4 itself for holes after multiple passes. If there's copper on both sides this won't work. A lower power laser or 4-sides can definitely work, for which I'd run a laser pass just marking the drill marks on the spray-paint in order to drill them in a drill press. You can then use laser alignment (just jogging the thing around with an indicator coming out the tip) to line it up using those holes to get it in the right place for both sides. That alignment may not be sufficient for dense vias, but locating dowel pins should work sufficiently to that end. Put the thing on a more rigid mill-style base if you're worried, not like the laser will punch through to the other side either way.
Getting solder mask out of it is a bonus.

Personally, I think an LCD lightbox is even more versatile, though you need to dabble in the particular chemicals like photoresists and developers. Gets sticky too. Lasering is just etchant and rattlecan/airbrush.

>>2391926
7805s only have ripple rejection around 100-120Hz for rectified mains, so don't expect much in the way of high-frequency attenuation. If you want that, then a BJT with a high gain-bandwidth product making a capacitance multiplier before the 7805 would definitely be my recommendation, to say nothing of the normal recommendations of chokes and shielding and restricted loop sizes. Optically isolate long distance signals if you're really in need of noise isolation, though twisted pair wirings should be good, particularly when used with differential signalling like RS485. A low enough termination resistor with a regular signal+ground setup might be fine enough.

>>2392122
If you're not doing it with a reflow oven/hotplate, you've made a wrong turn somewhere. Hot air on a BGA sound like a fucking pain.

>>2392148
>If you're not doing it with a reflow oven/hotplate
i don't have an oven, only hot air, plus i watched some luis videos and he always uses hot air for bga

>>2392126
>>2392147
https://youtu.be/sHO_rCXzI3M
Here's a guy that does mill with solder mask with the spring-loaded bit I mentioned. He also uses those hollow rivets, and does a tinning solution. I tin some of my boards using a tinning solution, see NurdRage's video on the topic for more on it. I add a little bit of sodium bicarb to the solution to fill my container with CO2 before loosely closing the lid, prevents O2 in the air from further oxidising the stannous chloride so it lasts longer.

>>2392149
>i watched some luis videos and he always uses hot air for bga
That's because modern appliances have components on both sides, so you can't use a hotplate, and even an oven won't work without a lot of preparation. You'd need to somehow protect the components on the bottom from reflowing and moving with the weight of the board atop them, and protect all the components that can't handle soldering heat to all sides, like electrolytic capacitors and the like. Hot air isn't used because it's good, but because it's the only option you have.

It can be ok when assembling if all your BGAs are on one side, since you can use a hotplate to do all those and hand-solder/hot air the rest. No clue how crapple computers are actually manufactured though. Directed infrared reflow maybe.

GPU oven reflowers are a special breed.

i ordered this camera so i can post pretty pcb pics here for you frens
i really hope it's not shit
>b-but anon why vga
it was the cheapest model

>>2390553
Yes this is what I was asking, thanks for answering

>>2392154
Why don't you just use the HDMI port on it?

>>2392172
i will of course, putting vga on it was retarded in the first place
it even has remote control, fucking why, who needs to remotely control that shit

>>2392174
>who needs to remotely control that shit
https://www.youtube.com/watch?v=sTsc9UIz5VI

>>2392174
>who needs to remotely control that shit

so you can zoom in on the money-shot without leaving the bed.

What retard decided to make black the hot in AC power in most of North America when black is usually ground in DC?

And why do we put up with this stupidity?

When are we going to come up with a new colour for hot?

>>2392295
green = ground because they both begin with g
white = neutral because Switzerland
black = danger because corn pops
Hope this helps.

>>2392295
Just use red, blue and yellow-green like any civilized nation.

>>2392306
Mauritius? lmao

I have this 72V battery charger for my ebike that committed sudoku. The battery has a BMS so I'm pretty sure this thing just had to shit out 84V and turn on a green LED when the current stopped, so likely it's barely more complicated than a transformer and full bridge rectifier. But NTC2 near the power input leads, which was a through-hole disc component similar in shape to the green RV1 next to it, though smaller, exploded. Any guesses what this is and if it'll be good to go if I can replace it? I also fished a dead spider out of there so it might be the spider shorted something and NTC1 sacrificed itself.

NTC2 is the diode-looking glass package by the transformer, by the way.

>>2392498
I mixed up the numbers between NTC1 and 2, sorry

>>2392498
>NTC2 is the diode-looking glass package by the transformer, by the way.
I think that's a thermal fuse and the thing that sploded might have been a MOV.

>>2392503
Oops, yeah NTC1 is a diode. Between the two smoothing caps is the thermal fuse.

>>2392503
>>2392504
The green thing has refdes RV1 hidden behind it, so from that, I'd assume that's the MOV. But I don't really do any AC stuff so I don't know what the refdes conventions are.

You really think NTC1 is a diode? I don't see any polarity markings.

I also checked continuity from NTC1 to the input and read OC.

>>2392509
Here's the guts of NTC1. It's sort of a dull, dark, thick material. I'm inclined to say ceramic.

>>2392511
New theory: negative temperature coefficient thermistor. I can't say whether that's a typical package for one, but apparently there are NTCs in diode packages to explain NTC2.
http://www.tndz.cn/En/ProductView.Asp?ID=1241
I don't see anything about thermal protection, though.

>>2392509
>You really think NTC1 is a diode?
No, you're right. It's a NTC in a glass package.
>>2392517
Yep. My bad.

>>2392504
>Between the two smoothing caps is the thermal fuse.

that looks more like a fusible resistor.
if the NTC blew up from over-exertion, then that's likely also dead.
you def need more info about the NTC, like a schematic, to be able to replace parts.

