/diy/ - Do It Yourself

this thread's digits brought to you in part by the Intersil TW2826 digital video multiplexer

>>1522826
Is www.allaboutcircuits.com good for learning electronics or are there better resources?

Yeeee boys 24hrs and I will have my new power supply and Arduino kit in my hands! And I can use my new 2,250 in-lb impact driver to assemble it! What could possibly go wrong?

Cant find a pick but my nephew was about 2 or 3 and would shove chairs under light switches and just turn ahit off and on all day. I was a little worried he might be "special" but he was fine. Made my day. I built a box out of 2x4 and wired up a bunch of fans and light and buzzers and shit in it. Put random switches all along the bottom. He spent 2 weeks standing on the lexan top screaming at it and anyone that came buy just flipping switches.

Hes old enough now to start understanding some of what he is playing with so we got him a modular click together electronics kit.

When he's old enough to solder we are going to build some serious shit!

Lol, pic related. Not him but another nephew. My sister sent that yesterday and said "he is REALLY excited for Christmas."

Nb4 reee. She photoshops for a magazine all day.

>>1522826
can i use a capacitor and a transistor to switch a servo motor back and forth?

>>1522869
that's cool man. be interesting to see what he becomes when he grows up, an Electrician, Electrical Engineer, or the guy the flips the switch on the electric chair

>>1522880
you can do anything you set your mind to

>>1522884
cool, this opens up a whole other realm of possibilities

>>1522832
Ooh, that's a nice one.

>>1522910
Outside of my skill range, both desoldering and programming, I'm afraid. If it's one of the extra-cheapies, it won't be reprogrammable anyhow.

>>1522849
Looks pretty good, provided you're going through the "education" section and not just browsing articles.

Do you guys use solder paste with no-clean flux? About how much paste do you use per assembly, i.e. is a 5cc syringe going to last me a decent amount of time or should I spring for a big tub?

I'm self learning electronics. where do I go (which textbooks to follow) after finishing learning circuit analysis?
fine with the mathematics by the way, and don't want to shy away from theory though it would be nice to get into the practical side sometime

Maybe this is a dumb question, idk. I am looking to simulate the heater element of a car oxygen sensor. For this, I am supposed to use an 8ohm, 25 watt resistor (some say 20 watt is ok but 25 watt seems to be the safer bet). Now it seems I have two options:
1) a chassis mount resistor. It big and bulky but this is what everyone uses. Its supposed to get pretty damn hot so you stick it somewhere inconsequential like on the firewall or on the body inside the engine bay.
https://www.mouser.com/ProductDetail/ARCOL-Ohmite/HS25-8R-F?qs=sGAEpiMZZMtlubZbdhIBIEidPY%252bx%2fLaaBVxe6QVE8Ic%3d

2) an SMD pcb mounted resistor. This thing is a fraction of the size, which would be awesome for me because then it would fit into a nice discrete package along with the actual o2 sensor simulator circuit.
https://www.mouser.com/ProductDetail/Bourns/PWR163S-25-8R00FE?qs=sGAEpiMZZMtlubZbdhIBICjW2gS1%252bJ0hqGE7vqvlvjM%3d
>pic related

However, it comes to my attention that something just aint right about this. How is something that tiny supposed to handle the same heat as a resistor like 100 times bigger? Will this thing melt or something? Why would everyone use chassis mount resistors that take up so much more space if they can have the same resistor in a much more compact package? Basically, whats the catch to using a SMD resistor? I know fecking feck about electronics, this is babby's first circuit project, so forgive me if my question is retarded.

>>1523073
Look at the datasheet for the chassis mount one. It's like 2x2 cm. SMT will also have lower thermal resistance to the board because the solder gives it good thermal contact.

>>1523077
Forgive my ignorance but could you break this down for me a little more? Are you saying the SMD wont be a problem, or that I should be using the chassis mount one, or..? Again I dont know wtf im doing, I just found a schematic online and soldered it up and this last part regarding the resistor has me stuck because I dont know which one I should use. I have no concept of what thermal resistance entails or whether or not thats good or bad.

>>1523081
If you're designing a board then you can put the SMT resistor on it and be fine. If you're using some pre-made board or protoboard use the chassis resistor.

>>1522882
Kek. His dad is a worthless sack of shit and was never around. Hopefully he hangs out with me hotrodding electric rides. He loves the hell out of his little john deere power wheels. I found one at a grage sale to strip down and hotrod with him. They let the kids run power wheels beside the pit at the mudruns before it starts. Everyone thinks they are cool shit buying a $700 plastic truck or throwing an led strio on it and bragging about their shit motorcycle battery.

>tfw 36v lipo 775 motor with metal gearbox. Fucking fight us.

