/diy/ - Do It Yourself

>> 2744497
> overclocked, hack, why USB 2
No need to overclock, the rp2040 bus already does around 2 GB/s.
Not a lot is known about the USB implementation, it was primarily for drag and drool programming.
USB is a block protocol, we’re not talking about bit-banging it. It would have something like a complex PIO-like assist. Notably, a hardware shift register. That’s an 10x speed boost right there.
The why is for anything that uses, reads or writes data, and a better match for the capabilities of the rp0.
My camera generates > 256K raw images for each photo for christs sakes.

>>2744630
>the rp2040 bus already does around 2 GB/s
Oh wow, even though the cores run at 133MHz, the DMA bus can go a lot faster. Guess that explains how it can fetch instructions from flash over a serial protocol without having to wait a bunch of cycles.

Really would be a shame if they didn't implement a high-speed USB just out of laziness. It would be cheaper and more versatile than an FT232. A nice controller chip for an SDR.

>>2744598
Modern MOSFETs can definitely run at >500kHz if you put some care into your gate drive stage. See high-power discrete class-D amplifiers. Not sure about IGBTs, but they're probably similar. But I think modern induction hobs use like 5kHz, even 50kHz would be a big step up.

>>2744676
IGBTs are very slow. MOSFETs suck for high power loads because of Rdson
> But I think modern induction hobs use like 5kHz, even 50kHz would be a big step up.
At 5kHz it would be too loud, so it is definitely running above 20 kHz.
But idk, somebody did a video on induction cookers and he got 16 kHz.

>>2744833
>At 5kHz it would be too loud
is that bad? an unattended fire hazard should have some sort of audiovisual indicator

Why would someone need to order expensive ASIC if a million times cheaper FPGA from a marketplace can do the same?

>>2744911
Because it's cheaper at very high quantities.

>>2744911
FPGAs weren't even remotely cheap for much of the last 20 years, and still hardly anybody knows how to program them

>>2744788
This sounds like a really compelling work environment. How would I go about setting one up, as a retard who's never used any of the things you've mentioned?

Got this fucker from like 5 years ago. Picrel audio connector was ripped off the motherboard. Not sure what kind of connector it is/if replacements are available. Luckily the thing itself was still in there.

1 more pic

>>2745015
This is what the undamaged connector on the other side looks like.

What’s my best option here? Think I can just heat the right pads up and stick it back on? I’m afraid of melting the plastic, anything I can do to avoid that? Should I just cut the cable and run jumper wires/add in my own connector?

>>2745015
>>2745017
https://www.amazon.com/BTF-LIGHTING-Connectors-WS2812B-WS2811-WS2812/dp/B01DC0KIT2
Solder one to the PCB and the other to the speaker(s).

>>2745015
Was it just the solder that ripped, or is there damage to the pads on the PCB or to the leads on the connector? If the solder is all that ripped, you should be able to reflow it without too much difficulty. I'd add fresh leaded solder to both sides first, then add a dab of flux to the board before reflowing. If your iron has the right geometry to put heat quickly into the joint, you should be able to get in and out before damaging the plastic.

>>2744911
> fpgas are cheaper
Also, you design for much higher speeds and you can even throw analog/linear circuits on an ASIC.

>>2744035
>>2744061
>>2744069
>>2744087
I'm back from microcenter. Decided to buy a variety incase I want to do some other stuff in the future. Wasn't too expensive I think. Now I just need to find a 3d printable case I like. Maybe buy a male micro USB to female USB c/b for strain relief as one of you mentioned

when should i use (or not use) grounded mounting holes on a pcb?

>>2745089
>use

circuits very sensitive to noise, static, or proximity - eg. iPhone6 that dropped wifi signal when held in a certain way

>not use

most circuits, incl those where radio signals are involved, unless have external antenna.

>>2745089
>untented vias
Don't see those every day. Makes it easier to test the board I guess, probes sit nicely in them and don't fall off like on normal test pads.

>>2745089
There's usually no reason not to ground them, connecting them to a ground fill means a more robust ground fill. Having copper there in the first place makes the hole mechanically stronger, which matters if you'll be disassembling often or using significant clamping force. Having the hole surrounded by vias also makes it robust against clamping force.
The only important electrical thing about them is it allows you to carry continuity from the pad to a metal chassis, washer/eye terminal, or screw, which you my or may not want. If you have a metal chassis then I would want to connect this to the PCB ground. Some holes may not have copper in the hole walls but have copper on the flat faces in order to prevent current flowing through a screw, if that's important.
You may also find that in an RF circuit you can't have any metal near a PCB antenna. Whether there's metal, AND whether or not that metal is grounded, both will influence how the part interferes with RF.

>>2745139
One screw isn't usually enough and if you plate both holes you might get a ground loop.

>>2745189
This could be an issue in some circumstances. If I were using a star ground, I'd be sure all the fills from the holes lead back to the star, instead of going to other parts of the grounding. But for a common digital circuit it would be fine, unless there's a nearby source of magnetic flux (i.e. a nearby non-toroidal inductor, choke, or transformer) to flow through the loop area. Aside from having magnetic flux through the middle, I find it hard to imagine a case where a ground loop like this could cause issues.

A lot of things I see just ground through one of their mounting holes, which is also an option.

hakko chad

>>2745306
Terrible interface that could be vastly improved with just one more button.
And why did they design it like a fisher price toy?

>>2745313
You can get a fancy silver FX-888D, but only if you're a bong. NOT ALLOWED in the US.

>>2745316

>>2744885
Yes. Houseburners must be silent.
And induction cookers are probably worse than gas, since with sensor malfunction (or when retards put silicon mats under the pan), you can heat up oil to the ignition temperature lol.

I'm looking to get a variable power supply kinda thing, so I can electroplate and use it also for other small projects like that, does the brand or anything matter? yall got any specific recommendations for power supplies?

>>2745450
I just use a thrift store laptop brick wired up to a CC/CV LM2596 module with a current/voltage panel meter. Apparently it goes up to 3A, but I wouldn't trust my chinky module. Still, it's like a $15 power supply. For electroplating you can definitely get away with a variable switching supply, no need to spend extra to get a linear one full of relays. How many amps do you need? Some guys doing electroplating and electrolysis will go all the way up to 10A or higher, but 5A should be enough in most cases.

If you're up for a fun project, I'd design and build a CC/CV buck converter with a TL494 and discrete MOSFETs bolted to a heat-sink. Then you can just hook it up to whatever fixed voltage AC-to-DC supply you have, be it a laptop brick or an ATX supply. Bonus points for using a half-bridge driver with twin FETs for synchronous rectification.

If you also need the output to be smooth enough for analogue projects (audio, radio, instrumentation) then you'll probably want a linear stage in there. You can just add a linear postregulator after a switching supply, a capacitance multiplier should do the trick.

>>2745450
>recommendations

electroplating can use a lot of current.
once worked on an industrial supply that did 12,000A at 12V.
to get a lot of current at reasonable cost, i'd recommend an old PC power supply.
you get 3, 5, and 12 volts around 20A, and -12V around 2A.

>>2745464
I used an old crypto mining supply, 12 V at 240 A.
They’ve junked millions of these things.

My spin-coater operating is as scary as I thought it would be, even though it's somehow faltering before it gets to full speed. Time to increase the current limit!

What mistakes could cause an implementation of this circuit to go hard on +rail no matter what V+ is plugged into? Vin is 3.3v, Vcc is 5v. Vref is 0.

>>2745794
>Vref is 0
>Vss is 0
unless you're using a 358 or the specific chip shown, this is bad practice. TL07x's slam positive when one of the inputs are too low

>>2745801
Forgot to specify I'm using a 4558 and not a tl07x.
Vin is at 3.3v when it's slamming at positive, but no matter what I plug V+ into, nothing changes, i.e. even if I plug it into Vcc or into ground.
I tried increasing Vref to 3.3v as a test. Sure enough the reading changes: now it is at about 1/2 rail, which still isn't what it should show.

>>2745801

After more careful testing it actually looks like you were exactly right. I just also had a flimsy connection that I needed to fix, then I could see that exactly like you said, if either of the inputs are too low it slams the +rail.
I've been trying to figure this out for a while now, you're my savior!

>>2745812
Actually there remains something weird about this. Despite increasing Vref, I'm reading Vout = Vin when Rf=Rg. It should instead be Vin-Vref. Something's still fucky here.

>>2745084
Do these have a status LED to show they're getting power?

>>2745822

>>2745824
I think that's controlled by the chip like how Arduinos have an LED on pin 13.

Just pulled a PC case and power supply off the curb. It was otherwise stripped of parts. Want to pull the PS (which is high wattage) for another PC.
Is there a good way to test these things aside from plugging into a MB? IIRC testing voltage on these can give wonky results under no load.

>>2745895
Attach a load to 12V and short the green wire in the 20/24 pin connector to ground (any of the black wires). I use an automotive headlight.

would it be fine if i connected the pads in the center to each other by a wide trace? or would it be better to put a copper fill underneath instead?
im trying to power 4 micro servos via a 1x4 pin header

>>2745897
He should also put a load on the 5V and 3.3V rails. -12V can't handle much current so it's probably fine if you just measure its voltage.

>>2745924
The only downside to a wide trace is it will be more difficult to solder. 1mm wide should be a fine compromise.

>>2745897
Good idea; I think I’ve got an old h4 bulb somewhere.

How do I identify these LEDs?
about 2.7x3.7mm in size
diode mode on my multimeter reads 2.5V across the contacts

>>2746094
I'd guess it's a 2835 warm white 3V chip. Look at some datasheets on LCSC or wherever, not all of these chips have the same dimensions or footprints, or power ratings. The voltage may also be off, since the 3V of a white LED is usually too high for a multimeter to read. It may have just shown 2.5V as its maximum voltage. Ideally you want to test it with a higher voltage and a current limitation to 10-100mA (e.g. a 9V battery with a 100Ω 1W resistor). Or just measure the voltage across it while it's turned on, assuming it still works.

Figure out the actual current going through it too, by reverse engineering the LED voltage, source voltage, and the size of any current limiting resistors. For example, if you have a 12V LED light with three 3V chips per one resistor, and the resistor is 15Ω, then that's 200mA. That way you can be sure what current they can handle and hence what replacements to buy / what power supply you can give them. They may not use any resistors if they run on a constant-current power supply however.

Are generic capacitors good enough for buck converter input/output filtering?
The datasheet of the AP63203 I'm using says to pick an input capacitor with a high enough RMS input current as well as a low ESR, but none of the capacitors I've looked at seem to have one listed in their datasheets. The reference designs using this IC use ceramic 1206 capacitors with no part number listed, so I don't really know where to start with selecting a specific capacitor to use. Am I just overthinking things? Can I just use a generic capacitor with the right capacitance/voltage rating, or do I have to find a specific one that would be right for this application?

>>2746127
I think all ceramic caps will have really low ESR, and hence should be able to handle big current spikes. The ESR and RMS current rating matters mainly for electrolytic capacitors, read one of their datasheets to see. In some applications you actually want some ESR, like some linear regulators.

>>2745895
>IIRC testing voltage on these can give wonky results under no load.

this was true maybe 10-15 years ago. not today.
and the rail that needed a load was the 5V rail, not the 12V.
since that was the only rail that provided feedback to the primary.
(the other rails followed along, due to fixed transformer ratios)

>>2746134
How do you really test a PSU if not for loading the rails? I'm fully aware of older PSUs wonky feedback without a load.

>>2746139
>How do you really test a PSU

the usual way. put a load and measure voltages.
my point is that you dont need a dummy load to prime the supply.
it's already primed and ready: legs akimbo and ready for your seed.

I got a fuckload of Phillips LED strips that have a drive voltage of 34V or something like that. I wanna hook up like 100 of them on some rig, but I don't know what to buy for the right drivers. Any tips? Can I get any LED driver circuit, including the "12v" ones, and just let things roll, or what? I thought "constant current" drivers let the voltage go as high as it needs to be?

>>2746221
Philips*

>>2746221
I'm guessing there's no resistors on the strips? You probably need to find out the maximum continuous current draw of the LED strips, then buy a constant-current power supply that outputs that current, and can go to at least ~36V, and can at least go as low as ~20V. For example, here's the datasheet for a Meanwell 150W LED PSUs, of which there's an 18-36V model:
http://www.meanwell.co/assets/data/ELG-150-spec.pdf
It will output a constant current of 4.17A by default, though by adjusting a pot or PWM or control voltage you can tune that all the way down to 0A.