>>2392498
It's a thermistor with negative thermal coefficient.
They're usually used to limit inrush current when you first plug the thing in so your breakers don't trip from charging the input caps quickly. When in use the thermistor heats up, just like any resistor, and its resistance drops (because it's NTC) so the current limit increases.

hi anons
i'm in the middle of a diy keyboard project and i have an idea i want to implement
i want to use this usb hub inside the keyboard case, and i'd like to know if i could split one of the usb3.0 connectors and connect it to the keyboard pcb
is this at all feasible? i'd imagine i have to hire a repair specialist with a microscope like you see in pcb repair videos, but i'd like to know from the electronics side, because i have no idea how usb daisychaining works, especially from 3.0 to 2.0

>>2392965
One port would be dedicated to your keyboard, and connected with small gauge wires soldered internally so you don't have a USB plug taking up space.

newfag here. I bought USB-C PD trigger and am using it to power a 60W delta fan. The fan controller I'm using uses the usual PC 4-pin fan headers.
I need to connect the wires to the screw terminals on the PD trigger. What crimped connectors are good for this? It's 12V@5A
Or should I desolder the screw terminals and put something else on (pcb holes allowing)?

>>2393079
ferrules

>>2393079
I see. can i split the voltage pins in parallel to power some LEDs?

>>2393085
nice
that works

It's me again. I followed your advice about replacing the jfet amplifier stage with an opamp implementation. I did some studying on opamps and also got ltspice, initially to model guitar pickups but it's proved more useful for more than that, particularly because I hate math.
Also, I wish anime girls would come and fix my ugly circuits.

After looking at various schematics of opamp input stages, I managed to come up with my version of the amplifier stage. I managed to sort of match the frequency response, as well as input and output impedances. Once in the circuit, the freq response curve is exactly the same, although with the opamp it's about 8dB louder.
In the original circuit, varying the gain also changes the bias of the jfet as well as the low end response, which I could not accomplish with my rudimentary knowledge. I do get some clipping with the opamp at higher gain as with the original circuit, but it's not the same. I can post pictures of my simulations

Original circuit in ltspice.
I also saw something in a video about buffering vref which someone here also mentioned I could do, but I don't understand what the benefit here would be.

I need to measure force over time for punches, kicks, and jumps. Ideally I would be able to display the results directly on my computer.
What kind of sensor or similar can I use to achieve this? I've looked around on Mouser et c but I get confused by all of the different options.

>>2393237
>to measure force

https://en.wikipedia.org/wiki/Strain_gauge

>>2393237
https://www.ebay.com/itm/231450467623?hash=item35e385dd27:g:PDwAAOSw6EJhp4zq
piezoelectric discs will generate their own voltages when increasing pressure is applied.

>>2393237
Polyurethane foam inside of a sealed rubber membrane with an air pressure sensor on an output valve, with a check valve to let the foam expand with air again.

How would I make a raspberry pi or any sound playing device play Croatian nationalistic music when ever it hears Serbian?

>>2393278
>raspberry pi
get nvidia jetson
and work through the ai code samples

Do the XTAL1 and XTAL2 of the microcontroller and the crytal have to be a differential pair ?

>>2388735
Those girls are not wearing the proper PPE.

>>2393319
>Do the XTAL1 and XTAL2 of the microcontroller
whats this?
is this 2 phase clock fetch execute cycle?
so it not a multiple T state operation?

>>2393088
Yes. Just make sure you use a current limiting resistor suitable to the voltage, and don't short anything. If you know what you're doing you could use zener/LED thresholds to get different LEDs on at different voltages, which could be neat.

>>2393113
>>2393116
Just so you know, the circuit you've ended up with is designed as a single-supply circuit (audio connector's ground is the lowest voltage).

I'd personally design it to be split-supply (audio connector's ground is halfway between). This basically means connecting the audio input grounds to the node between R6/R7 (Vref), and doing away with all the high-pass caps (C1, C2, maybe more) as doing so will reduce or even eliminate startup pop (assuming you mirror C4 up to the 9V rail). This also usually means redefining the "ground rail" triangular symbol to be that middle rail, running the ICs off ±4.5V.
It's less conventional with guitar circuits, but I prefer it. The one downside (I think) is that it means you can't use the audio jack's switch to disable power.

Also consider adding node labels to your digram, Vref, V+ and V- are ones I often use, helps get rid of long connecting wires. Wish the guys making the Silicon Chip magazine would catch on, picrel. Works just like how connecting them with the ground symbols does.

>>2393122
Buffering Vref basically means you can use far smaller caps on it, especially if the current going in/out Vref is high. Maybe 10nF or whatever, with a battery supply you can maybe even ditch the caps entirely. For a low-current circuit like this it probably doesn't matter., but if you have an extra op-amp from a quad/dual IC I'd do it anyway.

>>2393237
How do you want to mount the force sensor, ideally? Most of the time a few strain gauges on a load cell is the best method, but it really depends on your geometry. Pressure-sensor-based may be a bit too sensitive to changes in ambient pressure, though I guess a small vent hole would act like a high-pass filter.

>>2393319
You're not using it for a crystal I take it? Probably not, they're usually normal GPIOs, especially in a modern chip with an internal oscillator, doubly so if the pin-count is low enough to need the extra ones. Look at the datasheet, there should be a description of all the pins and what they can be used for. Here on the ATtiny1614 the TOSC1 and TOSC2 pins can clearly be seen to also act as parts of the USART hardware, and the timer A, and the analog comparators. Less clearly it can be used as general purpose IO.

If you are using it with a crystal, then one pin will be acting as an input and the other will be acting as an output. Basically like putting a crystal across a schmitt inverter. But that's only if you have the oscillator enabled.

