/diy/ - Do It Yourself

Has anyone done PCB design for >=DDR3 memory? I want to make my own FPGA board, but I don't know what I don't know about high speed signalling. I was thinking of doing a PCB design course on udemy or something. I've designed some small (<40 ttl parts) and low speed designs and hit some pitfalls, so being a bit more cautious this time.

>>2596827
All I can think of are trace length matching, proper ground plane shielding, and impedance matching. But I'm not sure about that last one, like if having the traces a certain distance from each other and from the ground plane actually makes a significant difference or not.

you know i think 3d printer main boards are overpriced and not sufficiently modular
all you need is a slot for a pi pico, slots for stepper drivers, and some power mosfets for switching the heater and heated bed, and connectors for sensors like thermistors and end-stops and height sensors
using an external dc-to-dc adapter it could be reasonably compact, you could even have an optional external 12v regulator for powering more common fan types, you'd just have a jumper to change which power rail a given mosfet uses
could even just socket the mosfets so you don't need to include them if you don't need them, shame that drain instead of source being the tab means you can't just clamp them all to the same heat-sink without isolating pads

i'd have like 8 possible stepper driver locations for any eventuality, 8 mosfet locations for powering different fans, leds, lasers, etc.
make it long and thin and put it in a forced air tunnel with x, y and e stepper drivers at the start

>>2596862
Then why don't you go ahead, produce and sell a cheaper one? Pic related is striped down to RP2040 already and schematics are all open source.

>>2596937
>"High saturation current core!"

>>2596937
Blocks AC (artificial coloring)

>>2596937
You could dissolve it in water then cast it in an iron rich resin.

>>2596960
You could dissolve it in your mouth and then have a nice retainer.

I found this old lamp body and want to make it my first /ohm/ project. I did lots of inhouse electrics and did some soldering and I guessed that making a diy battery lamp is easy but fuck I have some starting problems because of translation problems.
I want to use the original bulbholder wich is a ba20d. I thought using 18650 batteries and a usb-c charging port is a good idea.
What are the components I'm looking for to implement this project?

>>2596830
Ok thanks. I guess I'll look at those topics and see if I go down a rabbit hole.

>>2596937
>exclusive, super-low hysteresis material optimized for high-frequency operation

Does anyone here have experience with old gaming consoles? I recently got two Vectrex consoles and I'm trying to debug/fix them

>>2596827
Some pcb fab houses have impedance control as an option when making an order. Whatever EDA/ECAD software you're using should let you choose the single-ended/differential impedance for specific traces, so it'll auto adjust the trace sizes when you make your layout. I've worked with DDR4 and I believe the single ended impedance is around 40Ohm but once you actually get the thing built you can still muck around with the DDR4 driver settings and see what works best

>>2597156
>Vectrex
https://www.youtube.com/channel/UCODLTJtRV-jYvmTx9YScCVQ/videos

>>2597162
Yeah I've seen those, and I've already re-capped the thing but still getting the same visual. Thankfully at least the screen seems to work, it's showing a white dot in the center but the X and Y-axis control doesn't seem to be doing much.

If anyone has a working one, lmk and maybe we can do some A-B comparisons with certain signals

Would anybody be able to help me ID this component or atleast what it is? From a power supply from the 90s. The white one, it says “COL” with an arrow on the top, and on the side says

“LF-4N
502
28R”

Tysm, I’ve never seen something like this. I don’t usually work on old stuff

>>2597284
I forgot picture duh

>>2597285
Probably a common mode choke

>>2597291
This seems likely given the context, I'm gonna assume correct. Thanks

Most of my projects cannot be done unless i design a circuit board, get it manufactured and buy parts from digikey/wherever. I guess i should be happy that my projects are getting more complex and interesting but it's such a fucking pain. Hobby stores all have the same 30 year old parts.

I want to do some current sensing in my new project and need a few decent opamps with low offset voltage and low bias current. All i can find are BJT opamps with input offset voltage as high as 5mV. Shit like LM358 and TL0XX (only frontend is JFET based, everything else is BJT). What can I do?

>>2597347
You want something like the OP07. Frequency response isn't that great, but they're brilliant for low input offset.
Also dedicated current sense amplifiers like the INA180 (see texas instruments webpage for selecting these) are nice and simple.

>>2597347
>>2597373
The JFETs are the reason for the poor offset voltage, not the BJTs. A properly tuned BJT op-amp can have decent offset voltage. You also have the option of making a chopper amplifier, or buying a monolithic chopper amplifier like I'll be doing (all 3 of the COS855x series from LCSC). I'll also be using the INA181 myself, as the output can be offset to measure both positive and negative currents from a low-side shunt without needing a negative rail.

As for making a circuit board for projects, I'm kinda agreeing with you. But I'm also building up a toolchain to be able to laser out single-sided boards pretty easily. Maybe even double-sided if I make an alignment jig. The tech will be:
>KiCAD to design board and export as SVG
>Inkscape shell script to convert overlapping shapes into one big composite shape
>Lightburn to produce G-code with concentric offset fill
>Ender 3 with 5W diode laser bolted on
>burn away spray-paint in mask pattern
>Etch in regenerated ferric chloride
Also I'll probably bolt a rotary tool onto my printer to use it as a drilling machine.
I tried using a resin printer and dry film photoresist, but the contrast isn't sufficient to prevent the exposed mask softening in the developer or etchant, and the whole process was finicky enough to make it annoying and not very repeatable.

>>2597373
Can't believe I missed a jellybean part like OP07
Just had a look at the offset voltage, its breddy good from 20uV-75uV. Input bias current is decent too 1nA-4nA. No mention of gain-bandwith product though. Unity gain bandwidth is 600KHz, which should be okay for my purposes I think. Slew rate and output swing aren't too impressive either but I can live with that
Thanks anon this should do

>>2597391
>INA181
I looked into using these as well but the only way to obtain them is to buy them from mouser et al

>>2597406
GBWP and unity-gain-bandwidth are basically the same thing. So long as your desired gain at X frequency doesn't surpass a -20db/decade slope from 600kHz you should be fine (plus another 30dB or more for noise immunity).
Watch EEVblog's video on jellybean op-amps for more that you're missing, like the RRIO LMV358.

>the only way to obtain them is to buy them from mouser et al
Are you buying from local vendors? If you're buying from alibay I wouldn't trust a TL072, let alone an OP07. I've had fake TL072s from them before, add that to the TL432 pack with TL431 markings and I've never trusted them since. I'm just making a large list of shit to buy from LCSC, including jellybean op-amps among other things. Also GL200N06 MOSFETs, CH340Ns, RP2040s, etc.

How do I tell if these types of fans are any good or not?
Just want something for getting rid of the solder fumes while I'm working

>>2597347
single-package op-amps like the TL071 come with pins called "offset null", where you can manually fix the offset with a trimmer pot.

>>2597471
>solder fumes

205 CFM is massive overkill.

>>2597347
>current sensing
INA series plus shunt resistor and be done with it

Are there any guidelines or "industry standards" for the design and form factor of "Industrial" embedded systems ? Like I dont have a college education on this subject, the books i have read on this dont really tell you how to make a PCB for example. Like a PC motherboard has a standard set of connectors, HDMI, VGA is RS232 type connectors, USB for example. I was wondering if something like that exists for embedded systems ? like for power supplies - use a linear / switch regulator ? or for the power connector use a plug type or a wago or whatever else ? If anyone has something like, please share. I am lost here.
AND
I have seen people make PCBs where they use 5-15mm standoffs and 0.1" Arduino type female connectors to connect 2 PCBs together. Is it a good idea ? You know how they use shields ?

>>2597006
You'd either need a DC/DC converter to pump it up to 12V or power it off a small motorcycle battery as-is. 3D-print a case, maybe get a heavy-duty circuit breaker for shits and giggles.
https://de.aliexpress.com/item/1005002952767709.html

>>2597531
Can only answer on the second question:
Most manufacturers use their own standard for board-to-board connectors, see picrel as an example. I don't know whether there is a standard but if there is nobody's adhering to it.

>>2597531
If you’re making something on the small side, I’s go with the arduino form factor or the bigger raspberry pi formfactor.
For connectors, follow the rc crowd using xt60s for power jst-xh for lower power/signalling
…there is so much stuff out there for them it’s hard to go wrong.

>>2596830
>>2597156
Just for completeness I did some researching and it seems that this guy is a good resource for exactly what I want to do:
https://www.youtube.com/watch?v=5vPeSdU22ns

>>2597531
there is no industry standard for embedded systems. its all about the use case. basically do you need the device to be general purpose or do you need something small, efficient and only has one purpose. as an example, im working on integrating embedded device to a PLC, because PLC and/or industrial sensors dont do what i need. dor this particular use case ive decided to expand extra peripherals on their own connectors and power supply pins, in case something else needs to be added later

>>2597531
The standards which do exist and influence engineering are mostly performative/testing rather than design guidelines.

For example, embedded systems are often (perhaps "mostly") used in environments with moisture and dust contamination, so customers will look for an "Ingress Protection" or "IP" rating which is a reflection of how well the product tested against damage or infiltration. Power supplies are a good example of something you'd want an IP rating on.

Otherwise it tends to be whatever is common at the time and fits the purpose -- usb mini, micro, and now C for example, have carried arduino and etc for ages.

>>2597842
why did micro b take over mini b for arduinos and shit
same for charging random cheap things with lithium batteries in them
why would you use an inferior connector for it when cramming it into a thin cellphone isn't the main consideration

based original usb a and b are still here and have a 0.1" pin pitch so are highly suited to hobby projects, and can even be used with perfboard. buy yourself some nice and solid usb b connectors today.

What is the name of this connector? Seen on RC plane wings connecting to main fuselage.

>>2597974
It's called "blade connector".

>>2597284
those thingies are very common in the old crts. the black box is some type of capacitor, the white box is a coil. from what i've heard it's like a filter for electromagnetic noise coming from the power line. it makes sense cause they're always next to the ac and fuse

>>2598005
It actually is, thanks anon.

>>2598017
no problem anon

>>2596815
maybe not the right place to ask but, I just started a sparky apprenticeship, I've been working for a month and I've broken my ankle so I'm stuck in bed for a bit. Was thinking I'd learn PLC stuff while I'm out of work, where is a good place to learn it online?

Where would you even go to get something like this EMI shield made?
There's little cutouts in the sides for feddthrough capacitors and the like to pass through.

>>2598028
I would take measurements and fabricate it on the bench using a dremel and vice for low-volume production.

>>2596815
Can someone explain to a retard like me how higher voltage like in a power line results in less current therefor heat? If 220v and the wire has 1r then you have 220a. At 120v 1r that would be 120a so why do I keep seeing seeing something along the lines of this when I search on it.

>High voltage transmission minimizes the amount of power lost as electricity flows from one location to the next. How? The higher the voltage, the lower the current. The lower the current, the lower the resistance losses in the conductors.

Or Quora

>Current and voltage have an inverse relationship in a resistor. When the voltage across a resistor increases, the current flowing through it decreases, and vice versa.

This isn't true is it?

When making PCB transmission lines how much care needs to be taken when choosing the transmission line length?
I only know that the lenght of the line shouldnt be a 1/4 fraction of the wave length because... reflections are bad?

In my case I want to use a 2.45GHz chip antenna on a PCB that needs to be atleast a 3-4 centimeters long.
What do I have to take into account other than designing the transmission line so that it has an impedance of 50 Ohms?

>>2598042
>higher voltage like in a power line results in less current therefor heat?

it's super simple
power = voltage x current
if you need to transmit 1 megaWatt of power, you can do it two ways:
- 1 million amps at 1V
- 1 million volts at 1A

higher voltage means you need less current
thus thinner, cheaper wire

>>2597006
are you planning on replacing the bulb with an LED or keeping the original bulb? 60W off a battery is a lot of power, you won't be running it for long.

i've heard the market for repaired electronics is non-existent, anyone know any counter examples?