If you want to power multiple at once, you can either:
>put X strips in series with one another, and use a >34*X volt power supply rated at the same current as a single strip
>put Y strips in parallel with one another, and use a >34 volt power supply rated at Y times the current as a single strip (balancing resistors will likely be required)
>a combination of the above
You may even be able to do a mains-powered solution by putting 8-10 of them in series (4-5 for 120VAC) with some sort of high-voltage current regulator.

They almost certainly won't be visible with 12V.

>>2746232
Why do I need balance resistors in the Y option?

>>2746249
>Why

each LED has a unique breakdown voltage.
whenever you put 2 in parallel, they might share the current 50/50
or 60/40 or 90/10 or whatever.
so some could be dark while others could be on fire.

>>2746249
Some might share the current 51/49. Doesn't sound bad, but LED forward voltage for a given current has a negative temperature coefficient; forward current for a given voltage has a positive temperature coefficient. The LED getting the 51% split will get slightly warmer than the one getting only 49%. Because it's warmer, it will let even more current through, getting up to maybe 55%, putting the other one at 45%. That makes it get even hotter. This phenomenon is called thermal runaway. A resistor that drops ~5% of the total voltage should keep the current split sufficiently evenly. It's also an option to use a BJT current mirror, though I'm unsure of that circuit's tempco.

>>2746102
>>2746094
alright some more info.
25ma current draw
38ohm resistor in series
2.7v across the LED while running

Any RF guys here? I'm an EE and wanted to learn everything that there is to know about Antennas, RF, EMI, etc, on my own time, anyone has access to university level courses that I could access for free? I need some kind of outline of the topics to read, the textbook to use, and if there are lecture videos, the better.

>>2746270
>25mA
Is that the current draw it's supposed to have, or the current draw when you were testing it by injecting voltage? Because it sounds really low, I'd expect 100-500mA for those chips.

Is there a part number on the LED strip?

>>2746490
Ask the /ham/ thread. They do plenty of making their own antennas and stuff.

My haul from LCSC. The mystery items were:
>shitty hand-held packing tape dispensing thing
>single large empty SMD reel
>five smaller empty SMD reels
Fun. Anyhow I got my 50pcs GL200N06 MOSFETs, along with my TMC4671 FOC motor pre-driver ICs, so I'm happy. Time to start designing a PCB I guess.

>>2746157
>>2746134
OK, good to know. My knowledge on that topic is from mid-2000s.
> primed and ready: legs akimbo and ready for your seed
lol

>>2746094
Only thing you can identify is package, and it is kinda weird.
Looks like 2835.
2.5V = 1 chip
Current - well, you will never know, because you can have small die and it would be rated 20mA, or big die that is rated 1W. All in the same package.
Same for CRI, CCT, etc.
>>2746270
Generic 2835 LED. 0.2W one. With nominal current of 65 mA or so.
>>2746526
>Because it sounds really low, I'd expect 100-500mA for those chips.
I'd not recommend running 1 chip 2835 at this kind of power.

What kind of diodes do I need to hand wire a macropad. I'm looking at quite a few selections on Amazon as well as breadboard kits with little diode selection but a lot of other junk I may or may not use. I've seen videos use "4148".

>>2746575
>65 mA
Even though the package looks like it can handle half a watt? If it's sinking heat into an aluminium PCB, at least.

>>2746609
1N4148 is the bog standard for this purpose. It's the same as a 1N918, apparently. These diodes are fast and cheap, not rated for too much power but as a result they're nice and small.

>>2746666
>It's the same as a 1N918

Sorry, I meant the 1N914. I always have 8's on my mind in loving memory of our liebchen Führer.

>>2746688
t-thanks

>>2746688
Nice get, what RS232 line driver IC do you use?

>>2746712
>RS232 line driver IC

sorry to disappoint, but still havent used up all the 1488/1489's i bought in the 80's.
mostly coz i've favored wireless over wired.
it's closer to magic.

>>2746728
My workplace was getting rid of a bunch of old ICs they never sell, so I got a bunch of MC1488s and MC1489s. MAX232s also, and a large amount of TL494s which I'm quite happy about. They're all SMD, we're getting rid of all but a couple of SMD parts for whatever reason, guess hobbyists don't buy them. At least not from us for $20 for a pack of 10 ICs.

>want to import symbols and footprints and 3d models to kicad
>spend like 30 minutes trying to figure it out (the easy install tool is windows only)
>finally get them in
>import symbols to eeschema
>it's just these auto generated pin-by-pin symbols that just mirror the actual pinout
>vccs and grounds, inputs and outputs, on all four sides
just fuck my shit up

>>2746666
>Even though the package looks like it can handle half a watt?
It's not about the package, but about the diode chip itself.
1W would have really big chip, and really low voltage drop at 25mA (lower than 2.7V)
And most 1W 2835 are usually 6V or 9V (i.e. 2-3 chips in series).

But yes, its generic 2835.

>>2746743
that's one done, took like 40 minutes
someone should train a neural net to make a sensible symbol from diagrams in the datasheet
or just a normal algorithm that uses normal ocr i guess

>>2746762
>took like 40 minutes
KEK gg scrub.

A bit of a "different" question.
I have a tray of 3Dfx "Bruce" Texture mapper ICs I keep as a display piece. They're QFP208.
A while back, they got knocked over and there was unfortunately some amount of pin bending (see picrel). The damage isn't too bad except for one IC which is significantly bent.

I'd like to rectify this. Does anybody have any tips/ideas? I was considering just xacto blading the pins to straighten them out a bit, but I'm not sure how I can fix the legs that are bent inwards/outwards and not side to side.

>>2746728
> wireless
Wireless is bullshit. I have to do wired all the time now due to the neighbours wifi and bluetooth noise, and they’re using up like 8 channels each and multiplexing them. I get pop-ups on my screen all day long for bluetooth devices that want to connect to it. Plus, jammers are so common place now, they render your wireless security system useless immediately. Went back to wired and PoE since there were no outlets anyway.
>>2746733
> workplace getting rid of
They’re fucking NAND gates that can handle 12V with ESD protection. How is that not useful? I’m tired of companies that would rather pay money to burn shit rather than put them on sale.

Any soldering station recommendations? I'm looking at doing some DIY guitar pedal kits for fun.
Are any of the budget T245 clones good? Yihua 982, OSS T245, KSGER C245 etc

>>2746844
3D printed jig that you press them into.

>>2746982
>Yihua 982
Only 40W, literally no reason to get this over a T12 setup
>OSS T245
130W output
>KSGER C245
Somehow 180W, even though the PSU looks identical to the 70W PSU inside a T12 station.

The Aixun T3A is the only one I know of that can output the full 200W that the T245 tip can handle, though whether you need that or not depends on what you plan on soldering.

>>2747005
Aixun T3A is closer to 200$ here which is easily 3 to 4x the price of the others, how much would the power output difference matter?
In my case I'll most likely only do guitar pedal PCB kits, guitar wiring (pots, pickups) and maybe attempt a Picofly Switch installation.

Would an "underpowered" T245 like the OSS still be better than the older KSGER/Quicko T12 stations?

>>2747005
>3D printed jig that you press them into.
I was thinking something along these lines too. Do you mean where it's recessed and there are grooves for the pins?

>>2746259
>>2746258
Why isn't this accounted for in the power supply? I'll get some fkn photos or a datasheet pulled up soon

>>2747043
In your case, you can certainly get away with below 100W, hence I see no reason to go for a T245-tipped machine over a 70W T12-tipped machine. Unless you plan on getting into aluminium PCB design/modification, or RC soldering of XT60s.

Going for T12 also gives you the freedom to use an external power brick, since some of the built-in PSUs can be a bit sketchy. The only T245-tipped station I can find that doesn't come with a PSU is this:
https://www.aliexpress.com/item/1005005408604037.html
Which can apparently handle the full 9A, and so gives you room to start with a 4A or whatever PSU and upgrade it later if you need to. Assuming you can set a hardware current/duty-cycle limit, that is. Some assembly required.

>>2747088
>grooves for the pins
I wasn't thinking of getting the pins with the right separation to one another, just get them the right position relative to the edge of the package. Like up all the pins with a channel that runs around the IC.

>>2747126
The PSU will output a constant current. If you just have one string of LEDs, then there isn't a problem. But if you have multiple strings in parallel, your PSU will be outputting a constant current without any way to measure how much current is going into each string, only the total. The easy way is just to have a seperate PSU for each string.

>>2747126
>Why isn't this accounted for in the power supply?

it is.
constant current supplies handle this easily.
if you have the cash and the space for 50 to 100 supplies, you're golden.
but your requirements are way out there, so, to save money you gotta improvise.

could someone please tell me what resistor this is on the left? I’m reading it as orange grey orange red gold which I interpreted as 38300 ohms but I can’t find that resistor on google so I feel like I’m doing something wrong

>>2747235
Looks like 3.3k Ohms to me.

>all entry level EE positions require 3 YOE and pay ~70k
>this is in CA
lol
What a meme field

>>2747184
>I wasn't thinking of getting the pins with the right separation to one another, just get them the right position relative to the edge of the package. Like up all the pins with a channel that runs around the IC.
Thats a great idea. Then I can just bend them laterally with an xacto
I'll try that

>>2747272
YOE gained during school count too
they just don't want anyone who's never touched an iron

>>2746982
Get one with hot air.
T12 and similar integrated tips are nice, but if you can get a good deal on Hakko 900 non-integrated knock off + hot air, just go for it.

>>2747341
most EEs never touch an iron in their career

>>2747427
Pathetic. They should just become physics professors so they can tell their students about the need for communism for human space exploration.

does anybody know of any decent through hole socket connectors that will work with 3 pin servo connectors? preferably one with some sort of locking system

>>2747667
Molex micro fit

>>2747667
locking JST

>>2747667
Whatever the hell these are called. I only know them as 3-pin fan connectors. They have a 0.1" = 2.54mm pitch that a stock 3-pin fan connector will fit just fine onto. Though it won't be locking unless you replace it with the dedicated mating connector. I don't think there's a socket version for PCB-mounting, if you wanted such a thing.

>>2747672
Here's an image of the opposite connector, cable-mount only.

shit, i bought the wrong hall effect sensors

>>2747674
>>2747672
molex 2510, also available as chinese "KF2510"

>>2747690
> wrong hall effect
There’s only one hall effect, and we’re pretty much done with it.
The hall effect was cool back in they day, but that fad is over now.
We went back to using little coils of wire and sense amplifiers.
And, of course, now we also just sense the back-emf from the motor coils themselves rather than bothering with another part.

>>2747775
Back EMF and sensor coils have a speed-dependant amplitude, and hence are not suitable for low-speed high-torque motor driving. As much as I'd like to just wind additional coils around my stator and change the turns number as needed, instead of having to buy a try a bunch of different hall sensor sensitivites. Isn't 160mT way too small for the inside of a BLDC right next to a rare earth magnet? I can't find a linear hall sensor that goes higher, these ones have an asymmetric curve so I can't use them anyhow.

Maybe I should just give up and build a seperate encoder or magneto into the motor.

>>2747799
I collected a few dozen inkjet printers walking by a set of dumpsters everyday on my way home from work. I took them all apart as the require $100 of ink. Virtually all of them have encoder wheels and/or encoder tapes, and their sensors.
Maybe do something like that.

>with a constant pressure, the voltage needed to cause an arc reduced as the gap size was reduced but only to a point
>as the gap was reduced further, the voltage required to cause an arc began to rise and again exceeded its original value
h-huh?
>early vacuum experimenters found a rather surprising behavior
>an arc would sometimes take place in a long irregular path rather than at the minimal distance between the electrodes
>for example, in air, at a pressure of one atmosphere, the distance for minimal breakdown voltage is about 7.5 μm
I guess the short size of it at atmospheric pressure means we don't notice it in everyday life, but it is really strange.

I just want to put a beta emitter in the middle of a vacuum chamber with an extra electrode going to the emitter itself, and wait for the whole thing to charge up to the electron energy voltage, without it arcing over.

Why are modern batteries complete and utter garbage?
>buy phone
>1-3 years pass with little phone usage
>phone is still fine but battery is practically dead
>have to buy new battery and attempt to replace the non-replaceable battery which is glued in place and rather difficult to replace
Thank fucking god the EU has mandated user-replaceable batteries by 2027. Either way I now have to deal with replacing the battery with zero guides online because it's not a very popular phone.