What micro?

how do i solder solid core hookup wire to solder lugs like pic related and the more rectangular type? specifically, is there a preferred way to position the wire? any advice on the soldering itself is also welcome

>>2393325
they are the 2 pins you connect the external crystal to.
>>2393330
AtMega1284, trying to make my own 'arduino' out of it. I just want to know if the tracks from the microcontroller to the crystal have to be a differential pair / length matched ?

>>2393341
What i usually do
>Tin the soldering iron
>Tin the solder lug
>Tin the solder wire
>Put the soldering iron on the solder lug with some extra solder
>Touch the tinned wires to the solder lug and solder
>Wait for a few seconds

>>2393346
>they are the 2 pins you connect the external crystal to.
oh, its input pins....
Then, there is only one xtal, not 2.
was nostalgic drifting back to ye old 6502 beginnings.

>>2393346
>AtMega1284
Are those even in stock? That's a pretty high-spec machine as far as 8-bit micros go. I'd lean towards the more modern but slightly lower spec model that is the 4808/4809, or else jump to a 16 or 32-bit micro. 520kB of RAM on an ESP32 for like $2 man, that's fucking value, plus its castellated vias are kinda easy to solder. But that's just me. Maybe you're one of those "tht parts only" kinda guys who loves sockets and has never heard of ICSP.
As for the XTAL pins, they aren't labelled as PB1/PC5/etc. and the description for the pins only mentions the crystal, so I'm 99% sure they can only be used for the crystal. Not as GPIOs.
>XTAL1: Input to the inverting Oscillator amplifier and input to the internal clock operating circuit
>XTAL2: Output from the inverting Oscillator amplifier
Pic related is an inverting oscillator amplifier. It has its own internal RC oscillator so you don't HAVE to use the crystal, but you'll be limited to 8MHz instead of 20MHz.

Also note that getting a bootloader onto the micro (the one from MightyCore) will require an external programmer or using an arduino as ISP. I'd program it through a USBasp instead of through the bootloader either way, because not having a USB-to-UART module on the board saves space. Sure serial printing can help debugging noob codes, but once you're caring about discrete clock cycles you find that the delays and buffering involved with the serial printing become unusable. If you actually cared about debugging you'd use Atmel Studio or MPLAB X and use a proper programmer with in-circuit debugging. IIRC there's a GNU licensed tool or two that does debugging for AVRs, designed to work alongside AVR-GCC and AVRdude, but I can't say whether it will be good.
Also, external USB-to-UART cable things are a dime a dozen and pretty handy, be sure to buy or make one that supports switching between 5V and 3.3V.

Any particular features you want to shove on the dev-board?
>see also:
>>>/diy/mcg

>>2393341
I loop the wire through for more mechanical stability, and it makes them easier to solder without helping hands, though it does make them harder to desolder. Depends on what you value.

>>2393416
Yeah so far I have 2 USARTs (because it has 2 ports like the AtMega2560), 1 I2C port with 5x3pin headers, ISP header for programming (you can use those USARTs for programming too, they have a DTR pin), + 2 SPI headers (one will be used my the SD card). Down the line I want to add VGA support and add a PS2/USB keyboard to it. It has enough memory to support graphics. It should have onboard RTC, onboard EEPROM (I2C). I am looking at adding a seperate header for the PWM pins too. All in all, its just a project where you throw all that you have onto a board and see if you can make it work. I have 2 of these chips, have had them for 4 years with 0 use. I saw people make BASIC computers with them and such. I havent fucked around with pico power but i want to. I even saw these camera modules you can connect to microcontrollers and do some very basic signal processing so thats on the checklist too. I will probably have breakout pins with headers that let you use the pins as you want.

I have made boards for the AtMega328 before with relative success so i dont think this one will be that hard. I am still using the Arduino IDE to program all this because i fdont know shit about programming embedded stuff. Is there any place I can start ? I know C++ pretty well so its been easy to just write a program in C++ and let it run. I have been told that Arduino code is very very badly optimized.

http://www.china-zhongdi.com/product/36.html or http://www.china-zhongdi.com/product/38.html

>>2388735
Am this retard again >>2391025 and this time I need to snoop a RS485 connection between an arduino and a Laptop (adapter). Can I just hook up addtitional lines into the same connector and lead them to another arduino that has SoftwareSerial running and just reads what each line sends? Anything that I should look out for?

>>2393329
Any other considerations to a split supply, assuming I don't buffer vref? Large caps, current draw, etc. Why do so many effects units use bias instead of split supply?

>>2393526
https://jheyman.github.io/blog/pages/RS485Sniffer/

>>2393502
Look at the tips you can find locally for cheap before buying a soldering station.
I had a soldering station similar to the second one and I had to sell it because I couldn't source any tips locally and buying from overseas ends up being 6-7$ per tip after taxes.
Generally those screw in tips are harder to find and more expensive, the simple straight ones that lock in with a set screw are easier to find.

>>2393570
Thanks!

>>2393588
no problem, fren.

I have a 12V DC plug as a power source in a system and I'm trying to integrate PoE by ORing the two power supplies. The presence of the DC jack shuts off the PoE controller via an optocoupler, but I need some sort of current blocking in point A to stop the PoE from shutting itself off and to prevent any current from flowing towards the DC jack. I can't use a Schottky due to current leak and a normal diode has too much of voltage drop. Is there an easy replacement for a diode here? I was thinking of using two back to back PMOS.

>>2393612
Transistors, a relay with the PoE on the NC contacts, or a switched DC barrel jack.

