/diy/ - Do It Yourself

>>2676971
i've never tried these so you might want to do your own research, but something like this could work: https://www.amazon.com/Taidacent-Chargers-Battery-Controller-Charging/dp/B08B14MSD3?th=1

I want to pulse a heating element with around 8A and a duty cycle of 20% to get around 20W from a 12V power brick.

How powerful does the power brick need to be? Can I draw that much from one that is rated for something like 2A continuously without worries?

>>2677370
>Can I draw that much from one that is rated for something like 2A continuously without worries?

no, you gotta worries.
a whole lot worries.
like it going into shut-down, or blowing a fuse, or dying.
but there's no need to worries when you can get something like an xbox360 power brick from a thrift store under $10.
diff models exist but typically all can do around 12V at 10A.
get one and your worries are over.

oh, you gotta connect the Power_enable pin to ground or some other place top get it to turn on.
look it up.

>>2677370
The power brick will shut down if it sees 8A for even a millisecond, likely down to a microsecond but I've never tested them that low. It may be possible that a fast enough PWM will get past its overcurrent detection threshold, especially if it just has an undervoltage detection threshold. Adding a massive output cap may not reduce the plausible PWM frequency due to the higher inrush current, you probably need to smooth out the current as seen by the SMPS with an LC filter between the PWM circuit and the SMPS. You may need a freewheel diode, I'm not sure.

>>2677407
i wonder if you can use a tp4056 as a part of a feedback loop for a buck converter?

>>2677390
>with an LC filter between the PWM circuit and the SMPS. You may need a freewheel diode,
So basically a buck converter?

>>2677413
A buck converter has continuous current at the output and discontinuous current at the input. You need continuous current at the input, not the output, which presumably doesn't care. A buck converter is PWM followed by an LC filter. What you need is an LC filter followed by PWM.

You can also browse the wikipedia page for SMPS topologies, they list where the current is continuous and where it's discontinuous. Pic related. Though the unlisted "inverting boost" topology is simpler than a Cuk and has continuous input current.
Such an SMPS will act like a buck converter by making the output voltage be a constant voltage that's lower than the input, which may or may not be desired. If it's a 12V LED, then feeding it 3V or whatever isn't going to turn it on very well.

There are 5 wires attached to my light switch, one is the ground, one is a neutral, and the rest are black. Im assuming 1 is a negative, and 2 are positives going to the light.
I connected one end of a multi load capacitor to neutral, and one end to the 2 positives. The MLC isn't working.
Tried moving the MLC to neg + neutral, light wouldnt turn on at all. Someone help pls ;_;

>>2677713
>Im assuming 1 is a negative, and 2 are positives going to the light.

not even close.
the single black goes to the light (it's switched)
the double blacks are (1) power coming from breaker box, and (2) power going off to some other circuit.
so you connect your MINIMUM load cap across neutral and the single black wire.

>>2677800
thanks anon, apologies for the retardation. I tried it and it didnt work before, will try again.

>>2677713
>Im assuming 1 is a negative, and 2 are positives going to the light.

not even close.
the single black goes to the light (it's switched power)
the double blacks are (1) power coming from breaker box, and (2) power going off to some other circuit.
so you connect your MINIMUM load cap across neutral and the single black wire.

if that doesnt work, chuck it all away and use a normal flip switch like a straight guy would.

>>2677810
Aha, so I connected the red wire from the switch + MLC + single black and it works. It was connected to the black from the switch before and didnt work. Thanks fren :)
>>2677810
>if that doesnt work, chuck it all away and use a normal flip switch like a straight guy would.
I wish. Someone thought it would be a good idea to put the light switch across the room, so I have to use an RF light switch or install a 3-way switch.

I have a big bunch of enameled copper wire off of bathy devices from when I was in the navy. inb4 give it back, how do I solder with this shit, the enamel doesn't burn off, it's impossible to strip, and trying to use it just dirties up the solder

>>2677810
Have some gore :3

Im building a device that has two separate electric motors. Im assuming they are both going to be 12v. One pulling like 40 watts and the other is just a little motor for a rack and pinion drive system. The 40 watt motor is going to spool up and spool down at the press of a button and the pinion motor will kick on with a harder press of the same button.

The pinion is going to have basically a set speed that it always travels at and the same goes for the other motor. The electronics shouldn't be too crazy for that right? Shouldnt it be pretty cheap to have a control board for the motors, switches and power supply built?

Im probably overthinking this, but this shouldnt be to difficult to have running on a breadboard in a weekend and then have the circuit board ready to fab not too long after?

>>2677814
>it's impossible to strip

need to level up your skillz.
if it's thin and fragile, use the sharp end of an x-acto knife to scrape it off.
i put the wire against my thumb to ensure i dont use too much force.
if it's a bit more robust, you can use fine sandpaper.

the only enamel that burns off is on cheap chinese wire.

>>2677813
>thought it worked
>screwed everything back in
>it didnt
NOOOOOOOOOOO

>>2677829
nvm
it did work, just came loose when i was stuffing it back into the tiny junction box

>>2677832
Et voila
I should become an electrician

Moron here, what kind of wire this and how do i make it conduct electricity?
It's the cord from an old headset that I was going to repurpose for the plug, but I have failed at the first step.

>>2677840
Enameled.

You need to scrape it of or burn it off of the.tips so you can solder it together into whatever you're working with

>>2677840
just use water

>>2677840
Crush the ends with pliers and then scrape.

>>2677844
>>2677850
>>2677859
I did the sensible thing and lit it on fire.

>>2677840
Turn up your soldering iron, get a blob of tin on the iron, push the wire in and out of the blob until it gets tinned.

Kinda like a poor man's solder pot.

>>2677834
Nice alignment of all plate screws… all set to 2 o’clock.
If the electrician doesn’t do this you can get your money back and complain to the better business bureau.

Picture related, an old Xcelite tool kit they used to sell years back when we still made things here.

I've been repairing electronics for years but I have also been poor. So, my tool kit has been whatever I could cobble together and whatever was cheap when I needed something specific. Well, I have a bit of money now and want to buy a really nice multi-bit screwdriver set plus all the other bits and bobs I would need for this kind of shit. Problem is, everything is flooded with these cheap kits that look fucking awful. What doesn't suck? How can I tell? Half these fucking things don't even have names I can pronounce.

I'm mostly repairing electronics, computers, old toys, and consoles. So, lots of screwdriver types that have to be narrow and long it get into tight spots.

>>2677871
You know what they say, two is better than one ;)

>>2677819
>>2677814
Dip it in molten salt. I thing sodium or potassium hydroxides will work. Might be able to fill up a solder pot (>>2677870) with the stuff. I’d also make one for citric acid, that shit cleans oxidation off real well.

>>2677819
What sort of button? One with two distinct stages at different force levels? Constant speed is easy with electric motors, since for a given mechanical load, there will be a constant factor relating voltage and speed. That is dependent on the load though. If you need more precise speed control, you’ll have to go for some sort of closed loop control. That means encoders and microcontrollers, if not some self contained servo setup.

>3.3A through a breadboard
good luck

>>2677890
I've found that even cheap bit sets bought in a western brick and mortar store are made of a good enough metal to last well. But some of my hex bits weren't well enough made to not round over (one was so under-dimensioned that it was circular out of the box). I recommend going for someting with a standard bit size (probably quarter inch but maybe the smaller one that I think is 4mm), that way you can easily buy extra bits for them.

I'd check out the /g/ EDC threads for ideas on neat sets, but verify they're good brands via reviews and maybe asking on this board. This general in particular probably isn't the best place to find people with quality screwdrivers, but it's better than some. I've heard people shilling Wiha and/or Wera, though at least one of those is a reddit brand, whatever that means.

>>2677919
The 40 watt motor is wide open. Basically you will pull a trigger that spins up the motor. The trigger completely depresses a switch. Then while you pull that some trigger harder and deeper it hits the second switch that turns the tiny motor that drives the rack and pinion.

So one motor is wide open, and the rack and pinion motor is also but it will be at much lower rpms. Im assuming this can all be done with maybe some clever gear ratios and simple electronics rather than giving it a brain.

>>2677942
>some clever gear ratios and simple electronics

gears, belts and clutches are tried-and-true methods.
PWM control exists for the mechanically illiterate.

>>2677945
Not him, but I'm assuming PWM will have less speed regulation than a buck converter.

The chad move is to add a frequency-to-voltage converter circuit onto a hall sensor and use that as the PWM circuit's control voltage. This may be another excuse for me to shill the CD4046, though honestly I don't know if it will work in this case. Will the type-two phase comparator still give an appropriate voltage output even when it's running unlocked? My guess is no, but it might.
A more sensible frequency-to-voltage converter would be a monostable multivibrator followed by a low pass filter, or an LM2907.

>>2677946
>PWM will have less speed regulation than a buck converter.

neither of those has any regulation.
but a buck converter is a worst solution than PWM.
coz, when you simply lower the DC voltage, torque drops.
PWM keeps torque high because it's giving the motor some hard kicks of full voltage.

i'm trying to build a charger/BMS solution for a 1S2P 18650 pack. i chose TP4056 for linear charger and FS312F-G (functionally almost identical to the classic DW01/DW01A) with a high side P-FET and zener diode for power path management. i want these cells "as protected as possible." i understand that parallel cells will see the same voltage and whatnot, but differing internal resistances and other factors can introduce risks.

pic related is the schematic. i'm pretending J1/BATT1 has lower internal resistance so higher current capability (red trace) and J2/BATT2 has higher internal resistance so lower current capability.

am i correct in believing that there's absolutely no benefit to using two FS312 since it senses low side return current (the sum of BATT1 + BATT2)? if i'm correct, how are parallel cells typically protected in commercial designs?

>>2677958
>how are parallel cells typically protected in commercial designs
The same way single cells are protected: TP4056 and a DW01 for undervoltage protection.
You don't need anything else. Set the max charge voltage to 4V and they're more protected than most lithium cells have any right to be.

>>2677948
>neither of those has any regulation.
The motor has a maximum speed bounded by the maximum voltage, that's how back-emf works. I think PWM would let the speed creep up and up until it gets pretty close to the full voltage speed, while buck would cap the maximum speed lower.

>PWM keeps torque high because it's giving the motor some hard kicks of full voltage.
Torque is directly a function of current, and if you're PWMing it fast enough with a freewheel diode, the current doesn't spike at all; the current sits mostly constant with a bit of ripple.

>>2677958
>functionally almost identical to the classic DW01/DW01A
Check its undervoltage threshold, hope it's better than the DW01's shitty 2.5V.
>differing internal resistances and other factors can introduce risks
That's only an issue if you're drawing a lot of current from the cells. When you stop drawing current, any imbalance will balance itself out again as current flows between the cells. If anything, I'd consider adding a series resistor to make their ESRs constant.

If that FS312 was designed properly it would measure the current via components in series with the cell as a sense resistor. It says it senses the voltage of the CS pin, which I don't really understand, it must be somehow using the FETs ESR as a sense resistance. Or maybe it somehow measures the impedance between bat+ and bat-. If it's the former, then two of them will work independently. If it's the latter it probably won't. No clue how that would function though, see if you can interpret the datasheet's internal schematic diagram. It seems to use voltage thresholds on the CS pin, but I don't then get what the CS resistor is for.

Also if J1 and J2 are your cells, you probably want your off-board power connected to the GND node, not the battery's unprotected 0V node.

>>2677961
>Set the max charge voltage to 4V
Is that possible with a TP4056?

>>2677962
>Is that possible with a TP4056?
For some reason I remembered that you could set the voltage with a resistor but I opened the datasheet and it seems that's not the case.
An easy workaround would be to pull the chip enable pin (pin 8) low when you reach 4V.

>>2677962
>Check its undervoltage threshold, hope it's better than the DW01's shitty 2.5V.
it's 2.9V lockout and 3.0V release, which is lower than i want but still better than DW01. this is for a MCU project and i plan on using one of the (many) ADCs to do a software lockout at 3.0V, the FS312 will just be added protection.

>Also if J1 and J2 are your cells, you probably want your off-board power connected to the GND node, not the battery's unprotected 0V node.
do you mean like this?

>>2677964
>An easy workaround would be to pull the chip enable pin (pin 8) low when you reach 4V.
If you're going to throw in a comparator, you might as well just use a dual RRIO op-amp to implement a custom CC/CV charging curve.

>>2677966
I meant having the negative of the batteries ONLY connected to the FETs and the FS312s. Any connections to the battery negatives should go through the FETs, both ingoing AND outgoing. So put the ground nodes back where they were in the previous image, and connect J3 to the ground node instead of to the battery.