>>1523073
>However, it comes to my attention that something just aint right about this. How is something that tiny supposed to handle the same heat as a resistor like 100 times bigger? Will this thing melt or something? Why would everyone use chassis mount resistors that take up so much more space if they can have the same resistor in a much more compact package? Basically, whats the catch to using a SMD resistor?
Look up the data sheets, smd resistors (and other things, like mosfets) use the PCB copper cladding as a heat sink.

Also they might use different materials which can withstand higher temperatures. Through hole resistors with the same power rating can be different sizes depending on the material they use.

>>1523014
>solder paste with no-clean flux
yes, an activated flux would be hard to clean under QFN/BGA/etc.
>how much
I have 50g of SAC305 in a "tub" about the size of a golf ball. when I stencil a small-medium board, I scoop out ~10g, almost all of which remains on the stencil and goes back into the tub. if I were using a syringe, 5cc would last a long time. solder paste does have a shelf life, which some people claim to have successfully extended by adding a touch of alcohol and mixing well

>>1523007
>If it's one of the extra-cheapies
fug, also a possibility

>>1523144
>PMS150C
Well the pinout matches.

In other news, my SA602s arrived yesterday, and I can't help but feel fear that I'm holding one of the LARGEST SMT ICs out there. I think I'll need to grind down the tips of my tweezers to actually match one another.

I'm trying to figure out the intensity (mW/mm^2) of light emitted on a small section of a plane above an LED. The square is small enough that I'm assuming the intensity is constant and depends only on phi. I know that I(Phi) = dFlux/dPhi. But I'm having a hard time figuring out where to go from there. My current best guess is that the relative luminous intensity chart is showing the intensity relative to 2.5mW/sr. If that's the case, then at Phi, adding some dPhi adds a certain flux which I can find from the chart, and I also can find the amount of surface area of a partial sphere that's added when I add that dPhi, then dividing dFlux/dArea will give intensity... does that sound correct? Is there some shortcut I can take since the total emitted power is given?

>>1523160
Have you done any second year university physics? If not, you may need to learn how to do angular integrals. I think that circle is a r = cos(theta) curve. Figuring out how to integrate over solid angles is also probably a right pain. If I'm still alive after learning KiCAD I may help you out in a few hours, since I've done something similar to cheap LEDs.

Nice that the datasheet gives you radiant power, not a bunch of garbage in candela/lumens.

>>1523163
Yeah, I have. I'm doing this numerically though as part of a simulation so summation is closer to what's going in the code than integration.

I did a piecewise linear approximation of the curve in the datasheet. Cos is only a little off.

>>1523163
And if you need any help with KiCAD I might be able to help. I started using it a few months ago.

>>1523165
Oh is that Spyder? Neat. You've probably remembered more of the integration than I did, but I think the main thing is just to sum the intensity over the whole of an area and normalise it to equal the total radiant power. By selecting that circular shape as your surface you'd ensure that all of the power at those points would be equal, but I'm unsure if you'd have to multiply them by the cosine of their relative angle to the surface since the light will be hitting them at oblique angles.

>>1523166
I'm following this hour-long tutorial:
https://www.youtube.com/watch?v=qnaDnzK26D0
And am not having any troubles so far. We're on KiCAD 5 now though, so there may be some discrepancies since the video is a year old or so.

I think I've narrowed down the fault on the tv to somewhere in this area. There is 120 before the bridge diode, and then 30v after it. It's supposed to be around 110 (highlighted in red).

Anyone have any idea of what might've failed? I've traced out what I think is the path in purple. I'm going to try checking all caps for high esr, and measure the resistances.

>>1523194
The... bridge diode failed? If you plug it in long enough, anything shorting should cook itself. If it's not, then a charge-up circuit is failing, because a normally restricting component is trickling enough charge to prevent recharge (a cap not charging up possibly).

>>1523169
Yep, Spyder. I posted on sqt on /sci/ what I'm thinking right now. I got this intensity vs phi graph and it falls off a lot quicker than I imagined. I feel good about the math up to a scale factor but maybe I messed up the code. Tomorrow, I'll try simulating it with the LED right up against the plane and calculating the total power incident on the plane.

Merry christmas nerds

>>1523194
Where did you get your TV circuit diagram? What year is the TV? I've got a TV in need of a repair and no clue other than replacing the board.

What’s my next step? Buy a Hakko soldering station? And then I probably need a hot air station. Then an oscilloscope. Then I bet I could hook up that 9V battery.

>>1523249
> buys a bunch of shit he has no idea what to do with
> business as usual
Fuck off you cancerous tripcunt faggot. Plenty of info in the OP if you are actually interested in leaning as opposed to just posting more shit you'll never use.

>>1523252
>tripcunt
Excuse me sir, I think you are mistaken.