>>2388735
>Principles (by increasing skill level):
>Mims III, Getting Started in Electronics
>Geier, How to Diagnose & Fix Everything Electronic
>Kybett & Boysen, All New Electronics Self-Teaching Guide
>Scherz & Monk, Practical Electronics for Inventors (arguably has minor issues with mains grounding)
>Horowitz and Hill, The Art of Electronics
how much does everyone agree on these? downloading them now anyway.
I took some EE in college even though it wasn't my major but I feel like I learned absolutely nothing practical that I can use in my day-to-day life as a homeowner and tech consoomer. Just how to find out how much current is going across some wire in an abstract circuit diagram and blah blah blah.

hi /diy/ I have a mobile phone with a smashed screen. I've removed it and have a new one to put on. Is there anything to think about when gluing the screen on? Does it matter if there's a couple of spaces?

>>2393612
>I can't use a Schottky due to current leak
Couldn't you just add a resistor to ground to shunt leakage current and keep voltage below switching threshold? MOSFETs will also leak.
>>2393637
If you're interested in practical knowledge there is no substitute, but to actually do practical things instead of reading too much books. Schools obviously won't teach you practical things, because it's up to you to apply theoretical knowledge to practical problems.

>>2393637
>I learned absolutely nothing practical that I can use in my day-to-day life as a homeowner and tech consoomer

then forget all the books on the list.
all you need is youtube and a question.
''how to connect my jingle-jangle so it turns on when the razzulator reaches peak roundedness?"

>>2393637
put the books down and pick up a soldering iron

>>2393502

two equally dumb choices.
you want to buy from a company that will be around in 30 years to sell you replacement tips, heating elements, etc.
like Weller, who's been operating since 1952 (that's almost a millennium in metric)

>>2393637
All of my theoretical knowledge is directly related to the circuits I work with. In other words, I don't bother learning anything if there's no practical application I care about first. For example, I didn't know what impedance or bias was before I wanted to build audio effects circuits.

>>2393694
>you want to buy from a company that will be around in 30 years to sell you replacement tips, heating elements, etc.
autistic consideration that is completely irrelevant unless youre in a highly critical industry, and even then.
if an iron shits the bed you trash it and buy another.

as for the comfort of the iron, that is very relevant and worth paying a premium for.

>>2393502
how much are you paying for this chinesium? buying an iron is like buying a computer. you are probably gonna use it a lot, and theres nothing more frustrating than an iron that sucks. id avoid chinesium products for test gear cos youre gonna be using it everyday. chinesium flash chips, chinesium MCUs, chinesium resistors, thats all fine cos if they become completely unusuable you can just trash them and not feel bad about it. but for irons, get something decent with an address you can return broken shit to.

>>2393711
>if an iron shits the bed you trash it and buy another.

good point.
when the ashtrays on my Porsche fill up, i just junk it and buy a new one.

am i stupid or does dragging a "group" of components in kicad pcbnew not work? i can select an ungrouped component, use hotkey D, and drag just fine. when i try to do it with a group, it doesn't do anything?

>>2393443
>VGA support and add a PS2/USB keyboard to it
That's the sort of thing I'd do only as an addon board. Not really sure if I like the whole "shield" thing with the traditional Uno, but stackable I2C headers or a few different SPI headers with different CS lines sound neat and compact.

>am still using the Arduino IDE to program all this because i fdont know shit about programming embedded stuff
Same for me up until recently.
>Is there any place I can start ?
I jumped ship to AVR-gcc, or at least I tried to, a week or so ago following this tutorial:
>https://www.pololu.com/docs/0J61/6.3
Just changing the mega32u4 in the Makefile for my chosen MCU, and changing the pin assignments to the arduino nano's (read the datasheet for the PB2 etc. pinout). I also needed to change the AVRdude command in the Makefile to use my chosen programmer, should be fairly trivial to change it to use "arduino as ISP" as a programmer. Not sure about programming via bootloader. Worked a treat right out of the gate.
>I have been told that Arduino code is very very badly optimized.
Honestly, if you're doing direct port manipulation and shitty libraries, the resulting code is barely going to be a few clock cycles longer. It just checks for serial and a few other things that aren't included in the standard main.c with the while(1); etc.

>>2393562
As I said, using an audio jack as a switch is handy, but more I think it's just tradition. It's simpler with smaller effects too, which was the main case when transistors (or valves) were expensive. The more complicated the circuit is, the more high-pass caps you'll have, and the more you'd benefit from a split supply.

what battery chemistry has the highest charge rating? I have a solar system with a limited sun window and I want to be able to dump as many amps as possible into a battery store in that time, what should I go for?

>>2393838
Why not just more batteries in parallel?

>>2393502
Buy a T12 station or Pinecil or Aixun T3A. TS100/TS80 are acceptable too. Open source hardware meme irons are the way to go.

>>2393584
>>2393697
90% sure the tips in that one are common non-cartridge Hakko clones, the things you buy 5 for a dollar on alibay. Elements should be similar too.

>>2393656
Read and watch a bunch of tutorials for it to ensure you understand exactly what you're doing. That or buy a few extra preemtively.

>>2393838
Highest charge current? Batteries are often rated with a C rating, representing how many inverse hours it would take to empty/fill them at max current. For example, a 25Ah battery rated at 8C can be discharged at a maximum of 200A, taking seven and a half minutes to do so. Usually charge ratings are a lot lower than discharge ratings, but usually you'll never get close to maximum charge current anyhow. You first select how much energy you want your battery setup to store, and size it off that.
Only if the best option from an energy point of view is insufficient from an input current perspective (I very much doubt it since solar charge times will be significantly more than 1 hour) will you need to look at other chemistries. Deep cycle lead acid may get close to 10h minimum charge time, while lithium ion will almost always be less (perhaps 0.5C charging). Not sure about lithium iron phosphate, which is what I'd lean towards. FYI, it depends not only on chemistry but on geometry.
Make a spreadsheet if you're worried. Read some battery datasheets.

the cringelords in charge of the engineering department won't give me access to a lab so I have to keep using my portable lab™
with this being the case, how are portable soldering irons? Worth it or no?