Also is there a type of battery for phones that is actually *reliable*? What do I look for? It looks it's all more of the same shit that will die in a few years, they're all lithium polymer.

>>2746094
All leds operate at that voltage + or - 1.5 v

So the constant current led driver for my bathroom light died. It's rated 280mA+-5% but the closest ones I can find so far are 250mA+-5% and 300mA+-5%, all other parameters are equal (Wattage, Voltage range, etc.). If I go with the 300mA I might end up frying it? So I'm thinking it's safer to just go with the 250mA one?

I'm continuing to search for a 280mA one but it's looking grim ngl.

Is it ok to have a power adapter constantly plugged in the socket and connected to a device? I have a 5v ac/dc adapter powering this cool little 3D printed arduino lantern I made by combining 20 different random tutorials. Power to the arduino is controlled via an on/off siwtch, although the adapter has a little red light that's constantly on when plugged, which is giving me second thoughts.

>>2747900
>EU has mandated user-replaceable batteries by 2027
looks nice on paper, but I can forsee a lot of red explosions in the future

/o/ tourist here. Im looking for a new soldering iron as my $5 china one dosent get hot enough after a few years of abuse. Any reccomendations for a new one under or around $20? Dont need any LED lights or gimmicks just something solid. Will only be using it for automotive wiring and occasional electronics repair

>>2747900
Phone batteries are exactly the same chemistry as in 18650s but they are charged to 4.4V instead of 4.2V.
This nearly doubles the energy density, but makes it last... makes it not last fucking at all.
>EU user replacable
You know they would say "unglue back cover, remove battery, glue new in, here is a guide go fuck yourself". Seen what EU did do with tracking cookies? Exactly the same shit is gonna happen here because EU laws suck.

>>2747954
250 mA driver would work, and it would be dimmer but LEDs would be used in more efficient region and would last longer.
I'd go with 250mA.

>>2748005
For automotive wiring you want something compact, so all soldering stations are out of the question.
Ive seen a lot of car mechanics using gas irons, but soldering anything except wires is pain with gas iron.
I'd buy something like this, 220V (or 110V) mains iron with temp control.

>>2748005
hi rbdk

>>2748005
You shouldn't solder automotive wiring. Get a ratcheting crimp tool and pin/housing kits, and marine-grade crimp butt splices.

>>2747900
The chemistry used is just not that long lasting to begin with, and they tend to get abused.
I'd gladly take a LiFePO4 phone battery, but the average person doesn't want to pay for a lower advertised (initial) capacity even though a lithium polymer will drop below it in no time

>>2745318
Pine64 Pinecil.
RISC-V processor, runs on DC barrel plug or type c pd. 5-28v, can do 80w at max power. Also cheaper than TS100. type c pilled

>>2748025
Agreed. I'd even use a phone that was twice the thickness of an average flagship samsung if it meant better cameras that don't oversharpen the image due to paper thin lens elements. And the fruits of added thickness being battery, and type c pd capable output. It's not too much to ask for giving me back my SD card slot, fuck you apple. Plenty of room for a headphone jack too, maybe one with enough power for high impedance headphones too. Fuck you apple. And fuck you samsung for following apples lead after you mock them for no headphone jack.

>>2748003
>looks nice on paper, but I can forsee a lot of red explosions in the future
What do you mean with red explosions? Chinese batteries have a really bad reputation where I live so I expect people to prefer the manufacturers batteries.
>>2748007
>Seen what EU did do with tracking cookies? Exactly the same shit is gonna happen here because EU laws suck.
The cookie tracking shit was poorly executed because politicians are fucking braindead when it comes to anything computer related. They should have mandated a way to automatically give or deny consent through technological means, such as setting it up once in your browser and letting the website know on first visit (or just send a cookie-value with your decision in every http request).
I'm hoping they get it right with the batteries but I can of course see them completely fucking it up again.
>>2748007
>Phone batteries are exactly the same chemistry as in 18650s but they are charged to 4.4V instead of 4.2V.
>This nearly doubles the energy density, but makes it last... makes it not last fucking at all.
>>2748025
>I'd gladly take a LiFePO4 phone battery, but the average person doesn't want to pay for a lower advertised (initial) capacity even though a lithium polymer will drop below it in no time
Interesting, maybe we need a better metric than simply mAh for consumers. When I look for a phone I look for a large capacity, but maybe an additional metric like "total expected watt hours over the course of a batteries lifetime" would be good when they advertise their phones.

>>2748052
>I'm hoping they get it right with the batteries but I can of course see them completely fucking it up again.
nah, they won't.
Technically a glued phone with pull tabs on battery and manual is considered user-servicable.
>Interesting, maybe we need a better metric than simply mAh for consumers. When I look for a phone I look for a large capacity, but maybe an additional metric like "total expected watt hours over the course of a batteries lifetime" would be good when they advertise their phones
Idk, all we need is 4.2 or 4.1V cut-off, i.e. 80% charging. Some phones have it, some don't.
Also charging speed matters too. Fast charging (>0.5C) is potentially damaging.
And good explanation to the end user:
>if u charge 100% with fast charging: 100 cycles until 80% capacity
>if u charge 100% with normal charging: 300 cycles until 80% capacity
>if u charge 80% with normal charging: 500 cycles until 80% capacity.
And phones should come with 80% charging limit enabled by default. Then batteries would last

>>2746844
I have never felt this level of jealousy before

>>2748033
>And fuck you samsung for following apples lead after you mock them for no headphone jack
THIS
I went for a new phone last year and the desk guy was like 'yeah, none of them have ear jacks anymore, had to make room for 3 more cameras hahaha'

>>2748064
On my phone I set the max charge to 70% using "modes and routines" thing.

Does anyone know what manufacturer or model type this style connector is? i think pin pitch is 1/10th inch
i would be incredible grateful, a replacement harness with plug is like 500$ for low grade wire

>>2748094
I mean I kinda don't get it! First they take away my headphone jack then they take the damn SD card away! can't have shit on modern phones other than 9 cameras and like 32gb storage that isn't upgradable. fucking hell

>>2748007
>This nearly doubles the energy density, but makes it last... makes it not last fucking at all.
Sounds like a blatant case of planned obsolescence. Doesn't Phillips do the same thing with their LEDs - overdrive them so they die in 2 years?
The subtle difference between 4.4V and 4.2V makes it extra nefarious

>>2748147
>Sounds like a blatant case of planned obsolescence
Hold on to your butt my dude, that's just the icing.
The clock speed on iPhone processors isn't constant throughout its life. It decreases because it monitors the total capacity of the battery and underclocks as capacity is lost over the years. If you have an iPhone that is 3-4 years with original battery and it works like a drunk sloth, this is why.
They claim it's to "maintain battery lifespan".

>>2748094
Xperia still has them

>>2748147
I bought some Philips hue bulbs on sale in 2016, I'm still using them across my entire apartment. Stop putting them on fucking dimmer circuits.

Have used 6-pin (6p) power-only USB-C for anything? These are cool little connectors but I don't know what to use them for.

>>2748181
I think Andreas Speiss just did a video on USB C for hobby projects.

There's so much confusion on the web about balanced vs unbalanced vs differential signal.
My understanding is balanced doesn't necessarily mean differential signaling. There is "something" (maybe a twisted pair?) that makes the connection "balanced" since the noise signal is induced in both wires: the signal wire and the ground wire. But I am not sure what happens on the receiver side. Is it essentially a single ended signal that is fed into a single ended opamp but since the same noise is injected in both the signal and the ground, it gets cancelled out? Even though technically it is not a differential amplifier?

But all opamps amplify the difference between V+ and V_. In case of a differential signal, V+ and V_ inputs will accept the input signals on two wires of the same magnitude but the opposite phase. So the result will be twice the magnitude on either line. And the common mode noise will be canceled. But if it is a simple balanced line, V_ will be at zero except common mode noise will also be canceled
So differential inputs are used in both cases?

>>2748147
Phillips was one of the members of the Phœbus cartel, one of the most well-known examples of planned obsolescence in light bulbs back in the 20’s and 30’s. It’s probably deeply embedded in their corporate culture. How they are allowed to still even be in the light bulb industry is beyond me.

>>2748239
The environazis should focus their energy on slimy industry practices like planned obsolescence instead of wasting their time attacking energy sources that are actually useful and increase quality of life for everyone.

>>2748186
Yeah, he basically said it was a crap standard and gave plenty of examples why.
$150 for apple’s USB-C cable with built-in DRM chips?
It’s not something I’ll be adopting in the next 20 or 30 years.
If they were smart, they’d be like the crack dealer, and subsidize/giveaway the stuff to establish it, and then start jacking up the price.
Although they probably accounted for zoomers who will be tripping over themselves to get the latest “new” “tech”

>>2748241
>$150 for apple’s USB-C cable with built-in DRM chips?
I know for a fact that fleabay sells USB-C dongles for $3-5, so it can't be that hard to bypass

What's the most practical and common size of SMD resistor to design around?
going to be applying the solder paste without a stencil and heating with hot air.

>>2748208
>Balanced: signals will add to 0V (or a DC voltage)
>Differential: signals are being subtracted from one another

Differential signalling is a natural byproduct of some types of circuits. For example, measuring a current sense resistor or a wheatstone bridge will be a differential measurement because it goes into a differential amplifier/ADC, but it usually won't be fully balanced. For it to be fully balanced, the signals would be symmetric about some DC voltage.

Balanced signalling however is a technique used specifically for noise immunity, even from source circuits that are fundamentally single-ended in nature. For example, a common condenser microphone has a grounded chassis mesh and a floating membrane that changes its voltage with respect to the chassis, atop a bias voltage. This will be fed into a preamplifier and converted to a balanced signal before being sent down an XLR cable, maybe to be converted back to a single-ended signal on the other end. Though technically you could create an intrinsically differential and balanced condenser microphone capsule, where the membrane is at exactly 0V, while the chassis mesh on either side are held at approximately ±48V. Or have a membrane on either side of a grounded mesh.

>>2748239
You mean the guys ensuring all light bulbs had the same lifespan? That was almost entirely because all manufacturers had an intrinsic tradeoff between lifespan, and brightness/efficiency. They had already found the perfect trade-off point for common lighting, and decided they'd enforce that on each other rather than allow the free market to come to that conclusion. But it's not like there weren't bulbs with different characteristics, like the long-life bulbs intended for hard to reach fixtures. And they often gave these bulbs away for free, so it was hardly for direct profit. They were in cahoots with the power companies and wanted to reduce the high electrical load of people having lots of dim/inefficient light bulbs.

>>2748290
0805 if you have bad eyesight, 0603 for most people, 0402 if you really the density. 0402 is getting kinda small for a soldering iron, and while it's doable with a nice ~1mm screwdriver tip I'd prefer using a reflow plate.

>>2748308
i sometimes hand solder 0402 but i lose about 3-4 of them per one soldered. mostly because i have crappy tweezers

>>2748305
> guys ensuring all light bulbs had the same lifespan
No, a maximum lifespan, in addition to fraud, racketeering, profiteering, blackmail, extortion, violating the sherman act, patent trolling, etc, etc, ad nauseam.
> gave these bulbs away for free
Never happened to any significant degree. In fact, cartel members were fined for giving discounts, selling too many bulbs, or selling special bulbs, including fluorescents.
> worried about dim/inefficient light bulbs.
But thats not what happened. Some bulbs were already naturally too efficient (by more that two times) so they made the filaments thinner, or decreased the vacuum or argon gas so they burn out faster.
Some companies were running them at 230 volts instead of 240 to extend their life, so they made an even worse one that would get 900 hours to compensate (at more expense) but gave that up because of risk of running into the BSI minimum spec of 1000 hours.

That “dim” bulb/efficiency argument proposed by british regulators was bounced out decades ago by actual scientists, physicists, engineers and the Phœbus cartel themselves.

To this day that same cartel is still operating, the only thing they learned was to keep a lower profile and use secure comms.

>>2747996
It shouldn't be an issue, unless the power supply is getting really hot or something.

>>2745318
>>2745306
there can only be one

Rectumfrying 120v ac to 90vdc to power a motor. All is well. Setup mosfet for eventual pwm control, use bench ps to drive mosfet. If I tie 0vdc of bench ps and rectumfrier together the gfci trips the second I plug ac in. Is it just so dirty its trippin or what?

>>2748373
rectifying 120VAC (rms) gives 120VDC (rms), if you don't have a filter cap
if you do have a filter cap, it's 170VDC.