Also that PWR_FLAG is tripping me up, it looks like they're both the same node and hencein parallel, not sure if that's intentional.

>>2677972
>Any connections to the battery negatives should go through the FETs, both ingoing AND outgoing. So put the ground nodes back where they were in the previous image, and connect J3 to the ground node instead of to the battery.
got it, makes sense.

>Also that PWR_FLAG is tripping me up, it looks like they're both the same node and hencein parallel, not sure if that's intentional.
i did it because kicad ERC kept bitching at me. AFAIK it doesn't indicate any sort of continuity, it's just an ERC thing. when i highlight the PWR_FLAG nodes they are separate entities.

How does something like this work? Using plasma as a speaker?
https://www.youtube.com/watch?v=5c_jsJmAJgw

>>2677019
gibs me cheap soldering stations
there s no way that these century old technology hakko and weller stations cost anywhere around 200 bucks
it literally is just a transformer, a breadboard and a shitty resistor

>>2678011
>there s no way that these century old technology hakko and weller stations cost anywhere around 200 bucks
>it literally is just a transformer, a breadboard and a shitty resist

this. in 1923 you could buy the same exact thing for about $3. We all know who is behind this.

>>2677994
>How does something like this work?

I clicked on every link in sight and all I can find is guys lauding the "joy on Eric's face". Can Eric type, or just give presentations and make videos.

>>2678018
He calls himself a feral human and doesn't understand social media or doesn't want to understand.
He seems genuine but the fanboys are weird.

I'm just curious how something like this works? How is the discharge making noise?

I am thinking of using a double sided copper board like this one as a poor man's PCB: it would still be a dead bug style but both sides would be used as a + plane and a ground plane connected by wires through holes as "vias". Thoughts?

>>2678021
Depends on how much work you wish to put in. Diy boards are a bastard to get going.

>>2678020
Expanding and collapsing gas (air). Like thunder except modulated.

>>2678034
This is an actual PCB that looks more like a finished project. That's too much effort for a prototype. I was thinking quick prototypes, no traces, just sloppy dead bug style upside down parts, and point to point wires. Like this one, but with a twist: use the other side as a second plane, so you'd have + and GND planes and drill quick holes for "vias".

>>2678047
>too much effort

why would you use that when you have this?
so much less effort required.

>>2678021
i'm with >>2678052
aren't those sheets more expensive than perfboard?

>>2677919
the meme iFixit kit is pretty much industry standard. It's a little overpriced but the quality is fine and it comes with basically everything you need at each price point.

Also, if you trawl dollar stores you can at least find very cheap driver sets. The one I use came with torx and a few security heads and it was 4$ and came with the driver. Standard 6mm or whatever shanks on the bits so I can use them with other drivers/holders.

>>2678011
>>2678013
I got an iron for 12$ at Princess Auto and it's actually worked out fine.

>>2678064
The first time you use a better iron you'll immediately recognize the difference. If soldering is something you do a lot of then you shouldn't skimp on tools and solder/flux.

>>2677994
basically increasing or decreasing an arc's power causes it to expand and contract, moving the air, just like a normal speaker would do.

>>2678061
yeah i use them for making actual pcbs with toner transfer when i need them reeally quick

>>2678052
Yes, good question. There are many types of prototype boards but for some reason I see lots of RF prototypes done on solid copper boards. Probably to have a solid continuous ground plane to minimize interference at high frequencies.

>>2678061
You can get offcuts cheap.

>>2678021
Save those for permanent projects. Use copper tape on cardboard for throwaway/test prototypes.

And if you're gonna drill FR-4 (fiberglass), get used to buying drill bits cuz apparently they chew through those. Consider FR-2 instead, even if it's shitty.

>>2678047
Don't solder point-to-point like that, it's a complete waste of space.
Break your schematic into modules (e.g. class-A amplifier) and twist+solder all the leads together first. Then the whole amplifier can be tacked down at two points, through ground, and through the Vcc decoupling cap.

>>2678080
Do not waste your time with perfboard/veroboard unless you are a wizard at grounding. Save yourself the headache, ALWAYS use a ground plane.
Even your audio projects will turn out better, and for RF it's a no-brainer

>>2678067
I had that moment with one of those cheap 8w usb irons and fine solder and a bottle of flux. Solder would actually wet to it. I could finally do stuff on boards, let alone splice wires.

It was fucking tragic when the tip finally broke (it's a shitty cartridge type with a heating coil and stuff right in there, not a solid slug of plated copper or whatever)

The Princess Auto one is also pretty good and the tips are decent but the size of the thing makes working on electronics frustrating. Solder at least wets to it and the tip is cleanable, though

I don't have a connector or a soldering gun, should it work the same if I do it this way just temporarily?

Where do I buy more of this shit for cheap? I've got a DualShock 4 pad I bought and the cable that joins the main board to the USB daughter board is fucked up. I've also got a laptop with a similar problem but the connector at one end is also fucked too (it's the type with a clip but the clip broke)

Is it even worth trying to fix and is there anything else I should try before trying to find a small roll of this shit and an appropriate way to strip/terminate it? If I've got a bunch of fine enameled wire can I nigger rig some shit with that?

>>2678127
it probably won't make good contact and you'll get fuzzy/popping audio as it fails by just existing

I'd just splice the cable entirely, at least a rat tail will hold together for more than five seconds

>>2678129
>hey guyze i threw my game controller at the fucking wall how fix ???
https://gprivate.com/5yvm4

>>2678129
>anon mad he broke his controller over fortnite

>>2678072
OK I guess so how to you get that pulse of plasma without it discharging?

I'm designing a board that has a shitload of capacitors on it. I'm aware of inrush current, and I don't want to annihilate my voltage regulators/power supplies

This is where I would use an NTC thermistor, right? At the output of the voltage regulators?

im messing around with a uniden r3 pcb, if none of the power input pins are short to ground can i conclude that nothing in the circuit is shorted?

>>2678192
"shorted" is a load-bearing term with multiple possible meanings
what specifically do you mean in this case by "shorted"?

>>2678195
i mean a resistor in a parallel circuit opened or a capacitor is shorted and causing the circuit to fail shortly after power up, i suppose i would have to apply power and test the capacitors individually or remove them. there doesn't seem to be any continuity from the power pins to the rest of the circuit

>>2677945
>>2677946
>>2677948
Interesting. Will have to take a look at this when I start building the circuit. Fine tuning a motor speed to an appropriate gear ratio would be the most ideal scenario. If I could consistently get the smaller pinion motor to lets say 3k rpm with gear assembly , that would be the best option I think.

>>2678186
Putting the NTC thermistor after the regulator will mean your voltage will be a bit lower than it should ideally be. I'd put it before the regulator, even if that means the regulator will end up in a brownout state during startup. Though that's just my intuition, maybe the dropout once the thermistor is up to temp isn't much at all, do some research to what existing circuits do.

You can also put a conventional resistor in series with your power input and use a relay to short the resistor out once the input caps are mostly charged.

>>2678186
>don't want to annihilate my voltage regulators/power supplies

billions of regulators have been subjected to the same tortures and survived.
so you dont need precautions.
coz of good regulator engineering, transformer's internal impedance, and god's love of his silicon brethren.

>>2678186
Doesn't matter. The regulator will take time to ramp its output voltage up to whatever it's configured to. No regulator has instant turn-on response.

What would be a good idea however is to have a diode between the input and output so that there's no current going from the output back to the input through the regulator when the input supply shuts off. Pic related for linear regulators, same should apply for switching regulators.

>>2678323
some regulators come with such a diode

>>2678129
it's right in the OP you retarded nigger

>>I have junk, what do?
>Shitcan it
>consumer product support or PC building?
>>>/g/

>>2678129

>>2678392
>uxcell
complete garbage. never buy from these amazon chinks.

Is there any way I could make this overvoltage protection circuit work? I'm a retard and didn't think about the VGS exceeding the +-20V limit. Should I just go with something like a crowbar and a fuse as overvoltage protection?

>>2678441
add a zener from the fet's gate to source, a resistor from its gate to the bjt's collector is a good idea too

the fet gon get hot though, consider bonding it to a PTC thermistor that will block the power flow

or better still, make an overvoltage latch that turns the power off hard, and requires a manual reset. easy to trigger the same latch with an overcurrent condition too.

>>2678047
Yeah I made it. Etched the board using ink toner transfer method. Results aren't great but it worked.

Little off topic but where do you go to meet others into EE and want to work on projects like this?

>>2678529
u dont

>>2678392
before I pay 7.69 to repair a device I got for 25$ cash no tax, is there any nigger rigging I should try?

Even if I can't use the pad as-is at the end I wouldn't be adverse to using it for an arcade stick or something.

>>2678529
Deep in Himalayan mountains, meditating in temples, chanting "ohm" at 432 Hz.
https://www.youtube.com/watch?v=I_4UXdsW0h4

>>2678552
>any nigger rigging I should try?

you got a soldering iron with a good tip?
and some thin wire, like wrapping wire?
a magnifying glass?
and ninja soldering skillz?
then no, you still cant do it.
but you can scrounge lots of cables from curbside printers, monitors, etc.

>>2678567
>you got a soldering iron with a good tip?
>and some thin wire, like wrapping wire?
>a magnifying glass?
yes actually, I have an iron and a pack of different tips, I have enameled copper wire (less than .5mm) and I have a helping hands set with a magnifier, a "phone vr" headset i use for its magnifying lenses, and some handheld loupes.

>and ninja soldering skillz?
I can try my best

>>2678573
Kill yourself.

>upgrade to kicad 7 in march
>keep using kicad
>be me, today
>realize ive still been using kicad 6 for the past several months
>computer has both versions installed for some reason

>>2678651
Use a package manager

I want to make a forum related to in-between autisms, which includes electronics...
What do you think of this captcha?

>>2678727
>What do you think of this captcha?
The answer is 6914, both are ohmic devices. The lesson being, you're going to end up with more autistic discussions on how to interpret captchas than you get actual content.
I do like the idea though, it's hard enough to google that it's a good brainlet filter.

The chad move is to use the captchas to crowdsource training for a neural network, in order to reverse engineer proprietary schematics and/or circuit-boards. e.g:
>write the netlist for the following schematic
>read the part number off this component
>draw over the traces hidden beneath this black solder mask
I for one welcome the coming EE generative AI.

What kind of forum, just the classic sort with threads that lock after a long period of time and some sticky/permanent threads?
Personally I'd like to make a forum that's intrinsically linked to some sort of wiki. For example, circuits people design get tagged and stored on the wiki with links to their source threads. So if someone searches "differential amplifier + RRIO" or "split rail SMPS" they'll get a list of the circuits people have designed matching those tags, with specs on each circuit's wiki page. Maybe having to dive deeper into the original thread for more nuanced info. Ideally you'd be able to sort categories of circuits by their input impedance, gain, noise, etc. or even import their netlist. Having a database of subcircuit netlists that can just be slapped together into a full circuit sounds really useful.

>>2678727
Aesthetically, I like it. Functionally, I don't really know what makes a good captcha but it looks good enough to me.
It would also do a great job at filtering out retards/newfags, which makes it worth it for that alone.

>>2678749
>The answer is 6914, both are ohmic devices.
thats too autistic, even for my taste.
>The lesson being, you're going to end up with more autistic discussions on how to interpret captchas than you get actual content.
I've had a board in the past, that is correct, people discussed non-standard captcha, most people hated, so you know it is good.
>What kind of forum, just the classic sort with threads that lock after a long period of time and some sticky/permanent threads?
In-between autisms forum.
Think of it as programmers go there ask dumb questions about electronics, and vice versa, and idk... Photographers explaining chromatic aberration to the programmer.
Mainly inspired by fact that back in the day, programmers kinda knew how computer works at a very low level (wasn't uncommon to dick around with C64 for ex.), but these days they don't, not even unis teach them stuff about digital system, flip-flops, or idk, basic ohm law. But that's not limited to the coders and EEs, everyone suffers from overspecialization.
Because you know, imageboard style replies are good, they are fast.
But all imageboards suffer from same exact issue. You want to flash coreboot on thinkpad, where do you go? Ask /g/? ask /diy/? Would be much better if you could mark a thread with "computers" and "electronics" tags so people that lurk computers will tell you that you've compiled shit wrong, and electronics will laugh at crusty fluxless solder joints.
>>2678750
>I don't really know what makes a good captcha but it looks good enough to me.
From bot protection point? anything non-standard.
>It would also do a great job at filtering out retards/newfags, which makes it worth it for that alone.
Filtering out newfags isnt good, because you have to start somewhere, but filtering out retards it will.