And yeah I need to work on the blinking light. I’m overwhelmed and I don’t know where to start. I’m waiting for that power supply in the morning so I can fux some more. My mind is blown that this Ridgid 1/4” actually broke lugnuts loose.

There’s a CD and website for that kit too. I need some purpose and then I will get off and running.

I might have to wife this girl.

>>1523252
The fuck will you faggots learn.

Use Altium regularly at work, want to start designing my own PCBs at home for home projects and potentially making my own products. Is Circuit Studio worth purchasing?

>>1523262
No. Use KiCAD. I used Altium at work and got KiCAD a few months ago and the transition has been fine. Having independent symbols and footprints was the hardest thing to adapt to but it honestly makes more sense this way.

>>1523257
Ridgid is actually a good company, I've seen many professional wood workers use their tools. Just like Kreg, so I gave their stuff a try.

>>1523194
Actually, I'm stupid. The what is supposed to be 110 is 130V DC, (and also 25-30V AC apparently).. I think that's close enough, though the 15.7V is 18 or so V on the PWM reference. This circuit is a bit over my ability. I had some trouble figuring out not to mix up the two or three grounds. I blew a few fuses... Since all I need to do is make sure it outputs a clean -15V on one of the rails (which is currently -22V) I'm just going to throw a zener in there and try to call it a day. I think it's used for op amps only. I only have so much patience for a tv I plan to get rid of.

>>1523248
I have a 1997 Sony PVM 14M4U. It tears vertically, and horizontally, but otherwise works. The service manual is on the internet archive. https://archive.org/details/sony_PVM-M_14_Service_Manual This things a beast to service because you have to dismantle it, and have the boards hanging off each other (with the tube nearby). Tough to repair while running (I think the 10" PVMs are a bit easier to service). Though to be fair, I'm able to disconnect the tube and just service the power board, for safety sake.

I looked at the power rails, and it's all labeled. One of them is clearly off, so that's the first thing to fix. I wasn't able to find any bad caps on the video board, but you could try with an ESR meter (on just the electrolytics). Beyond that... I don't know. Get a book on TV repair.

Oh, and this video was helpful:
>https://www.youtube.com/watch?v=iyJDam0blYA

First KiCAD PCB design, not too bad if you ask me. Though I think I got the front and back copper layers mixed up. Not going to be able to plate my thru-holes properly, and my vias are going to be whatever copper wire happens to fit, hammered and peined over into the hole with minimal soldering. I may also try to work a cartoon into the ground plane (Gondola?).

>>1523294
actually that trace on the top of R2 is a bit iffy

>>1523294
You can easily make that on a single side. One-sided pcbs are cheaper you know.

>>1523312
assburgers guy doxxed himself kek

>>1523194
i bet its a diode

>>1523294
LMV339x4 doesn't seem to exist and the LMV339-N is not available in PDIP14 format, only TSSOP14 or SOIC14. There's an adapter for TSSOP14 but I would consider using the standard LM339 if possible.

>>1523313
you'd think people that act like this would have been bullied enough to know better

>>1523294
You're making it yourself? How?

>>1523057
Horowitz and Hill might be the book for you, then

>>1523149
>pinout
also matches 8-pin PICs, which, if you were more of the programming sort, could be dropped right on the board for extra fun
>LARGEST
Philips/NXP has some much bigger SOICs that are as wide as DIPs, Pic related
>not buying the whole 8pc set of chinkshit anti-static "Vetus" tweezers
I can tell you it's nice when there's always a spare not far away. sometimes it means the difference between ten solid minutes of "where the fuck" or "GET ON THE BOARD YOU LITTLE CUNT" vs. grumbling and grabbing one of the others
SMT pro tip: if a small part like SOT-23 or some passive is upside down, pick it up a few cm and drop it, 50% chance it'll land desired side up

>>1523294
did you run the design rule checker?
>trace crossover at the bottom of U1
nope
also, if you're using a shitty process, avoid acute angles such as are connecting the middle pot to the two commons

>>1523337
>somewhat annoying to remove
oh kek

>>1523312
>I seriously hope you guys don't scan this.
is that George Costanza face in the middle intentional or just a happy accident?

Hey /ohm/, I'm doing this pedal effect for my Analog Circuits class. It's an inductorless wah-wah pedal. The circuit is pic related. The linear pot is for modulating the effect. The log pot is for volume.

I have to explain theoretically why/how it works. I know the wah wah effect is produced by a variable bandpass filter. If I had used the version with the inductor, it would've been easy to explain (here's an explanation on how it works for reference: http://www.geofex.com/Article_Folders/wahpedl/wahped.htm).). But this inductorless version is a bit tricky and I have no idea how/why it works.

I'd appreciate any help.

I hate electricity so much.

I’m gonna throw this thing in the trash and go play with my sawzall some more.