>>2393918
You should use intimidation tactics backed up by physical violence to push your way past them and use the lab. When you're finished, take their pocket protectors and give them wedgies. Then, format c:

>>2393838
just build a higher voltage system

Have some esoteric autism thought I'd share
Used to be really into microcassettes for some reason
Had some old tapes I still needed to digitize
Wanted to see if I could bypass some of the circuitry
Since I know you can get better quality than expected from these
From a time I respooled a microcassette onto a "normal" cassette
And got way better quality than playing it as intended
Somehow found a datasheet for one of the ICs
Tapped after first preamp
Got this working first try, surprising for me since I'm retarded
Need to enclose player though since more susceptible to interference from phone ect
Kind of a test run since at some point want to see if could get even better quality
by tapping directly to the tape head
Would be harder since I imagine direct tapping tape head is very low output
Could apply that to all other analog tape formats if it worked and was better quality

>>2394012

is there a part 2 to this story that contains a conclusion, or the point you wished to make?

>>2394015
not really just a project I wanted to share, was happy my idea actually worked
working on getting some tapes digitized with this method
probably have one somewhere I've already done I can compare with
to see if it's actually worth it
all recordings are just going to be my own audio journals and other esoteric autism
did at least find this old recording of Reviewbrah on shortwave I forgot I recorded
https://www.youtube.com/watch?v=zxs_Z9FugnI

>>2394012
>>2394034
That is very cool.

>>2394034
>working on getting some tapes digitized with this method

not sure if you're aware that tapes are recording using an RIAA-specified frequency response.
so you gotta undo this with a reverse filter to get it to sound right.
schematics can be found online.
maybe it could be emulated using equalizer software.

>>2394052
Yeah there's lots of consideration if I try direct tapping the tape head
not going to try it quite yet, baby steps
probably not worth it for most tape formats since most sound good enough using standard methods
seeing if it's worth it for microcassette since it needs all the help it can get regarding fidelity
tried it for the hell of it, stole the idea from the Laserdisc Domesday Duplicator project
and the ongoing sister project of people doing direct RF captures for VHS to decode in software
I'm too brainlet to try video

>>2388735
I need to ask a really dumb question. So, when a power brick says output: 12v dc 3a for example. Does that mean it can provide up to 3a or will it supply 3a always?

>>2394102
up to

>>2394102
This >>2394106
and go heavy on the PSU. If your device requires 3A, use a 5A supply. Just because you see 3A on the label doesn't necessarily mean it can shit out 3A continuously.

>>2394102
>>2394106
Thanks fellas.

>>2394122

Is the guy I talked to about test equipment on /g/ yesterday here?

Any guide for putting together a lab? I was thinking about how I'll need stuff like ventilation and I was wondering what other safety/useful stuff I'll need besides my soldering iron itself

>>2394143
what do you want to do, what do you already have, and how much space and money can you spare?

let's say i wanted to make my own ultrasonic cleaner power supply. i understand that most designs, at least the cheap chinky ones, use some sort of a self-tuning analog loop that i don't really understand. i found this design which uses a phase locked loop that is manually tuned (with potentiometers) to resonance. http://www.imajeenyus.com/electronics/20110514_power_ultrasonic_driver/index.shtml i've never worked with phase locked loops before so i don't understand the specifics, but i think i get the gist of what he's doing here. is there any reason why this couldn't be done digitally with a MCU? MCU drives a MOSFET driver (with dead time insertion / overshoot protection) with the output MOSFETs in a half bridge configuration. resonance frequency could be determined with a current sense resistor on the output, with some sort of control loop continuously adjusting the MOSFET driver frequency until sensed current reaches a maximum. would this work or am i completely misunderstanding/missing something?

>>2394162
3D printing and electronics repair. Old consoles and various tinkering. Don't have anything at the moment as I'm starting out. I can spend in the couple hundred dollar range pretty often.

>>2394167
since were talking 20-40khz here id just use a 555 as a base frequency generator and convert it to a sinewave, then use some amplifiers to get it to the desired output

>>2394168
well, first thing is the biggest bench you can find.
then another one to put the 3d printers.
id get a cheapish oscilloscope in the 20-40mhz range (cheap as in used, not bad), a good digital multimeter, and some way to test capacitors (if you dont just plan on tossing them, I dont know how pricey they get in your sizes).
next level would be a logic analyzer for digital stuff, and a signal/pulse generator.
if the smoke/smell of soldering bothers you or your room is badly ventilated, id just use a fan to blow it towards a window/exit. more really isnt needed.
all in all you shouldnt be out more than 500us to get a really great start.

>>2394168
and if youre interested in dealing with discrete electronics, a curve tracer for an oscilloscope is useful, and a great diy project

>>2394168
Get a good soldering station with appropriate tips, not a chink iron that breaks if you breathe on it too hard. You'll want a hot air station too. Lighting, magnification, fine point tweezers, blu-tack or helping hands, various chemicals, stackable parts storage bins, security bits and precision screwdriver set, variable bench power supply, and maybe a silicone work mat. I missed a bunch of stuff.

>>2394191
Stock up on jellybean components, different types and gauges of wire, copper wool or soldering sponges, desoldering wick, adhesives and tapes, liquid dispensers, connectors/terminal blocks/crimp tools, and heat shrink.

>>2394196
also a (good) desoldering gun is worth gold for ICs

>>2394196
Protoboard/perfboard/breadboard, pin headers of various pitch, breakout boards for high pin count packages, UV light to cure epoxy or expose photoresist, and other stuff. lmao
>>2394198
Yes, if he can afford it. Some solder suckers work pretty well so maybe he could get by for a while and save money for the desoldering gun.