>>2678765
>because you have to start somewhere
Well they can start out somewhere with non-permanent post storage, like 4chins.

>>2677019
Greetings.
Sorry to bother, I have a couple of old monitors based on CCFL and I'm roaming if there is so much trouble moding it to be able to use LED.
I know every manufacturer make his own drivers and that, but it is viable to do so?
Thanks

>>2678819
It's usually pretty simple from what I've heard, but I'm not familiar with the details. Some caveats I can think of may be:
>having to use an external LED power supply if the internal rails aren't the right voltages or can't provide enough current
>being unable to disable the HV power rail
>a lack of a straightforward method for dimming the LEDs via the monitor's own control hardware
But I'm pretty sure they're all doable. Though that first one is contingent on the PSU creating a LVDC rail from which it steps up to the HV for the CCFL, if it converts mains directly to HV then the internal LVDC rail won't be enough to handle the LEDs. Maybe someone here will have experience doing this, maybe people on /g/ will have done so with their pinecil so I'd ask about there too. Otherwise there should be plenty of online tutorials and videos on the process.

If you find that there isn't a straightforward method for converting your monitor in particular after doing that research, then post the insides and chip numbers and such here. Maybe a solution will appear.

>>2678828
Thanks for your support and guidance.
>>being unable to disable the HV power rail
> PSU creating a LVDC rail from which it steps up to the HV for the CCFL
I think this is the hardest part of the process and it is likely to need a sort of step down or a external PSU, It would be unelegant but perhaps any LED with is own driver will do, but there will be no control on dimming or other options.
>/g/
Doubt about it, they are good at consumer and ready to use things and coding, but they are not DIY capable.
I will disassembly and try to use. I mean no disrespect for /g/ either.
Now over there I have too broken tv's (all works except the displays), in theory coul It be swapped the PSU and the LED lamps? (Perhaps it will attain a basic/generic level of PWM or whatever it is needed to control the LED former CCFL).
Also
Is there a way of bypass the mainboard of the "smart tv" and connect then analog/digital inputs directly to the display?
I the eMMC it is removed or erased (aka programmed). it will still work?
Really there is nothing of value added in it, so much bloatware and nothing more.
And there is no use (for me) to connect it to the LAN.
but since the firmware is in the same package or it is the same thing with the OS, is there a generic firmware to enable to basic level (aka "dumb" operation).

Thanks a whole lot again.

>>2678835
Or perhaos to use some board like the arduino, etc, to basic functionality?
Thanks.

>>2678819
>there is so much trouble moding it to be able to use LED.

oh yes, so so much trouble.
1) taking the screen apart to get at the bulbs = 60% chance you'll break something critical.
2) 80-100% if you're an LCD virgin.
3) LED strings typ need 48 - 90V. you may have to stick 3-4 laptop chargers in series to get that.
4) LED strips get very hot. you'll absolutely have to stick 'em to a metal frame with heat-conductive adhesive.

>>2678819
Chinks make conversion kits
https://www.aliexpress.com/item/4001224383234.html
It is kinda straight-forward to install, but PWM behavior might get weird on some boards.

Problem is that LEDs they use in these are junk, and your monitor will always look like cheap trash with those, you'd want ones that closely mimics CCFLs to keep all colors right.

>>2678835
>here will be no control on dimming or other options
I think it's possible if you find the line going from the control board to the CCFL inverter and figure out what kind of signal it's sending. I'd guess it's just PWM, but it could be anything analogue or digital. Worst case you just need a cheap little microcontroller to interpret it and spit out a value that your LED PSU can handle, or directly to a MOSFET.

>in theory coul It be swapped the PSU and the LED lamps?
Usually the LED PSU is on the same board as the other power rails, so I doubt you'd be able to seperate just the LED PSU. If you're lucky you might be able to fully swap one PSU board for the other, but chances are the voltages, current ratings, and/or connectors won't match up properly.

>Is there a way of bypass the mainboard of the "smart tv" and connect then analog/digital inputs directly to the display?
I doubt it would be easy, but I've never taken one apart properly. There will probably be a dedicated video processor chip on there, which you can maybe control directly if you cut through the traces going to the processor (processor = locked down android arm soc) to solder on wires that go to a microcontroller. Read the datasheets and see I guess, with any luck it just needs a SPI command or whatever to tell it what video input to use.

That said, there should be generic firmware you can boot onto it. There's definitely a base model android for TVs which I hear you can sorta degoogle and enable 3rd party apps, but you can also put on generic android, which won't have all the integrated adware to the same extent. A bit less convenient without a better control scheme though. Might even be able to flash armbian.

>>2678765
>Filtering out newfags isnt good, because you have to start somewhere
They just need to lurk more.

>>2678838
>oh yes, so so much trouble.
>1
>2
>3
>4
Thanks for helping me realizing the nature of my approach and the mistake of the difficulty assessed.
Very kind.
>>2678840
> you'd want ones that closely mimics CCFL
even in my dreams would be able to found them. kek, but I think the PWm business would me more effective to address (first).
>>2678841
>it's just PWM
I would think the same, but ...
the MOSFET way seems to be more reliable.
>the LED PSU is on the same board
you hit the jackpot, I fast roamed the one at hand and it is this way; while possible it is unviable because it would need to whole PSU side by side, and outside the case of the monitor, and it is not having the led's yet.
>(processor = locked down android arm soc)
this, but perhaps, I could swap for the one of a mobile (at least the eMMC) with the firmware preloaded or to be.
>There's definitely a base model android for TVs
Do you have a name or some other info to fetch out there?
>Might even be able to flash armbian.
This too.
Thanks again, I will post what I have founded.

>>2678871
Cut the 12V trace that feeds the CCFL transformers and use that to power the DC-DC converter for the LED strips. The PWM dimming is in the feedback loop.

on the subject of CCFLs vs LEDs, check out these two TVs thrown out by a local school yesterday.
note the 2020 model is 10x simpler and lighter than the 2007 model.
took just 20 mins to fix the new one (dirty connector), and to diagnose the old one (bulgy cap).

>self-made perfboard circuit is constantly turning on the MOSFET
>figure the FET is dead
>desolder it
>part tester reads it as good
>solder FET back in
>same issue
>trace more with DMM
>figure out there's somehow a short from gate to drain
>desolder FET again
>measure between gate and drain pads
>still a short
>but no short to ground or any other node
what the hell is going on here
pic related is the circuit, the FET in question is Q4.

>>2679168
on closer inspection there was a microscopic solder bridge
i apologise for wasting your time

>>2679171
I couldn’t tell what this circuit is for. It has a xlr connection, is it a “disco light” that flashes in time with the music?

>>2679023
> TVs are light
I had an original sony 720p lcd tv. It’s so heavy i can’t really lift it. It was $3k in the early 2000’s.
New TV is bigger, 70” and weighs about as much as the box of the old one and uses a wall wart for power and cost $600.
I think they made them heavy originally to persuade boomers that they were getting good value.

>>2679178
I'm just using mini XLR plugs because they're available, small, and have a good assortments of both line and chassis connectors. I'd use mini DINs if they had both male and female chassis connectors, but they don't. At least not where I shop.

It's a "controller" for a laser and drill for my 3D printer setup. The digital logic is mainly to ensure that power never gets sent to something that isn't ready for it. Mainly, it won't send a PWM voltage from the printer to the laser if the laser doesn't itself have its external +12V PSU plugged in. But it also ensures that while sensing the +12V supply, it won't backfeed into the 3D printer if it's turned off, through the use of Q1 as an isolated source follower. And there's a schmitt trigger oscillator (U1D) that produces a low-duty-cycle PWM when the button is pressed, for manual focusing of the laser with a low-power pilot beam. U1A,B,C makes a sum-of-products logic array using only NAND gates with the logical equation:
>PWMout = (OSC AND 12Vgood) OR (PWMin AND 12Vgood)
The PWM signal also goes to the MOSFET gate driver circuit (Q2, D3, Q3, R7) and the main FET Q4 and its freewheel diode D4. This FET is meant to drive loads that don't have a PWM input, specifically my mini drill. It would arguably be better if the pilot button also turned the drill on slowly for testing, but that would have required another NAND gate and I couldn't manage to make one gate in the combinatorial logic stage also act as an oscillator. It was probably possible but I couldn't manage it.

There's also an RGB LED to tell me what signals are live, which shines through the holes in the casing, pic related.

>>2679180
>It’s so heavy i can’t really lift it
Is it one of those rear projection TVs?

I'll begin making the electronics for a whack-a-mole type game this weekend. Hitting a "mole" turns its light out through a SR lath and eventually the AND gate makes sure they are all whacked before ending with a random melody out of twelve melodies.
Perhaps I should have picked a NAND gate instead of the AND one but I might be able to negate the diode state to still activate the AND gate with NOT:s from loose transistors... Seemed so logical and easy after watching Mr Eater! Got some caps for debouncing as well. I feel confident but still nervous from anticipation.

>>2679186
> designs own mult-use general PWM circuit
> doesn’t have option for gating the frequency to AC/DC’s “Thunderstruck”
baka

>>2679186
> designs own mult-use general PWM circuit
> doesn’t have option for gating the frequency to AC/DC’s “Thunderstruck”
śmh

>>2679214
I'm more of an "Electric Avenue" kinda guy

Ey yoo so i got this old floor fan its working ok but i want to clean it, i suspect its axle sits on bearings which i could also replace for improved smoothness, to do this i have to dissasemble its whole body which happens to be tied to the magnetics part of it, so i am afraid of fucking up the internal copper windings too, any advice? Has any of you serviced something like this?

>>2679419
probably uses self-centring bearings

>>2679419
There is a commutator attached to the shaft. After you remove the steel shield, pull straight out. Use grease on the bearing(s), not oil. Wipe the commutator with a damp cloth, and clean the transformer.

>>2679424
lmao induction motors do not use commutators you dunce

I’m an EE student trying to build a simple LED circuit with a BB. For some reason, the circuit will not function unless the LED is connected across the valley/center divider of the BB. If I connect the LED on the terminal strips and connect a resistor across the valley then it won’t turn on. Why is that? Is it just a peculiarity with my BB?

>>2679453
What is BB?

>>2679459
Took me a while too, I think it is a breadboard

>>2679459
Breadboard

>>2679453
Post picture of the BB. Could be your BB or could be a poor contact. You can test it with a MM. Do you have a MM?

>>2679453
if you can't figure it out EE is not for you

>>2679462
The configuration on the left works. The configuration on the right does not work. I don’t have a multimeter on me sadly.

>>2679464
Anon, how do you think the pins on a breadboard are connected?

>>2679464
Shit is probably just defective.
>>2679419
ah man so cool dissasembled the thing, i see the squirrel cage, the shaft does not use bearing but sits on rings, i will put some grease i suppose its more of a 'hydrostatic' bearing , wanted to find a way to mod the speed but i see its using a capacitor to provide phase difference between windings, i will just leave it as it is.

>>2679464

>>2679466
those "rings" is a self-centring bearing.
>Shit is probably just defective.
dumbfuck

>>2679465
>>2679468
Oh. I see now.

Trying to have a product built. Sourced brushless motor and ESC going to a custom built PCB, or scrap the ESC and build it into the PCB?

>>2679471
why are you mad about some shitty chinese breadboard ?

>>2679452
wat's a dunce

>>2679471
>self-centring
A ring that made itself out of one cent? Help me I don't understand and I'm scared.

>>2679516
basically poofter aka spastic aka nonce

>>2679536
>basically
wat
>poofter
wat
>spastic
wat
>nonce
wat
>aka
wat

i want to drive a string of 20 RGB LEDS from a single cell 18650 (1S2P). these are low current LEDs, 20 mA. is it practical to boost a single cell to 60V? or do boost converters not handle ratios that high.

>>2679575
Better off using LEDs in parallel. For high voltage ratios, transformer-based converters (e.g. flyback) are preferred to boost converters, they'll be more efficient.

>>2679581
would something like 2 parallel strings of 10 series LEDs be more sensible? or is that still too high?

Whats a good bench power supply?
Do any of em get the job done or will some blow up and burn down my house?

>>2679593
Ultimately there's a compromise. Each series string ideally has some form of current limiting, but if you're using resistors for each LED anyway it flat-out doesn't matter. Make it 20 parallel strings and run it directly off the battery, though you will notice the LEDs dim noticeably at lower battery voltages. An ok compromise would be to run the LEDs at 10V-30V, boost converters that work up that high are much easier to get. If you are using active constant-current regulators, then only needing two of these up at ~30V would be fine in my eyes. Though 3-30V won't get you the most efficiency.