>>2597285
>>2597291
>>2597297
>>2597284
Since it's from a power supply, and all i see is a capacitor, 2x 2 pin connectors and a fuse, it looks like the white box might be a 60 hz transformer to step down 120/240V to probably 12V since those were common for power supplies back in the 90s Is this all there is to the circuit or is this a section of the entire power supply? I doubt it's a CMC.

>>2598105
The market for repaired electronics is the people who repair electronics.

>>2598074
>how much care needs to be taken when choosing the transmission line length?
For the length, not much unless the timing and synchronization is important. A perfect example of this is if you have two data lines going to say an XOR gate. If one line is longer than the other, and the state changes from say +5 to 0V on both lines, then at the end of the line, one of the traces will show up as 5V then 0V, and there will be a delay before the other trace will transition from 5V to 0V. That means your output of the XOR gate is initially 0 (1 XOR 1 = 0), then the state of the data lines change (0 XOR 1 = 1), then the second trace catches up (0 XOR 0 = 0), which will show up at the output as a glitch where the state is 0, then jumps to 1 and back to 0. This is not an issue if you account for the delay in your design and "read" the value only after a certain time has passed, but if you're switching the data back and forth a million times a second, there might not be enough time between two cycles to wait for the transistion to stabilize, so you need to match the length of both traces so both signals arrive at the XOR gate at the same time. What you should be more worried about is reflection, so focus more on impedance matching than length matching. Most PCB software can also do length matching automatically.

>>2598130
>switching the data back and forth a million times a second
What the hell is the slew rate on your XOR chip?

>>2598105
It's kind of true, the only exception is for
1. discontinued products, such as oscilloscopes, old consoles (ATARI, SNES, etc.) or old computers (Commodore, etc.) as well as industrial hardware that are no longer available and expensive to retrofit.
2. temporary fixes, such as to fix an apple laptop/phone to backup data from a dead computer
3. Restoration of vintage tech (somewhat related to 1.)
Apart from those three cases,I don't see why anyone will try to fix anything when the fixes aren't very professional (jumper wires, non-matching replacement components, etc.) and will most likely fail or cause trouble as soon as the repair warranty is over.

>>2598138
No particular XOR chip, was just intended as an easy to understand explanation

>>2598042
You're assuming that the resistance of the transmission line (1r) is all there is in the circuit and is connected directly between live and neutral. This is not the case, the transmission line resistance is actually in series with your load. So if your load is 200r, and is drawing 1 amp (200W total of power, say a heater), redo the math:

current through transmission line = current through the load. Say 1A.

Pline = 1ohms*1A^2 = 1W
Pload = 1A^2*200R = 200W. What about the voltage? The voltage across the load is 1A*200R = 200V.

So to transmit 200W to the load, you lose 1W on the transmission lines if the voltage across your load is 200V. What if you were to do 100V?

Well the current through the load at 100V to keep 200W is:
P = VI => 200 = 100I => I = 2A.

What about transmission loss?
P = i^2*R = 2A^2*1r = 4W

You see how to transmit the same power (200W) to the same load with the same resistance at 100V results in 4 times the power loss (4W vs 1W) than transmitting the same power to the load at 200V.
.

>>2598138
> xor slew rate
Better than average.

>>2598169
I'll say.

>>2598105
I got my oscilloscope from some guy that repaired it. He repairs stuff for the city, like those deep pipe / buried wire detectors.

Audiophile market actually supports repair/upgrade (even if nothing wrong… like discrete op amp replacement).

That autist mr. carleson repaired his fridge and put in a zero-crossing detector on the relay to it would would click over at zero current. Better than it was!

So, in many cases, the repair can be better than the original. Like capacitor replacement.

I had a buffer that kept burning out it’s bridge rectifier, so i replaced it with some beefy diodes and never had the problem again.

Big clive fixes LED light bulbs to give them indefinite life by under driving them. This is very useful if you have a bulb, like, 30 ft. In the air.

>>2598042
Yeah, and if you use 1V mains then you lose even less power in your wires! But then your 2kW heater is now a 0.1W heater. If you want to actually do useful work with that electricity, then you don't fix the resistance of your load, you fix the power consumption of it.

>>2598109
Way too tiny to be a mains frequency step-down transformer unless it's using some obscure high-perm core material.

>>2598028
Just wrap whole thing with kitchen foil

>>2598028
I think some of the fab houses like JLC or PCBway do sheet metal fabrication, they probably have the right metal for a shielding can, though if you’re talking mumetal I couldn’t tell you. For those I think people just buy ready-made cans and run traces beneath it.

>>2598028
use copper tape on thin cardboard if you need a more ghetto solution

Hello /ohm/. I have built a tube amp, but the HT voltage is higher than expected and outside spec for the tube anodes. Is it reasonable to just use a dropper resistor to drop 370V to 275V (dissipating about 5W). I would use thermal glue and mount it to the chassis for heat transfer. Is there a better way to do this?

>>2598321
Grounded copper tape is fine for electrostatic noise, but it won’t do anything about magnetic noise. A soup can is the chad shielding. Pre-tinned for easy soldering, it’s like it was made for shielding.

>>2598337
Should work, but I’d be more of a mind to use a linear regulator of some sort, since chances are the voltage will drop enough when under load anyhow. Maybe a shunt regulator? Whatever would make less heat and be reliable.

You could also try putting some backwards windings on the transformer to lower your output voltage.

I want to build a device with the same footprint as an auto relay (pic related). It's easy enough to design the PCB, but where do I get pins of suitable size? If I was making just one or two I might considering gutting an off-the-shelf auto relay, but this is for a device that will need thousands of pieces and it needs to be fairly robust so hacking third-party hardware is a poor solution.

Naturally, each pin will need to be soldered to connect to the rest of the circuit, so whatever metal I use needs to be solderable.

>>2598467
>
oops, pic related

>>2598467
>>2598468
You can buy PCB-mount spade terminals, I've got a packet in my parts bin.
Interesting idea to copy an existing footprint, what's the project?

>>2598475
Thanks.
I want to build a secondary system that can disconnect the circuit that the relay operates. My intention is to keep the existing relay in the loop, so if my component fails, the circuit fails open rather than closed.

>>2598476
Do you plan on having a relay socket atop this device, and just add an extra switch in series with the contact?

>>2598478

No, my application is space constrained and I wouldn't be able to fit my device + the original relay. However, it looks like pre-fab relay harnesses are cheap & easy to find so my plan is to run the harness wires from my device to some other place where I can plug in the original relay.

>>2598467
> pcb mount spade connectors
Salvage them from old microwave oven mains filter boards.

>>2598505
Not everyone has a shitton of microwaves disassembled. For those who do however, what do you do with all the beryllium bits that come with the magnetrons? Kinda want to make an image macro of apu grinding up and snorting it.

>>2598510
>what do you do with all the beryllium
Break it into chunks, which go into the pepper mill I use to spice my wife's food. Aint no way that bitch is gonna outlive me and get the house to herself.

How many of you here can make RF transmitters out of BJTs and passives?
I wish to be knowledgeable enough to be able to do so

>>2598565
You might be better off asking that in /ham/

>>2592486

>>2598670
eh, those guys seem more about bands and events than actually building stuff

>>2598670
The guys in /ham have no clue what they are doing and how things work. They just buy SDR and similar off the shelf components and never do anything with hands (except putting antennae on roofs).

>>2598565
Here is a good playlist of basic RF design.
https://www.youtube.com/watch?v=r_p7AHsSOdw&list=PL9Ox3wpnB0kqekAyz6blg4YdvoEMoJNJY

Things get more difficult with higher frequencies (GHz range) because you're constantly fighting against parasitics. If you stick to low frequencies up to HF range and maybe VHF with some experience creating a basic transmitter is not that difficult. Just chuck your solderless breadboards in bin because they are useless for anything above few MHz and instead solder on copper clad boards manhattan style. RF is also more accessible nowadays because of cheap SDR dongles and NanoVNA for testing. Just be careful that you don't turn into a ham boomer and start erecting massive dipoles in your backyard.

>>2598684
>They just buy SDR
>>2598690
> SDR dongles
Haha classic. HAM retard can't post an answer without mentioning SDR.

>>2598691
I don't have a HAM license though. What's wrong with using SDR for testing? Not hardcore enough for you? Or it doesn't count unless you do it on 50 year old tektronix gear? We aren't living in 60s anymore grandpa.

>>2598690
Thanks anon, my interest is more on the academic side. I'd like to be able to design everything on paper and know just enough to recognise real world problems like parasitics etc. I'm not really building anything

>>2598691
it's a broad spectrum frequency analyzer for $25, you'd have to be retarded to sleep on that

>>2598717
>>2598691
Can an SDR function as a VNA? Of course I wouldn't expect to find the return losses but if I know the TX power of the circuit, could I know just how much power is actually being transmitted

>>2598690
Guy says this is university material, however it’s clearly something like a “technical collage” so HAM hobbyist stuff is what you’ll find. Be warned.

Started watching the video, and immediately saw that the transistor they show is drawn (copied from somewhere) incorrectly. This is likely to be the “tip of the iceberg”

>>2598751
(((iceberg)))

>>2598751
>and immediately saw that the transistor they show is drawn (copied from somewhere) incorrectly.
I don't see anything wrong with it. It's just standard common base amplifier with NPN BJT as an active device with divider bias.

>>2598755
The base, collector and emitter are all shorted at a point in the center.
The problem with doing that is I’m gonna think this person hasn’t seen enough transistors to know that it’s wrong.

>>2598761
are you autistic? the slight misdrawing of the symbol doesn't confuse the schematic in the slightest, it's obviously an NPN transistor and couldn't be anything else, therefore there's no problem. your "it's gay if the c and e touch" conventions can eat ass.

>>2598761
>The base, collector and emitter are all shorted at a point in the center.

>>2598761
>obsessing over tiny autistic details
Are you german by any chance? Do you get skin rashes if schematic doesn't conform to DIN Schemedickenschwanzshaftug standard? /gd/ is 2 boards to the right.

>>2598766
>it's gay if the c and e touch
it's not gay, it's just a diode, it should cost half as much.

>>2597156
So one sounded like it booted fine but had screen issue, and the second one didn't seem to boot at all but the screen had a white dot in thebcenter. After doing some A B comparisons, I tracked it down to an MOS6522 IC, and after swapping them the issue filled the IC so now I have one working vectrex.

Does anyone here have experience with vector displays? I can provide a pic of it's current status if requested

>>2598827
issue followed* the IC

Instead of getting a NOS replacement I think I'll cheap out and get a Chinese clone in case the screen is fucked. If the circuitry for the screen is messed up, it's fixable, but if the screen itself is messed up, then the thing's basically only good for parts. I think

>>2598829
>MOS6522
WDC W65C22N is the equivalent

>>2598829
>get a Chinese clone

you dont have $10 for a known genuine part?

>>2598860
More like $60 where $50 will be delivery from those stores.

I have a 12V lead acid battery that sits around

>>2598869
Buy $50 at once and shipping is free. If you can't think of $50 worth of shit to buy from Mouser or Digi-Key you're not trying hard enough.

>>2598871
*I have a 12V lead acid battery that sits around 7V even after connecting it to 12V. Can this be recovered?

>>2598875
sure it can!
https://www.youtube.com/watch?v=afwO_MZjRjA

Wanted to post this on /vr/'s CRT thread but I feel that you guys know better here. Is the NJM2244 a good drop-in replacement for an NJM2534? It seems like it has the same internal DC bias and same pinout. I can only find the former on eBay and Digikey, while the 2534 is either unavailable or in an incompatible foorprint.