>>1523363
>contemplative Jazz music ensues

>>1523383
>I’m gonna throw this thing in the trash

por que

>>1523385
do not talk to avatarfags

>>1523385
I probably shoulda learned the basics.

So constant current means the currents will stay at the set max, but voltage can vary. And constant current means the opposite. So when I have an LED hooked up to the thing wirh voltage set at 3.000V and current set at 0.100A, but the LED can go brighter, why does it sit ar CC 0.100A and 2.600V?

There’s something simple in the laws that I haven’t grasped yet. I shouldn’t have smoked so much pot in college, or maybe should’ve smoked more. Or done more acid. I’m not sure.

>>1523389
>voltage set to 3V
>current set to .1A

>2.6V .1A out

If you're in CC mode it's because that's the target current.

If you're in CV mode, it's at the current limit (.1A). If you raise the current limit, it'll increase the voltage until it hits either the new current limit or the target voltage. Since it's a diode it's going to suck down a lot more current and barely change voltage, and you'll probably release the magic smoke.

>>1523389

as a beginner, I think voltage mode is way easier to understand. like setting 5 volts and using a suitable resistor in series with an LED.

Electricity/electronics is harder to learn compared with woodworking and mechanics because very little is visual. Stick with it, and realize that it could take a while to feel comfortable.

>>1523391
See, I think the diode thing is messing with me because I have been playing with LEDs. I need a goddamn incandescent bulb

>>1523392
This is my problem too, was a ME major because I hate learning about stuff I can’t hold in my hand.

>>1523393
>stuff I can’t hold in my hand.

and the best way to make electronics visible is a multi-channel oscilloscope. even if you are doing arduino/digital it is still a great tool.

like for example an arduino sketch that has an interrupt routine that seems to not work. in that routine, raise a pin at the entry, and lower it at the exit. do something similar in your main loop or somewhere, and on the scope you will have great visibility as to how long and how often the interrupt is running.

good scope start at about 3-400 bucks but they are worth it.

a cheap but excellent tool is pic-related which tells you tons about I2C, SPI, whatever, and costs about 10-15 bucks and works with free software.

>>1523376
look for the bjt small signal model, chug it in and see what gives

>>1523376
wild ass guess, but consider the two RC high-pass-filters to ground, C2/R6/R7 and C3/R6/R7. the C2 filter attenuates highs in the input, straightforwardly. the C3 filter attenuates highs in the negative feedback, so that lower freqs are fed back more and so receive less gain than higher freqs

>>1523389
>tfw third worlder brazilian fag making my own psu and going to do a signal generator next if everything does good
>stupid fat american doesn`t even know the basics of the basic stuff buys a psu out of the bat
>tfw when I graduate and have enough money for my shit I`ll not have much time to use it
>you`ll never be american and have so much disposable income and still be poor-medium class

I wanna make a backlit sign of like 6x4 inches, how hard would it be to get good uniformity if I make something myself with LEDs and perfboard?
As I understand each LED would have different brightness if using just one resistor, but it might not be a big issue in practice

>>1523415
LCD screens also have an LED backlight. To get uniformity they use a thin sheet of semi-opaque plastic that diffuses the light. If you can tolerate a decrease in brightness due to absorption and reflection from the diffuser, that's the way to go.

Anyone know of a way to diffuse UV light?

>>1523418
with a UV opaque material

>>1523396
Scope is on my list, was looking at those Rigols and Siglents but still recovering from Xmas.

>>1523403
That 3rd part is true for me as well. I barely have time to use the shit which is why that one magnetic LED thing was on my workbench for over a month before I got it working (somewhat). I was 75% done and then had to go out of town with the girl for a romantic weekend. I work a bunch of hours so I have a little extra money and then most of my weekend is taken up by pussy and maybe I get a couple hours on Sunday morning to fuck around with the toys.

Anyway buy your toys and try to have fun before you have kids. Plus this power supply and the Arduino kit think were both Christmas presents, I didn’t pay for shit.

Also you could always get out of that shithole. A few years ago my girl and her family were in Venezuela and you couldn’t even think of shit like this because you were too worried about eating. Now they’re getting diamonds and PS4 games and art supplies from a blue eyed American boy.

>>1523403
Also good luck on the degree nigger. I wish I had that piece of paper that entitled me to 6 figures in my mid 20s. Now I’m busting my ass to have health insurance and a few grand in a 401k. I’m trying to think of something else without sitting in a classroom for 4 years or taking a big pay cut so I can attempt to work my way up in another career.

>>1523422
Opaque doesn't work.

I'm asking for materials recommendations. I know the general idea. I don't know materials that are semi-opaque to UV.