Anyhow, if you're using a bridge rectifier, you've got this problem, pic related. The negative output has a non-zero voltage with respect to neutral, and hence non-zero with respect to ground. By connecting ground and this negative output together, you create a short-circuit for half the mains cycle. I recommend not treating this negative output as if it is a safe 0V, but treat it like a high voltage negative output.

You could use a single-diode rectifier, though then your average voltage would be even lower. Maybe that's what you want though. That would mean you'd be connecting neutral and ground together, though I'd avoid that (at least not without a ~>10Ω resistor) since it will probably trip your RCD.

>>2748376

fuck that's an ugly/unclear drawing.
suggest 10 cent pencil + paper.
10x faster.
10x more aesthetically pleasing.
costs $1500 less.

>>2748376
What bench power supply has negative tied to ground?
The chinese ones have plastic cases and no ground lead at all.
> all my bench regulated supplies are mercury arc rectified, tube regulated with live chassiæ.
Ok then.

>>2748391
In my experience chinese ones have a ground lead that’s floating around connected to nothing. It’s often the case in chinese electronics. I don’t know why they feel the need to include a ground lead but not actually connect it to anything. I’ve seen several led lights with metal casings that have grounds that just float around. Some were really high wattage too (though like 40-50% claimed wattage, as is custom for chinese lying nonsense)

>>2748391
>What bench power supply has negative tied to ground?
Any non-Chinese one. Chinese ones rebranded and UL/CE listed might, I'll check the ones we sell at work tomorrow if I remember. If they don't ground-reference part of the power supply's low-voltage end, it's quite easy to get a capacitively coupling buzz off the exposed metal output, which could be kinda scary to an electronics noob. Though that depends on the PSU design.

Considering it's his GFCI that's tripping, there's got to be a ground connection somewhere in his circuit. Unless he's really stupid it's not going to be on the bridge rectifier, and the only thing less is the benchtop power supply.

>>2748404
Yeah saw that a lot on Big Clive's channel too.

>>2748391
>What bench power supply has negative tied to ground?

commonly they have a separate 3rd terminal for ground.
this lets the user choose whether to ground the negative (for safety) or leave it floating (to prevent shorts on live circuits.)

>>2748404
ground is bad in lights. they get wet inside and it trips the breaker.
>>2748414
Ground is good in chink crap power supply.

>>2748421
This looks like an AI generated image.

What electronics knowledge do I need if I want to try making the Ben Eater 6502 computer?

>>2748423
Have you ever put a jigsaw puzzle together?

>>2748424
Yes, but not one where I might have to debug the puzzle pieces

>>2748423
>What electronics knowledge

not really about knowledge as all the steps should be well-documented.
the most basic requirement is infallibility.
but noobs make a mountain of mistakes, nothing ever works, and they have zero skillz to catch them.
- put some chips in backwards
- forget to wire the power rails coz not in schematic
- miscount the pin numbers
- use wrong capacitor/resistor values
- reverse polarized components
- plug in chips with a pin leg bent underneath
- cant recognize when shitty breadboards have shorts or loose springs
- use under-powered supplies, unregulated supplies, or mix up plus and minus
- forget to strip some wires
- 1000 other things

if it's your first time, it's important to start with a single chip, like a 555, and progress from there.
should take less than 10 years to reach Ben Eater levels.

Is there an easy way to remove the silkscreen from components in kicad?

>>2748505
edit their individual footprints by right clicking on them
if you have a lot of the same components that you want the silkscreen gone from, you can save a custom footprint without a silkscreen in a project/global library, and then apply it to all the components

>>2748442
there's no fucking way it would take ten years, that's absurd. Maybe like 2 years. Everything you listed are pretty easy fuck-ups to get over.

You're working off of someone else's template, so if you're focusing on results, it's super easy to do. easier than say, focusing on developing your own understanding to where you can problem solve the same solutions on your own.

>>2748702
>>2748442
I'm a mechanic, so I know 12 volt systems reasonably well and know how to use a scope and multimeter, but troubleshooting or designing anything more than a simple electromechanical circuit is above my pay grade. I figure there's a reasonably large knowledge gap between what I can do and what is required to fuck around with microprocessors, I'm just after a good way to bridge it - either a book or a series of simpler projects, or both - so any recommendations would be appreciated.

>>2748376
Thank you. And your drawing is perfect. Fuck that other anon.

It is indeed a shitty chinese ps. I will hopefully report back with good news.

I have some pin headers on a PCB that I use for controlling servos, but usually have nothing plugged into them.
How dangerous is it to leave them exposed? They're connected to a 2s LiPo, and the positive and ground pins are right next to each other, so the possibility of getting shocked while fucking with the various connectors near them seems pretty concerning to me.
Should I cover them up with some sort of plastic piece when I'm not using them?

>>2748352
Meh, its customers who are dumb this time, buying more watts, not lumen/watt.

>>2748763
>shocked
If it's low voltage that's not a concern.
Covering them may not be a bad idea just to reduce the chance of accidentally shorting, but that's about it.

>>2748763
2S is 8.4V maximum. You won’t feel anything below 50V unless you stab the wires through your skin.

>>2748089
I got them pre price gouge from someone on falconfly for 20 bucks too haha
Should've bought two in hindsight.

I have an old VHS camera that a buddy of mine handed me to try out and see if I could get working. I found an appropriate plug and the camera itself works, and I found the exact battery charger and the charger seems to work, but it seems the battery is not working. It's an EP096FS battery, which is either hard to come by or way too expensive to buy online. For the record this whole camera, plug, and charger was only 12 bucks. Three questions:

>1
I thought it was just the battery contacts not working but after using some copper wires to test it wouldn't charge then, either.
>2
is it possible to repair the battery as-is?
>3
Is it possible for me to make a reproduction battery? I have a 3d printer so I can make a housing for presumably some lithium battery I get online. Can I just make sure the volts and such are the same and make a contact housing and make my own? Or is there more to it that I'm nor privy to?

>>2749011
>is it possible to repair the battery as-is?
It's not worth it.
>Is it possible for me to make a reproduction battery?
Yes, but you'll need extra circuitry to keep the battery from discharging too low while it's being used. You'll also need a new charger and BMS board depending on the configuration/number of cells in your DIY battery.

>>2749018
>Yes, but you'll need extra circuitry to keep the battery from discharging too low while it's being used. You'll also need a new charger and BMS board depending on the configuration/number of cells in your DIY battery.
Sounds doable. I'll try it. Thanks.

>>2749021
Keep in mind that when you connect batteries in series the voltage adds up. So if your camcorder requires 10V and you have two batteries in series (full charge=8.4V), you need to boost the voltage to 10V with a boost converter. Leave room for it in your 3D printed battery casing.

>>2749022
...and if you have 3 batteries in series the full charge would be 12.6V which would need a buck converter to reduce it to 10V.

>>2749023
unless there is a voltage regulator inside the camcorder.

>>2749022
>>2749023
>>2749024
thank you for being thorough, have a meme as payment

>>2749028
Thanks, Anon. Good luck.

>>2749011
>I have a 3d printer

if you can make a null battery like the pic, then you can use a couple of lipo power banks in series to generate 10V.
this would be a solution involving minimal work, which is my favorite type of solution.

>>2749022
But it won’t take just a fixed voltage, because it originally ran in batteries in the first place. It will have some allowable voltage input range through its battery terminals, it should be pretty easy to just use the same chemistry it originally had and not have to worry about any of that. Plus the camera’s internal battery meter will be correct. You may still need an overdischarge protection circuit in the battery though. That said, if the original cells were NiCd or otherwise some chemistry you don’t want to deal with, going for lithium with a voltage converter might be required to remain within the intended voltage range.

>>2744526
Fuck EMI
Finally fixed it, later of kapton, layer of copper, chopped up shielded cat6e cable, added more copper, added even more copper to encompass the fucking SPI wires.

Protip: don't operate SPI on wires/prototype above 4Mhz with a GPS, you will lose your fucking mind.

Apologies in advance, complete retarded beginner.
Any chance of getting a thermal sensor ripped from a flir security camera to work separately from the housing unit circuit board?
I have the sensor itself and the image processor board (I assume) bolted on the back of the sensor house itself with a white flexcable coming off of it that was hooked up to the big ass circuit board inside the camera housing but I dont want the IP video or AI functionality, just "basic" sensor image is enough.
The camera was tossed away (a fucking 6k surveillance camera, FC-344ID) for scraps. I thought I could salvage the sensor for a project. Could I check the sensor via the white flexcable somehow? Unsure of what to connect it with.
I started with some basic electronics projects to learn the various components and it's a lot of fun so far but I am getting very itchy over the sensor. I really wanna figure it out.
I can post pictures of the board and modules a bit later if that helps.

>>2749577
You may be able to find a datasheet, but it’s possible they’re kept secret because muh US foreign policy. If there’s no datasheet but the camera still works to some degree, then you should be able to snoop the comms lines from the sensor chip using a logic analyser while it’s working, then reverse engineer the pinout and protocol from there. With any luck it’s a standard pinout. If it doesn’t work, then it will be pretty tough to reverse engineer it. Maybe trace the board and try to repair it, try to read off the MCU firmware, that sort of thing, but you’d likely be better off selling it to someone who specialises in this. Or to the CCP.

Anons, I want to set up a basic electronics lab, the main items being an oscilloscope, digital multimeter, and power supply. What's the kind of minimum budget I should be looking at for each item that will get the job done without being hot garbage?
Alternatively, are there any particular brand or make of scopes/PSUs on the used market I should keep an eye out for?

>>2749811
>set up a basic electronics lab

https://www.youtube.com/results?search_query=electronics+lab+setup

one giant Aussie faggit even did a 37-part series on the subject.
enough to keep you busy watching for a week non-stop.

>>2749811
There's tons of garbage advice about it so here's speaking from experience and assuming you don't have any specific project in mind: you can get a $50 scope and it'll do exactly what it needs to do until you get into complicated high-frequency applications. It won't have dual+ inputs but will be able to do hold-and-add, it'll work at relatively high frequency and in a good voltage range, etc. That'll last you forever if you will be doing lots of more standard projects. Yet, the only advice I've ever seen on the subject was 1k+ scope or bust.
Not that I disagree with all advice, don't get a usb scope for example.

If you're going to deal with ICs a lot, you should probably get a logic analyzer as well. Again you can get one on about $50, no need for fancy $30k+ ones. Same idea with programmers.

For PSUs I do recommend spending a bit more. Maybe in the $200 range if you can. But you might not need a proper PSU in the first place. You can get shitty chink power modules for pennies and 9/10 of them in the strip you receive will work fine, assuming you don't have noise constraints and need 5v + 3.3v out and nothing else.

For multimeter I find that basically everything works since it's basically only useful for a quick check here and there, and you'll be using the scope and logic analyzer for anything serious. Continuity/diode test, current with the right range for your application and voltage is all you need and pretty much every multimeter has those.

For bands, just get cheap chink shit on amazon for everything. Even the used items are way overpriced most of the time. You can google around for chink brands that are less bad.

>>2749849
>For bands,

what?

>>2749871
brands, duh

>>2749811
> I’ll into electronics once my toolbox is complete
This is part of the toolbox fallacy.

Anyway, you can make your own power supply.
I suggest a linear one.

You will probably need a meter, they’re all roughly the same. I prefer LED bench meters. You can get 2 decimal chinese meters for a few dollars but they just measure DC volts (good for your DIY psu with an LM317). Most people tout hand-held field service/linesmans meters but you don’t need that shit—get a benchtop.
Don’t get one with a VFD—those are irreplacable garbage.

I got a used mint CRT Tek scope from the 80s, but you can make due without a scope for audio-frequency projects since you can use your built-in scope which detects signals from 20 Hz to 20 kHz.

Two more good projects is a signal generator and frequency counter.

You can get a logic analyzer together with a rasberry pi zero or pico.
All of my old logic analyzers are useless on modern high speed stuff, so I usually have to build something with the matching chipsets and/or front-end drivers anyway (e.g. USB 3.0).
In the old days, logic probes and analyzers used 74LS logic since you were probing 74LS logic boards. Those days are gone.

When using multiple lm317 (for higher current, load and thermal spreading), for a single load, single power supply setup, should I use no diodes (left circuit) or use them (right circuit)

Anyone got pinout for rts5803?

>>2749981
I'd bet dollars to donuts that Realtek has it.