>>2598856
The actual part was labelled SY6522 but I got what china had to offer.

>>2598860
I swapped the working part from a victrex with a potentially broken screen. If it ends up being unfixable, then the whole thing was gonna end up being for parts anyways and I didn't wanna end up paying thrice as much for an IC that was gonna go to a parts board

>>2598905
>>2598892
Aw snap, I wasn't aware that they had a /CRT/ thread, thanks anon. Idk why I didn't think to try a video game forum for video game console questions

>>2596815
Well I did it
I went out and got me a couple electron guns.

I want to do vaccine chamber sputter disposition for lenses and other tomfoolery

Should I grab any other components?
There was crt specific transistors but they were only rated for pussyfooting 1.5kv.
Some schmuck harvested the flyback and copper, but that's fine I've got other flybacks and a box of mots

Also why does the election gun have three holes in the barrel end?

Explain it all to me like I'm both retarded and too young to have every done crt

>>2598996
Three holes in the tube are the three electron guns that account for the red, green, and blue parts of a TV's display. Each gun can be controlled individually, and emits either a "triangular" pattern on regular displays and "in-line" patters for Trinitron displays.

Can I use this to regulate current though a load
In words its a differential amplifier across the said load which feeds into a opamp where I can set the reference voltage (after gain)
https://imgur.com/xX2m74Y.png

>>2598886
i'd prefer to avoid both lead poisoning and eczema
also it's a sealed 4.5Ah battery

>>2599052
No. You should put the differential amplifier across a fixed known current-sensing resistor, just putting it across the load only sets a constant voltage across it instead, which isn't desirable if the load resistance changes. But with the sense amp across a sense resistor it should be possible with the following caveats:
>ensuring that the common-mode input voltage range of the op-amps is respected
>ensuring that your MOSFET isn't operating in an area of thermal instability or replacing it with BJTs instead
>ensuring that your feedback loop doesn't oscillate by limiting high-frequency in your sense amplifier or limiting the total gain of the error amplifier
>performing the thermal calculations to ensure your transistors won't overheat

Further reading:
>https://www.infineon.com/dgdl/AutomotiveMOSFETsinLinearApplication-ThermalInstability.pdf

>>2599051
Thanks
Did I get the wrong kind of electron gun for coating lenses

this mpu6500 shows no signs of life, doesn't respond at all. I've used various libraries/example codes to communicate with it using an esp32 but no go
I checked the connections on the board using a multimeter, the i2c pins correspond to pins on the small IC there (actual MPU), the VCC pin goes to a voltage regulator, and the output pin of the regulator goes to one on the mpu chip
Thoughever, what puzzles me is that there seems to be nothing connected to the gnd output pin of the regulator, there is no ground pin on the mpu.
How do I further test it? The chip looks crooked. Can this be repaired or should I buy a new one?

>>2599176
It's most likely a soldering problem. There is no way it's going to work properly at that angle when pins are only tenths of mm apart. If you have any heat gun that can reach about 230°C then you could wiggle it back in position.

>>2599176
update
looks like the mpu has a ground connection, can't see the problem, just doesn't say anything over any of the lines
>>2599188
unfortunately I have only a soldering iron (25w). Can I use it and a desol pump/wick to fix this?

>>2599189
>I have only a soldering iron
Another option is to desolder pin header and then use cooking hot plate assuming there aren't any components on the bottom side.

>>2599176
That chip is mounted at quite the angle
Check to see if the pins are

What back alley did you buy this in?

>>2599176
Make sure you use an adequate power supply, and probe the regulator for proper output.

>>2598239
>Way too tiny to be a mains frequency step-down transformer unless it's using some obscure high-perm core material.
It's hard to tell by the picture, but what size is it? 60Hz transformer can be surprisingly small for low power applications. I don't see what the purpose of a common mode choke be on that specific board, unless it's one part of the entire power supply design. Is there any top and bottom view picture of the board? If there are rectifying diodes or a rectifier IC behind that white box with traces running to the white box, it's 100% a 50/60Hz transformer.

>>2598761
(you)

>>2599189
That chip looks fucked, honestly wouldn't be surprised if some of the pins are shorted or even if it's mounted upside down. try not to power it until it's fixed, you could potentially short power rails and smoke the traces. Like everyone said, you should get a hot air and reflow it, use flux. If you don't have a hot air station, your best bet is to return it for a new one or refund.

>>2599209
i have nothing to lose might as well try this
>>2599254
online order
fucking chinks i tell you
no not aliexpress
>>2599262
did that
it puts out a proper 3.3v
i had to improvise because those are the only leads i have
quite proud of myself
>>2599282
too late lol i've been tyring to power it since a few hours now
and i have been sitting on this for a month (i ordred this, it came in a week before my exams, i opened the box after a literal month)
still i will try the stove thing if i get time, if not then this project never needed an accelerometer/gyro.

Quickly thrown together after Clive dismantled that 12V auto tester. Added a fuse to that aux output.

I have an arduino nano wired into a motorcycle's ignition and would like to keep it running for around 30 seconds max. while the bike is cranking and it has no supply voltage. It is currently being fed at the 5v pin through a linear regulator. What size capacitor would be good? I don't really know how much current it draws, but I wouldn't mind going with an upper bound and buying a bigger cap than I need. How should I wire it and what other components would I need? I imagine a diode somewhere?
I imagine I should put it before the regulator so it has 12v instead of 5v and it holds more charge, right?

>>2599318
If this is for a university project, check if your school has a soldering station, I know a lot of universities do.

>>2599393
Would something like this work?

>>2599393
No, a lower voltage cap can hold more capacity at the same size.
Eg. 100uF 450V cap is quite big but a 10V version in the same container hold few thousand uF.

>>2599402
I was thinking more if it was the same cap. A Farad is one coulomb per volt, right? So the same cap would hold more at 12v than at 5v. But yeah I did't consider the fact that caps are rated for a certain voltage.

>>2599406
A cap with double the coulombs at the 12V input side would be wasteful since you're a using linear regulator anyways.

>>2599399
the cap probably shouldn't be shorted out
the arduino goes....?
normally people put + at the top and - at the bottom fyi

>>2599462
Oh yeah... Something like this then? The arduino would be the resistance pointed to in red. Loads are commonly drawn as resistors, right?

>>2599410
Would having it at the 12v side be beneficial to keep the 5v side at 5v for longer?
If the cap was at the 5v side, as soon as it started discharging, the voltage would drop, right? But if it's at the high voltage side, the regulator would keep the arduino at 5v until the high voltage side dropped to 5v, wouldn't it?
Or is that something I should not worry about?

>>2599399
Capacitors are not the best solution for powering a device while the main power source is off, you use batteries (rechargeable if possible) for that. If your arduino uses maybe 1 mA at 5V you could get away with using a capacitor, but if it draws something like 1A, there is no way you can power that with just a regular capacitor, you'll need a pretty big capacitor that would be significantly larger than a simple battery with a charging circuit would be, and most likely more expensive as well, and you still have to deal with the inefficient conversion. There is a better way to accomplish what you want, which could be a nice project: ditch the voltage regulator and use a buck converter with a rechargeable battery backup system with a charging circuit. It's a bit more complicated but would be the ideal solution. Should you still want to use a cap and diode, here is what you do:


1. Measure the current that the arduino draws (preferably peak current to have error margin), this is important. You could try to measure this with an ammeter or you might get lucky reading the datasheet
2. Measure the voltage, this is also important (voltmeter or datasheet)
3. Determine the resistance = Voltage/current
4. Determine the capacitance using the RC time constant by solving equation shown in pic related where:
* Vs is the initial voltage of the capacitor at the interuption of the battery power (Vs = Vbattery(12v) - Vdiode(~0.7V for silicon).
* Vc is the voltage at which your capacitor voltage is no longer enough to output 5V across your regulator (check the datasheet, different for each regulator, usually 0.3-0.7V above the output voltage, i.e. 5.3 to 5.7v)
* R is the resistance found in 3)
t is the time for the capacitor to output 5V, in this case you stated 30 seconds.
* C is the capacitor value you need in Farads ( convert to uF).

Not the best way, but very cheap with minimal components. Critical: Diode required to prevent cap discharge during cranking.

>>2599393
A 5.5V supercap is like a dollar or two. If you remove the power LED (and any other LEDs that might light up during use) from the arduino, then operating it in power-save mode will draw a couple of µA. Say you can handle a ∆V = 2V as the cap discharges, we get:
>I = C*dV/dt
>I = C*2/30
>I = C*15
>C = I/15
If I=10µA, then:
>C < 1µF
Considering there's probably a 10µF MLCC on the board, you're probably fine without an extra cap if you design your circuit properly. Note that having a negative voltage across your linear regulator is likely out of spec, so you'll probably need a diode on the output. The 0.6V drop will be a bit of a pain, but the arduino should run just fine regardless. Assuming the I/O that activates the ignition is something like a MOSFET that doesn't have a significant drain current and will run just fine at 3V or whatever.

Also consider a CMOS 555 timer, it uses 250µA quiescent current but it's probably a lot simpler.

>>2599393
>I don't really know how much current it draws,
"According to this page, the Nano takes 20mA when running, 17mA with the blue power led removed. That is not standby/power down current, which should be a lot lower. The FT232RL has between 15mA full operational current, to at most 2.5mA or minimum 70µa (0.07mA) in USB suspend mode, or with reset pulled low."
https://electronics.stackexchange.com/questions/68938/power-consumption-of-arduino-nano-current-draw-low-power

what's your power supply? a battery? consider placing a switch in series with the supply (FET or relay), and getting the arduino to count 30 seconds before sending out a shut-down pulse

So many projects, so little time on the weekend

just ordered generic first surface mirrors from aliexpress, wish me luck
the intention is for the laser itself to be mounted static while the light is guided by mirrors on the gantry, but for that i first need to collimate the light out of the diode so the changing distance of the gantry doesn't impact the optics, so i ultimately need an additional lens in the system
hope this lens from a 0.5W laser will cut it

>>2596827

>>2599686
Why do so many EE textbooks use the word magic?

>>2599689
>Why do so many EE textbooks use the word magic?

They're either gaytard fags or they don't actually understand what they are writing about. Any time I hear someone say that something happens "automagically" I nearly puke.

>>2599689
Because even though all EE is based on physics, the theory is so far removed from actual real world which has so many innumerable factors that you cannot control that its often best to learn just the most significant subset of the said physics which will let you design stuff even though you don't have much knowledge of the underlying theory.
If you don't understand the underlying theory 100%, it might as well be magic

>>2599697
>>2599705
Nice duality.

>>2599689
Electricity is IRL magic. Nobody really understands why or how it works exactly, we have approximate models that can be very precise, but exact nature of it is still unknown. We're summoning strange energy from another dimension that is able to make rocks think and make things move by themselves.
https://www.youtube.com/watch?v=5LOtawvIGZI

I need to make an electrical wiring diagram but i cannot find any software that lets you do it. I think i will be forced to use Paint to make it but it makes rerouting or editing things ass. Or do it in layers with GIMP idk.

I am trying to slim my amount of possessions. One thing I have is a "power conditioner with light module". It looks like the same one as pic related but without the cool LEDs in the middle.

What does the light module do? Also curious about anything else this thing might do in general.

>>2599883
autocad or qelectrotech

>>2599883
KiCAD

>>2599883
gEDA

>>2599883
There's a fucking section for it in the OP.