>>1523428
Man, from my experiences of removing infra-red and making houlsehold shit filters, I`d start with plastics you have arround you. Like garbage bags and stuff, there are millions of kinds of them near us. Other than some factory made, special purpose UV difusers I don`t know anything else.
>>1523427
>Also good luck on the degree nigger.
top kek, Electrical engineer is hard, thank god I`m not indian. Everytime I google a highly pajeet theme for studying like motors I cringe with the results.
But anyway, all I got from university was autism and depression. Drop the trip pls.
>venezuela
TOP KEK

>>1523403
have you considered that being brazilian was a poor choice on your part?

>>1523398
And try to calculate the Thevenin resistance on each capacitor, as to get the low cut-off frequency? I'm trying, but I just end up with a circuit too complex to solve by hand.

>>1523435
The capacitors usually behave as short-circuits in the frequency you desire (the coupling ones), if you leave only the ones related to the filter it`ll simplify things.
>>1523434
yes,very.

>>1523435
>>1523438
For small signal analysis you have to ground the power supplies, leaving only the signal input. You don`t need the thevenin eq, just find the transfer function.

>>1523432
>drop the trip
Que?

Yeah engineering school was really fuckin gay. I got out of there and transferred to a college with a big teaching program and abour 2/3 girls and half of the dudes were fags and/or hipsters. That shit was nice... until you realize what a fucking waste it is paying $30k/yr to take liberal arts classes.

Oh and Venezuela is topkek. My main goal is to convince all of these Venezuelans that the Democrats want to turn the US into the same shithole that Venezuela is.

It sucks for them though because that country had a real strong middle class half a generation ago. So they grew up doing pretty well with middle class jobs, and then shit went to hell. Everything is imported so anything beyond the basics was impossible to afford. They had to sell their cars to eat. Last night I gave her mom a sterling silver necklace with some babby diamonds and she was looking at it all skeptical because for so many years over there any jewelry was likely fake unless you were a gov’t elite. Took her a couple minutes to understand that this is Trump’s America and we can buy .925 silver with tiny diamonds for less than one day’s wages.

>>1523415
>a resistor
kek
use a series connection and a constant-current converter
also consider SMT LEDs to maximize the distance between source and target, and/or "straw hat" LEDs designed for more even dispersion

>>1523403
they'll get theirs soon enough
so will you if you don't get Lula back sooner than later

>>1523427
to be honest with you. Look up the requirements for jobs for electrical engineering or Embedded programming jobs. Most people who apply to them have done very few projects, so if you can show them a bunch of projects working on your resume with like youtube or a website showcasing them, you'll be a major candidate over the majority.

>>1523499
I got a long way to go with that.

I’d like to snag a 3D printer one of these days too and start fucking with CAD some more.

Upside- at least I think I figured out how to charge muh 18650.

>>1523507
Hey, I'm the guy who helped you out a few threads back in return for dropping your trip. I also helped you in this thread with your power supply. Can you please drop the trip permanently? Otherwise I'm filtering you and never helping you again

>>1523539
>Hey, I'm the guy who helped you out a few threads back in return for dropping your trip.

how naive are you

>>1523304
>>1523353
I've got some dry-film photoresist on the way from china, along with some solder mask, carbide PCB drills, and some sort of transparency that's supposed to absorb inkjet ink particularly well. I'll be following Big Clive's method, using an old clothes' iron to press the photoresist onto the copper. I'll make my own lightbox too, once the IEC socket and capacitors arrive.
https://youtu.be/CrdheP3Dwyc
I'll run some tests with my inkjet to see how fine I can get the traces, but it looks like it will at least be able to handle 0.1mm details. Still, I think I'll increase my trace width to ~0.8mm or so.

>>1523324
It's just a TI LM339 I'll be using. The LMV is just the one they had already inside KiCAD. I made the pinout from scratch based off the LM339 datasheet so there shouldn't be any issues, though changing its name would be advised. I also realise as of writing this that I forgot the pullup resistors, so I guess that's more layout for me to do.

>>1523355
That crossover is intentional, all 4 IN- pins are commoned. Should I do something fancy with the crossover? Make it a swastika? Oh also I realise I can't solder both on the top and bottom copper on the inside of the IC socket. I might need to buy a row of machined pins and use those instead, lest I change the design.

>>1523539
That's not a trip, he's cool now. Hell, I've even posted and had people think I'm him. There are lulz without the trip, win-win in my book!

>>1523376
>no idea how/why it works
Click on 'Twin-T Wah Circuits' in the article you posted for a qualitative explanation. Deriving the transfer function of the complete network is mathematically intensive because your circuit has a gain-floor resistor (R5) in the R-C-R branch and the C-R-C part is asymmetric which prevents the usual reduction. See also www.drp.fmph.uniba.sk/ESM/twin.pdf

>>1523608
Thank you, that was super helpful. I don't know how I missed that on the article.