>>2749980
Could't hurt.
Also I’ve seen ppl put reverse protection diodes across the in/out leads of the regulator for good measure.

>>2749830
>one giant Aussie faggit even did a 37-part series on the subject.
Channel? I'm in Australia as well, so it might actually have some relevant info. Our secondhand market isn't as flush with the kind of used electronics gear that burgerstan or yurop is.

>>2749849
I want to do some home automation with microcontrollers and play around with wireless communication, see where it takes me, but this is all good advice, thanks.

>>2749950
>you can make your own power supply
Would be a good starting project I guess, I assume there's some good resources in the OP to follow for project builds?
>You can get a logic analyzer together with a rasberry pi zero or pico.
I've got a Pi 4B gathering dust, that'll probably be fine for now.
>toolbox fallacy
Another good point, hence why I asked - it's good to know I can get started with this hobby on a fairly tight budget and minimal equipment. Bootstrapping the "lab" will probably be a good learning experience too.

I have a pair of these speakers I've been using forever. Often one or both speakers will go silent, and if I repeatedly press the power switch on the back eventually the speakers will both come back on. How would I figure out what's wrong with these speakers? I don't have much experience with audio electronics. I've opened them up and tested the the obvious things for continuity. Since repeatedly toggling the power always works does that stand out to someone as something I should check?

Literally any diagnostic advice is helpful because I'm well out of my depth but really want to repair and continue using these speakers.

>>2750111
The three things I'd check first are the DC power input jack, the on/off switch/potentiometer, and all the electrolytic capacitors. Examine all solder joints and reflow anything that looks bad.

>>2750132
The wall wart might be a problem too, so try another power supply just to eliminate it as a possibility early on.

I accidentally built a spark gap receiver, and I couldn't be more confused.
It's not connected to the kart in any way shape or form. This would be a cool feature, but the reported RPM isn't anywhere near accurate.

I need to replace these LEDs from porch lights. They are 2.5 in diameter ( 60mm maybe ) I can't tell are these supposed to be 12.5 W, or 5W ?

I'm seeing many options, but they all seem to be 12v. This doesn't appear to have a driver or anything - it's just 120 wired straight to this flat disk.

>>2749811
For a PSU, there are three types that you might want:
>fixed voltage 12V, 5V, 3.3V, etc. for powering digital/power projects
For this I just buy barrel jacks and plug them into thrift store switching wall adapters. Salvaged ATX PSUs are also good for this. Fault prevention isn’t a necessity, though a fuse or latch circuit isn’t a terrible idea. Same for a current readout.
>variable current and voltage 0-20+V for testing loads
Because noise isn’t an issue when you’re powering up a motor or an LED, I just use a powerful wall adapter with a CC/CV DC-DC converter. A voltage and current readout is a necessity. You can make this for under $40. A capacitive multiplier on the output for noise reduction would be nice, but likely isn’t needed.
>split-rail low-noise supply for audio projects
I just use a centre-tapped transformer to make +/- 12VAC, followed by rectifiers and caps and 12V regulators to make it +/-12VDC. This is also pretty cheap, the most expensive part is the transformer, which you might be able to get as salvage. Doesn’t need to be adjustable, though having multiple taps for different voltages could be nice. Voltage and current readouts are optional.

You can merge all these power supplies into one expensive split-rail, adjustable, current limited, low-noise, high-power unit, but I don’t really see a point. Build your own relay-autoranging transformer linear power supply, or switching with linear postregulator power supply, if you want.

>>2749980
Diodes aren’t necessary. It’s a constant-current circuit, you don’t need to worry about parallel outputs fighting.

>>2750097
> .au, power supply
I was just on sound-au.com, there is some resources there, too. And here, like the anon that just posted >>2749980, and I was reading on sound-au that they use LM317s in audio because they are lower noise. I did not know that.

>>2750213
I’ve spent maybe $10 on all my power supplies over the years. Maybe.

I also like to have an AC output on them sometimes… for example, I have a US to UK transformer-based(!) up/down converter that I can put on it and get 24 VAC from the original 12 VAC, or 6 VAC the other way around.

any random shit worth buying for a newbie in digikey? i need something to fill my cart to get to 50 bucks or ill get charged 18 bucks...

>>2750464
Spool of Kester 44 in your preferred gauge.

>>2750465
Seconding this, bad quality tin will drive you mad more than any component especially if you solder often

>>2750176
Just buy cheapest 5W or 12W or whatever amount of light you need lightbulb and shove it in.
Maybe disassemble it and shove it in.
>driver
it is built in the PCB with LEDs. That black IC probably regulates current, and LEDs are connected in series to match mains voltage.

You can buy driverless LED PCB from aliexpress, but for some reason they are all for 220V, not 110V.
Maybe in COB form, idk.

>>2744526
What parts are salvageable (if there are any) from a lcd tv broken display?
I have a couple of it in the garage, and want to take to the thrash, but I don't know if the display in itself has anything interesting to sell or to do diy things.
I mean, aside of the power source and the logic board, I see something holding from below the display, once I heard this is called T-board (or like these), this is the display driver?
There is some flat cable, this can be unplugged?
And the IPS or whatever is a bunch of sandwiched layers, but I don't know if it can work to anything.
And the whole plastic case, takes too much space. That's why I want to throw it off.
Thanks in advance.

I got a pair of pic related because I want to cut the earbuds off and just have one earbud wired to aux. That and these have a horrible built-in volume slider I don't wanna touch on accident in bed. I thought the wiring was just 4 wires from aux being split giving each earbud 2 wires. I cut the aux and earbuds off, stripped the wires, and tried both ways to connect the wires but no sound. I made sure music was playing and that windows didn't automatically mute my audio or anything of that sort. What went wrong? Did something change with how cheapbuds are wired?

>>2750605
It might be the enamel coating on the wire.

>>2750578
Here’s what I salvage.
- the remote, and IR module
- the speakers, some are surprisingly good
- the plastic TV case, if this has large flat areas you can cut it down to make face panels or project cases. It’s ABS usually.
- the power supplies.
- the wire goes into the jumper wire box
- the stand is used for other monitors, or to clap work lights on to
- the power cord (obvi)
- the plastic layers of the screen. Used to protect circuit boards in metal cases
- sometimes there is a thick clear plexiglas piece in the screen (monitors usually)
- the LED backlight strips. (again, obvi).
Sometimes you can (mostly) make an LED driver from the PSU parts but it’s computer controlled.

Plasma TVs had all kinds of good shit in there. The older the better… applies to LCD too.

>>2750605
No, it’s never been like that since the 1950’s.
Stereo needs only three conductors, they share a common ground.

>>2750097
> Pi 4B gathering dust
Probably a good option, or even an rp2040, its GPIO has these independent state machines on the pins. This is not unique, but still a brilliant addition.

There is a common misconception that, if you get an embedded PC with GPIO, and the CPU runs at 5GHz, you’ll get something like 5GHz GPIO.
Many of them actually just use a Super I/O chip to do the GPIO (like NCT6126D) and use LP bus or eSPI to communicate with the CPU at only 33 MHz or 48 MHz respectively. Which is crap.

If I were an internet personality I’d probably do some speed tests.

Another thing you can do is use the A/D converters to make a simple oscilloscope. Just be careful with the input pins. A low freq (like maybe 44 or 128 kHz) oscilloscope can be made with audio inputs) which I used to do back even in the 90s.

One thing I never figured out how do do was make a waveform capable (i.e full sine waves) frequency divider. I can only do it in software. I was toying with some ideas about stretching out the pwm signal on these newfangled class D amplifiers I’ve been seeing recently, but that’s not fully formed.

>>2750610
Thanks, it's interesting, I didn't think in all the stuff you write.

>>2750605
>one earbud wired to aux

not at all clear what "aux" you're talking about.
if you mean the aux (blue) on the back of your computer, that's an input, not an output.
and, even if you use the green "speaker" output, you may not be able to hear it if the headphones are low efficiency, or the volume in less than max.
coz the speaker output is meant to drive an amplified speaker.

I’m trying to build a hex schmitt trigger inverter oscilator to use in a counter but it’s not working.
The LED is just always on when it’s supposed to be turning on and off by itself.
Blue thing is a 1uF capacitor, resistor is 10kohm.

>>2750705
Here’s what I’m attempting to replicate btw.

Need help with something very specific.

I have a seedling mat (just a big patch resistor) I use to keep my chicks warm.
Plugged directly into the wall, it can heat the air below it to over 50°C if left to run.
To rectify that I attached it to a digital thermostat set to 31-35C.

That works but during the day, but I noticed at night the mat lacks the power to reach 30C. Holds like 29.7C... Even though it could before I wired the thermostat...

I think the thermostat switch can't provide level of current the heater needs.
I'm thinking of building my own thermostat that CAN provide that but I need advice. I've never made anything running on mains power.

The most ghetto solution I can think of is an Arduino with a thermostat and some kind of mechanism to flip the switch on the heater from the extension cord (hence, the heater current doesn't have to go through it) but there's definitely a better solution than this

>>2750705
led is loading the inverter output. you should put a transistor between the output of the inverter and the led. something similar to this.

>>2750717
Thanks!

>>2750714
>That works but during the day, but I noticed at night the mat lacks the power to reach 30C. Holds like 29.7C... Even though it could before I wired the thermostat...
>I think the thermostat switch can't provide level of current the heater needs.
If somehow your thermostat is not providing enough current, use the thermostat to control a relay instead. You already have a thermostat, no reason to reinvent the wheel with a microcontroller and temp sensor. I've gone the microcontroller route before and life is too short to bother with that shit when an existing thermostat + relay can do the job.

>>2750717
>>2750705
Yeah, I get a signal through it now but it's apparently still too noisy.
I connect it to a 4 bit binary counter and it only counts in increments of 2 (Still have to press twice so I'm assuming that's the reason).
Dont know what the hell is wrong with it

>>2750705
>10kΩ
>1μF
You made a ~100Hz oscillator, no wonder you can’t see the LED transitioning. Try 100k and 10μ.

>>2750717
Or just use another of the inverters as a buffer. With an appropriate current limiting resistor to suit the specs of the LED and of the inverter IC, of course.

>>2750757
Don’t forget bypass capacitors on your ICs, and a big enough cap to buffer your power supply.

>>2750767
I had misread what I had to do; I don’t need it to oscillate at a visible frequency but I do need to stabilize it for the Schmitt inverter and then the counter (Trying to make a FSM).
Can you see what is wrong with it?
The part obscured by the red LED is a diode.

>>2750773
Im watching https://www.youtube.com/watch?v=tmjuLtiAsc0
but I don't quite understand the neccesary relationship between the resistor and capacitor.
I have 2 1uF capacitors and 8 10kohm resistors

>>2750778
If you're lazy you can use CircuitLab to draw the graph online and just screenshot the result. Don't just draw shit in paint like that.

>>2750796
https://youtu.be/e1-kc04jSE4

>>2750773
why the cap?

Getting started at reading industrial wiring diagrams, i can see whats connected on one wire and how works but does anyone have advice on how to look at a wiring diagram and paint yourself a picture of how the whole circuit works? On paper it’s not linear. Where do you guys like to begin when looking at drawings?

Is this any good?

>>2750908
I'll give you $20 for it

>>2750847
Wiring diagrams are for safety and troubleshooting, not for teaching theory of operation.

I want to build a small chargable lamp with LED strips. Therefore, I am looking for a powerbank/powersource that charges with USB-C and has and outputs through a barrel connector 5,5/2,1. It would be great if the on/off switch of the power source was flat on the bottom, so I can reach it through a hole in the bottom of the lamp case, on which the power cource should rest. Is there something like this on the market?

>>2750962
>outputs through a barrel connector 5,5/2,1

very very small chance you'll find that.
you should make or buy an adapter.

>>2750840
Trying to follow pic related but I only have 1kohm resistors, LEDs with unknown internal resistors and 1uF capacitors.
Ive been told that an LED as R4, A 10kohm resistor as R5 and a 1uF capacitor as C1 should work but for some reason it doesnt.

>>2750994
>for some reason it doesnt

- (red) move the wire north to other side of switch to make sure it's getting 5V
- (orange) you keep using pins 1 and 2. possibly you destroyed this gate so try another one. FIY, CMOS gates are way less likely to die than TTL.
- (green) why is there a cap in series with this wire. makes no sense unless you're half-way thru making a differentiator.
- (yellow) please use current limiting resistors in series with LEDs. even if you think they have internal resistors. it hurts our souls to see that.