>>2598467
I'd use crimp connectors. Brass solders well.
If you want to learn a useful skill, order some resin and pot the electronics. Takes care of insulation and enclosure at the same time.
https://www.youtube.com/watch?v=h8ycmroFQSs

>>2599925
found some answers
>2 slide-out, swiveling light fixtures for rack illumination
>Dimmer control for lamp brightness

>Triple-mode varistor spike protection
>RFI filtering

so theyre just to illuminate the equipment on the rack apparently.

not sure exactly what triple mode varistor spike protection is

>>2600058
it's audiophaggotry, shitcan it or resell it to someone equally retarded

>>2599189
Are you blind?
It's not connected properly.
The top is off to a degree where the pins could literally be shifted one full pad to the left, which would mean the ICs SDA line is NC.
Friendly reminder that this is /ohm/, not /csg/

>>2600058
>>2599925
>triple mode varistor spike protection is

>>2599689
Same as with classical mechanics.
The smaller and faster a circuit becomes, the less your knowledge about electronics is worth and the more you have secondary effects take over that you can barely predict. A high speed circuit can work in theory but produce gibberish when actually put together and it can take anywhere from a few hours to literal months to figure out why.

>>2600067

>>2600071

Hello /ohm/, /g/ tourist here.
Thinking of making my own PCB's to house 200+ 18650/21700 cells for e-moto battery packs in 14p20s config.
I need to pull about 300A from the pack and my current way of doing it is spot welding nickel strips for the 12p, soldering thicker copper strips for the 20s, and shrink/kapton wrapping them up into packs and chucking them in a box
It's pretty tedious and labour intensive, I was thinking of making a PCB-box that I can just throw the cells in and spot weld them, but not sure if there's any PCB's capable of handling 300A in this config without ridiculous pricing.
I see some guys doing smaller packs with thick copper PCB's - 48v 50A kinda stuff, but nothing 100+ amps, not sure if it's even possible/feasible.
what you you guys think?

>>2600098
Usually, you should avoid using PCBs for very high voltages or currents, but if you absolutely must, you usually use a PCB bus bar. It would probably be easier to use wires.

>>2600098

You'd need some very heavy copper in order to support 300A with the amount of space you're going to have available between cell groups, or some unusual construction perhaps involving multiple stacked PCBs. There are more practical ways of doing it, including just straight-up soldering wire/copper sheet to the traces.

Most high-current packs just don't use PCBs as anything more than an intermediary between the cells and bus wires, if they use them at all.

The most practical construction method I've seen is just a basic holder for the cells made of wood or plastic. The holder has an open grid on both sides to both hold the cells and allow space for bus strips. Pic related. They're common and inexpensive, and I believe do exactly what you wanted your original PCB idea to do. It's what I did for my ebike, except I just printed the holders myself and the cells had screw terminals, so no welding required.

>>2600131
I can smell this image

>>2600138
What do you smell, Anon? I can't imagine it.

>>2599983
>>2599986
>>2599992
Can any of those do wiring diagrams? I think they’re all just for schematic diagrams.

>>2600058
I think it’s more likely to mean variation protection on both the AC input lines, and the AC output lines, and after the rectifiers. Power surges kill things, when your multi-thousand dollar magic crystal powered equipment is at risk I don’t see why not spend another $5 on varistors to cover extreme edge cases. Though with any luck they also have an appropriate array of chokes and filter caps.

>>2600131
Very nice image. Makes me want to do the Great Scott LED soldering intro but with backwards elcaps instead of LEDs and have them all blow up when you plug it in.

Where to get cheap vacuum variable transistors 2-10kw 1-50pf

>>2600153
Damn i meant capacitors, not transistors

>>2600142
exploded capacitors are among the worst things ive smelled in my life

kinda like a mix of burning plastic and sulfur

>>2600225
It's like a cancer patient pissing on a wood stove.

>>2599992
i dont think any of these can do that
KiCAD 6+
Circuitmaker
Logisim Evolution
>>2599983
KiCAD sucks because you need to make footprints for parts, otherwise you have to use 'connectors' to act as stand in for parts you want to connect to but yeah i can try that.
>>2599986
this sounds promising.

>>2600267
>because you need to make footprints for parts
Do you mean symbols instead of footprints? You could use sub-sheets instead and just make them empty, and while that's a bit of a hacky fix by itself it does give you room to flesh out the internals of each block.

>>2600309
yes symbols, footprints are for the PCB. But yes thats the idea. Looks workable, if i can get nothing else to work, i will have to go with this.

>>2600064
i am a baka mechnigger what can I do :(

Is putting a 1A/230VAC switch fuse before a 150W fan motor an acceptable idea or useless?
Last summer heat hit a lot of shed fans/"ventilation systems" hete, causing quite some fires. I would like to counteract before it gets really warm again, but am not quite sure how yet.

>>2600629
If motor is made from low grade chinesium then it should never be running unattended. Fuse will make it a bit safer, but you have to make sure that it won't trip due to inrush/turn on current. If you really want to protect against fires then maybe add in AFDD (arc fault detection device), but those are expensive.

>>2600647
Fans are italian, but old. I guess 90s, still running though. My fear is a bit subjective maybe, they have been running fine ever since, but last summer it struck several of closer farms and it was always related to fan motor shit. AFDD sounds like a great idea for a whole barn instead of a single fan. Frankly, anything else I might throw in? My knowledge is mostly how to repair machinery as needed, not electrical installation.

>>2600657
thermal fuse

>>2600143
>Can any of those do wiring diagrams? I think they’re all just for schematic diagrams.
I don't see what the difference between wiring and schematic is. You can place the components and wire them any way you wish, you can create your own symbols as well, so I'd assume yes, it does wiring diagrams unless you mean something else by it.

>>2600131
where do you find those images? Any collection? I've seen similar images posted before, would love to have the entire collection

>>2600267
>KiCAD sucks because you need to make footprints for parts
or you can download pre-made ones. Most parts have KiCAD footprints that can be downloaded from places like ultralibrarian, and if they don't have it you can make a request to have them made for you for free.

>>2600757
saucenao

>>2600657
>always related to fan motor shit

best way to keep motors cool is to make sure they're well-oiled, or even better, well-greased.
and adding a thermal fuse (as mentioned above) to the coils, or the metal case will increase safety in case the oil dries out.

>>2600756
>You can place the components
The thing is, a wiring diagram doesn't feature op-amps and resistors. It features entire blocks like an ECU and PLC. Each should be defined as a box with proper labelling and positioning of pins as a function of purpose, not indicative of actual location. The box can be subdivided into areas denoting key sections like a CAN bus or dotted lines denoting electrical isolation. While it is possible to make these as custom symbols in KiCAD, it's a tedious process that requires you go into a seperate window whenever you want to make even a minor edit to the symbol. And KiCAD flat out doesn't support things like colour coding groups of wires.

>>2596815
>12V Motor, max at 11A
What parts do i need to control this machine with a low power signal from a pi pico? Now first found upper pic related which does exactly what i want, but isn't available in my area, at least not with paying additional import fees. Now aliexpress to the rescue suggests me lower pic related. Now am i just too mistrusting or does this thing not look at all to handle 130W? As i feel like i'm clearly doing something wrong, is it just the keywords "H-bridge", "single channel controller", "Mosfet pwm controller" or am i looking for the complete wrong thing?
>inb4 it's easy and you are dumb
I am literally despairing over this shit. This should be easy.

>>2600783
Is it just a simple on-off load? Or do you need to invert direction? Inverting direction is what an H bridge is for, otherwise you only need a single low-side MOSFET. In all cases putting FETs in parallel will increase current carrying capacity, but it's usually cheaper to just buy a higher spec transistor (e.g. not a 50 year old design from aliexpress). Current carrying capability is what matters, not load power rating.
If you need to switch it fast then you want a dedicated high-current MOSFET gate driver, otherwise just a logic-level-transistor and a pullup should suffice. Cascade two of them if you don't like the idea of it being normally on.

>>2600794
>Is it just a simple on-off load? Or do you need to invert direction?
Invertion would be a plus, but is not needed for now.
>otherwise you only need a single low-side MOSFET.
>In all cases putting FETs in parallel will increase current carrying capacity, but it's usually cheaper to just buy a higher spec transistor (e.g. not a 50 year old design from aliexpress).
Oh, ok. So basically one mosfet, pullup resistors and this could be done. I'm a bit confused what all the cap magic on my first pic was for then, but I keep that for later.
>If you need to switch it fast then you want a dedicated high-current MOSFET gate driver, otherwise just a logic-level-transistor and a pullup should suffice.
Defining fast, i would've gone with 10khz/adapt to whatever solution i would find.

I know practically nothing about elextronics, but I wanted to be a technician until I learned that wages and jobs are not in good shape. Is this something you can make a good career out of outside of being an engineer?

>>2600818
10kHz could definitely be fast. I’m assuming you’re talking about PWM? You can either make a high-current gate driver with discrete transistors like this guy:
>https://www.circuitden.com/blog/11
Or use a dedicated IC like an IR2148.

Whether you actually need a high-speed gate driver is a function of your switching losses, and how significant they are. Calculating switching losses isn’t that easy, but what you can do is calculate the time it will be transitioning (some form of gate charge put into the I=C*dV/dt equation) and see what that’s like compared to the cycle time.

>>2600898
If you work on expensive equipment your pay could be high. Hi-fi or home entertainment stuff, if not commercial or military equipment is what I'd aim for. Repair is going to be more intellectually stimulating than assembly, while design/prototyping is going to be better than both. If you only have assembly jobs available, aim for a position that has the potential for upwards mobility, smaller companies without dedicated HR teams are probably more likely to let any old unqualified yobbo climb up the ladder compared to a big company. Larger companies are also more likely to have automated everything besides inserting THT components and packaging odd things, I ended up in a job doing nothing but tying cables in knots and putting them in plastic bags for hours. Handed in my two weeks of notice a week into that job, now working retail again but it's at an electronics shop so it's fun when we don't get schizos.

what are some ways to dim LEDs other than PWM?
Particularly, a way that could work for LED displays.
Being able to change from full on to full off and vice verse in 1/60th of a second would be good, faster preferably, but even like 1/24 would be good.
For starters and as a proof of concept, I wouldnt mind say only a few brightness levels (ie 4 levels:0%, 33%, 66%, 100% for example)

>>2600969
>Being able to change from full on to full off and vice verse in 1/60th of a second would be good, faster preferably, but even like 1/24 would be good.
you can easily do that with pwm

>>2600969
If you PWM in the low or mid kHz then that's easy. There are some reasons to avoid PWM, mainly just aliasing with fast shutter speeds on cameras and optical/IR receivers, the other option being basically a current-mode buck converter.

Are you designing an LED display from scratch? With so many LEDs it's infeasible to add an LC filter to each, so constant-current switching is probably out of the question. Varying current linearly using BJTs (or JFETs lmao) is an option, but you'll effectively need a DAC for each one, with as many bits as needed for your full colour depth (e.g. 8). A monolithic resistor ladder (apparently these exist) as a current-mode DAC is probably the cheapest way to do this, otherwise you could have fewer DACs multiplexed via sample+hold circuits, or just analog switches and only turn on certain pixels at once. It may be possible to multiplex analog methods with rows and columns, but I'm not sure how (please tell me if it's possible with 0-100VAC).

But I think PWM is the way to go. Dedicated display driver chips can likely do this reasonably easily, but you have to know what you're looking for. You can kinda multiplex a few hundred LEDs just with 74HC595-like shift registers driving rows and columns (probably also higher current driving chips like ULN2003s) and a microcontroller, as is done in LED cubes. I'd do the math on how many sub-pixels you want to see the required bit rate, and see what would be required to piss that out ~16 I/O pins at once. Pretty sure a ~100MHz MCU can handle that kinda stuff.