>>1523539
We had a deal, I drop it in this thread if I’m allowed to post Newports

So this lead acid doesn’t charge the same as the Li-Ions I see... 18650s go constant current first and then switch to constant voltage at 4.20V and the current dives off. Baby AGM is charging at constant voltage from the start. Any explanation?

Google had lots of people charging Li-Ions with bench power supplies but there wasn’t much good info for people doing the same with lead acid 12V batteries.

>>1523661

charging batteries is not fun, bepper. somebody has to do it, but it's not us.

FIND FUN STUFF AND WE'LL TALK

damn

>>1523665
I’m trying to figure out how exactly the power supply works before I fry it. And maybe understand what Ohm and Kirchoff mean.

Pic related came out of a cable signal booster that was making a god awful buzzing noise when it died. It stopped buzzing when I put a magnet up to the case and grabbed whatever was vibrating. But there are no motors or anythint inside so I’m curious what it was.

I wanted to power it up but not sure where exactly the + and - inputs are since the power adapter used a coax cord instead of a regular plug. The power input is just to the right of that vertical metal part.

Can anyone recommend me a good in-circuit ESR meter?

>>1523669
Sheeeeiiit my apologies

>>1523665
I’m trying to figure out how exactly the power supply works before I fry it. And maybe understand what Ohm and Kirchoff mean.

Pic related came out of a cable signal booster that was making a god awful buzzing noise when it died. It stopped buzzing when I put a magnet up to the case and grabbed whatever was vibrating. But there are no motors or anythint inside so I’m curious what it was.

I wanted to power it up but not sure where exactly the + and - inputs are since the power adapter used a coax cord instead of a regular plug. The power input is just to the right of that vertical metal part.

>>1523672
How about you fuck off and stop being an avatarfag instead you obnoxious cunt.

>>1523608
How can I cite this Anon in IEEE format? I'm thinking:

[1] Anonymous, "/ohm/ - electronics general, seasonal edition", /diy/, 4chan, 2018.

I specifically want to cite this part: "Deriving the transfer function of the complete network is mathematically intensive because your circuit has a gain-floor resistor (R5)"

>>1523674
It’s impossible to please some people

Also... those Wally World multimeters are still laying around a couple stores. Even it I have 3 meters, should I buy every one I see simply because they’re only $7? I saw one about a week ago and regret not getting it because I could’ve given it to boomer neighbor as an Xmas gift in hopes that he would electrocute himself.

>>1523576
consider Adam Seychelle's method of applying the photoresist. he used a wet method (boiling water heating the board to 90°C) and an acrylic/rubber squeegee that held the resist flat while applying. you could also try using a credit card the same way, for smaller boards
>4mil trace/space
that would be impressive, and would require a well-controlled developing/etching process. even without such fine tuning, I think you'd be perfectly fine at 0.25mm and still have room enough to route a trace between DIP pins and between 1206 SMT
>intentional
ohhhh, now I remember. nah it's good
a swastika would work too

Can someone find an equivalent of this on Mouser or DigiKey and explain how you found it? This top and bottom terminal packaging seems rare.

>>1523685
I've upped it to 0.5mm traces and added all of my pullups. I've got a felt vinyl squeegee thing on the way too.

>>1523686
first I looked up the datasheet at Sanyou, to get the pinout, pin spacings, and coil/contact ratings
http://www.sanyourelay.ca/public/products/pdf/sfk.pdf
then I went to Digi-Key's "power relays" section, set contact ratings between 20 and 100 amps, set coil voltage to 12VDC, and just looked at the results and the pretty pictures until I found a couple of little boxes with two Faston terminals on top for under $10 (took two or three pages)
then I looked at those datasheets and compared the pinouts and pin spacings on the board side, and both matched
https://www.digikey.com/product-detail/en/omron-electronics-inc-emc-div/G4A-1A-E-DC12/Z952-ND/202598 (if the contacts being inline is important)
https://www.digikey.com/product-detail/en/te-connectivity-potter-brumfield-relays/PCF-112D2M000/PCF-112D2M000-ND/1128659 (if not)

>>1523692
Please do not encourage this faggot.

>>1523693
>no trip
>no name
>no Newports
did I miss something?

>>1523693
Came here to say this >>1523697

>>1523692
So why such a huge contact rating range? Curious...

I don't know why the hell this didn't show up in my searches. I started out entering a reasonable range for each of the ratings that I actually understood from the data sheet and then started removing filters until there were too many pages. Are their filters just not to be trusted? This looks like it should have shown up right on the money, and it isn't chinkshit so I expect it to have been properly entered in their system...