What kind of current regulation and brightness control are you planning on using?

>>2679665
Really depends on what you use it for. My guidelines are:
>audio stuff - make (or buy) a transformer-based linear fixed-voltage split-rail power supply, maybe with multiple fixed voltages
>load testing - buy (or make) a cheap converter that uses a variable CC/CV DC-to-DC converter running off an AC-to-DC power brick
>low power digital stuff - a USB socket breakout board
>really high power stuff - atx psu or microwave oven transformer
very little reason to use anything else unless you're into instrumentation or whatever

>>2679575
Why not drive them in parallel? That is just 0.4A in total no?

I've basically finished upgrading my 3D printer to have drilling and lasering features for making custom 1 or 2-sided PCBs. Just need to figure out the finer details of applying the soldermask and riveting vias. It works via a dovetail quick-change tool holder that I slot each tool into, pic related.

But watching some youtube recently got me thinking, it would be pretty possible to upgrade this setup to also have an EDM tool on this holder. The most expensive part is the PSU, but I saw one guy literally just using rectified mains, and another guy using a microwave oven transformer rewound for 60V or more, followed by a rectifier and big filter caps. Since I live in a 240V country and have a couple of spare MOTs lying about, I could definitely do the latter, and the isolation would be significantly safer. I'll need big caps and high-current diodes, but it's definitely possible.

The first youtube video I saw on the topic is Rack Robotics, where he basically has a constant 72VDC power source, followed by a large RC filter, that allows for a large amount of energy to be shunted into the workpiece with current limiting afterwards. For a resistor he uses an entire ~20-30cm PCB covered in a hilbert curve, and for a capacitor he has a bunch of MLCCs in parallel. Switched to the workpiece by a beefy FET. See:
>https://github.com/Rack-Robotics/Powercore-V1-Hardware

But I've also found this guy:
>https://tenebryo.github.io/posts/2021-04-07-edm-power-supply.html
While I don't really understand his use of an H-bridge to allow current to flow through the workpiece in both directions, his use of an inductor instead of a resistor is quite interesting. I was thinking that an inductor won't work due to its resistance to start current flowing, but he keeps current flowing in a loop through the inductor, and dumps it into the workpiece after it's already built up current. I suspect this is more energy efficient but I'm not informed enough to say if it's significant.

Thoughts?

>>2679665
Bunch of batteries.

>>2679676
>What kind of current regulation and brightness control are you planning on using?

i'm not sure of the exact part numbers yet. but the general topology will be a boost controller to boost the output voltage to just above the minimum voltage required to drive the series strig of LEDs, and then multiple PWM dimmable linear LED drivers, one for each series string and RGB channel.

>>2677019
can someone explain how these ends are supposed to fit in the chuck? I bought a couple of these bits, good quality ones, no matter what I do they wobble like crazy. i thought my chuck must be worn out and not holding them well since they're name brand and brand new, so i replaced my chuck and they wobble like crazy with the a new chuck

>>2679768
Do you have a collet for the smaller bits?

>>2679786
i don't know, it's just what came with the drill, it's a bosch hammer drill if that matters

>>2679789
https://dengarden.com/appliances/How-To-Replace-a-Cordless-Bosch-Drill-Chuck

>>2679790
i replaced it already and it still wobbles, but only with this >>2679768 style bit. regular bits are rock solid

>>2679791
Those are SDS (slotted drives system) bits. The shaft geometry is different and your chuck isn't grabbing it right.

What capacitor kit do I need to buy to cover all my needs?

>>2679793
What are your needs? Electrolytics can burn a hole in your wallet.

>>2679792
how do you use them? is there an adapter or something?

>>2679802
Check if one of these will fit your drill:
https://www.boschtools.com/us/en/boschtools-ocs/adapters-and-chucks-22556-c/

>>2679793
10-20 of everything from 1pF ceramics to 200uF electrolytics, plus a hundred extra 10uF (EL), 1uF (EL), 100nF (cer), and 10nF (cer)
E6 series should be plenty

>>2679796
Both digital and analog. It seems like I only need a few basic values of high value caps: 0.1uF and 1uF and maybe 10uF. And I'd also like to do some RF / filter projects and those are all pico and single nano farads. I notice that most kits are heavily skewed towards the lowest values, a typical kit might have up to 60% of the values in the lowest range 0-100pf, then another 30% in nf range and only a few values in >0.1uf range. It is cheaper that way but probably that's what I need? Nobody needs precision in the uf and even 10's-100's of nf range. Plus the capacitor tolerances are pathetically low and capacitance also varies with frequency and DC bias so any arbitrary capacitor should work in a very wide range and it is probably also trial and error. So a bit precision may only be needed in pf/low nf range.

>>2679811
I was just thinking about this >>2679814
Could you explain why you need an equal distribution of values all the way to uf?

>>2679814
>SMD ceramics
The AOL keyword is "capacitor sample book". They come in all sizes and ranges.
>electrolytics
Stay away from electrolytic -kits- and only buy Jap/US caps from reputable dealers.

>>2679819
> Jap/US caps from reputable dealers.
$$$$$$$$$$$$$$$$$$$$$$$$

>>2679820
Yes. Go ahead and "save some money" and find out. lmao

>>2679575
>>2679593
>>2679745
ok i feel like a retard because i just realized that most RGB LEDs are common anode or common cathode. so how do you drive a "large" number of dimmable RGB LEDs? you'd need dozens of linear controllers which seems like a pain in the ass.

Turns out nerve cell resting potential is 75mV
https://en.wikipedia.org/wiki/Resting_potential

I thought it would be much smaller. Does that mean I can calibrate my multimeter this way by measuring my nerve cell voltage? Any idea how to measure it safely? Also 75mV with respect to which ground?

>>2679872
>Any idea how to measure it safely?
You ever heard of an ECG?

>>2679882
But it is probably some kind of inductive measurement. They are not sticking probes in your brain. I am asking about measuring 75mV DC with a multimeter.

>>2679872
squid

>>2679817
for resonators and filters, and applications where you can't just cram together 10 caps in parallel

>>2679902
but resonators and filters are usually RF work in the pf territory. So yes it is nice to have E12 or even E24 in that range but it is unlikely that your resonator would require 0.1uf to 10uf and if it does you could probably get away with an approximate value.

noob here, can i use this TP4056 to make a solar powered lora module? i already have the 18650 battery, the lora module, the solar panels. but i dont understand the module very well. the plan is to solder the solar pannels in the left sockets (i assume they are IN +-), the module in the right out sockets, and the battery in the B ones. is that right or am i retarded. first time doing that btw

I am thinking to save space on my desk I will use my scope as a low current power supply. There are probably 3.3v and 5v buses there somewhere, I am going to find them and wire them to the panel and drill a hole and put a USB jack in. Any problems with this approach?

>>2679949
>Any problems with this approach?

yep.
dumbest idea i've heard this month.
you're risking the health of (let's say) a $200 gadget, when you can get lil chargers at the thrift store for $1, or $5 at the pharmacy.
risk in 2 ways, (1) taking apart and soldering inside instrument, and (2) shorting the supply and thus killing your scope.

>>2679817
On occasion you'll need particular values of large capacitors, but usually it's a part of some low frequency RC circuit that you can tune via the resistor instead. Maybe you'll need them for an LC circuit instead, but they often go to negative voltages so electrolytics aren't suitable. And for audio purposes you probably don't want electrolytics (or monolithic ceramics) anyhow.

So yeah, wider spaced values past 1µF should be fine. Definitely stockpile on a lot of 100nF and 10nF ceramic caps though. Make sure you write down their voltage rating when you buy them.

>>2679859
Yeah if they're common-cathode or common-anode, as opposed to seperate diodes in a single package, there's basically no practical method other than resistors and PWM. Arguably you can replace the resistors with current mirrors, but I wouldn't bother. Make the whole thing parallel, running off the battery voltage. Include some PWM algorithm that increases the LED duty-cycle as the voltage goes lower, in order to keep the effective brightness equal.

>>2679872
It's a compromise between noise attenuation and sensitivity, I think the DMM's 10MΩ input impedance is often swamped with coupled mains frequency noise.

>>2679927
If the maximum open-circuit cell voltage of the PV panel is greater than ~8V or so it won't work. If it never goes below 6V then you could use an LDO beforehand.

>>2679949
Terrible idea for multiple reasons:
>if you fry something it's expensive to fix
>your voltage rails are constantly referenced to your scope so you can't make floating measurements
>drawing power from these rails may introduce a voltage drop or noise into the scope's own circuitry
Tie one of those automotive panel USB sockets to the underside of your desk.

I'm scared of leaded solder, bros...

>>2679969
stop sucking your thumb

>>2679969
its fine.

>>2679922
what the fuck are you talking about
>but resonators and filters are usually RF work in the pf territory.
false and false
> it is nice to have E12 or even E24 in that range
I said E6
>but it is unlikely that your resonator would require 0.1uf to 10uf
false
>and if it does you could probably get away with an approximate value.
true, but all values in the E6 series could be defined as "approximate" as they are within 20% of each other

you are stressing out over maybe $10 worth of parts that will save you a lot of headache in the future

>>2679885
https://en.wikipedia.org/wiki/Squid_giant_axon
squid

With the a4988 stepper driver it says the min pulse is 1us. I dont intend on driving the motor crazy fast. But I need to be able to have the code for stepping it be as short as possible and nonblocking. So my question is will there be any downsides to having the step pin drive high for 1us and then low (leave it there) and then come back to it when its time for the next step. Should I do it the other way around. Drive low for 1us, then high and leave it there until the next step?

>>2680049
I don’t think it will matter. If it’s active high, then I’d leave it off most of the time and turn it on for the microsecond. Probably doesn’t matter but it’s better practice.

>>2677019
One of my hard disks suddenly starting to make clicks. I have other over there, the same model, manufacturer , etc.
imagine that I can carrectly dissasembly the spinning disks, and fir in other? is this doable? (it would be needed to recalibrate the lector arm).

>>2680139
>is this doable?

you can certainly do it, but it wont work.
when the disk is made, certain info about the geometry of the disk is collected and stored in a memory chip inside the disk.
if you replace the platter, you also have to replace this chip.
and if there's more than one platter, another hurdle is introduced: the two platters must be re-positioned exactly as before wrt to each other.

since the beginning of time, the only solution is backup, backup, backup.

>>2680139
>is this doable
no, you need a clean room and those hard drives are alligned together, so if you have a multi-platter hd, removing them without the proper tool might make them misaligned, and then it's almost impossible to realign them.
https://aesonlabs.ca/blogs/misaligned-platters-data-recovery-or-why-you-should-never-dismount-platters/

Instead of trying to fix the hard drive, just back it up before it fails (you didn't mention it doesn't work so I assume it's just clicking for now). Use your other donor hard drive to copy the failing one instead of trying to repair it.

>>2680146
>>2680150
>it's just clicking for now
I just bought it, started to make clicks and the os stopped recognizing, it was not too much that info in it. But It wpuld be nice to know what was lost.
Since I just get it, I requested the warranty, but it was refused by the vendor to accept it.
>Small scratchs on the bottom (the side that does not have the sata/power/jumper connections)/the paint in one of the bolt connector scratched a bit...warranty status:void.
I wasn't told/explained hot these things in the cabinet/cover ot the disk caused the clicks or the missing in the bios/os; the outside scratches was the excuse, I doubt it was the reason though.
I was pondering to check with the manufacturer (seagate), but I don't know if it's worth it.
Thanks for replying . I didn't know about the geometry chip details, I was going to try an old eide disk that the board went bad , first on the pins bent and next I was trying to exchange the board from another disk (same model,etc).
But my dad used that disk on his pc, so I has been laying there for years.
Cheers!!!

>>2680166
I wonder why even when I check my orthography still appears errors in the posts done.

>>2680166
>but it was refused by the vendor to accept it
go to small claims court / ftc / consumer guarantees / whatever applies in your country

What's the real reason not to connect different-capacity 2ndary cells in parallel, provided they're only charged/discharged at low C-rates? I could see how a high charge/discharge rate might cause current spikes when the external voltage/load is removed, but only at high currents.
Most info on this topic sounds like FUD. Half the people 'providing answers' don't seem to understand the difference between parallel/series config, or between capacity and state of charge.