For example, if you want to drive a 360x480 display at 60fps with RGB pixels, that's 518,400 sub-pixels. For 8 bits of colour depth, you need to give 2^8=256 clock cycles per frame, which is 15360 clock cycles per second per sub-pixel. Multiplying by the two and we get 8GHz. That's not really feasible without an FPGA or ASIC. If we scale things down to 150x100 and 4 bits of colour depth, then it's only 43MHz, which is doable with DMA.

>>2600922
>use a dedicated IC like an IR2148
See >>2600783
I would like to have the tolerance to open all up without blowing up in between.
>make a high-current gate driver with discrete transistors like this guy: https://www.circuitden.com/blog/11
That's a lot closer to what i had in mind purchasing at first, but given
>>2600794
Shouldn't be pic related enough for my purposes, as same thing as aliexpress is offering, but checking mosfet specs first?

>>2601017
Without Dir. It was just the best N-mosfet depiction i could find in kicad.

Quick rookie question,
what dedicated tool do you use to apply Soldering Paste on board ?

>>2601103
syringe
toothpick
pin tip
stencil and card
...to name a few

>>2601150
ok thanks
I did yesterday with some tiny desoldering handtool, the one you push into pinholes with help of the soldering iron, but I felt it wasn't very precise as it's some special metal.

>>2600985
>>2601003
>what are some ways to dim LEDs other than PWM?

>>2601003
>Are you designing an LED display from scratch?
Yes

>Varying current linearly using BJTs (or JFETs lmao) is an option, but you'll effectively need a DAC for each one, with as many bits as needed for your full colour depth (e.g. 8). A monolithic resistor ladder (apparently these exist) as a current-mode DAC is probably the cheapest way to do this,

>otherwise you could have fewer DACs multiplexed via sample+hold circuits, or just analog switches and only turn on certain pixels at once.

>It may be possible to multiplex analog methods with rows and columns

All great ideas, thanks very much

What is my best sensor option to detect motion/human presence in a 1m radius which will work outdoors under direct sunlight? Pir and other infrared sensors are no-go and I'm afraid a mmwave sensor will be triggered by leaves/moving branches on a windy day.
I just want home assistant to send me a notification if there's someone out of my door.

>>2601207
Use a camera when it triggers?

>>2601207
https://www.youtube.com/watch?v=WVYAR1xwHKg

I'm a newbie when it comes to audio stuff and I'm practicing making PCBs with this headphone amp design. My question is, assuming all optional parts are populated, which of the caps need to be film and which are okay being electrolytic?

At the moment I think C2 and C4 should be film, but what about 3a and 5a since they also touch the audio signal? C3 and C5 would be prohibitively expensive to get in film

imagine the current

>>2601251
google ssmh bom and it'll even give you the mouser links of each component so you can check for yourself

>>2601191
logarithmic current source. just don't exceed their max current.

I have a mains to 12+12V center tapped transformer laying around, anyone has some good ideas of what to do with it?
It's from an old linear PSU project i did for school some years ago.

>>2601336
A new linear supply. For a chip amp.

>>2601251
film is unpolarized, electrolytic is polarized. see the little + sign? that means you use electrolytics.

>>2601017
For an actual MOSFET I type in NMOS or N-ch or something like that. Assuming you’re implying no driver at all, you have the problem that you’re feeding it your I/O voltage instead of 12V or whatever (FETs turn on better at higher voltages). Also the gate drive current will be lower, and so your turn-on time will be longer, meaning more switching losses. If you’re switching quickly the extra losses could definitely kill your FET.

If you did imply a driver with that symbol, use an external one instead.

>>2601207
Why no PIR?

>>2601352
Film caps are going to outperform any electrolytic cap unless the ESR was explicitly taken into account. Generally there’s no downside other than cost in replacing electrolytics with film caps. For these high voltage uses, you may even find that X or Y caps make the circuit much more failsafe.

>>2601207
When the “star wars” movie came on TV, it went all snowy again when we went back to the couch after adjusting the rabbit ears. To fix it, we had my sister stand right next to the TV and antenna and that cleared it up. I guess that’s what commercials were for… breaks for the snow remover job.

So, probably you could detect a person put into an RF tank circuit between 50 and 200 MHz.

>>2601287
Thats a 15A vsb (very slo blo) fuse.

I have shit ventilation in my apartment, and the window is impractical to use. Will a small fan with a piece of carbon filter be enough to remove excess solder fumes?

>>2601207
depending on your required fov you could use ultrasonics but they arent weatherproof

>>2601207
Consider a doormat pressure sensor, or even a vibration sensor fixed to the steps.

>>2601368
Yeah a large-scale theremin could be an interesting project. Probably really prone to outside noise though. I'd make it frequency modulating as opposed to amplitude modulating, for noise immunity reasons. Circuit would be similar to an RF microphone.

>>2601400
Should be, so long as you can see the majority of the fumes being drawn towards the fan, and the fan forces all the air through the filter. Adding a duct hose might be worth considering.

>>2601368
Wonder what they did with a BILLION little tv tuner modules after the “digital revolution” <spit>
You could makeuseof.com one if those.

>>2601207
theres a simple program for linux called "motion" available in most distros default repositories. It will start recording via a webcam when it detects motion, and stop when it detects it has stopped. theoretically you could get a $5 thrift-store webcam and a $5 rPi Zero (i mean this t theoretically) and youd have what you want for $10. of course running an SBC running Linux is probably overkill for what you want, but figured Id share

Holy shit this active pull-up resistor is genius. The instantaneous sinking and sourcing current is way higher than the steady state current when the output is low.
Does anyone know who came up with this idea? I first saw it on this blog: >>2600922

>>2596815
Dumb Question:
I have a 12v DC PSU and a SBC that needs to be powered with USB-C PD. What is the saftest way to get this done?

>>2601477
5V regulator, "7805" or similar

>>2601477
>12v DC PSU
>>2601478
>"7805"

for better efficiency and less heat you can take apart one of the dozens of car chargers you've been hoarding for the last decade.

>>2601456
>1 picosecond rise and fall times
Kek
I don't think even most micros can go as fast as 5 nano

High ohm esd wrist strap, or low ohm, /ohm/ies? I've seen both.
Cable on mine just completely desintegrated

will i get lead poisoning if i dropped molten lead solder on meself?

Am I doomed to become cognitively challenged

>>2601550
>will i get lead poisoning if i dropped molten lead solder on meself?
no
>Am I doomed to become cognitively challenged
absolutely

>>2601550
Yes to the first question.
The second question is a red herring. You were born cognitively challenged. You're doomed to become cognitively dead. Thank you. Please drive through.

>>2601514
>High ohm esd wrist strap, or low ohm

if it's gonna accidentally touch 120Vac, would you prefer 1M in series or 1K?
(personally, i've shocked myself so many times, it wouldnt matter coz i'm immune)

>>2601554
I thought wrist straps were meant to protect unpowered/unconnected devices from ESD damage.

>>2601557
He said he shocked himself many times, so there's that...

>>2601492
To put that into perspective, ⅛" is how far electricity travels in 1 picosecond. Put another way, this is 1 THz, and is well on it’s way to becoming light.

Doing that stuff in a component model simulator is a lot different that building and testing it where fractions of a millimetre count.

>>2601557
>wrist straps were meant to protect ... devices

you're attaching a wire to your body.
actually held in place quite solidly.
so you have to be mindful of what that wire is gonna touch.
intentionally or by accident.
just common sense.

uncapping an Amiga. A pad on C304 came with.
it's audio decoupling area.. shit
how could I fix that Anon ? Wire with the circle above ?

>>2601684
Yeah see if you can solder a jumper wire to the nearby via, otherwise follow the trace back to the nearest intact solder pad and put your jumper wire there. Nice magnet wire is what I'd be using.

>>2601684
Did you rip the via out too?

>>2601730
sorry, what is the 'via out' ?

Maybe you mean the burned aspect ? Cap had a leakage, I put kapton all around and removed caps with hot air.

I've brushed more with alcohol then isopropil since, and it seems to be okay, maybe cleaner than the pic. At least I hope. Maybe the phone crisped the picture.

>>2600575
Learn about electronics for starts.
Consider taking courses or reading up on one of the books in the OP.
Electronics and mechanics go hand in hand and being fluent in both can open you a lot of doors.
t. 3.5y training in electronics and currently getting a degree in mech.eng

>>2601737
A via is a hole leading to another layer of the PCB with a conductive path running through it.

>>2601716
So I think the road leads to some SMD capacitor in which I will plug a strip. I just tested continuity and it bells.
Thanks for the tip.

>>2601747
I understand. Thanks for the information.
Leading the capacitor into that hole could be then "less intrusive".

>>2600225
>among the worst things ive smelled
Ever ran a lab power supply through a pencil lead?
Also hydraulic oil. Rotting corpses ain't as bad as fucking hydraulic oil. God, I hate that smell.

>>2601753
Yes, you could use a through-hole cap with one lead anchored in the via hole if you don't want to repair the pad. Carefully scratch the solder mask away from the via until you see copper and tin it with solder.

>>2601753
>>2601763
How about just making a new via?
I never used picrel but I heard good things.
Obviously only works on two-layer boards but I doubt the Amiga has more than that.
They're quite cheap on aliexpress, too.

>>2601550
I could answer that, but you wouldn't get it; so take that as an hint...
>>2601768
Or you can literally just solder a copper lead wire through it like the old pre-smd days

>>2601777
>Or you can literally just solder a copper lead wire through it

Rivets are so much faster it's not even funny. If you have a lot of vias, rivets save a ridiculous amount of time over fiddling with a tiny scrap of wire.

>>2601768
How would this help? Are you suggesting he finds some space that’s free from traces and fills and just drills another hole or two? He’s still need to make the electrical connection, arguably he can just make an unplated hole and anchor a THT cap there, with the leads bent taught and soldered directly onto the correct pads. This would require that the cap is on a different side of the board compared to the jumper wires, but the same basically applies to using rivets if you don’t want your capacitor floating in the air.

Also via rivets are kinda massive, and that via hole probably isn’t big enough to fit a capacitor lead into. Maybe has solder mask filling it too, idk.

Rivets and soldered copper wire vias makes your board lumpy, hand-riveting solid core wire is where it’s at.
Also electroless silver plating is pretty neat.

>big uni club project involving a $300 IC 6 months in the making
>be a co-lead for the project
>finally got all the parts in
>finally got all the boards in
>finally got all the parts soldered on, including the $300 IC (we have fuck all for funds)
>first test of functionality with all of the components on
>the other co-lead is taking the reins and testing it
>instead of easing up to the working voltage from a lower voltage (preferably 0), he sets the power supply to the IC maximum voltage and connects it to the IC
>3 amps flow through the IC
>the IC is toast

>>2601934
What is it? Some high end opamp? ADC? FPGA?

why didn't anyone tell me you could build modulars like this? this is all your fault /ohm/

I've put some wire as it is for me the easiest way to do it.

>>2602017
Nice.
Why not enamelled copper wire though?

>>2602042
Well, I took what I had in stock. picrel.
Since I'm learning and don't have that knowledge, "Enameled Copper Wire" is a good tip about what to order for useful use.
I feel Electronics is like woodworking, you always need something.

Any resource on how to build a speaker ?

>>2601811
Honestly, never used, so wouldn't know; how do they contact intermediate layers (assuming multilayer)? Seems like it's not as easy for solder to permeate through to those.

How do I design an impedance matching circuit if I don't know the output impedance of my RF tranceiver IC? It's a differential output.

The datasheet says its load impedance is 15Ω+j88Ω but I am not familiar with that term.
The datasheet shows a matching circuit but I would still like to know how they came up with it and if maybe one with less parts can be used.