>>2394202
I dont know if im stupid or got bad stuff, but I never got that copper stuff to work good.

>>2394208
>copper stuff
Exposing copper clad boards? You need a real UV light that's dangerous enough to damage your eyes and give you skin cancer.

What are good repairing Youtube channels, with explanation, why is this the fault item etc.
"My Mate VINCE" seems pretty chill.

>>2394210
i meant the copper desoldering wick
never tried my hand at pcbs, but I thought about using a pen plotter for it
>>2394211
my favourites are carlsons lab at curious marc, but those are a bit more esoteric than "repair videos"

>>2394217
Next time you make an order, try NTE #SW02. It's good stuff.

>>2394202
for small series prototyping wire wrap is a serious contender too, especially in digital circuits. Much better than soldering for interchanging and much more stable than breadboards.
>>2394217
thats supposed to be "and curious marc"

>>2394222
im still on the stuff from my grandpa, I think its from the 80s. That may contribute to its less than desirable function.

>>2394223
>wire wrap
I always forget about the Kynar. 30AWG.
>>2394225
Flux the old wick and hit it with a quick blast of hot air to remove the excess before using it.

>>2394229
thanks mate, Ill try it next time.

>>2394229
Oh, also: Do you have a tip on where to get wrappable pins/sockets? All the stuff i have/see available has little stubby rounded legs.

>>2394245
The machined sockets are all round pins now as far as I can tell. The cheap sockets are flat pins. I haven't seen any square pin sockets in a while, just pin headers. Guarantee someone here knows of a source for the old school stash.

>>2392174
I like VGA.

>>2393892
You have to make sure they're the same type and have the same amount of usage on them.

>>2394284
i like turtles.

>>2394284
VGA best video output
VGA>DVI>DP>HDMI

>>2394294
How do you feel about mini-DVI?

>layout circuit on perfboard
>put long side of headers through the top
>not long enough to penetrate breadboard
>push short end of headers up through board
>solderable part of pins on wrong side of board

>>2394296
Never dealt with it since I dont use apple, but I only rate DVI over DP for the VGA pins, and mDVI has those, so its all good.
Still prefer screw-attachable connectors, so slight knock for that.

Is this the thread if I want to turn my vehicle into hybrid?
Basically, in the event gas becomes too expensive or scarce, I have solar already setup and wanted to somehow install electric traction motors onto my axle.
What do I need to make this happen?

>>2394327
>What do I need to make this happen?
more money than you'd spend in gas before you junk the car.

>>2394327
You wanna build a friggin solar tank?

>>2394306
Do you have any funky VGA projects like RF transmission or are you into old arcade games and computers?

>>2394327
>What do I need to make this happen?
Tesla parts.
Not joking, I've seen quite a few people convert regular cars to electric by dropping in parts from Teslas that ended up in the junkyard. The parts are still quite expensive like this but it can be done.
You can also opt for cheaper parts from other EVs but you sacrifice power and/or range.

>>2394338
>>2394328
Money isn't a problem.

Tesla parts being salvaged and used I've kind of seen, I'd have to do more modding than what I wanted as I was hoping to do something similar to:
> https://www.xlfleet.com/assets/Uploads/XL-XLH-Ford-F250-FLY.pdf

Where the gas engine is still intact and usable at times.
The new ford hybrid is very similar to the link I shared.
Although, I was planning on doing a much bigger battery pack, for at least 20 mile range a day (it's the range I usually use daily on the farm).

I was looking at diesel electric trains, but I can't find any schematics of them.

I have a 12k solar system (12k per hour, for consistently 4 hrs a day) and a battery bank large enough to charge another battery bank for the vehicle (assuming it's under 50kWh battery bank).

>>2394334
I like the old CRT monitors, they just look better to me.
Other than that, Im also into old computers, but most of that stuff there is composite or some weird TTL, if not integrated tube ala HP85.

>>2394349
>Where the gas engine is still intact and usable at times.
If you get something front wheel drive you will have the back axle completely free for modifications while the front remains powered by the original engine. Still, there's plenty of other stuff to look out for in your base vehicle to make sure you arrive at a drivable result. Having an electric power steering pump is one thing that pops into mind right now.

>>2394349
Making a hybrid is very hard, since youre trying to cram everything into a single body. Building an EV conversion on the other hand is easy. Seems silly, but maybe try to build either a "swap in" instead of a "plug in" hybrid, IE build a kit that allows you to change your car from gas to electric. The motors arent the real issue here, but finding space for the battery is a bitch. Instead just build an EV from a mid 90s car with a junk drivetrain, and use it to do small tours.
What kinda car are you driving?

>>2394377
Pickup, but I can get a low profile "big rig" that can hold easily 1-2klbs battery bank.

It can't be that hard, as the link I posted those guys are basically just dropping in the traction motor right on the axle. I don't know how they are wiring it to the controller though.

>>2394377
>>2394379
To add for reference, even the new ford 150 hybrid has a similar setup, but it's built into the clutch it seems, or it's still on the main drive axle and just encased with the transmission.
>https://www.youtube.com/watch?v=Ts-3RU-J8cQ

>>2394366
Analog does look good and still has its place. CRTs are just a pain in the ass to keep up and take up too much space. Do you have a preference between curved glass or flat glass? Gateway 2000 bundled pretty nice flat glass Vivitron monitors with their first Pentium-class PCs in the early/mid 90s. NEC and Sony made some nice ones that people just gave to me in trade for fixing their computers.

>>2394379
>>2394382
sorry for the wait but Im on mobile and cant type out big messages right now

>>2394414
Ive used normal curved and classic trinitron, but only just bought a flat crt (GDM-500R) so I cant really opine on those. I dont really mind the curve, but CAD does look a bit weird on it.