>>1523706
the contact rating and the coil voltage are the primary parameters of a relay. I probably could have done just fine with 50A as a top limit. just didn't want to risk excluding something that might have been a functional but not pin-compatible substitute, if it came down to that
tbf their default sort order is basically random. I might have just gotten lucky. for that Sankyousomuch relay that's a step or two above chinkshit, sort by price could have saved a lot of scanning

>>1523676
Selective citations of short-lived web sources are frowned upon because the validity may quickly become unverifiable. Sources like archive.org fare better in this respect or in your case, warosu.org/diy/thread/S1522826#p1523608

>>1523457
Thanks for the SMT/strawhat idea
>series connection
The voltage needed for ~100 LEDs would be a bit impractical, no? Not to mention unsafe since I'm clueless about designing such a converter

Ok I'm a little stuck trying to import an image into KiCAD. I can successfully convert a PNG into a black+white .kicad_mod file, but when I try to import it into the KiCAD PCB layout editor it just shows up too large. I tried an online image resize software but it spat out a .mod file, that isn't recognised even when the name is changed. So I figured out it was spitting out an error inside the text file in html that I had to delete, but once deleted I got left with the same, massive image. Changing the PPI of the input image does nothing. And it's all still in silkscreen, not the layer I'd selected.

So basically I need to make a python script to go through an entire file, divide all the coordinates by a given value, and replace all instances of F.SilkS with F.Cu. The latter shouldn't be too difficult, but the former will definitely require some nasty reformatting.

>>1523786
In kicad 5, you can change the DPI when on the bmp import, and it will resize the image. What version of Kicad are you using?

>>1523796
5, lemme check that.

>>1523796
>>1523808
Oh, that's really handy. Sucks that MacOS dark mode turns half the font unreadable.

>>1523809
Success so clearly in view!

>>1523813
Completed test board. Willing to accept suggestions on other things to add (for the next month until my etching tools arrive), but I don't really have much variety of packages to try out. Suggestions for graphics to add on a board would also be appreciated.

>>1523818
check your spacing between traces, too. a poorly collimated UV light source can widen traces or spaces

>>1523822
Any ideas how to improve collimation outside of using a laser?

>>1523818
Could you please post your files somewhere? I want to try making this too

Anyone have recommendations for high quality helping hands?

>>1524036
I did not find them helpful and reduced mine to 1 clip and use that along with a small movable desk vice.
>>1519407

>>1523676
>4chan
cringe, get with the times grampa

is using parallel discrete diode bridges a bad idea for increasing capacity?

>>1523822
I'll print a positive and a negative of it, that should work for tuning. I don't think collimation will be a big deal as I'll be putting the printed transparency directly atop the photoresist with a piece of glass on top of that, but if it turns out to be an issue I can try to put each LED in a directional enclosure, perhaps made of the tubes within ~4cm of corrugated plastic.

>>1523920
It's kind of a pain to upload it anywhere, and I really would recommend learning how to make your own PCBs with KiCAD as that's just 2 days after me downloading the software. What I'm doing isn't too picky since I'll be hand-etching the lot, but if you intend to send it off to a PCB house you'll likely need to pay a lot more attention to the particulars.

>>1524057
The load will act like a current sink compared to the voltage drop of the diodes. As the diodes heat up their forward voltage will drop, and will pass more current. Have enough of a difference between the diodes and you'll get one bridge hitting thermal runaway and taking most of the current from the other bridge. That said, using two identical bridges should work fine unless you're really pushing their stats, but it's probably cheaper to just buy one higher-spec bridge rectifier. Possibly a Schottky bridge, I don't know.

>>1524084
good start. I'd add some actual lines tho, simulating the case of parallel busses running in close proximity (which may not etch the same as more isolated lines), such as Pic related. also consider some small TSSOP/QFP footprints, the usual cascade of SMD passive footprints
also add spurdo under the DIP footprint saying "Hei älä koskaan ikinä muutu :DD"

>>1524108
>spurdo
I'm more of a gondola guy but if it's Finnish then why not. Assuming the monospace font can handle those umlauts. I figure the positive and negative versions of that thin wedge will give me the kind of parallel line test values I'm looking for (as I know the wedge is 1mm by 50mm in size), though I could always alter the ancap wedge to split into multiple wedges after the graphic. I guess I'll change the size of the whole thing to fit int the TSSOPs and QFPs, then print it on ordinary paper to see what it's looking like.

>>1524112
shouldn't be a problem, at all

I have been working on circuits with PIC18F4550 and Zilog Z80.
I want to make an interface for them with C++ but first I need the driver.

I'm sorry for my English, I used the google translator.

>>1524175
I forgot it.
How to program drivers? (for W10/GNU-Linux)

>>1524176
why not libusb? is there some OS behavior in particular you seek to interact with?

>>1524176
also, if you make your device USB class-compliant, often you can use the OS's class driver and won't need to write one

So I bought a third party power supply for a sega genesis (9 volts) from a local retro gaming store that has turned out to be cheap chinese crap. There's 1.35 volts of voltage swing on the output. The 5 volt rail gets about 1 volt of swing. The swing is at 60Hz.

would adding a large shunt cap be enough to fix this, or would I want to make a LC low pass filter? Or should I troubleshoot at the switching supply itself?