>>2601934
>putting a fuse on your board: 10c
>destroying your club project: priceless

>>2602177
It's a complex impedance thing. The impedance of a C farad capacitor is -j/(2πfC) while the impedance of an L henry inductor is j*2πf*L. Replace 2πf with omega.

>>2602283
> -j/(2πfC)
What do you plug in for “j” though.

>>2602177
Is that the application datasheet? looks like they designed it for you.
Standard RF stuff is 50ohm

>>2601990
a DRF1200 RF amplifier

>>2602283
I told him we needed a fuse but he overruled me. Shit sucks.

>>2602404
>he overruled me
Remind him of his expensive failure as often as possible. Print pictures of fuses and scatter them everywhere he looks. Wear a shirt with a huge fuse graphic. Nickname him Bernie.

does anyone have a recommendation for a high speed (at least 13.56 MHz) DAC IC?

>>2602348
That’s sqrt(-1) = j = i = the imaginary identity
With the identities j*j = -1 and 1/j = -j plus some extra techniques, you can basically just treat it as a variable and do algebra as normal. Presumably you want the output to be some whole number purely real impedance, so you calculate the real and imaginary segments as a function of some matching network components and calculate back. It’s fairly easy to modify a pi or tee matching network to also cancel out an arbitrary imaginary component. Since the output impedance is 15+88jΩ, you can just add -88jΩ to before a conventional matching network in the form of a series(?) capacitor. So for a CLC Tee filter you can just augment that capacitor into the first one by changing its value a bit with the series capacitance calculation. I think. Maybe it’s better to use a parallel component here, idk.

>>2602423
A monolithic SMD R2R resistor array plus an RF op-amp of your choice.
Otherwise look for chips used in producing VGA video signals.

>>2602430
I forgot r2r dacs existed. If I just want to generate a 13.56 MHz sine wave from a microcontroller DDS, do I just directly output the PCM onto one of the mcu pins into the R2R dac?

>>2602452
Discrete R2Rs want parallel digital signals. You'd need to either use the MCU to present the full N-bits on a bunch of its I/O pins (e.g. using one or two direct port writes like PORTB = 0b01011010) or feed them to a serial-to-parallel converter like a latching SIPO shift register. That first option will be much easier and very lightweight on processor time, assuming you have the I/O to spare. If it's an 8-bit MCU then you may need two port writes, but chances are you're dealing with a 32-bit MCU at these frequencies, so it's basically a single clock cycle. Not sure if you can DMA parallel pins, maybe.

In addition, your clock rate will have to be at least twice that of your maximum desired frequency (see: Nyquist frequency) and you'll need an anti-imaging filter for proper wave reconstruction that fits between your passband and stop-band.
For example, if you want 10 bits of resolution, that's a minimum attenuation of 60dB. If you picked 13.56MHz as your maximum frequency and 100MHz as your DAC frequency, then that's 0.87 decades, so you'll need 70dB/dec ≈ 4 poles of filtration to jam all of that attenuation in there for an ideal anti-imaging filter. I think. You may be able to skimp on that somewhat, or cheat it by using a notch filter. An elliptic or chebyshev type-2 both have notches in them you may be able to use to stamp out certain present frequencies. R2Rs tend to have issues with code jumps (e.g. 0011111 -> 0100000) where the bits don't update at quite the same time, so you could need more or less filtration than I described. I think current-mode resistor ladders (1R-2R-4R-8R etc.) are the same in this regard. See also:
https://www.analog.com/media/en/training-seminars/design-handbooks/basic-linear-design/chapter8.pdf
https://www.analog.com/media/en/technical-documentation/tech-articles/guide-to-antialiasing-filter-basics--maxim-integrated.pdf
https://www.ti.com/lit/an/sbaa001/sbaa001.pdf?ts=1681970604394&ref_url=https%253A%252F%252Fwww.ti.com%252F

>>2602430
> identities j*j = -1 and 1/j = -j plus some extra techniques, you can basically just treat it as a variable
Thanks for the answer to get me started there, this is worthy of knowing a bit better. I can’t be the only one this helped.

>>2602054
i dont like enameled wire. what you did is my preferred way, except I would have went with a smaller gauge.

i dont like enameled wire because enamel is harder to remove than normal wire insulation

>>2602508
light a candle and burn it off

>>2602017
You superglued it down, right? Since that bottom side isn't bonded to anything it might rattle around a bit and be in danger of fatiguing its other solder joint.

>>2602506
There's probably an entire chapter on complex impedance and another on impedance matching in Art of Electronics, same for any other electronics textbook. Find that in libgen and you should be good. I often just learn stuff from wikipedia or equivalents, but in particular here there's sorta an order in which to learn things, so the section-by-section style of a textbook might be better.

>>2602508
Enamel wires are good for really thin jumper wires, like computer motherboard bodges. It's so small you can run parallel wires from a 0.5mm pitch footprint without trouble. For larger work, my only concern with conventional insulation would be it creeping back and loosening under the heat. With silicone or PTFE wire it would be fine.

>>2600131
Sex with cute japanese schoolgirls who smell like flowers

>>2602396
>>2602430
Yes but I still want to know the math behind it
I have seen people use smith charts and there are online tools that use smith charts like https://www.will-kelsey.com/smith_chart/
But still, I would need to know the input imepdance to solve this and I dont know how to deal with this differential output of the IC in question

>>2602547
Bowick's "RF Circuit Design" has a good chapter on it. Great book, highly recommend if you're getting into this stuff.

Having a hard time getting into this book right now.
I really thought my english knowledge would be sufficient at this point, but this does not account for scientific english apparently.
Right in the first chapter the book throws so many formulas and mathematical terms at me that I'm completely overwhelmed.
My math is also quite rusty at this point. Should I read the recommended book in the math appendix to refresh this knowledge and get familar with this kind of language?
>The authors recommend Quick Calculus by D. Kleppner and N. Ramsey

>>2602580
Its a reference book, don't treat it like a textbook
A reference book means you're supposed to look at it if you need to reference something

>>2602582
Thats's very disappointing.
I got it as a present years ago and was intending to read it from cover to cover to become some sort of electronics wizard, but never quite knew how to approach this doorstopper.
I guess I miss a lot of the knowledge the book preaupposes. At least I'm learning the trade right now and am getting lessons at a school, so I'm not completely on my own.
Which book would you recommend that I can work through cover from cover?
I mean I can always download stuff for free from libgen, though reading on a computer kinda sucks.

>>2602580
the book's written like shit, don't waste time on it. read something more beginner-friendly

how do i operate a relay using a microcontroller? i'm a midwit who can kind of understand how the relay works but not intelligent enough to understand what to use to switch the relay. i want to make an h-bridge. i am thinking of doing picrel. but i don't know anything about transistors and diodes and resistors.
which ones should i use? mcu pins put out 3.3v
>>2601742
i am trying. i really want to get into autosar programming, and automotive software development (even though i absoultely despise anything more modern than the Ford Sync 1 system, no one needs that bullshit, but i suppose it must be fun to work on.)

>>2602627
If you are using a small-ish relay, usually you can get away with a metal-can 2n2222 transistor, a resistor, and a flyback diode. Metal can is for street cred.

>Elon Musk has been launching star ships and I am trying to figure out a stupid 10 part schematic
>and GPT4 is here
Why live. Why DIY

>>2602585
I agree with others and I am never tired of mentioning this: that book is pure evil, it has done a huge disservice to the beginners because it is being pimped so hard for some obscure reason. This is NOT a beginner's book, this is NOT a textbook, you will NEVER learn anything from that book as a beginner, it assumes that you already have some EE background, the authors are just flexing on the unsuspecting beginners and show off their guru skills without any systemic approach. It is just a bunch of random building blocks without any through explanation of the theory behind them.

>>2602585
I'd be able to post parts of my professor-assembled uni textbook on the section, though first I'd catch yourself up on solving algebraic equations with complex numbers. I don't think calculus is necessary for this part, though basic calculus is behind why capacitive and inductive impedances have their values.

>>2602627
Firstly, use NPN transistors in common-emitter configuration. Secondly, flyback diodes are to be placed across the relay coil. Thirdly, you can get away with two SPDT relays instead of four SPST relays in order to reverse a motor and turn it off. With your and my circuit though you can't have it freewheel, only magnetically brake, in which case you'd need a third relay, or ditch the ability to magnetically brake by using a single DPDT relay for reversing and a series relay for turning the whole thing on or off. See "switching motor direction with DPDT relay" on google images.

Also you'll likely want an arc suppression capacitor if not a TVS diode across your motor.

>Repairing a faulty generic atx psu
Does anyone know about a UTC7608D ic(its in a DIP-8 package), i've seen a bunch of listings for it but i couldn't find a datasheet. Any help is very much appreciated.

>>2602743
SDC606P

>garage door keypad battery is dead
>keep procrastinating and just use remote from the car
>some time later
>completely forgot battery was dead
>enter code
>it suddenly works
>???
>disassemble
>9V battery measures 4V with no load
>apply a bit of load, drops to 1.2V
>basically completely dead

Explain this, einsteins? How the hell did it manage to open the door twice? Free will?

>>2602743
What's it connected to? With only 8 pins I doubt it's the main switcher, but it may be an auxiliary switcher for powering up the main converter. The name has hints of 8V linear regulator about it.

>>2602746
Where'd that come from?

>>2602747
Battery voltages can creep up over time as internal chemistry happens, though they still store very little energy. It's possible the voltage creeped up past 6V or whatever and was enough to activate the garage door once. Might happen again after another few months.

>>2602747
Is the battery for holding the code in memory during a power outage or powering the keypad?

>>2602748
>Where'd that come from?
UTC7608D is a clone of SDC606P.

>>2602749
It powers the keypad including the backlit keys
But it is a good question. When I replace the battery, I don't have to change the code. Oh wait. Maybe I do? I don't remember. But I doubt there is an additional button cell hidden anywhere. Must be EEPROM.

>>2602748
Hmm, possibly chemistry yes, it was a fairly new battery, but it is alkaline so it probably froze in winter and died and finally started to thaw out or something.

So why does the keypad need such a beefy 9V battery if the remote door opener works on a single puny 3V button cell and both just send some wireless signal to the receiver inside the garage? The backlit keys don't take that much power do they? And there must also be some kind of a little 8-bit micro (probably) that reads the input and compares it with the stored code, but it shouldn't require much current / voltage either.

>>2602755
Does it power a relay?

>>2602758
No it is wireless just like the remote
But since it is a very old keypad I suspect it has TTL chips or some other parts that may require high voltage, perhaps.

Need help to understand how they got the equivalent resistance on the right. I found rpi//R1, but they have rpi//R1//(1/gm), where does the 1/gm term come from? I assume it comes from the dependent current source, but I don't know how to solve that, what are the steps to do this?

>>2602582
>>2602586
>>2602695
Well, it comforts me that it's not incomprehensible just because I'm a retard.
What other book would you recommend then? Something with a hands on approach?

>>2602686
>>2602779
(You)

Anyone here know if it's possible to drive a diode laser at a really low power level? Like 100µW out of a normally 5W diode. Problem is my laser just has two wires going to its driver, no seperate PWM or constant-current dimming input, so there's probably a maximum speed I can PWM it.

I want to use it as a pilot laser. I plan on trying to modify or replace the laser driver to have a constant current dimming control pin, but I don't know if the laser will even run like this. I've heard of a "threshold current", but don't know what it could be for this diode. I'll be testing with this 0.5W laser until my 5W one actually arrives, which is when I'll really need the pilot laser. Some datasheets suggest threshold current is under 10% of maximum current, others suggest more like 30%. If not constant current, being able to PWM it sufficiently fast will be what I'll do. The laser is on a CNC that will probably be moving in the ~1m/s range with 0.1mm accuracy, so I'll want the frequency to be at least 10kHz. Not sure what an ender 3 v2 mainboard uses, I'll have to measure it.