>>2751002
Oh shit I didnt even realize I put a capacitor on the fucking counter. I didnt sleep yesterday at all, in my defense.
But I have now tried changing the gate; Same shit.
I'm beginning to think there might be some kind of thing ive done to my counter (74HC161N) because the output works fine for a d flipflop

>>2750978
Thank you. Yes that is a reasonable workaround, since it will be hidden, anyways.

>>2750994
An LED will not work well as a pull-up resistor, as its effective resistance will be far higher when the voltage across it is lower than the forward voltage. Say it's a white/blue LED, they have a forward voltage of ~3V, and let's say you've got a 5V Vcc. The cap will charge up to 2V (5V - 3V) fairly quickly, but it will really take its time getting above that. As far as I see, you have two options that don't include just going outside and buying/salvaging different values of resistor:
>remove R5, R4=1k, C1=1µ: should work fine though your switch may not last that long
>use two 1k resistors in parallel for R5, and two 1k resistors in series for R4, and C1=1µ: should work flawlessly so long as you get the high-side voltage above the maximum input threshold value for the schmitt inverter
That threshold is ~3.5V for a 74HC14 at 5V, using two resistors for R4/R5 each will get you to 4V.

>>2750962
I assume you want a powerbank that can put out while charging as well… not all of them do that.

>>2750978
I’d have just opened the case and soldered a some wires leading to a jack for the output, and let the wires leave thru a hole, or, if room, put the jack on the case.

>>2744526
Fried my laptop, I was getting a 10% battery although it was plugged. I pulled out the cord to re-insert and that was it. I opened it up to see R363 and r688 fried. I replaced them only to have the give off the white smoke again. Not sure how this is possible, DCIN_PWR20_F is the psu output (powered on from a different adapter after resistors were replaced.

>>2751291
Double check the diodes D78/D87, it's a long shot but if one is dead then it could be current flowing in a direction it shouldn't, and way too much of it. More likely is that C450 died short, MLCCs can definitely fail from thermal cycling. If it's neither of these, then I can only conclude that the PVCC pin on that IC is sinking current from a fatal fault.

>>2751291
Maybe that cap is shorted?
I’m not sure what the typical failure mode for caps is, electrolytic vs ceramic smd ml caps.

>>2751291
Did you remove the battery?

thank you so much!
>cap?
ya its shorted actually.. gonna replace, gotta look for 1 brb (after this i will check diodes)
>battery out
yea, just the motherboard on the bench with PSU

>>2751301
...white smoked the resistors, new cap now shorted

>>2751301
new resistors.. smoked
new cap now shorted
the diodes are fine and both testing/showing similar results in diode mode. as for the power in, the connector looks fine. maybe it is the PMIC (BQ24742) somehow

>>2751312
You replaced it with a 25V+ capacitor? I see no reason it would die, unless it's being killed by a voltage spike. In that case, replacing it with a higher voltage cap might actually be a bad thing, as the current cap could be acting like a spark-gap or MOV.

Do you have a scope to measure the voltage transient when it's plugged in?

>>2751358
I am not sure, it could easily be a 50v, if that is the case i will pull one from something i can confirm the spec
>scope
yes i do, i will check in a little bit and most likely I will pull the PMIC off and see if they burn out again in a little later, if not i will assume it is the IC

>>2751358
I am not sure, it could easily be a 50v, if that is the case i will pull one from something i can confirm the spec
>scope
yes i do, i will check in a little bit and most likely I will pull the PMIC off and see if they burn out again, if not i will assume it is the IC

>>2751367
I wouldn't pull off the PMIC before figuring out if you can buy a new one.

>>2751312
>new cap now shorted

very likely the cap is wired across the big chip at the bottom.
so i'm predicting you'll see a short across the pads even AFTER you REMOVE the cap.

>Order T12-958 soldering station from AliExpress this morning
>Now evening, soldering with 20eur soldering station
>It makes a worrying buzzing sound and stops heating
>If only you had lasted a month longer
Heating element resistance measures 1.2k, should be 2 to 3 ohms according to internet.

>>2751571
Get another element for it. It will be worth it so you can use it to solder a DC socket to the back of the T12 station, and fix it if it ever breaks.

>>2751667
I found a replacement iron it for 7e (PCB revealed that it's a ZD-99) so I'll fix it up.

Can I trust an insulating thermal compound to actually prevent my transistors shorting against my heat-sink? In particular, a nano-diamond-based compound. What about when involving extreme vibrations? What are insulating thermal pads like in comparison?

>>2751778
use thermal rubber, or mica or ceramic insulators.

>>2751778
there are little mica sheets and plastic bearings you can use to isolate your trannies from the heatsink

>>2751571
with chinks you buy twice

>>2751785
>>2751837
So paste is a bad idea? Each TO-220 FET will be dissipating like 15-30W. I see values like:
>silicone: 1.25 C/W
>phase-change pad: 1.15 C/W
That would mean 20-35°C between the transistor and the heat-sink, I'd rather have 10°C. It's hard to get good values for some other materials, as they don't state the C/W value but rather the W/(m*K), and thickness isn't stated often. I found one place where it quoted 0.05mm thickness, but that sounds extremely thin. Also not sure how to compare it to thermal paste, since the thickness of it depends on how much is applied.

Is there risk of the corners of the transistor pressing into the silicone and shorting against the heat-sink?

>>2751870
Down to 0.9C/W for Henkel's HI-FLOW 300P, at 0.001" thickness under 200psi pressure, which isn't much better than the values I got from Aavid Thermalloy. Maybe I'll try to use a really damn thin layer, after lapping my heat-sink and transistors to a mirror shine.

>>2751870
>So paste is a bad idea?
paste is not insulator.

>>2751870
why not just isolate the heatsink

I need help fixing an 80s stereo receiver with output issues.
No pots or switches to clean, all microwitches.

>start receiver in AUX, no output
>start receiver in Tuner, one channel will work, change volume or input both channels die

>fiddle enough, sometimes AUX input comes through at ONLY 34db volume (drops out when changing up or down)
>when this happens, change to any other input and the screen shows you changed input but the actual AUX audio still comes out of speaker.

No capacitors look bulged, I cleaned all the in/out plugs and the ribbon cable connectors. The AB selector is analog, but I pulled them and cleaned them and it didnt work. When you do have output, fiddling with the AB never caused scratchy noises or dropouts so its not that.

This receiver is a garage receiver, heavily used almost daily the last decade. I unplugged it from the wall, didnt go in my garage for almost a year, plugged it in and it was acting funny for a day. Worked fine for the last month, swapped in some new speakers and was moving it around and now this is where we are.

This is the service manual for it, but im not versed well enough in electronics to really dig deep into it.
https://archive.org/details/manual_KRA57R_SM_KENWOOD_EN

>>2751917
Not feasible, I have 6 MOSFETs that make up a BLDC driver. The three high-side FET drain tabs are all tied to the positive rail, but the three low-side FET drain tabs make up the three output nodes to the motor. I briefly looked into using heat pipes as bus-bars and some sort of higher surface area electrically insulating interface from the heat-pipes to a heat-sink, but dealing with heat-pipes sounds like a massive pain.

>>2751939
I assume you don't have an oscilloscope? I highly advise you either get one, or make something that will function as one.
A high-input-impedance amplifier with input protection feeding a little speaker would work. Battery powered with an op-amp IC and some probes should make it convenient enough. Could even make it feed a stereo headphone output with a different gain going to each ear (with clipping) so you can listen for a greater range of signal amplitudes.
You could also use such an amplifier to feed a sound card, though I don't think it will be as convenient.

You'll want to be able to inject a signal to all parts of the unit to test, aux is pretty easy to just hook up a phone/laptop with a sine wave generator app, for FM you'll probably want to get an FM transmitter. AM and other inputs might need a bit more work, but start with those two at least.

It's also possible that the issue is digital. Old EPROM got too many cosmic rays, modern electronics are injecting noise on the comm lines, etc. This could maybe cause a digitally controlled amplifier circuit to cut out your aux signal when you try to adjust the volume. To troubleshoot a digital problem you'll want a logic analyser, they're like $5-20 online and quite useful. Though before that I'd look at the keyboard matrix with a multimeter to ensure that's not sending dodgy signals.

>>2751949
Vishay Thermawick thermal jumpers are pretty good at heat transfer especially if you are using multiple of them. I used them with a fully isolated heatsink connected to the high side of a forward converter's primary switch. I added 4 1206 sized one parallel between the tab of a TO263-7 and the heatsink and the tempereture difference between the FET and the heatsink was about 5°C

>>2751949
why is it infeasible to isolate 6 aluminum heatsinks

>>2751984
Having the heat go from FET to PCB to thermal jumper to PCB to heat-sink is never going to be as good as just using a thermal pad.

>>2751994
Because they're going to be all clamped to one heat sink. Mechanically speaking, making them seperate would weaken the chassis a lot and require that they're held in place solidly somehow.

sup /ohm/, retard CS major here.
I want to make a jogwheel for DJ shit, can I just use one of these high-ish precision magnetic encoder chips
>https://www.amazon.com/Magnetic-Encoder-Induction-Measurement-Precision/dp/B094F8H591
and a neodymium puck?

if so, why would anyone use the (cheaper, sure, but not by that much) physical or light-gate encoders that have 4bit precision or worse?

>>2751949
>Not feasible, I have 6 MOSFETs that make up a BLDC driver.
Chinks use kapton tape + insulator bushings
Also, TO220, BLDC driver... You wont be generating much heat if designed shit right. So you can use any silicon, idk, mica, kapton, whatever, insulator. And silicon rubber has to be thick, so you can't use idk, soft rubber like for NVME drives.

But 'silicone TO220 gasket' is a good keyword for aliexpress or whatever.
Also, consider using insulated FETs?

>>2751870
>>2752308
>Each TO-220 FET will be dissipating like 15-30W.

That's a bad design. Stop now. Rethink your life choice. Total losses should be less than idk, 2-3W. It should run fine without heatsink

>>2752286
They’re less stable with regards to external magnetic fields. Can’t imagine they’d be more accurate than a high-density encoder ring. Double check the price of those ICs on DigiKey or LCSC to check if they’re likely to be fakes.

Also the chad method is to use a brushless sensored motor, that way you can add force feedback.

>>2752308
I considered kapton tape, I’ll probably do some testing with a bunch of setups.
>You wont be generating much heat if designed shit right
I’ll be pushing past 1kW, and possibly only at 24V. I’ll find out how significant the thermals are after I build my test-rig, and see if I really need 10A of gate drive.

>>2752309
I bought 200A MOSFETs for a reason. 2.4mΩ. Surface-mount FETs don’t come close to good old 220s in ease of thermal management, though the ones with thermal pads on top come close. Really don’t want to rewind my motors for a higher voltage, but it’s technically still an option.

>It should run fine without heatsink
>a package specifically designed to be bolted to a heatsink
Brainrot take

>>2752119
why is it infeasible to isolate one aluminum heatsink

>>2752313
>I’ll be pushing past 1kW
With 1 kW it might get a bit toasty. Chinese usually push only around 600-800W with 6 FETs, and they are usually 80-160A FETs.
>I’ll find out how significant the thermals are after I build my test-rig, and see if I really need 10A of gate drive.
What are you gonna use for gate driving? A dedicated IC, or bunch of BJTs with big bootstrap capacitors and such?
>I bought 200A MOSFETs for a reason. 2.4mΩ.
Seems reasonable enough.
>Brainrot take
I did run chinkshit controllers without a heatsink, FETs did get hot, but somehow it survived more than entire contraption.

But yeah... Re-do the calculations. No way you're gonna generate 15W.

>>2752318
TMC6200, with ZXGD3006(? Can’t remember name) 10A gate drive totem poles.

100A^2 * 2.4mΩ = 24W

I have a 20V 150W solar panel
I want to use it to drive my electronics.
Can I do this by making 240VDC (it should be fine, right) by using a boost converter?
I don't want to fuck around with a flyback transformer or a bulky mains frequency transformer.

>>2752447
A ratio of 12 will make your boost converter pretty inefficient. I'd make a ZVS switcher + rectifier instead, no need to worry about air gaps in your transformer, plus the thermals will be less stringent. Though a ZVS circuit's lack of feedback will make MPPT difficult to achieve.

Check the input voltage capabilities of the things you're plugging in. If they have switched-mode power supplies, they can often go down to 100Vrms (140VDC equivalent). If you're powering any sort of dumb resistive element that's not going to work, though I can't see why you'd do that.