>>2677019
>company buys Chinese Li-ion cells for our outdoor IoT devices
>winter comes, Chinese cells get damaged from the cold and never recover (internal resistance increases a lot), even though their datasheet said they would work fine
>1000s of devices in the field don't work because the cells can't provide the peak current for comms
>boss forces me to design a fix ad-on in 5 days
>doesn't give me a faulty battery to characterise it, doesn't give me more time
>design a basic low Iq boost converter board with tons of decoupling so that the current surge can be supplied
>board gets manufactured, I test it, it works fine
>boss gets it, its not working fine
AHHH I HATE THIS BULLSHIT SO MUCH

>>2680271
E-scooter IoT module? Which company

>>2680276
>E-scooter IoT module? Which company
Not automotive, but the devices are used in industrial settings. I'd like to keep my job, sorry anon

>>2680271
> Things you should already know
In c-c-canada they use nicd batteries since they are much more cold tolerant. Also generally more robust with higher charge/discharge cycles.

>>2680286
We use primary cells, the devices are supposed to last for 10 years on the single cell. One site visit eats our operational profits for 6 months
Life is beautiful, isn't it?

>>2680314
smoke detector? forest fires? trail cams?

>>2680318
Its a small device which is related to metrology in European district heating systems

Hey, /ohm/ I don't know if any of you anons remember me, but I was the anon that posted about getting a really old oscilloscope that I basically inherited from my granddad. It's a Tektronix 475A. It turns on, but it does smoke when doing so and I was wondering if any of you guys had any ideas as to what it could be. I mean, it could be anything, but it's fairly clean on the inside and I just opened it. Most of the screws are fucked to say the least as I can't get the internal screws off, but do you think it's a good idea to plug it in while it's open so I can see where the smoke is coming from?

>>2680339
You should be able to find it by visual inspection. If it was working until suddenly not, look for blown electrolytic caps, tantalum caps, and burned resistors first. Look for discoloration on the PCB that is indicative of heat and check the components in those areas. Touch up any sus solder joints and check wiring for bad crimps or connectors.

>>2680339
>any ideas as to what it could be.

of course not.
but these old scopes had very comprehensive service manuals showing all voltages and waveforms.
so find that manual, and another scope, and compare waveforms with the manual.
then realize you now have a working scope, so you can trash this dead one.
as a side benefit, you avoid certain death from the 15kV flyback voltage.

>>2680381
>>2680395
I spent about two hours looking at it and I think I found the issue. Decided to say 'fuck it' and plug it in. There's a cap back by the power supply next to a cable that is probably about to explode, looks like a balloon. I can't make out what the amount it is, because it's burnt, but I at least know where the problem lies. Hopefully that's the only issue, I've heard that some of the transistors can go bad, too, so I'll have to check some of those. I'll rip it apart further at some point and try to get a new one in, then see what happens. In my retarded actions, I also cut the cap and lost it inside of the thing, so I'm gonna have to take it apart anyway (or shake it out). Luckily, my granddad kept the manual and a bunch of different probes, even the bag it came in, sadly not the service manual.

>>2679969
Just don't eat while soldering and wash your hands after
If you have to solder day in and day out you can use gloves if you are really worried

>>2677019
where do you buy CRI>90 LED strips?
any chinese seller that isnt a scammer?

>>2679969
You're fine, just don't inhale too much.

>>2680173
I tryed that, requested a review of the request but it was dismissed under the same "argument".
Well that's it.
Thanks for your kind support.
>>2680199
I don't understand well about what kind of cell you are speaking.
But overall it has to do with the power concept (measured in Watts), but it has very much more than meets the eye(tm), since there is really no standard for making whatever cell's you refered, nor the procedures to get it going.
>but only at high currents.
Sometimes it is, sometimes it's the high/low voltage, power, temperature and even the quality of the circuit design or the components.
>Most info on this topic sounds like FUD.
Who knows?

>>2680434
>where do you buy CRI>90 LED strips?
When they hit the market. No it's a joke.
Really IDK, but do you know, AFAIK, they don't really get that far, the top notch ones from a couple of years ago, were arount >70, and they were supposed to be bosch, osrom or the fuck.
If you really needs the cri level, perhaps the usual incandescent bulbs are the option to go, and have care for anyone selling you leds stips of cri 90+ (even for 80+).

>>2680199
I tend to agree.

>>2680271
Shoulda just shoved a supercap in there to buffer the charge while it’s transmitting.

>>2680398
You can often find the service manual online, I did for my 2nd hand CRT scope. Should tell you the capacitor value and voltage rating.

>>2680398
Decided to shake it out, damn look at that thing. I'm surprised it wasn't the first thing I noticed, then again I was mostly looking on the two side boards and not the large main board since I thought that's where the smoke was coming from. Also, I can't find a label on the board itself for what it is, but it goes right next to the +15 input right next to the power supply and has 10 and 25, I couldn't see the label I guess the way my flashlight was shining on it. So I guess it's a 10uF 25V? If it even is a cap.

>>2680548
Use your favorite search engine to find the service manual. We are not your favorite search engine.

>>2680550
Yeah, I'm looking at a parts list and schematic right now.

>>2680548
Could well be a tantalum cap with those values. Could also be a varistor or polyfuse or whatever though. But tant caps definitely die. If it has a stripe indicating polarity of one leg then yeah it’s a tant.

I'm currently building a cheap oscilloscope using my laptop's sound card. It's an average sound card from an old laptop. Thing is I'm afraid to blow up the sound card because I can't find anywhere how much current and voltage it can handle. Help.

>pic related

>>2680613
Kinda useless, as sampling rate is really low.
For reference:
>500 MSa/s: typical 50MHz oscope:
>1MSa/s: shitty 200KHz (doubt) toy oscope
>384KSa/ : your sound card

> I can't find anywhere how much current
Uh, ideally you don't want any current flowing.
>voltage it can handle
+/- 1.7V but I think you can go up to 2.5V, dunno, read datasheet for your audio chip.

So I think you should buy $10 oscope toy from China. Or make one using STM32 and its built in ADC.

>>2680434
>where do you buy CRI>90 LED strips?
Usual suspects.
>any chinese seller that isnt a scammer?
I didn't buy strips, I bought LEDs, a fancy ones, sunlike ones.
store/1669096 on ali.
>>2680493
They mimic daylight better than incandecent bulb ever would.
>OSRAM
They don't make much high CRI LEDs. They do have CRI90 and 190 lm/w though.
>Bosch
doesn't make LEDs at all.

Thing with LEDs is that high CRI isn't hard (all manufacturers, from big brands to szhenzhen "we buy taiwan LED and put it in SMD package or COB" have figured out what kind of phosphorous to use). It is pretty hard to have high efficiency and high CRI at the same time.

>>2680620
Initially I'm planning on using it for some specific sensors. I've read their frequencies are very low.

>Uh, ideally you don't want any current flowing.
What if I use a buffer, a voltage follower

>So I think you should buy $10 oscope toy from China. Or make one using STM32 and its built in ADC.
I'll buy/build a better oscilloscope eventually. I just want to build this cheap one out of curiosity.

>>2680613
Better off getting an external ADC and using optocouplers for isolation. But if you're only using it with a certain sensor, and know there won't be dodgy voltage flowing into it and your power supplies are sufficiently isolated, go for it. Watch for AC-coupling caps in the sound card, you may not be able to get any net DC signal through. Common-mode noise might also be a problem.

What sort of sensor?

probably a retarded question, but i need to make a 1S2P 18650 pack. it needs to be arranged end to end (so an 18mm diameter x ~130mm length cylinder) due to geometry constraints. the cells can be permanently mounted/installed and the current demands will be minimal, maybe 200 milliamps.

what's the best practice for doing this? my initial thought is to weld nickel strip tabs to the cells, place some sort of insulator between the cells, and then make the parallel connections by soldering wire to the nickel strip tabs.

>>2680669
>my initial thought

shoulda been this.

>>2680629
well, it's actually a handful of sensors of different types. I was thinking about measuring some signals from sensors in my car.

>>2680682
damn i'm retarded.

>>2680595
Checking the schematic and the parts list, it's definitely a cap with a polarity, so you are correct. I'll see if I can find anything or if I have anything on-hand to replace it.

How tf does something like this work? They have like 100+ fans that each can have it’s own speed? It’s a windshaper for reference.

>>2680747
>windshaper
CONSOOM
What a stupid fucking gimmick. I'm surprised it isn't covered in rainbow LEDs. lmao

>>2680754
? sure it’s not needed over a normal wind tunnel, but it’s a neat concept for gathering more data? Being able to do scale wind tunnel tests and thrust tests would be rly neat. If I could diy one for under 300$ it’s a no brainer for me desu

>>2680763
computer fans create turbulent flow nowhere near laminar you would need for this even if it worked like beam shaping which it doesnt or it was useful which it isnt

>>2680766
Do what other people do for small wind tunnels and do the wall of lil tubes to help create laminar flow
Idk it should work I don’t see why not

>>2680613
use diodes boy

is a second edition art of electronics any good in 2023? i have a feeling i'm going to be tinkering with mostly 1980's tech or older anyway.

>>2680815
Yes, even 3rd edition has examples with mostly obsolete parts that you can't get anymore. But this doesn't matter because principles stay the same.

hi i have some old caps on the board of a electric fence trying to figure out what the codes mean since the manufacturer Epcos was merged into TDK Electronics and they dont have documentation for old stuff online (so it seems)
the one blue block cap (N115409693 67D) reads
>6u8 K 300 H6
and the other one (N114781153 67D) is a little smaller and reads
>1μ K 750 FD
i have never seen this type of notation before
especially the lowercase u is annyoimg
anyone got an idea what this capacitor encryption is called ?

>Trying to buy a test probe
>Search electrical probe
>First result 'electroejaculation'

>>2680857
>6u8 K 300 H6
6.8uF, 300V

>1u K 750 FD
1uF, 750V

>>2680872
>double check to make sure
>it was the pornhub search box

>>2680873
thanks

if an oscilloscope and your testing circuit are hooked up to mains power/mains ground, why would you short the circuit through yourself to ground if you touched it (unless the circuit itself was already shorted to the chassis ground)? don't get why people say to use an isolation transformer in between the powersupply if there is no ground fault in the circuit.Why can't you hook your oscilloscope ground to chassis ground and have them both hooked up to mains power while testing the signal assuming there isnt a ground fault?

>>2680747
Same principle as a phased array.

>>2680766
The vortices of adjacent fans kinda cancel each other out. The initial flow near the array will definitely be turbulant, but 3, 5, 10 fan widths away from the array and the flow should be pretty laminar. You just lose some energy to that initial turbulence.

>>2680887
Ground != neutral. If you're measuring a mains circuit, it will either be neutral referenced, or be referenced to the negative rail on the output of a bridge rectifier. Neutral can vary by a few volts with respect to the ground that the oscope uses. Clipping your ground clip to a neutral line will divert some current that would normally go through neutral to go to ground through your scope. This is pretty bad practice as the grounding wires in your scope aren't meant for much current at all, and it's definitely a source of internal noise. If it doesn't just immediately trip your RCD. Just measuring with respect to ground instead will result in your measured value having those few volts of noise on it, not a big deal for a high-voltage circuit but if it's some low-voltage logic circuit it won't be very nice at all.
That negative rectifier rail will vary by the entire full scale of the mains with respect to ground, so you definitely can't measure anything with respect to it with your scope.

Even clipping the ground clip of your scope to the grounded terminal of a mains appliance can route some of that common-mode noise via the scope and give you less-than-perfect measurements. Ideally you only ever want one ground connection.

So either shell out for a differential probe (lol), use differential measurements via your scope and halve the number of effective channels you have, or just wrap kapton tape around your scope's ground pin. If you do this often enough, it may be worth installing a switch to connect or disconnect the ground wire from your scope's internal circuitry and chassis.

>>2680887
PE (protective earth, round prong in the US socket, or side contact in schuko) is connected to oscope ground, thus there is a chance of some dumb retard connecting a clip to the negative (0v ref) rail of SMPS and short the mains via diode bridge via oscope to PE, which will trip RCD and/or pop a fuse.
Which is why you need an isolation transformer either for DUT or the oscope... Or just break off ground pin of oscope ;D
Or differential probes. Or even optically isolated probes. So many options.

And if you don't have the bridge rectifier, it is probably a 100% passive circuit and you can simulate in in ltspice or whatever without any issues. Or hell, even on paper.

>>2680920
>Or just break off ground pin of oscope

Nah, do it the way jesus would.

>>2680913
>>2680920
Not that Poster but I don't think I get it
I need to trip the fuse or whatever myself to understand what I did wrong

>>2680929
Lemme draw something.