>>2602794
Turns out an ender 3 uses 7.8Hz PWM. Gonna have to figure out something better than that, hopefully there's just an output enable option on the laser driver.

>>2602780
Well even practical electronics for inventors is better. Never mind the cheesy title.
I'd say use first year college books on EE. Basically two groups of books: circuit analysis (Alexander & Sadiku or Boylestad or Irwin, they are roughly the same) and electronic devices (Sedra & Smith or Boylestad & Nashelsky or George Floyd).

>>2602755
A lot of old wireless things (mics, rc car controls, etc.) have old designs that use a 9v battery. Or new things using the old design.

>>2602686
This applies to a lot of programming books as well. Most infamously it applies to K&R.

>>2602695
Thanks this is a good start
about the motor, it is supposed to drive a threaded shaft and does not need to freewheel, it is a Johnson geared motor at like 200 rpm. I don't need braking
I was thinking, if you switch both relays to states such that both are connected to negative, would that work? i mean sure there will be some back emf but this motor stops no more than 2 seconds after you remove power from it

Why is this faggory? Why are lots of opamps drawn like generic ICs not as a classic triangle symbol?

>>2602902
>if you switch both relays to states such that both are connected to negative
Yeah this is the magnetic braking I was talking about. By connecting either end of one coil to the same rail (i.e. connecting them together) you short-circuit any emf, causing them to produce some amount of stall torque to slow them down. This sudden torque may be an additional mechanical load on the system, but since it's a screw driven one that shouldn't be much of a problem.

If you don't care whether or not it has magnetic braking when it isn't moving, then the following two methods are basically equivalent:
>my dual SPDT relay diagram (braked)
>a DPDT relay for motor reversing and an additional relay for turning power on and off (not braked)
Both cases means two relay coils to control.

If you ever need to turn the screw by hand while the machine is still turned on the braking will make it rather difficult, but if not who cares.

>>2598892
Installing that into (I presume) your tv will kill it since the 2244 has a 75ohm driver, while I think the 2534 has just a voltage buffer.

>>2602794
>>2602803
Ok turns out there's a pin specifically on this laser driver for connecting to a dimming method. It can do both PWM dimming and analogue dimming, and all it needs for the latter is a 0.5-2.5V analogue voltage. Analogue dimming seems like a nice way of getting smooth brightness settings without worrying about jittering, while PWM dimming is probably best for a pilot laser since I'll be able to set the duty-cycle to pretty low levels with an MCU. Though to be honest it probably smooths out that PWM for constant current driving anyhow.

So the plan will be to have a constant 24V going to the laser from the printer main board, then 5V coming from the board too into an external MCU board to output some PWM and maybe analogue values too. I'll see if there's an extra pin to automatically run the pilot light with a pause in the g-code, though honestly just doing that with a button on the MCU might be more convenient anyhow.

>>2602909
> everything is a rectangle
Probably has something to do with the IEC.
As big clive recently said about it, “idiots were involved”
It’s just a kind of newspeak to form a society of retards that only they understand. Just ignore it, and it will go away. It’s wrong.

>>2602909
presumably so you can fuck around with the offset nulls, which aren't part of the standard model

>>2602952
You can just add more pins to a normal triangular symbol.

Even if you did want to go all IEC and have procedurally generated structures (e.g. a certain shape of box or thing means an offset input) you could still keep the damn triangle.

>>2602809
>George Floyd
lel
So practical electronics could be considered a textbook?

Anyone here have experience with Tag-Connect ? Are they any good ?

>>2603015
Never used one, but I've heard good things about them. Except for the price of the tool. IIRC someone 3D printed an equivalent.

I just use a clothes peg with three pin headers stuck in it, works fine for 3-pin UPDI.

>>2602995
Well It is sort of like a dumbed down textbook that covers a lot of stuff in a condensed form. Basically each chapter almost covers a separate course which you would need to take separately and use a dedicated textbook if you took EE in college. But I think it is a good starting point for beginners.

>>2603017
>3D printed an equivalent.
You can set leaded pogo pins in hot glue -- just put them in place with a properly spaced pcb or breadboard, glue blob, pull the whole thing off and you have a "connector." Just make sure not to use the conductive hotglue

How come multimeters have a "2"00mA range?

>>2603103
A E S T H E T I C S

>>2603103
electronics use binary which is base 2
dumbass newfag

>>2603130
>dumbass newfag

>>2603130
don't know why that made me laugh so hard

>>2603130
so there are no non-binary multimeters?

>>2603087
Ok. I'll look into it

>>2603145
>non-binary
I'm sorry you can't say gay on this board, there are under 18s here

>>2603145
Why aren't there any Roman numeral meters?
>MM Ohm resistor
>C uF capacitor

>>2603172
I want a meter that scrolls the reading across a retarded slow character LCD with a horrible backlight.
>probe resistor
>wait ten seconds
>one thousand five hundred Ohms

Actually, why are modern multimeters so chunky anyways?
I have this btw (yes really) and it fits in my palm

>>2603177
That one is smaller because it's analog and only measures voltage, current, and resistance. Analog meters are cool.

>>2603177
I don't know what you're talking about. My cheap chinesium ANENG fits into my palm as well, and it has all the functions.

>>2603170
ok, Ron, we just need your signature

>>2603203
Hmm I have a tiny ANENG too but I am going to buy a bigger one, more counts too.

>>2603215
I have a tiny Wang.

>>2603218
Same, that's why I want to compensate with a massive gargantuan bright yellow ANENG

>>2603219
So you can see it in foggy weather?

>>2603222
that, and also helps with mating.

I'm trying to build a simple spray booth for model cars and stuff (lacquer paints). I ordered this bilge blower that was suggested, and realize I dont k ow how to wire it to a wall socket:https://www.amazon.com/SEAFLO-Line-Marine-Bilge-Blower/dp/B01GDK3OTO/
Then I'm told to get a regular 12v power supply. I went digging and found these in some junk. Will either work? And if so what do I do to wire it up?

>>2603227
The link you posted is a 24V fan.

>>2602896
Can attest to this. Learning ANSI C, and half of my learning was finding out what the fuck the book was saying. There's a particular section in the first chapter where you pass a char array pointer to a function, when pointers are not even explained until the last few chapters. It just says "pass the array's name into the function" without saying that an array is just a collection of variables where its name is a pointer to the actual variable.

>>2603273
I learned C on my own, and I can confirm that my first contact to C/C++ felt alien. Like what is void supposed to be? It's nothing, like wtf? Now I think learning C first is useful but I don't know if these textbooks are a useful contribution.

>>2603227
>Will either work?

you need both in series to get 24V.
wire them up like in the pic, connecting + of one to - of the other.
then wire + output to red wire of fan and - to black wire.
if no voltmeter, just try diff combos until it blows (as good as your sister.)

>>2603229
fuck, I linked the wrong one, I have the 12v version

>>2603354
It's possible that the inrush current of the blower is more than a DC PSU like that will like. If it's a simple brushed motor you can calculate the inrush current by measuring its DC resistance with a multimeter, and dividing 12V by that. If it's more than ~1.5 times the current rating of the power supply, I'd probably want to add a soft-start circuit, or an NTC thermistor in series.

I'm interested in making an ultrasonic cleaner for giggles.
If anyone else has given it a shot, what kind of transducer would be appropriate? I know piezoelectric is the standard but is there a specific kind of piezoelectric transducer or limitations I should be aware of?

I'm having to design an RF preamplifier now, that needs to have its inputs and outputs matched to 50 ohms.
Now, from what I remember of my active electronics class, finding the input and output impedances of transistor amps was the most difficult and tedious parts of the class.
On top of that, the impedances are dynamic and vary as the input signal varies. How am I even supposed to begin to design the matching circuits for the input and output of the amplifier given the dynamic nature of the amplifier input/output impedances

not to mention, can a basic two-stage BJT common emitter + emitter follower amp even be used at 13.56 MHz?
There's so little information on this shit that my intuition just can't reach consensus on what to do

>>2603357
What frequency you aim at has a fair impact on the kind of work the cleaner does. I hear higher frequency machines are more expensive but do finer work. The resonant frequency of your system is a function of the mechanical load on it, the less rigid your system is the lower the frequency. I think you also want to match the electrical resonant frequency to the mechanical one to make driving it more efficient, but I'm not sure about the interplay between mechanicals and electricals. I think you get a phantom inductance from the transducers due to their linear motion. Different transducers may be designed to run within a certain frequency range.

You want to use a resonant driver circuit, I'd want to use an H-bridge or half-bridge circuit with feedback to make it self-oscillate instead of just feeding it a square wave from a MCU, that way it should hopefully be at its ideal operating point. They're kinda high voltage though, so a good PSU might be hard to find (use rectified mains lmao).

Power ratings are probably the most important thing, size the rated power output of the transducer to be similar to what existing machines use.

>>2603394
I'm also a noob at that stuff, but I think W2AEW has some good videos on radio circuits and on and transistor amplifiers. With any luck he has videos combining both. And as always, look for existing solutions.

In a tuned resonant amplifier, we're essentially switching a transistor with a square wave to generate a high-power square wave, that we then shove into a tuned resonant tank circuit, that filters out all of the harmonics but (ideally) the first
this leads to a clean sine wave, but all of the other harmonics are effectively burnt power.

Is it possible to have several impedance matching/tuned resonant circuits in parallel, at the first few harmonics, to try and capture more of the power if an exact sine wave isn't mission critical?

>>2603398
I've found some real good videos from Keysight on youtube for RF amplifier design. They're not 1-to-1 what I need to do, but they're a great starting point

>>2603427
>all of the other harmonics are effectively burnt power
They're not burnt as heat though are they? At least in a push-pull design, an ideal one is lossless, minimal current just flows at the other harmonics assuming a high-Q tank circuit. No heat is burnt by the reactances, and no heat is burnt by the transistors since they're either not conducting or not dropping any voltage. Ideally. If you want to pump more energy into the tank circuit, then just increase your voltage or increase the energy capacity of the tank circuit. I think there are LC or RLC or coax networks or coax networks you can use to convert that "zero impedance" tank circuit output into an actual 50Ω.

Also I think that LTspice guy "Fesz" did some videos on resonant class-D amplifiers recently. His practical implementation isn't brilliant but the simulation and theory behind it is solid. IIRC he talks about current-mode and voltage-mode drivers, and there's some interesting details behind those I hadn't thought about. I don't know if he gets into impedance though.

should everyone have a MOV in front of their home to suppress voltage transients (not current transients)?

>>2603394
>can a basic two-stage BJT common emitter + emitter follower amp even be used at 13.56 MHz?
Yes. Common BJT's like the 2N3904 can go up to 300MHz or something.
The main thing you need to be careful of is the BE-junction capacitance, which is about 25pF and has an annoying habit of shunting all your input signal to ground.
One way around this is to buffer all your inputs, so the RC LPF formed by your input resistance and junction capacitance has a higher cutoff frequency.

>>2603459
Would be a very big MOV. But maybe yeah, and it would tell you when it needs replacement.

>>2603482
Isn't that why you use common-base and common-collector stages at high frequencies? Cascading the two together would give you voltage gain without much change in impedance.

>>2603493
Cascode amplifiers address B-C capacitance (4pF * gain), but I have never seen anyone address the B-E capacitance (25pF), which has been plaguing me greatly

>>2603459

>>2603567
>contains no surviceable parts

>>2603568
>My God! it's full of sand!