You can also get a 150W modified sine inverter (i.e. square wave AC) for like $40.

>>2752350
>100A
Id go with 12 FETs instead.

Is ML2032 safe to use in place of an ML2430 for keeping an RTC powered on?

>>2752637
Yes. They're both 3V cells, just different dimensions.

>>2752561
I always design everything by looking at the
ABSOLUTE MAXIMUM RATINGS table.

>>2752786
Zheng does too, but his 1kW controllers use 9 fets, or 12 fets. He is profesisonal

>>2752867
Zheng uses knockoff FETs or tiny SMD FETs with insufficient heat-sinking. He also copies open source designs and makes them objectively worse, and sews cat skins into vests.

a while ago i stumbled into a video of xbox360 "repair" where someone did a reflow. looked easy and i have a broken xbox, so i tried it, bought a bunch of shit (heatgun, multimeter, flux, isopropyl alcohol) but sadly in the end it didn't work, although previously it didn't even show the error code (e74). is there anything else i can do at home? i've seen that one other thing that people did for that issue was reballing, but that looked like you needed special equipment.

i also have a broken chink roomba that i'd like to try to repair (replaced the battery a while ago but didn't help), any general tips on that?

>>2753072
If you bought the xbox used then there's no telling how many problems there are, but e74 is probably a bad GPU.
For the roomba, take it apart completely so you can clean it and replace the lubricant on any moving parts before you focus on the electronics and base station.

>>2753093
actually i bought it new back in 2007, i was the only owner, it just stopped working when i moved to another place like 4 years ago.

>>2753072
So, I had a roomba for a while until it died.

The thing with these is, you have to spend more time cleaning the fucking roomba than you would just using a regular vacuum!

I’d spend like an hour or too disassembling the thing and unwinding long hairs spindled around everything so the little brushes wouldn't even be turning, and cleaning out the crap from this tiny plastic tray. The thing makes no fucking sense.

>>2753150
You could try to reflow the HANA chip if you have a hot air station, but since it's probably a Zephyr motherboard it isn't worth putting money into it. You can use the PSU for other stuff, and you can salvage the RF module to use on a PC, but pair your controllers beforehand otherwise you'll need a microcontroller to stand in for the pairing process. The hard drive, fans, wifi adapter, optical drive and case can be sold on ebay.

CE student here. I’m in the market for an oscilloscope. I’ve been looking at the Rigol DS1102Z-E since it’s the cheapest one that Rigol sells, but people online have been recommending one from the new DHO800 series. I don’t know shit about technical specifications of scopes, but I intend to use one for home projects. That said, is it worth it to spend an extra $30 for the DHO802?

>>2753194
Get something 4-channel if you’re bothering with a DSO. Something hackable like the 1024z is a popular choice. If you want to cheap-out, get a 2nd hand CRT scope and a logic analyser, assuming you don’t need single-shot analogue capability for viewing power-on transients or whatever.

is there a way to simply and passively rescale a resistance range, like on a potentiometer? ie i want to go from 0-250 ohms to 50-200 ohms without a microcontroller or anything like that. assume that the component cant just be replaced with the appropriate one otherwise i would have just done that.

>>2753709
Adding extra resistors around a potentiometer can change the range, yes. Commonly this is used to turn a linear pot into a log pot, trying to keep your pot linear likely won't be possible. If the pot is just being used as a voltage divider, there are more options (usually involving op-amps) to change the effective gain and voltage offset after the divider.

>>2753711
damn, seems like a lot more than i was hoping to get into. thanks anyway.

>>2753709
If you put a 375 ohm resistor in parallel with the 250 ohm pot, you'll get a 0-150 ohm pot with a log taper. Add a 50 ohm in series and you'll get 50-200 with a curve

>>2753736
I have a veritable plethora of log pots I got out of an old mixer.
Is there any way to “remove the log function” and make them linear pots?

>>2753740
>“remove the log function”

easy.
gang two log pots together using belts, pulleys, shafts or gears.
wire them in reverse parallel.
you now have a linear pot with half the resistance.

>>2753222
Did you mean the 1054Z? I’m looking it right now, and the modification options look very interesting to me, but are the extra two channels worth it?

Quick sanity check with you guys.
This is a very simple battery checker (Q2 lights LED if voltage on Q1 base is too low) that's floating around on the net. I've been putting it in a lot of projects.
Is there any reason I shouldn't replace the LED with a relay and use it as a low voltage cutoff for a 36V Li-ion battery pack? (Yes with appropriate resistor values, not those on the schematic.) I don't see why not but these things make me uneasy.

>>2753809
Yes I’m retarded. I think I used one myself ages ago, can’t remember about the extra features. If you’re doing digital shit be sure your DSO has a protocol decoding capability, since it’s more convenient than buying a logic analyser. For comparisons, read some reviews with proper comparison tables, watch some video reviews too. See if there’s hacked firmware for it too.

>>2753906
>Is there any reason I shouldn't .... use it as a low voltage cutoff

yes, coz it's the completely wrong circuit.
to do what you want, you need a voltage reference chip, or a regulator, or at least a zener.
this would give you a precise fixed voltage to compare against the variable battery voltage.
all you have now is the Vbe of Q1 and of Q2 acting as a very inexact and temp-sensitive reference.
not good enough when it can mean killing a bunch of expensive lipos.

>>2753906
Use a TL431

>>2753906
So, let me get this straight…
When the LEDs are on, that means low battery?
This thing is going to kill-off whatever is left of the battery, assuming the LED is drawing around 20 mA. This is not good.
Instead, it should be a “Batt OK” indicator.
Furthermore it shouldnt be continuously on, it should flash with, say, something like a LM3909.
When it’s dark (off) that should be the indication battery is dead. As luck will have it, this will also be the state of your circuit when you’ve completely exhausted the battery and it’s destroyed.
The other thing we commonly do is a “push to test” button.
And yes, below a certain voltage, it should completely disconnect it from any current draw.

How do I make a linear position sensor with a range of one meter, a resolution of 0.1mm, sample rate of 1kHz, and a latency of 1 microsecond that costs $10?

>>2754093
Firs step: kill yourself for posting a frog.

>>2754095
How do I make a linear position sensor with a range of one meter, a resolution of 0.1mm, sample rate of 1kHz, and a latency of 1 microsecond that costs $10?

Is that better?

>>2754096
An ultrasonic method, while cheap enough, will yield:
>6ms latency
>170Hz sample rate
>9mm resolution

So you either have to go laser (maybe some aliexpress laser rangefinding modules are cheap, even then 0.1mm is gonna be tough) or some sort of encoder. A rotary encoder on a wheel/gear that rolls along the linear surface would work, but wouldn't be an absolute encoder; you'd have to keep track of location in memory. Also backlash or slip might be an issue. A 1m long (absolute) linear encoder is the most foolproof method, but you're looking at something a lot more expensive.

Now tell us the project so we can say that your design requirements are stupid.

>>2754093
Ever looked inside an inkjet printer? Do like that.
Get a 1m encoder tape.
It’s the most common thing in the world, and the specs are better than you need.

Putting together a CNC that uses 3 different power supplies for different things. Can I tie the input all to the same mains plug instead of having three separate ones? Just chuck them in parallel with fork terminals?

I made a RetroSpy for my Playstation 2. ᓚᘏᗢ

>>2754136
>the specs are better than you need.

Thanks. Just placed an order for pic-related. I guess my $10 budget needs to be increased somewhat.

>>2754095
Why do you hate frogs? Are you some kind of effeminate nigger?

Electronics noob here. I've been working on a prototype for a bit and it's going well, but the last (and only) time I did electronics was 20 years ago in high school so all I know is breadboards. It seems I've hit the limit of that and I need to lrn2solder now.
I'm planning to get assorted perfboards and veroboards but not sure what to get for soldering equipment. What cheap chink shit isn't too bad? And what's the list of things I need realistically, given I'm starting from nothing with this? I'm guess I need (this one's obvious) an iron, some kind of work surface, some kind of vice or clamp to hold the board in place, probably a magnifying glass, and solder. I don't have very good aeration here so what lead-free isn't unusable?
Am I missing anything and what should I look out for in all these?

>>2754159
>dumb nufrog poster
>accusing anyone of being an effeminate nigger
That's as rich as you are poor.

>>2754162
So you ARE a dickless faggot. I think you'd more at home on /g/. What's your hackadong handle? lmao

>>2744526
Does anyone have a recommendation for an electronics vacuum cleaner?
I'm in the market for one to clean the dust filter on my projector.

>>2754186
Shop vac + dry paint brush. For everything else that doesn't contain optics or filters, use compressed air.

>>2754161
>Am I missing anything

the 200,000 beginner soldering videos on youtube.

>>2754161
if you are serious then find a temp controlled iron you like. theres a push for the heater being part of the tip so you have to replace the whole thing which is a fucking gimmick to charge you more. ignore the hype. if you are poor then say a 25w iron is absolutely fine if you aren't retarded.
the ubiquitus magnifier helping hands are dogshit, the goosneck arms are hit or miss, a pcb holder that you can spin it around on one axis is like 10$ on amazon.
i use bog standard SnAgCu 96.5/3/.5 solder rosin core never had a problem with a temp iron at 350c. there was a big hoohaa about it all at the time but its mostly blown over now.
some people use fume extraction i dont bother. some people need magnifying glass i would say you shouldn't for .1" pitch stuff but depends on your eyes. good light is important.
don't really need flux if you have core solder. solder pump might be a good idea and they make solder wick(get it with flux if you don't have flux separate) for mistakes too can be useful.
the other thing is used to always be a wet sponge but a dry metal 'sponge' is all the rage now although a bit more boring.

>>2754199
>nonleaded

>>2754202
>25w
>no flux

>>2754208
>don't worry, the lack of fume extraction will make it all fine after 20-30 minutes

>>2754159
No, I'm just not a newfag or election tourist. On my home board frogs get banned on sight.

>>2754245
>On my home board frogs get banned on sight.
This isn't your home board. Maybe you should go back if it bothers you so much, dickless.

>>2754245
>On my home board frogs get banned on sight.

can you give us a hint as to where fellow frog haters might find this frog-free environment

>>2753998
I see. I hadn't thought about the temperature sensitivity. I'll use a comparator with a TL431 as suggested by >>2754005

>>2754015
>"batt ok" indicator
Certainly not. Having a fault indicator that un-lights when at fault is horrible design.
>push to test
Yes, I do implement some of them like that.
The point of this circuit is the minimal parts count, current draw and footprint, I put it in 9v-powered projects like instruments, effects or preamps. Draining the battery further is no object, the goal is to indicate to the user that they have to change it because it's bad already.

Thanks guys.

>>2754161
>I don't have very good aeration here so what lead-free isn't unusable?
Lead isn't what needs ventilation. Ventilation is to prevent you from inhaling flux vapors, which will fuck your lungs harder than chain smoking.
AFAIK lead-free solders have much harsher fluxes

Greetings ohm
I have a coil and a cap in parallel connected to a rising voltage
Lets say i also have an LED with the fitting resistor for 12v inside a pushbutton.
How would a primitive low power circuit look like (or how is it named?) that sends power to the LED as soon as a set voltage (above 12v) is achieved? Bonus if the LED survives up to 30v on the rising voltage, i cant really control the upper ceiling

>>2754295
standard rosin flux is chemically not that far off from weed smoke

>>2754130
>>2754136
>>2754156
It's for a special type of low cost 3D printer that doesn't use steppers. The encoder tape looks nice, but it seems expensive.

>>2754287
> indicator that un-lights when at fault is horrible
It’s mandated by law for pure battery-powered equipment.
Even car batteries which had a built-in “good bat” indicator used a green LED, and going dark indicated “bad”…. Don’t you remember that?
I can see how someone might think the opposite, but a completely dead battery will always go dark anyway. Unless you’re continuously looking at it.
Dig out a 20 year old battery… Oh, no red flashing fault indicator? … must be good… put it in some life-critical emergency equipment… death ensues.
See the problem?
Yes, mains powered stuff (like your mainframe) has a “fault” light for when your tape didn’t autoload. man, I hated when that happened, but it didn’t run on batteries.
Hell, just detect when the battery is down to 80% and make it chirp or project a princess leah hologram telling you to change it for a month before dying… but note it has a green LED indicating that it’s working.
This is your PSA to change your fire alarm batteries, and vacuum out the dust. Also change out your washing machine hoses if it’s been 5 years.