A revision of the speaker simulator amplifier stage.
I followed your suggestion and re-designed the power supply as a split supply, and got rid of the +in cap.
I figured out how to recreate the bass response of the original, which I did by locating the gain control pot (simulated by R10 and R11) between ground and -in and wiring the input cap in series with a resistor to the "wiper". In retrospective, that seems like it should have been the obvious solution if we see how it's analogous to the jfet amp version.
Gain control response looks good but needs fine tuning of the resistors and capacitor. Maybe I'll use the equations I know to figure that out rather than sink hours fiddling about in ltspice.

>>2394563
Neat.

>>2394563
>hours fiddling about in ltspice.

while you're doing that, kindly learn the command to turn off the grid.
coz having a splooge of black dots littering your screenshots marks you as a big dummy.
and you're probably just only a little itsy-bitsy dummy.

how does a wall switch disconnect power in something, like a light or a fan?

i'm powerwashing my porch tomorrow, and there's a fan which has a bunch of cobwebs and pollen and shit on it, but i don't know if it'll be cool to get it wet or not.

i understand that with electronics, as long as it's fully totally 100% dry before you put power back through it, it should be fine.
the wall switch on the inside cuts the power to the fan like you'd expect. but, is there zero electricity flowing to the fan when the switch is off, or is the circuit just interrupted?
like, is there any way for the water to still short the fan in some way with the switch off? like, should i flip the fuse to be safe?

i assume that a wall switch either disconnects both the neutral and the hot wire, or at least the hot one, so the fan should be fine, but i'm not sure.

>>2394683
>ceiling light
They are always live, even if they switched off at the wall.

I discovered this when I replaced a usual light- hanging-at-end-of-cord with a fixture that takes multiple bulbs.
The light was switched off at the wall, but I still got a shock.

Turn off the power to circuit from the electrical fuse box.

>>2394683
The switch normally cuts the live. But if the electrician fucked up and switched the neutral instead, you won't notice until you try to work on it while it's "off". Some countries have two "live" wires, in which case switching it off only prevents current flowing but doesn't make it not live.

>>2394688
>>2394754
i appreciate it, dudes. i'll just cut it at the fusebox to play it safe.

First time designing a circuit do you guys think it's bad? I did my best trying to manage wiring path etc but I'm a total noob in electronics so I have no idea how to actually do stuff. It's supposed to be a two bit binary full adder but using many interconnected logic gates through 74x ICs turns it into a bit of a mess. Also I keep getting "Clearance violation" errors in the DRC all caused by the pins (Error: Clearance violation (netclass 'Default' clearance 0.2000mm; actual 0.1500mm).

>>2394789
do you pan on printing it or do you just want to play with the simulation?

>>2394823
The final purpose is technically to print it, once the whole circuit is finished (Which includes many other things I had in mind). But it's also important to learn about Simulating it in the end since being able to make a final netlist is in the scope of the project.

>>2394826
I dont know about the function, but itd be a lot cleaner and less interfering(?) to have the 3 ICs in a staggered parallel arrangement, to get rid of those tight clearances around the pins.

>>2394949
I investigated a bit and I found out that it gave error because the default clearance (I guess standard deviation/margin of error) for the position of the copper pads was too big and interfered with the other clearances of the other pins pads. Overriding the clearance to 0.15mm instead of 0.2mm in the footprint options fixed the error. Does this affect the gerber files? Do pcb printers consider this clearance value when printing the pcb?

>>2394972
I think the printers just develop what you send them, judging by the fact you can have them print tiny smd paths. The standard distance is probably either to make sure the sockets will fit, or to prevent solder bridges.

Is there any current/enthusiast use for 6 pin DIN plugs? The ones that look like a chevron?

>>2395017
>Is there any current/enthusiast use for 6 pin DIN plugs? The ones that look like a chevron?

nobody uses that shit anymore coz too big and old-fashioned.
only the 5-pin MIDI ones are still relevant.
still, if you have a bunch of males and females, they're solid performers you can use on all your own projects.
but nobody's gonna buy them from you.

>>2395048
too bad, I got maybe 500 plugs, but not a socket in sight. I thought maybe the guitar or electric piano guys might want them, but the either use 5 pins or big quarter inch jacks. But the joke is, if I want to buy a socket, Ill have to pay 3 bucks a piece.

>>2395050

the only possible market is some old guy who repairs 1980's vintage computers.

>>2395053
I havent found any that use that kind of plug yet. the classic DINs and the weird stuff like atari video or commodore ports, but no 6pins.

>>2394826
You should mess about rearranging the gates (say replacing OR gate A with OR gate B) to minimise the size and the number of overlaps. Can almost certainly even get it single-sided. Also, with Demorgan's laws, AND gates leading into OR gates could be replaced with NAND gates leading into NAND gates. Should make layout even easier, though you'd still need the same number of total ICs. I set my trace thickness to 0.5mm, keeps me making good decisions and it's easier to home etch.

Also those aren't standard 0.1" pin headers, in case you didn't realise. Don't forget an edge-cuts layer.

>>2395017
Hobbyists use them for all sorts of things. Battery charging with balance wires would be one suggestion.

>>2395050
>maybe 500 plugs, but not a socket in sight
Maybe there's a cute girl out there somewhere with 500 sockets.

>>2395073
dont give me hope, anon

>>2395067
>I havent found any that use that kind of plug yet.

there exists at least one for sure.
for the TRS-80 coco joystick port.
of which i own three.

>>2395098
well, thats a start. If you need replacements (and are willing to pay for shipping) get in touch.