I still don't understand conceptually why reactive power (transformers) let you lower voltage without dissipating heat as opposed to active power where for example a resistor converts electrical energy to heat. I know that reactive power is stored and then released, so I think I kind get it but it is still mind boggling that a transformer let you lower 1000V to 1V without dissipating any heat. I guess it it is probably because AC is a fiction, something is pulsating back and forth but then net result is zero.

>>2603615
Maybe think you got a thousand hands with a crank on one side of a mechanism, and o e output shaft on the other.

All thw hands crank together so the one outlet shaft has a ton of torque.

If you have 1000 cranks and 1000 output shafts it's 1:1 so it's much weaker.

Tye mechanism in-between isn't particularly burning up a ton of energy to get anything done either.

Like I guess using a resistor to power the power would be one crank and one output but you have a belt wrapped around the shaft to slow it down and it just burns up as heat

Thoughts on a zener followed by a voltage regulator as a transformerless offline voltage regulator for low current loads? Any issues apart from lack of isolation and overall inefficiency and having to have a beefy series resistor?

>>2602941
Hell even a bridge rectifier is just a fucking square. Absolutely insane. Totally unreadable. I only notice this faggotry in EDA symbols while various SPICEs typically use conventional schematic symbols as found in datasheets.

>>2603645
it's fritzing for chinks

>>2603647
Yeah I would strangle the arduino fags who popularized shit like fritzing but what about sources like SnapEDA or UltraLibrarian or componentsearchengine that provide symbols for most EDAs. All of them have this shit, too. Perhaps it has become the trend, as it is easier to draw quickly and they provide those schematics for free. So they just copy and paste a fucking rectangle lmao

How those motorized controls are called? This knob has "haptic" feedback and inernia, when you spin it it continues rotation like it has real mass to it. I'd like to use such control for one of my projecta.

>>2603645
>Totally unreadable.
that's what the actual component looks like lmao gtfo zoomer

>>2603672
can't tell if this is bait or you are this retarded?

>>2603672
What I normally do is I print out pictures of the actual components and cut them out with scissors. Then I glue stick them on construction paper and draw the traces with a white crayon. I graduated summa coom loud at Kenya International Community College.

I have a question that I posted earlier on /mcg/ but this thread seems more active
> Anons, I want to make a led strip for waking me up in the morning. I plan on designing a pcb for an esp32 with a phone charger for power supply.
> However soldering on the pcb is worrying me, I just have a cheap chink iron, is it enough if I buy some flux ?

>>2603715
Get a decent iron, MG Chemicals RMA flux, and Kester 44 60/40 or 63/37 flux-core solder. Avoid lead-free.

Add desoldering wick, something to hold the PCB, light & magnification, isopropyl, tweezers, etc.

>>2603639
Why not put the resistor before the rectifier instead of after it? And then while you're at it, replace it mostly with a capacitor instead. You still need a bit of resistance to stop inrush current.

>>2603657
I get results from "force feedback rotary encoder", but I'm not sure if they're any more specific than "haptic knob". Basically they're a BLDC, and I think the 3 hall sensors found inside some BLDCs would give you enough resolution if you use trigonometric/sinusoidal drive currents (e.g. PWMing the three channels with a certain ratio to get it between an angle).

>>2603715
Are you soldering SMT? Or THT? For SMT I'd definitely want non-shit tip geometries (read: not conical) and to practice your soldering technique a little. If you do design an SMT board, you can just chuck it on a reflow plate with paste and let it do the hard work. If you're going for THT, I'd also advise going for a USB B socket. They're nice and large and easy to solder. Even mini B isn't trivial to solder, let alone micro B or C.

If your technique is decent, you shouldn't need external flux when soldering a typical joint. With heavy ground planes and large amounts of solder it may be needed since the flux in the solder wire itself can burn off by the time you finish the joint, but this shouldn't be the case for common THT and SMT work. If your soldering isn't reliably leaving you with shiny well-filleted joints, just practice a bunch.
It's still nice to have external flux though. And wick and a sucker for when you mess up a joint.

>>2603715
>led strip for waking me up ... soldering ... worrying me

timer from the thrift store.
LED strip from the dollar store.
total cost $7.
zero work, zero soldering.

>>2596815
the fox electronics logo looks like a vagina

>>2603482
So, Common Collector -> Common Emitter -> Common Collector?
Do you have any tips on impedance matching the last stage to 50 ohms? I'm assuming I have to go through the entire analysis of all three stages, determining their input and output impedances at 13.56 MHz, and then finally using that impedance to craft a 50 ohm matching network

>>2603430
oh whoops lol I totally forgot inductors and capacitors just store energy, not convert it to power/waste heat
I love Fesz. I'll take a look at his RF amplifier vids, I see he has one on the Class C amp already.

well here's take 1.
a 13.56MHz preamplifier capable of driving a 50 ohm load, with a power gain of 2.93 dB. Real low THD too.
The new amplifier this thing drives needs the preamp to have a power gain of 34 dBm. I'm not at all familiar with dBm so I don't know if I've met that goal. The other issue is that the circuit is about as stable as a trainwreck in terms of component variability

I 100% expect this thing to shit the bed on the test bench when I get around to testing it

>>2603731
I don't want to be serious about electronics but just do a quick little project where I could learn a thing or two. So buying +100€ gear seems a little overkill

>>2603743
I want it to be compact, so probably SMT. I'll look a bit into THT if it's easier. Thanks anon

>>2603771
Same reply as first anon

>>2604117
SMT is definitely more compact. For a beginner though I wouldn't go for anything smaller than SOIC ICs. Whether to go for the common 0603 size passives is hard to say, you definitely need a fine soldering tip for that (1.5mm screwdriver tip maximum size, definitely not a straight conical). You'd have an easier time soldering 0805s or 1206s, but they start eating at your space savings. MELFs are cool though.

So zeners are virtually indestructible as long as the max current / wattage is not exceeded, but at least there is no need to worry about the reverse breakdown, since all the excess voltage will be dropped by the series resistor and the zener itself will never see more reverse voltage than it is rated at?

>>2604214
>zener itself will never see more reverse voltage than it is rated at?

by definition.

>virtually indestructible

nothing is.
sometimes they short out, just for fun.
i once had one get so hot, it desoldered itself.

>>2604132
speaking from experience, 0603 are fine for a beginner, 0402 on the other hand turned out a pain in the ass. i lose 3-4 before i solder 1 cause i can't fucking see them and i have shitty tweezers so they always run away from me. but if you have a choice fuck anything smaller than 0603. or even 0805.

is it economical to buy toroid cores suitable for mains frequency? i need a transformer for a linear power supply project and custom wound transformers seem pretty $$$$.

I need a quality set of ESD safe tweezers shipped from within EU.
At the very least I'd need straight ones one with a fine tip and one with a flat tip, and curved ones with a fine tip.
What reputable companies make good stuff? Anyone tried Piergiacomi ones?

>>2604132
>>2604247
So I did some research, I will use this strip : https://dephen.net/collections/led-strip-light/products/led-strip-lights-warm
For the control I will use pwm (i want a dimmer) and this little mosfet, PSMN013-30MLC.
For the ESP32 I will use the dev board with some female pin headers. It's not compact at all, but I will have the possibility to reuse it somewhere else if I want.

Hi, I'm looking for picrel but have no idea of its name. It's a push button with 2 states.

>>2604305
Is that kind of 5V strip bright enough? Also check that your FET will turn on with a 3.3V gate voltage, you may need a level shifter.

>>2604376
Looks like a computer power button, though the ones I have are momentary.

>>2604376
Maybe "latching plunger switch"?

>>2604376

>>2603678
Place I work for hired a guy to reverse engineer a board before I started working there and I still have the schematic he made.
It's basically pictures of the actual components.

>>2604669
Page 2 of the schematic is a bunch of disjointed circuits. He literally filled it with labels instead of just connecting shit together

There's a question in drone thread, which has not been answered yet, maybe you can.

A drone may have same brushless motors from same manufacturer but with different design RPMs. Is it because of number of windings? What's the relationship, the more windings the faster RPM or the other way around?

>>2604673
The key words are "motor Kv". The Kv is a term referring to the ratio of voltage to speed of a motor, as the two are roughly proportional for a given load. It's more of a result of the fact that torque is directly proportional to the current through a motor, and the back-emf being proportional to speed. More specifically, the torque is proportional to amp-turns, so doubling the number of turns while keeping the current the same will double the torque, and so the motor will run at a higher speed since there's more torque to fight against friction. But in reality, having more turns in a motor means having thinner windings, and so less current-carrying-capacity, and a higher resistance.

So if you double the number of turns of wire in a motor, you:
>double the series resistance
>halve the maximum operating current
>double the maximum operating voltage
>keep the output torque and power at these different voltage and current points equal
>decrease the speed and torque and power usage when at the same voltage as before
>increase the speed and torque and power usage when at the same current as before

Generally changing Kv values is just something you do for running different battery voltages, but it's also done to put the motor in a different operating point if you need more speed or torque, or more efficiency. The harder you run a motor, the more voltage will be taken up by resistive heating instead of EMF, but the better power you get out. You can just tell your ESC to put less current into a motor with less windings so you have the higher torque when you need it, but that does make your system a bit less efficient.

At least that's what I learnt by looking into it a few years ago. I may have some things wrong.

>>2604430
Thanks you saved my day.

>>2604437
not the right model, super expensive, but thanks for the help.


Continuing on that Equalizer, it's from 1976. Since I started servicing, I'm now wondering if it would be wise to recap it from A to Z.
I removed 2 capacitors, they have correct values, one feels soaky under the thumb, and I think it's not a bad idea to change them. Am I wrong ?

>>2604692
70s is old enough that it's worth considering, but I'd check what other people say to see if it could have a bad batch of them, like the 90s capacitor plague. If it feels soaky I'd want to replace all of that brand of cap at least. I'd first ensure that replacements are available, and price it out before going for it.

>>2604689
Thank you anon. I know about KV but I assumed you don't here, so I kinda tried to change my question accordingly.

Basically I want to try to control miniature BLDC motors with Field Oriented Control board. However when I search for tiny motors they all have huge KV numbers (like 20000KV) because they need it to generate thrust on drones. So when I will spin them slowly with FOC the resolution will be very bad (smallest variation in voltage would change RPMs a lot), which i obviously don't want. So I was thinking maybe should buy the motors anyway and rewind them to decrease KV and increase the resolution.

By reading your post it looks plausible, I should take the smallest wire possible and do as many windings as I can and thus decrease RPM at the same voltage.

>>2604689
The only thing I feel is kinda wrong in your post is about torque. RPM and torque are not connected, it's the power which connects two terms. You may have huge motors that spin very slowly yet with massive torque or vice versa and still have the same power.

I feel in my case the torque should not change drastically.

>>2604669
>>2604670
I thought this was retarded until I thought about it for a second. He was hired to reverse a PCB for people who can't read schematics, so he made the Sesame Street version for the non-electronics people. At least that's what I'm assuming. lmao

Is making a hybrid electric bike as simple as strapping a gas generator to a commercial e-bike?

>>2603989
I'm a retard. I forgot to update the output node in the plot and it was plotting the wrong node, not the output node.
It doesn't work, it has subunity (~.91) gain.
I guess I need to figure out something smarter than having two common collector stages and one gain stage, since those CC stages intrinsically nuke the gain by at least .8

I think I'll just start learning about class C amps instead of fucking around with the basic bitch class A amps I learned about in my electronics 1 class

new
>>2604810
>>2604810
>>2604810
>>2604810

Herro Anon. Foubd this nice pooking 80s radio alarm on the dumpster. Working fine, the 7 segments are pretty dark though. Can I adjust it with that potentiometer?
Don't want to fuck something up turning random nobs

>>2605270

>>2605270
>>2605272
Before you turn the pot, take a reading with your meter. Also, it's probably the caps that need replaced.