>>2680920
So an AC circuit wouldn’t have that problem then, im talking specifically about something like a guitar amplifier which already has a transformer in it, it should act as an isolation transformer,

>>2680920
If the osc probe is connect to the PE ground via the grounding clip then you cant use it on a component hooked up to the same main then without creating a ground fault then right. Im talking specifically about a guitar amplifier ac circuit it has a transformer in it even. But if i hook up the probe to the circuit then put the ground on on the chassis ground, i just made a ground fault loop between the mains i believe. How do you determine a good ground spot for your oscilloscope probe even with an isolation transformer?

>>2680968
If it already has an isolated AC-to-DC converter of some sort, there's no problem with using a scope on it. If it's a grounded AC-to-DC converter you may have issues with common-mode noise and ground loops, but it's not dangerous at all. You should get perfectly normal readings, though if the device draws a particularly large amount of current or is otherwise electrically noise you may see some funky noise. Some appliances don't have isolated converters (night lights, mains power meters, smart switches, and LED light bulbs come to mind) which is where the danger is.

Bought a fuckton of common-anode displays on ebay. I counted how many I would need for all my projects, and I'm left with 18 displays that I won't use. Any ideas for a project I could do? This is after I set some aside for an alarm clock, pulse counter, microcomputer display, temperature sensor, and current meter.

>>2680747
>>2680763
How much does this cost? And the value of something like this is in software and the support.

>Benefit from the full potential of every wind pixel by controlling the windshaper through the powerful Python 3 API – an object oriented intuitive and performant code that controls WindShape equipment and manage test data efficiently. Implement your own time variable wind functions in just a few lines; define your wind speed regulation loop using your own flow sensors; and so much more…

That's all fucking bullshit. Wind Pixel? Gimme a break.
Just get a big fucking fan out of industrial supply catalog and invest in sensors for your drone instead.

>>2681096
Probably a lot given it’s for like,, institutional use
But like could whip up some basic software pretty easily

>>2681082
current and voltage displays for a CC/CV PSU

>>2680622
>>Bosch
doesn't make LEDs at all.
IDK what trademark was, only that was a big one.
Do you have more info like your picrel, it seems good to read.

>>2677019
One of my boys smashed to bits their laptop, I was checking if this could be fixed to make it work again. But I find a connector of the board showed broken, can you tell me please, how it's these named, I dont work often with that, I only know it must be a flat flexible cable from hp (in the service manual it says its a ZIF connector, yet they don't describe their pins or nothing).
Are these things propietary?
How must I interpret the description of the cable, it seems to have 22 lines/pins(?), and how to search for a connector to weld it.
Thanks in advance.

>>2681121
The connector itself is broken? To replace one of those you'll likely need to use hot air and be somewhat skilled to not delaminate the board. There are a few computer repair people who do this kind of repair, you may be able to get it fixed for under 100USD. Possibly.

I'm designing a mechanical keyboard, and I'd like to use thick, enamelled copper wire structurally as well as electrically. I want my keyboard to fold, and I'm imagining a hinge where one section of exposed copper wire is coiled around another. My question is whether this hinge can be made electrically conductive enough for the keyboard to work. Is it a bad idea? Will the copper eventually oxidize? Are there ways of making hinges out of conductors? Are tiny electrically conductive hinges something I can buy? I don't want to ruin my aesthetic with ribbon cables.

>>2681121
First of all, from your picture, it looks like that cable it connecting the USB ports to the mother board. You could just not fix it, you won't have USB ports but the computer should work fine. Now to answer your question.

Those are "flat flex cables", they are usually customized in length, amount of wires and shape for specific applications. The connector itself are usually generic, but finding the exact connector that matches the footprint on your PCB might be hard to do and is very important.

You can try to go on digikey and search through the connectors, it has a good parametric search function that allows you to filter, otherwise we're gonna need a picture of the PCB. When searching for the zif connector on digikey, make sure to select the right one. There are connectors that connects the cable at a 90 degree angle (vertical to the PCB) and others that connects parallel to the pcb, so make sure to select the right one. There will also most likely be different amount of pin connectors, so make sure to select the one with 22 pins. Also make sure that the footprint (dimensions of the connector + pads + height if the space in the case is extremely limited) match the original one. If you post better pictures of the PCB and connectors, I could try to find a suitable one.

Finally, remember that those connectors are made out of plastic, surface mount and a bitch to solder. It might be easier if you could somehow solder the wire directly to the pad, bypassing the connector. Whatever you do, careful not to solder the wires or connector upside down.

>>2681154
>My question is whether this hinge can be made electrically conductive enough for the keyboard to work. Is it a bad idea? Will the copper eventually oxidize?
Yes. But you can probably solder a wire between the two sides to bridge them.
Or even better, a short piece of desoldering strip. It's tougher than a wire, very flexible, more than conductive enough, and will probably fit the "bare" copper theme you are going for.

>>2680775
Please explain to me like I'm 5

>>2681244

>>2681154
>>2681163
Why would you want to touch an exposed energized circuit with both hands all day

the more and more I (try to) lay out my PCB the more and more it feels like I'm fucking it up royally despite all the videos I've watched. I can't get shit to consistently align to the grid; KiCad itself is gaslighting me by subtly changing the grid origin while I work; it feels like nothing is aligned with respect to the other components despite my best efforts (this I suspect is because of the component sizes/pads not being aligned to millimeters or even multiples/factors thereof so it's not my fault this time(who the fuck thought that would be a good idea)), and then I feel like I haven't made the board the correct size and that I don't know how to place the mounting holes a "pleasing" distance away from the edges of the board... the list goes on. What's worse, the further I get along and the more experience I get, the more I'm realizing the routing I did at the beginning of the board is unsalvageable, and as I fix it, it causes knock-on problems with the shit I've already routed

assuming the rat's nest is all gone and everything is routed to where it needs to go, even if the board is laid out like shit, it should still work, right? I don't have anything high frequency or too finicky on here

>>2677019
is there a cheap way to probe high and slow side FETs on a low voltage high bridge driver (<24 volts)? research has lead me to believe that most differential probes are for higher voltages and will suck at such low voltages.

>>2681269
>even if the board is laid out like shit, it should still work, right?

circuits laid out too neatly create compound waves of electromagnetic infetterence.
chaotic designs, such as this Macintosh prototype, eliminate said infetterence by cancellation and mastication.

i have no idea about electric engineering.
i broke my snes in an attempt to fix the AV out port as well as the power plug.
its not turning on anymore and my first assumption is to swap out the fuse.
i need a 125V 1.5A fuse.
i found this amazon listing that has me a bit confused. it says its a 250V(125V) 1.5A fuse.
is it 250V or 125V now? or are they interchangeable?
https://www.amazon.de/Sicherungen-Glassicherung-Glasrohrsicherung-Feinsicherung-Ausgeschnitten/dp/B083R3VYJ5/ref=sr_1_21?crid=29XSIRBT24OKN&keywords=125v+1.5a+sicherung&qid=1694625769&sprefix=%2Caps%2C113&sr=8-21

>>2681277
>or are they interchangeable?

yes, it's their way of informing noobs they're for use in NA and Yorup alike.

>>2681281
great, thanks!

>>2681277
That really sucks. I hate to say it, but if enough current was drawn to trip the fuse, you probably have bigger problems to worry about.

Putting that aside, for fuses, the voltage says "this fuse will tolerate up to X volts before failure". But, fuses aren't for voltage - fuses interrupt current. The current rating says "this fuse is designed to burn out at Y amps". You want Y to be AT MOST the first fuse's current rating. You want X to be AT LEAST the original fuse's voltage rating.

it looks like the 250/125 V thing is trying to say "these work with either European 250 V AC or American 125 V AC", so it should be fine

>>2681269
Things are not aligning properly because components use a mix of imperial and metric systems. You care too much about how things look though. Making PCBs is not art, it's engineering and everything should have underlying reason for why something is done in a certain way. How PCB looks should be on the bottom of priority list when it comes to designing a working board. You shouldn't even be concerning yourself with how things look, just let things develop naturally.

>>2681282

to add: the current is the most important spec you need to follow.
but just as important is size, make sure they match length and diameter.
most common lengths are 3/4-inch and 1-inch.

>>2681275
>>2681284
Alright, point taken.
If you guys don't mind another question, the only thing I'm really confused about right now is how to route power. I have 4 different power rails, +3.3V, +5V, +12V, and -12V. Let's say for now I'm using a two layer board. When I'm routing power to the various ICs on the board, do I route the power just as I would the signals? As in, use (thick) traces and vias to connect the power pins of the ICs to the regulators? I know of power planes but they seem untenable with 4 different voltages. I just don't know when to use one over the other.

>>2681291
You don't need power planes unless you're distributing >20A of current to multiple devices. You should read
>Bogatin's Practical Guide to Prototype Breadboard and PCB Design
You can also use kicad's track width calculator to get required widths. 0.3 mm trace can handle 1A of current. Good decoupling and solid ground plane are a lot more important things to consider. If you're routing 2 layer board then bottom layer should be sold ground to reduce noise, but again read the book because it will explain a lot of things better than I can do in a post.

what is the best way to fix a cable like this? i fucked up and did cut the own the cable of my electric hand saw lmao

>>2681253
The wires are on the back side. And even if that weren't the case, 5v is nothing.

>>2681320
Best thing is just to replace the cable or use whatever is left of it and connect the good end directly to where old cable was. You could solder it or twist and tape, but that will come apart eventually and be a hazard.

>>2681320
there are some shrinking tubes that includes solder, not sure how to call them in english, should work for that use

>>2681269
take the manhattanpill

>>2681320
crimp it with heatshrink butt connectors

>>2681277
Your first assumption should be tested. Checking a fuse is super easy with a DMM. If you don't have a voltmeter or digital multimeter, walmart sells super cheap ones (15-30$) in the automotive section. If you don't feel like buying a multimeter, you can test the fuse by removing it and connecting it to maybe a battery and a resistor and led in series see if it still conducts or even perhaps physically examine the wire inside the fuse if it's a clear glass case and see if the wire is burnt. If you don't have a DMM though, you won't be able to fix the snes at all if it isn't the fuse.

>>2681301
Thanks for the book rec, this seems fantastic so far. I wish I had time to read it all before the board has to be sent out, but I'll do my best.

>>2681320
Strip, put on heatshrink, twist wires, solder, shrink the heatshrink.
But it is better to just replace the cable.

>>2681320
A lineman's splice if you're feeling fancy

>>2681301
>Source is Wikipedia
chad book

>>2681273
subtract two scope channels

>>2681250
what are the diodes for?

>>2681374
lmao what the fuck ahahah

>>2681423
>diodes for?

they act like zeners, ensuring sound card gets at most +/- 0.7V.
but they can start conducting as low as 0.2V
so gotta keep signal less than that for signal fidelity.

>>2681423
Clipping the voltage so it never exceeds +/- 0.7v. CMOS interfaces can get fried at some pretty low voltages (I think around 10v), so it's good practice to clamp the inputs when you're pumping unknown signals into your soundcard.
Also useful for piezoelectric microphones, should you ever choose to experiment with those. Striking one directly can produce a pulse in the kilovolts.

And you might want to try one of those cheap $10 usb soundcards off ebay instead, so you don't risk frying your motherboard

Man, single-sides boards are a pain to route. And the ground plane always ends up really shitty. If only my laser and drill would stay fucking aligned to one another.

>made all my components 0402 without checking what currents would be flowing
they're op-amp filters
surely there won't be large currents flowing through them
it'll be fiiiiine

>>2681430
>>2681427
thanks.

>>2681126
>>The connector itself is broken?
Yes, it is half smashed.
>> Possibly.
I didn't think this would be so complicated and expensive to get fixed.
>>2681157
>the computer should work fine
Yes I tested it without and worked fine, now the thing is the arm connecting the processor to the fan its broken and it seemed to have some kind of liquid inside, and the plastic socked for the dimm memory just fell down from the pcb and I don't know what piece to request/but since it does not come in the manual. And the screen its a no go too, but it can work with a external monitor.
>I could try to find a suitable one.
Thanks friend, please do not bother. Very kind.
I want to learn it. It is safe to think for every line of the cable it is a pin? I counted the lines, but since there are so small sometimes counted 22 and sometimes counted 24 (one or two times 20), but how can I know it for sure?
>solder the wire directly to the pad, bypassing the connector
That's what I'm going to try.
Thanks for your valuable support.
Both anons are very kind, and thanks for the kwnoledge.

>>2681468
> single sided board
> ground plane
Not very compatible concepts unless you use the term “plane” very loosely.
Most things don’t need a ground plane, it’s just some faddish shit they read about on the internet, like “low esr caps”

>>2681427
> keep under .2 V
wat? Normal line-in design is around 1 V pp.