>>2395105

maybe an ebay store. look at these crazy prices.
seems to me there still exist coco parts sellers who might be interested.

>>2395115
seems like tandy actually used them, at least for peripheral inputs. Thanks for the hint, we never had those here, so I would never have thought of it.

>>2395367

forget it, dude.
it's a synchronous motor.
it'll cost you at least 1 kilobucks to get a proper VFD controller.
find something else, a universal motor, like in an old drill.
(i'm seeing like $40-$200 on Aliexpress but i wouldnt trust that shit.)

>>2395367
Maybe I'm retarded but I can't tell whether it's induction or brushed universal. If it has a lump on one side to hold a start/run cap then it's an induction motor, if it has ports to service brushes then it's a brushed universal. I think the RPM being almost 1800 means it's probably induction, but that may just be coincidence. The writing "AC Motor" also suggests probably induction, though I thought those were usually called "asynchronous motors" to differentiate them from the synchronous sort.

Anyhow if it's induction you'll need a VFD. Not too cheap, I'd say it's probably not worth it over salvaging a washing machine or treadmill brushed motor. With a brushed motor you can just use a TRIAC-based speed controller, which are a lot cheaper.

>>2395450
>it's a synchronous motor.
That runs off 60Hz AC without any sort of drive? At just under 1800RPM? Pretty sure it's asynchronous mate.

i have a furnace motor connected to a blower via belt that i am repurposing, but my application needs an adjustable speed, a dial element is ideal, but i dont know what to use, any suggestions? pic related, will underpowering such a motor have a bad effect?

>>2395457
>>2395450
hmm i figured it was brushed because it was last OK'd in 1951

it does have a temp sensor

>>2395466
Brushes could be on the sides behind those two(?) circular ports, open the back wiring hatch up and see.
The lack of a cap bulge is really throwing me off. Do you see sparks in the back when you turn it on?
>inb4 shaded pole motor

>>2395474
ill give it a peek tomorrow, meanwhile i found this, it looks exactly like it down to the stickers https://www.packardonline.com/products/gf2034d/

>>2395474
>Brushes could be on the sides behind those two(?) circular ports

- a 1/3 HP would be much much smaller if it was brushed.
- a brushed motor would be very noisy and that noise would be radiated throughout the house thru ventilation pipes causing all occupants to suicide
- it is therefore not brushed.

>>2395493
well it it is from a 1950's oil burning furnace

ha it is brushed, heres the contact pad. can someone rec a dial triac with an integrated "click" off position?

ah so the auto hot shutoff isnt a temp sensor, its a thermostat that pulls the rotor contact away a little bit, i guess so little that they needed a second set of contacts to leverage a big enough gap to reliably cut the power, thats wild

>>2395504
>ha it is brushed

doesnt look like it one bit
the big diff is the brushed rotor is full of coils, whereas the other is just metallic lines at around 15-degree angle.
there may be a slip ring there for some reason, but without coils on the rotor, it's not brushed.

>>2395514
ok, well maybe you are triple or even quadruple nigger

>>2395521

hehe.
a BLDC needs a bunch of electronics to generate and feed it the appropriate signals, which, using 1960 technology, would require a suitcase full of transistors.

>>2395504
Doesn't look like a commutator. Maybe a slip ring, but not a commutator.

>>2395508
That spring mechanism may be a centrifugal-clutch for the start winding, indicative of an induction motor. Still can't see a start cap, maybe it's inside the housing.

>>2395528
>a BLDC needs a bunch of electronics to generate and feed it the appropriate signals
Or just hook it up to 3 phase AC like Tesla intended.

>motor requires AC
>motor requires 3 different phases of AC
>call it brushless DC
Who the fuck came up with this genius name anyways?

Is lightning AC or DC?
Why does it make so much RF if it is DC?
Or is it neither of these?

>>2395686
>Is lightning AC or DC?
It's DC because it's caused by the accumulation of opposing charges in the ground and the clouds. When the arc forms the current only flows one way.
It's like shorting a capacitor.
>Why does it make so much RF if it is DC?
It's a huge burst of high voltage and high current so in that brief moment it can create electric and magnetic fields strong enough to cause interference in the low regions of the EM spectrum.
You know how the spark in a spark gap transmitter produces a lot of EM noise? It's basically that but fucktoupled in size.

>>2395697
>It's a huge burst of high voltage and high current so in that brief moment it can create electric and magnetic fields strong enough to cause interference in the low regions of the EM spectrum.

So isn't it at that point semantics?

Lazy question, will these be fine as schottky diode in switch matrix/keyboard pcb? Controller is a RP2040.
mouser de/ProductDetail/ROHM-Semiconductor/DA221FHTL?qs=4v%252BiZTmLVHHp9u2466A%2Fyw%3D%3D
Long question, how do i choose diodes in general? So far i just pulled some from my kits based on forward/max voltage, but looking at these tables now, i do feel a bit overwhelmed.

>>2395698
Take a fourier transform of a sharp DC pulse. Just because it isn't a repeating wave doesn't mean it doesn't have a bunch of frequency components.

>>2395750
Yeah should be fine. It's a keyboard, it really doesn't have to be anything fancy. Schottky diodes are more than fast enough for it. Not sure if common 1N4001 series diodes are fast enough, but 1N4148s and anything better are fine. Diode selection isn't very important for a keyboard, only when you get into designing switching power supplies and that sort of thing does it get important. Schottky diodes have good speed and low forward voltages, at the expense of high leakage current.

>>2395115
hah, I found a guy that sells sockets at 50ct a piece. Now with that I can do something.

What is the jack of all trades Fluke Multimeter I have to own ?

>>2395933
whats your application?

>>2395933
>jack of all trades
Fluke 117

>>2396121
>>2396121
>>2396121
NEW THREAD