>>2754372
Look anon, the low battery warning is there to remind a user to change the batteries if they're bad before they go on stage or whatever. They see it when they turn it on or push the check button. The rest of the time, the device is off. There is a pilot LED when the device is powered on. Of course that won't light up if the power source is dead. Are you good?

>>2754199
>lead-free
>cartridge tips are a gimmick
I got 4 more chinky T12 cartridge tips for $10 free shipping, a bunch of different tip shapes since I've never had one go bad. They heat up in 10 seconds, and can easily be swapped out while hot.

>>2754359
For 12V, they'd simply use a 500Ω resistor or larger. That way, the LED consumes 20mA (maximum continuous current) when fed 12V. The LED will be visible at 100µA and probably even lower if it's dark, so it will still be visible down at 3V. Above 12V, the current will get higher than 20mA, which is a bad thing. I advise that you pick an arbitrary maximum voltage (maybe based on other maximum values in your system, or based off a zener diode somewhere for safety) like 50V, and put a series resistor inline with the button's LED pins such that the LED current will never get above 20mA. The maximum current might be lower or higher than that, I'd probably measure the current through the button at 12V and figure it out from there. It will shortly die at 30V.

>>2754367
Cheap 3D printers use steppers because they're cheaper than using encoders.

>>2754393
Steppers + stepper drivers + lead screws are also expensive. Are there no other suitable position measuring technologies? Do I need to invent one?

>>2754156
>>2754367
> encoding tapes expensive
I never thought they were expensive since I salvaged all mine from discarded printers.
You can also salvage the encoder wheels, and then your carriage is infinite.
I have a few longer tapes that are like about ½ m from big printers, but they all came from the dumpster. I’d try and use the encoder wheels, or gang a few encoder tapes together.
Also, you could use two shorter tapes and move the bed as well as the carriage. And/Or have two carriages, and print at twice the speed, so the kitchen table you’re printing won’t take a full week of continuous printing.
Also salvage the sensors.

>>2754393
>I advise that you pick an arbitrary maximum voltage
its a hand generator powering a spark coil , i need to build that part first before i know what a drunk spastic might achieve lol
Do Zener diodes typically have a hysteresis like say i have a 13v rated on, will it limit the voltage to exactly 13v or fire at 13v and drop some below?

I'm looking for a 4-7V to 3V regulator, and I can't find one that meets specs.

Output voltage is probably sub 100mA, I'm not sure though. It's meant to run for 66 hours off 2 alkaline AAAs, so it should be low draw.

The first one here, the LDO, has the best IQ, but there's no efficiency rating and I don't know how hot these things get. Maybe it wouldn't matter at such a low draw. The second one, the buck, has an output voltage that's a little too high, but otherwise it's pretty good. The third one is only a 5V max, and because I'm using 4 NiMh cells to drive it, that could be too much if they're freshly charged. The other 2 are 6V max, which is more leeway. At the moment I figure the LDO is the best pick, but I don't know about these things.

>>2754381
> on stage
I guess a piercing chirping sound is out of the question then, eh?
I’d still pwm the LEDs to cut down on the draw.
I think you can probably get some high-efficiency LEDs that draw like 1mA nowadays that are still bright enough to see.
Green LEDs are the most visible due to the sensitivity of your eyes at 554 nm.

>>2754359
You’ve got the same problem as the low battery indicator guy, except worse, at least he has batteries. You’ve got a capacitor. That probably has leakage to begin with.
So if you have excess voltage, the only easy way using passive components is to burn it off as heat. But the LED and the resistor and/or zener all drain the capacitor more.
There are fancy oscillator-based, PWM, buck, transistor “active” solutions you could employ which would do a better job.

>>2754404
> regulator
The first on takes the voltage and turns it into heat to cut it down. The efficiency is dependent on the input voltage and vout difference, and it’s terrible.

As for the second one, you can often alter the output voltage by tweaking the support components slightly.

Doesn’t need a boost converter, really.

>>2754397
Don't use lead-screws, it's cheaper and more reliable to do belted-Z. Use a bed level sensor to get the z height each time you power the thing on since the bed/gantry will sag when the power goes off.

>>2754401
Zener diodes are linear, they don't have a precise threshold. Whether you set your "turn on threshold" at 10µA, 1mA, or 100mA will determine what voltage it "turns on" at.
The TL431 has a precise threshold, with no hysteresis. You can probably give it hysteresis, though I think you'd need to use an additional transistor, as opposed to just adding some resistors. Maybe you can have a resistor on the negative pin of the TL431, idk. Simulate it.

>>2754404
>there's no efficiency rating and I don't know how hot these things get
Efficiency of a linear regulator is just Vout/Vin, minus some small amount based on quiescent current. Don't forget to check the minimum dropout of the LDO though, it can be high at high currents, preventing you from dropping 4V (or less if you drain your cells to 0.9V each) to 3V.

Also stop buying modules and look at ICs. Browse Texas Instruments' website and use their parametric search tool to find the right part for you.

And consider using parts that just have a wide enough voltage range to handle 3 series batteries regardless of state-of-charge, that way you don't need a voltage converter in the first place. A decent number of MCUs can go from 5V down to 1.8V, if you're careful with their clock source.

>>2754359
>sends power to the LED as soon as a set voltage (above 12v) is achieved?

use this type of circuit.
a 10V zener in series with a 2V LED will only turn on when voltage is above 12V.
size the resistor so you wont surpass 20mA at 30V, or whatever is the max.
ex.: for 30V, do (30-12) / .02 = 900 ohms

>>2754453
Low voltage zeners especially have a pretty shitty voltage/current curve.
Better off with a TL431.

>>2754295
Thanks, I'll look into this.
>>2754199
Yes, I'm very serious.
I lurked for a while now and saw the debates about temp-controlled vs non-temp-controlled. I was planning to go for temp-controlled for sure. Thanks for the tip about the minimum I should aim for. I'm not poor but since I know I'll be doing this for quite a while, I figure I might get cheapo starter kits while I figure out what I want/like and then splurge on good stuff once I know my dick from my arse.
For magnifying glass: I unfortunately need one, my eyes can't focus at night even though eyesight isn't the problem per se.
Thanks for all the tips, that should cover what I need.
>>2754393
>>2754208
>>2754213
>>2754202
Also taking these into account.
Thanks all for the replies.

>>2754496
Saw last thread I assume? I'm the T12 shill, and also the one whose T12 station's power MOSFET died short after 3 years and took the tip up to 500C. These stations are good value and outperform most conventional stations, but I recommend getting a backup iron to fix/modify it with if you ever run into any problems. If the anti-cartridge tip anon wasn't just making shit up, it suggests that chinese tips aren't intrinsically trustworthy.

For a beginner though, I'd recommend something like a Yihua 928D or 9936. It uses Hakko-936-style tips, which are cheap and come in a bunch of different geometries. Has digital temperature control, which should be sufficiently better than the analogue temperature control on a Yihua 907/908/947 style of iron. You could also go for legit Hakko tips for not too much extra.

Though look at reviews before you buy anything. Teardowns especially.

>>2754509
Was looking at a quicko t12-956 actually but will checkout the hakko too

>>2754496
I’ve soldered more shit in a day than you will in your lifetime. This is what we use on assembly line amplifiers. I cut the leads before I’m even done letting the solder wick through the thru-hole joint.

> T12
The ‘T’ is for toy.

>>2754496
I’ve soldered more shit in a day than you will in your lifetime. This is what we use on assembly line amplifiers. I cut the leads before I’m even done letting the solder wick through the thru-hole joint.

> T12
The ‘T’ is for toy.

>>2754542
Thanks, but I am not a professional. A toy would likely serve me far better than whatever professional tools you prefer. I don't know if this is also the case here for obvious reasons, but typically, industrial-grade tools have higher "overhead", i.e. they're just inconvenient at best, if not awful, if you work at lower scales, while they shine in that regime.
For example, I sure as hell don't need a $10k bench scope, a $50 handheld wireless shitscope does 99% of what I will want it to do for a long while without restraining my space or setup.

>>2750773
1. Check the push button, you sure it's in the correct orientation? You might have connected the two jumpers to the two corners to prevent from accidentally shorting it, but you didn't place the button on the center of the bread board, you actually might be shorting the button there, bypassing it entirely. Reposition the button to be on the center line.
2. You haven't posted any detailed schematics or details on what you are trying to accomplish, FSM with 4-bit counter doesn't give us much to work on. You've also posted multiple posts that makes it hard to track of all the info. Repost your question with a complete schematic, number of all ICs, your power supply voltage and what you are trying to do.
3. Check your ICs, might be damaged. Check if they are hot, indicating a bad IC. Check your LEDs to make sure you didn't burn them by running them without current limiters.
4. Your left capacitor is connected between ground and to the input of the schmitt trigger, with a resistor connected to a button that is itself connected to ground and that same capacitor. That means this push button ain't doing anything useful, it cannot charge the capacitor, and the capacitor being connected to an input is not charged that way either, and that's assuming point 1) didn't apply in the first place permashorting your capacitor to ground. This means the left chip acts as a simple inverter, it gets 0 and outputs 1, which is then fed to a capacitor connected to the output of the second chip (which you haven't specified)

Overall, based on the pictures, the circuit is nonsense, you need to first come up with the full schematics, only then can you start building the circuit on the breadboard and us helping you.

>>2754522
The Yihua 928D is a budget option, it is inferior to a T12 oled station.

>>2754542
>commercial THT soldering grunt
If you only solder THTs, then sure. But that thing isn't going to be any good at power SMT packages, or really small SMT packages. Less thermal finesse in general, compared to what the board repair guys use.

I worked at a company a bit like that, but there were only a couple of THT soldering jobs at all. Almost all the THT soldering was done by a machine, it had some sort of computer controlled rotating hollow soldering tip that had solder fed into it from below, that moved beneath the board. The only manual soldering was done for bodge jobs, IIRC it was adding a daughter board with supercapacitors on it that wouldn't hold at the right angle in the machine. There were probably more people with oscilloscopes doing troubleshooting than there were doing THT soldering, though that may speak to the quality of their engineers more than their degree of automation.

>those thermal reliefs

>>2744526
Guys I'm a total amateur in need of some help.

I have two SDRAM memory chips (DDR1).
Want to make them piggyback on each other. ( dual rank /or whatever its called)
All pins connect to each other,except chip select.
Where does the chip select of the top chip connect to?

Example here
In the forum post pin 28 is lifted and is connected to a different spot.

https://www.ogxbox.com/forums/index.php?%2Ftopic%2F7548-16-128mb-ram-upgrade-tutorial%2F


I don't intend to to this mod but I want apply it to something else.
What should the second CS pin connect to?

>>2755231
you're in over your head
read more. learn the basics of how RAM works, practice with simple cheap ($1-4) RAM chips on breadboard before you do anything fancy to your memebox

I am looking for my first oscilloscope. Does anyone here use FNIRSI-1014D ? I have found a great deal in my area for that model, but I thought that it would be a good idea to ask if it works fine.

>>2755233
>you're in over your head
Yes I'm aware of that.

As for reading I'm reading internet articles including Wikipedia and still don't get it.
Can you at least point to a resource that will give me the knowledge I need?

>>2755248
you can't just connect sdram to a micro. you will need a sdram controller inbetween (its still dram so will need to have dram tyoe of processes done) to do all the module selection.

i've had these 2 for awhile. i use the small one a lot more than the big one. battery powered is really convenient. thinking about getting a dual trace higher bandwidth batttery powered scope.

>>2755248
no bro. you are WAY in over your head. the guys in these tutorials are massively understating the complexity of this project.
>Wikipedia
you need to be reading datasheets, at minimum. and not the nice, clear, 5-page ones, I'm talking 200-500 page manuals per chip. not to mention basic skills like soldering 0.5mm pitch leads, and compiling with the proper flags or whatever tf computerfags do.

The best you could do is practice soldering on some old fried motherboards first, follow these guys' instructions EXACTLY (no nigger rigging) and pray to god you didn't miss some tiny obscure detail every engineer implicitly already knows. actually understanding this project would take years of experience working with obscure hardware/software

>>2755461
Not everyone is as dumb as you. He'll be fine.

>>2755459
I wonder what PSU those digital scopes run on? If everything is converted from a 12V or 5V rail or whatever, it might be quite easy to convert them to run on a battery.

>>2755510
>>2755510
>>2755510
NEW THREAD