>>2681536
>That's what I'm going to try.
just be careful not to solder it the wrong way. Make sure that the direction of the pins on the cable match the ones on the connector. Also solder one pin first, then heat up the join to allign the rest. Then solder the other end and if everything is okay you can drag solder the rest. Good luck! Here is a video if you want to watch someone do it.
https://www.youtube.com/watch?v=Y1cKtGnw5dE

Use leaded solder, proper tip shape and make sure to use flux! If you got junked electronics around, you could try to open it and see if you can train on it before doing it on the laptop.

>>2681468
Single sided boards are easy.
0 ohm links to the rescue.
>>2681545
I've always thought that TV boards are awfully designed. I'm pretty sure they could have routed all the stuff on one side, especially with through-hole components.
But I guess 15x15 cm board would look like a toy inside a CRT TV, and FR2 was cheap af.

>>2681573
is it really single sided if you use 0 ohm jumpers

>>2681661
if you use SMD - yes.

>>2677019
i want to make an electronic load using a DC SOA rated FET. the usual way to do this is with an analog op amp PID loop. is it at all plausible to do this in software with a MCU running a PID loop, outputting to a 16 bit 192khz or 384khz audio DAC? i really want the ability to tune the PID loop under various conditions (apparently even $$$ commercial electronic loads are prone to oscillations) without dicking around with resistor and capacitor values.

>>2681777
It depends on what exactly are you trying to accomplish with your load. You can't expect it to work properly if you want to generate fast transients because DAC bandwidth will limit speed at which it can respond. You should also understand analog side of things before slapping on another layer of digital complexity. Is load oscillating because of whatever is connected to it or because of it's impedance transfer function that is limited in bandwidth and is not ideal? If you connect under compensated power supply to a load and response is oscillating then is it really the load's fault?

>>2681777
you can digitally control an analogue pid loop:
>digital potentiometer
>switched capacitor
>DAC + vactrol
>DAC + transconductance amplifier
>DAC + any other VCA/VCF method used by modular synths

How do i check the polarity of a speaker?

>>2681813
Listen to it in stereo with another speaker. If it sounds strange, flip it around.

Are there any esp32's that work with 5GHz yet?

>>2681837
Maybe you can try cooling one with your liquid helium setup, keep upping the clock until it fails and report back.

>>2681325
>>2681326
>>2681333
>>2681370
>>2681372
fixed it with those weird tubes + tape and now it's working and the cable is very strong, thanks bros

Cheaper alternatives to this fucker?
https://www.digikey.com.au/en/products/detail/hvm-technology-inc/SMHV05100/10707943

I'm making an IED... I mean an e-scooter battery, and im thinking.. If im soldering battery anyway (don't have welder), why would I use some hot glue and garbage if I can fabricate PCB for cheap and use it to fasten, insulate and fix cells mechanically, and provide a nice place for BMS tie ins?

Also, is it worth copying Tesla battery fuses things, or it is a meme?

>>2681904
Actually... This is something people do with screw-mount cells, so maybe.

>>2681905
I see people do this with board-mount spring-loaded battery contacts too. It requires tie-bolts going between the two sides in order to provide the tension, and likely isn't the best solution for a vibration-prone thing like a scooter.

If you're fucking about making PCBs for batteries anyhow, why not just make a lithium ion-powered spot welder?

>>2681865
>10kV
Actually making a push-pull transformer driver that goes that high is pretty easy, just keep winding. The tough part is feedback, since even a 10MΩburden resistor is eating up 10W of power at that voltage rating.

If you're doing secondary-side feedback, 1GΩ is probably the minimum resistance required to not unnecessarily burden your power supply. Probably 10GΩ to be safe. If you want a 2.5V feedback signal from that, the low-side resistor is going to be 2.5MΩ, which you can buffer using a relatively common op-amp, so long as you're careful to isolate leakage currents from the high-side circuitry. Potting the whole thing is a very good idea. That said, the cheapest 10GΩ resistor I can find that's in stock on DigiKey is over $60.

You might be better off trying to do primary side feedback, which might not be quite as accurate (theoretically it should be fine, especially if you're using a resonant converter) but the feedback will be easier. Just measure and rectify the primary coil voltage for feedback. A sample+hold+release circuit or two would be more accurate than just a rectifier and bleed resistor, but accuracy might not matter that much.

You could even go open-loop if your load current will be constant and the voltage drop isn't an issue.

>>2681918
>If you're fucking about making PCBs for batteries anyhow, why not just make a lithium ion-powered spot welder?
Problem is that i'm using salvaged cells. I had two ninebot es2 batteries with BMS that constantly had floating issues, and when I tried to fix it, I released some magic smoke out of the BMS because it is super fucking fragile and sensitive and impossible to work on.
Cells are 100% good, but they have spotweld remains and i'm not really feeling comfortable grinding them off or trying to pluck them with sidecutters, and for spot welding you need flat surfaces.
>and likely isn't the best solution for a vibration-prone thing like a scooter.
Sometimes spot welds fail, and yeah, that won't last..

But im thinking of smth like this. It will make everything look neat, and I won't need plastic spacers or hotglue to hold the cells.

>>2681923
>i'm not really feeling comfortable grinding them off or trying to pluck them with sidecutters
You can shear them off really easily. If you ensure there's just shear force and no tension force away from the cell's end cap, there's basically no chance for damage. Get a sharp corner like a scissor blade and give it a couple of whacks with a hammer. A knife or chisel or side-cutter would pull too much away for my comfort, but it would probably be fine.

>Sometimes spot welds fail
I was talking about the spring contacts getting jolted and causing a cell to go open-circuit for a split second, which could cause it to arc. Spot welding is far more vibration resistant than brittle solder. Especially if you're soldering directly to a load-bearing PCB, I'd want to solder to wires if at all.

>>2681927
>Get a sharp corner like a scissor blade and give it a couple of whacks with a hammer. A knife or chisel or side-cutter would pull too much away for my comfort, but it would probably be fine.
Its too late anyway, as all cells are now tinned and soldered in parallel, so they all have exactly the same voltage when I assemble the thing.
And it would stay like this for a month or so, because there were couple cells with damaged heatshrink (apparently a design fuck up), and I decided to just re-wrap all the cells, and inspect them, as some have rust on them which is odd, and I'd have to order the heatshrink.
>I'd want to solder to wires if at all.
Wires are ugly and get in the way when assembling entire thing, They add like 3-4 mm on average per side. PCB would add no more than 2 mm.

Idk really. Spot welding would be nicer, but I've neither a welder, nor new spotless cells.

>>2681921
Hmm, OK. Lots to consider
I might make a dedicated thread for this but basically:
My load will be far from constant. I'm aiming for about 10hz of bandwidth.
Accuracy is not super important, as long as I'm within about 20% margin of error at worst. I also posted the wrong thing, my current max is 1mA not 1uA so I want her to top out at 10W. My best bet rn is looking like a DC oscillator into a transformer

>>2681799
thanks, i managed to find this as well. https://pubs.aip.org/aip/rsi/article/88/12/123712/362253/Digitally-controlled-analog-proportional-integral

>>2681964
...although i don't quite understand why they have 8 different derivative op amps when they could just use a single op amp with relays to switch different combinations of cd1...8, rd1...8, and the negative feedback resistor.

>>2681975
I think a mux might be cheaper than 8 relays...

>>2681964
>tunning
lmao

>>2681975
Are all the Rd1 Rd2 values equal? I think they are since the DC gains are all the same. So then yeah you could just put the analogue switch ic before the op-amp but after the caps.
>relay
lmao

>>2681980
I assume S1 to S8 are floating analogue switch elements, shouldn't be too expensive even if he does need to swap out all the derivative resistors and capacitors.

>>2681975
Analog switches/multiplexers aren't ideal components. They have parasitic resistance of several 10s of ohms or even more that is non-linear and changes with applied voltage. They also have non-negligible parasitic capacitance which severely limit usable bandwidth. This forces you to drive them with low impedance sources (buffers, outputs of op amps) to minimize errors. Relays are closer to ideal switch but they're expensive, op amps are cheaper.

These 550mAh batteries look rechargeable, how can I try to charge one?

>>2682005
Do you have a PSU with CC?
Then just set the current limit (0.1A should be pretty safe) and put 4.2V on it.

This looks like it has a protection circuit on it which might protect from overcharging, so you might not even need CC.

>>2682007
I have a lab PSU and a FLIR, what max temp should I allow during charging? How long should it take?

>>2682036
The slower you charge it the better. If it gets hot, you’ve failed. At trickle charge, most devices say so something like 10 hours or so.

I an electronics noob and I'm trying to build a VCO for a synth project. I read online about using a voltage follower and high order filter to improve the output stage. I have a MAX293 IC that I can use for a filter. Are there any examples online that illustrate something similar that I can learn from?

>>2681865
>this expensive
Which products actually use this?

>>2682110
it looks to me that the MAX293 can be slapped onto the output of my VCO, no voltage follower needed since the signal will pass through the op-amp within. I don't know what to use for the clock signal, though.

I have a little design that I want etched into the solder mask on my PCB. The mask and copper go off the board a bit, and it also intersects two of my mounting holes. I don't mind either of these, but I know the board manufacturer will.
How do I properly clean this up so that it passes DRC?

>>2677019
Anyone work with embedded displays?
Trying to make a full color race dash for a gokart
Nextion displays are super limiting
So I also bought some $20 LVGL esp32 board with a screen, but this fucker is slow as dog shit. Are partial refreshes faster? This gives me zero hope.

I'm leaning more towards the nextion even tho setting it up is gonna be a massive pain.

I need to also log GPS/engine data at 25hz and I think lvgl is going to eat too much into my CPU overhead, never worked with anything other than i2c displays before and not sure the direction to take here.

>>2682187
you dont need to have it pass drc. doubly so if you give them the gerbers and not the pcn file. just make sure your board outline is correct and the fab house will cut it anyway. the point of drc is really only as a sanity check while youre designing the board

>>2682206
you need a display and graphics accelerated mcu such as those in the stm32f7 or f4 lines. the keywords are lcd interface, lcd tft, or parallel fifo. or you could use a mpu with a hdmi interface. esp32 specialty is 2.4ghz radio, not cameras and lcds. get the right tool for the job. standard i2c/spi displays will be way too slow for the size you need. a qspi interface or an octo spi interface would probably be fast enough, but the display has to support it

>>2682187
I wouldn’t put it past the chinks to put solder mask atop your pad. Can you put a mask layer keepout zone? Or add a mask layer polygon to enforce no mask?

>>2682297
that's what I'm afraid of. I've been trying, but kicad doesn't make it easy and even though it seems like something that should be easy to google, I'm not getting to much info on how to do so

>>2682237
everything else on the board passes DRC for pcbway, it's pretty much just this design that causes kicad to freak out. I've gotten PCBs printed with silkscreen off the board and they turned out fine, so I'm hoping this turns out the same way.

>>2682241
Yeaaa but I want wifi for data downloads because racers dumb from heat exhaustion at the end of the day

Honestly this nextion is gonna be great... but since it's a "product", trying to upgrade to a brighter display isn't exactly "easy". I could upgrade to a faster chip like the f series you mentioned, datalogging is a priority as well, but then I need to add back in wifi, and fight with SPI, I'm logging data at 25hz right now... so that makes me lean more towards the serial nextion as the esp32 has 3 serial interfaces

I should have gone to school for this...

>>2682323
Wait your decal is just excluding solder mask. So it’s not like it will place solder mask over the hole. It’s just taking mask away from something that already has none. You’ll be fine. Ask the fab house via email if you’re worried.

>>2682323
>>2682344
Is that also enforcing copper? It may just be reacting to the copper overlapping the hole. If it's just the hole desiring thermal relief, you may be able to turn that off (or set the minimum clearance to 0mm) in the hole settings.

>>2682335
>data logging
>data download
unironically might be worth looking into utilising an LTE modem that has gps. some modems even have wifi hardware as well. this can all be run through a single uart interface. this could work for you because you dont need to "stream" the data, and even if you did, like i said, a lcd accelerated mcu will have a much easier time running the screen in addition to other peripherals

I'm trying to repair my sister's PS4 slim. I've tested the power supply and it seems to be the source of problem to me as before the PS4 died, it was showing a red light to mean it was overheating. It eventually died as it was continually used even after getting the red light without being cleaned of dust. It then showed a blue light then a white light before not working at all completely. Could overheating blow the fuse in the PSU or is my problem likely on the motherboard (particularly the CPU, due to the overheating)? Appreciate any advice anyone could spare.

testes

>>2682428
Take it apart and clean it with a brush and vacuum. Then start probing for voltages.

what's more important: breaking up the ground plane as little as possible, or making as few vias as possible

>>2682755
Nothing really wrong with vias, unless you're making PCBs at home that is. I don't think fab houses charge for more vias. Via stitching around ground planes is actually an improvement for noise reasons. If you are making a high-frequency circuit, you still want to ensure there's a return path for each signal that minimises loop area, often this just means putting a via right next to the signal input and output pins. If you have this signal go through a via or two, you may also want to put return/ground vias connecting to the plane right next to these signal vias.

I'm a moron though, if there's a reason why vias are bad feel free to educate me.

First actual PCB I've soldered in a while. I do recommend these PCB vices.

>>2682760
It's a half-bridge that I designed as a generic fill-in for any project that needs it. The name is a bit of a pun, since it's the name of a situation in WWII and the film based after it, but also it's a half-bridge that suffers heavily from feature creep. Yes there's a 100µF electrolytic capacitor in parallel with a 100nF ceramic as a bootstrap capacitor. And that is a darlington capacitance multiplier as a voltage regulator. No clue why I used a THT LM358 and TL431. This time I'm using it to PWM a big brushed 12V DC motor for my PCB spin-coater.

Damn were these Power-SO8 FETs a pain to solder, I really should fix my diy reflow plate.

>>2682763
Here's the spin-coater: >>2682773

>>2682763
>>2682774
This brushed DC motor has current measurement but no back emf measurement. I still want to do speed feedback though, I wonder if I can do it via measuring the time between peaks and troughs in the current ripple?

How the fuck this balancer works?
It has just 6 discharge resistors. Usually, balancer on the BMS has 10 resistors for 10 cells.
It is based around PT6010 IC, which... doesnt have datasheet in English. But from google translate it does balancing feature, and in the datasheet, there is no circuit with transistors and low-resistance resistors (so probably it doesnt balance jack shit without said contraption, or it does it really slowly and constantly, like an IC from the Texas Instruments? ???), but the application guide has this thing, which is really similar to what I have on my BMS.

But yeah, after reading the datasheet, I've figured out that IC has many variants, and that made me think that e-scooter battery that I have right now (not the one im trying to build) is actually working fine, and it uses different IC that is programmed to cut off at 4.1V per cell instead of 4.25 like PT6010AA. And that all range issues are just due to the fact other chinks fucked up and shipped wrong charger with scooter that charges battery to 43V instead of 41V and when I contacted battery manufacturer, they told me to throw away that charger lol.

my scrollwheel click died, so I ordered new button.
Question - how do I solder it to the PCB? The solder joints are on the back of the button AND on the PCB and they're tiny as hell.
Soldering paste? Heatgun?
Issue is that the housing of button is plastic so I'm also afraid to melt it in the process

>>2682841
Hot air, i'd asl put kapton over the top of the button so it doesnt melt.
Since contacts aren't exposed on the side, you can't use soldering iron.
To avoid melting the button, use leaded solder (like dilute the solder on pcb, wick it up and apply leaded paste or leaded solder with flux)
You can also try heating from the bottom of the PCB to avoid burning the plastic, but I suspect they would survive hot air just fine,, as they are meant for reflow soldering by looks of it.

>>2682832
If that actually is a bunch of dual BJT packages, then it definitely needs to be designed for it. Otherwise you'll get fuck-awful hysteresis from all those Vbe drops. Can't see how it would have feedback either, since there aren't any other pins to speak of, though maybe it just pulses the transistor, and measures the cell voltages via those 1k resistors. Maybe there's something in the datasheet about that kind of timing diagram. Though chink datasheets don't usually have timing diagrams.

>>2682841
>computer mouse has tactile domes literally anywhere
throw it away man, that or 3d print a bracket so it can use an actual microswitch

>>2682842
you're right, the film is held by a glue according to reviews so it's probably the only thing I should be aware of damaging

>>2682859
>If that actually is a bunch of dual BJT packages,
In my board it has discrete separate pnp and npns. But I couldn't really reverse-engineer the circuit since PCB has soo many vias for no good reason, and I didn't really want to desolder everything just to see wtf is going on. Pic looks really similar to what I have, I think.
> Can't see how it would have feedback either, since there aren't any other pins to speak of, though maybe it just pulses the transistor, and measures the cell voltages via those 1k resistors.
It probably does. But thing I don't really understand is how 6 resistors is enough to balance 10 cells. On other BMSes and separate balancing circuit you have smth like this. 10 resistors (for 10 cells) and some sort of smart IC that does balancing per cell by turning on some transistor and wasting power on resistors, or TL431 with some transistors.

Also way circuit is drawn is really confusing. I think I will re-draw it. Maybe then it would be more obvious how did they do balancing with less resistors.

i need help with a circuit design, basically i need to design a circuit which outputs 2kilovots from a 12V power supply, but at the same time is safe to touch by limiting the current to a maximum of 1ma, so if someone accidentally touches it they only get a small zap
same way how touching your car can give you a static shock, those can be up to 10kv but are harmless
how hard would be to make something like that?

>>2682884
>and I didn't really want to desolder everything just to see wtf is going on
Can always probe about with continuity mode to see what's connected where.
>But thing I don't really understand is how 6 resistors is enough to balance 10 cells
There's a resistor in series with each odd numbered cell. And there's two transistors per resistor, effectively one per cell. This circuit only ever moves current from one cell to an adjacent cell, so you can just enable a transistor to send current between two cells via one of the resistors. Makes sense in a wonko world where chinky transistors are cheaper than SMD "power" resistors. The one resistor per cell method is easier, and I imagine those ICs they're used with basically have the transistors built inside.

>>2682888
Just add 20MΩ in series with the output, like what Big Clive does with his ioniser air purifiers.
>1mA
Probably too much, that's 10W of power, so probably more than whatever dinky circuit you design can output anyhow. You're better off aiming for the 100µA range, if not less.

Also if the output can be high frequency AC instead of DC, you can make it a lot safer at RF frequencies. Like what a tesla coil does.

>>2682889
>Probably too much, that's 10W of power,
actually the design will only be using 100mW to operate, 1ma of output power is the safety net i want to have where if someone touches it by mistake it will never output anything more than that so it can't kill anyone, however for normal operation the power consuption will never go above 100mW

>>2682890
1mA is arguably dangerous, any reason not to limit it to 100µA?

>>2682891
according to literature i read you have to get juiced by more than 10ma to get dead so 1ma seems like a good margin
but no, i don't need anything more than 100mW, as that is what i need for the device to operate, anything above that value means some dumbass touched it and and is getting zapped so the less it delivers the better in that case

>>2682892
The only issue with a series resistor is it will reduce the voltage slightly at normal operation. A factor of 100 between operational current and short circuit current means it will only drop by 1%, which is almost certainly fine. That would be a single 10MΩ resistor. As for an active circuit, arguably you could have a floating BJT current measuring circuit that powers an optocoupler when the current goes too high. The opto would turn off the switching converter, hopefully your output caps aren't too big. Maybe triggering a spark gap crowbar would be the way to go instead.

>>2682896
>As for an active circuit
i'm just gonna find some design and rip it off online, i have design many low voltage circuits for digital stuff but i am not comfortable at all designing one with voltage that can kill me

>>2682763
Shit, I killed the cap-mult by shorting the output with my scope probe, I'll fix it tomorrow. Something tells me a conventional voltage regulator would be more resilient. Maybe I should just increase the base resistance so it can't source as much current.

>>2682889
Not gonna lie, whatever they did is weird.
Especially since output of balancing pins has to be floating, pull-up or pull-down.

Love chink electronics, they always use weird circuits that don't make sense at first glance, but then you realize that it works, and that they saved every sheke- renminbi possible.
Makes me wonder, why didnt they integrate transistors inside the balancing IC...

do any of the handheld chink oscilloscopes have inputs with separate grounds / grounds that aren't common to one another?

>>2682928
No, almost every scope has common ground. If you need to probe with respect to different potentials you'll need a differential probe.

>>2682949
> scopes have common ground
Why? Isn’t that a major cause of blowing up your scope? I see no “pro” sides and all “con” sides to having a common ground.
My guess is it helps them sell more scopes.
Perhaps there was some TV/Radio reason that made more sense back in the tube era.
Is there a way to prevent your scope from blowing up, like perhaps plugging it into a gfci?

>>2682994
It's better to blow up your scope than to blow up yourself. If scopes were floating every retard would start poking around on mains and connect reference to live wire making exposed BNC plugs live. It's also a lot harder to make floating amplifier and problem has already been solved with use of differential probes that also address safety issues. If your product that plugs into mains has exposed metal parts, those parts must be grounded or you can't sell your product.

>>2682997
This, I'm one of the retards. I actually connected my probe to a ground removing adapter to be able to measure live circuits, then flipped the ground and live by accident. Had I not realized my mistake immediately, I would have accidentally touched the oscilloscope's metallic body and would have gotten zapped. Do not ever measure high voltages with a scope without proper grounding and proper safety.

>>2682888
2kV dc or AC? 12V power supply DC or AC? From the sound of it, what you want to do is design some kind of voltage doubling ladder. By selecting the right type of capacitors, you can limit the amount of charges that it can provide in short circuit condition, something similar to https://www.youtube.com/watch?v=jW3_txSfIAQ

Without knowing more about your specs, can't help more.

We have an issue where we can't use multiple things at the same time (e.g microwave and toaster oven, dryer and anything, etc.) because it'll trip a breaker. I finally sort of mapped out the electrical wiring for these outlets that have been causing us trouble. I'm very confused as to why this is the way it is. Why is the #/3 wire even being ran? The dryer apparently runs off of 120V, so why even have this? Also apparently every outlet in the kitchen is run off of breaker #7, but for some reason has the red hot from breaker #9 in there for some reason? Are the neutrals being tied together causing this issue because things are in different phases?

Also, the reason none of the ground wires are depicted is because none of them are connected to anything.

>>2683088
>The dryer apparently runs off of 120V, so why even have this?

this is 100% obviously wrong.
also, your diagram is stupid.
it should look more either more realistic, or more schematic-like - using standard symbols, or realistic drawings.
and why are white wires going to yellow boxes with the letter Y?
all these quibbles mean one thing: you're not reporting the situation accurately coz you are a crazy person.
and nobody's got time for that.

>>2683074
> ground and live reversed
I would think a GFCI would definitely solve that problem.
Back in the day—like in the 80’s—manufacturers started making ungrounded power tools with plastic bodies. They used to have metal, grounded shells, you can still see old guys with their metal black and decker drills and skil saws. Pretty much everyone in the industry moved to “double insulated” tools without grounds.
Perhaps the scope makers should actually learn about electricity and electrical standards instead of living in the 1940s.
I guess it’s part of their “let’s keep selling cheap lcd screens and smart-tv class ‘computers’ to people who don’t need them at 1000% mark-up”
That shit only made sense in pre-70s when you could get analog scopes with real vector CRTs.

>>2683105
>the industry moved to “double insulated” tools without grounds.

you talkin' out your ass, Paco.
when you use a scope on a lot of things, like a TV, you're working in close proximity to deadly live voltages.
grounded instruments offer considerable help in keeping you out of hospitals and morgues.

>>2683105
Oh the mistake happened on a philips pm3217 from the 1970s scope, modern day scopes are actually made of plastic, I don't think you can get a shock from them unless you accidentally reversed live and neutral and touched the bnc connector while energized, which is probably why the scopes are grounded, so it goes poof and trip the breakers instead of electrocuting the user.

>>2683121
> ass talk
I’m glad battery-operated portable scopes never existed, and apparently never will.

>>2683149
ah i guess that explains why these don't use bnc connectors in favour of doubly insulated connectors

>>2683101
It's definitely 120V, it says so on the plate. It's a natural gas dryer.
This is just an excerpt of trying to map things to what using pictures I took as reference. Yellow is there because that's the sheath of the wire. I would put white on the white ones, but then it might be more confusing because it didn't look like they're connected. Just using draw.io instead of LTSpice for this project.

>can kinda see ripple atop the high-frequency switching noise on my motor driver
woo, speed measurement should be doable
but the filtration will need to be digital in order for me to also measure the current spikes

>>2683121
> we can’t just stop using bnc connectors
Yeah, no progress has been made since the 40s.
Doesn’t matter, the EU will probably force manufacturers to use USB-C for probes and test leads anyway to justify their existence.

>>2683331
>>2683331
>>2683331
NEW THREAD

>>2681536
Watch out anon, the "arm" is most likely a heatpipe, you should order a replacement for your pc model because otherwise you're going to fry your cpu