Pic related are the readings from my scope

>>1524116
And here we are!
Comparing the QFN with pitch of 0.4mm from both sides you can definitely see that the ink flows out a little more than it's meant to, but as I'll be printing the one on the left, this will end up making my traces smaller, so I'll not have to worry about shorts appearing. Though running out of ink does become a problem. With the inkjet settings I chose, the dark black color is actually CMYK all atop one another, which should give enough density to block out UV. The 0.1mm trace at the top is obviously not going to work, but the 0.2mm looks fine so I may be able to get away with 0.15mm, and 0.1mm gaps between traces should also work fine going off the right-most image. The 0.5mm writing also worked in the right image and not the left, not that it matters too much. There was also some color bleeding with the magenta, so I guess I'll need to align my heads before trying again. The pictures look perfectly fine, perhaps with the exception of Lain's chin on the left image, which could be because of ink bleed or because the image I picked was a bit shit to begin with. I'll definitely stick to manually cropping.

To print the thing on the right I opened the SVG with my browser (inscape isn't working for me) and that worked fine, but to inverse the colors the only thing I could think of was to open it in GIMP, select all the white spots with the magic wand tool, and fill them in with black on a new layer. The process of rasterising and wanding left the thing with blurry outlines, which I believe is primarily the source of the short wedge on the left side (and the lack of holes in the DIP-14), so I'll have to figure out a better vector-based method of flipping colors, perhaps within KiCAD itself. But otherwise it worked pretty well, likely thanks to my selecting 1000ppi on the import screen.

I think I'll include a GIMP pepper on my next board.

>>1524190
is the output actually isolated? you could have your ground clip shorting to neutral. i'm not going to give it much thought but i guess that sort of lines up with the waveforms.

since it's third party i assume it's lightweight and therefore an smps. if so this isn't what flyback noise looks like so something is wrong. i'd open it up and if nothing obvious is wrong wrt contaminants or poor joints i'd reverse engineer the circuit.

>>1524201
>but to inverse the colors
Colors > Invert
>vector-based method of flipping colors
did you miss the "Negative plot" checkbox in the Plot dialog?
>inkscape isn't working
that's weird. I generally have no trouble opening KiCAD SVG plots in inkscape. kinda have to if I want to make use of le ebin press'n'peel toner transfer method (which, for me, is only good down to 0.3mm trace/space)

>>1524206
I'm on MacOS, where Inkscape requires Xquartz/X11, and I'm not too sure what I'm doing when it comes to that. I opened it up and it's not doing anything, but it worked after leaving it for a minute or so so I guess it's working now. Probably loading fonts for an age like GIMP does.

I was using the export to SVG menu, not the plot menu, but now I see that you can plot to SVG I guess I'll be doing that now. Plotting to PDF is also a potential idea, since I'll be able to print it via the Preview application.

Now if only someone knew how to turn a bitmap into a schematic directly as a copper or mask layer instead of only being able to select "silk screen" and "User" layers and having to open it in a text editor. That was one positive that the online object resize site did have.

i'm making a PCB that uses an AVR and can accept different modules. say for instance a button matrix that act as inputs for the AVR , or some LEDs or something. those modules all come with 4 wires : 5v , GRD , and two DATA lines. problem is , how do i let the AVR know what kind of module was just plugged in ?
the main PCB already has an init function, so i could send a signal to all the modules during that phase and wait for an answer. problem is, i have no clue as to how to implement that answer. do i just stick a microcontroller on all the modules and program it so that it send a serial through one of the data line ? do i need to synchronise both micro controllers to the same clock ?

ideally , what i want is pressing the start button and seeing that there's 2 button module on port 1 and 3 and one LED module on port 2 without me manually telling the AVR that.

i've never done anything close to communication between two micro controllers so i'm a bit lost, but i also don't have a lot of experience designing circuits so there might be a better way to handle signatures because as far as i know using a microcontroller for that is a bit overkill.

>>1524190
>>1524202
The main chip is a DK-1203.
I reverse engineered it a bit, found that it's pretty closely matches the example circuit in the data sheet I found ("FLYBACK TYPE SWITCH MODE POWER SUPPLY"), pic related.

D5 measures around 700 ohms in the direction that current can flow, so I suspect it's bad. Unfortunately there aren't any markings. I'll get a replacement and (try to) solder it on.

Can anyone help me identify the package size?
It's a 2 pin diode that measures 2.85x1.8mm, 3.85x1.8mm including the leads

>>1524232
i also forgot to mention, the data lines are used by the modules so i can't just use those two wires to replace most of the module's logic with an I2C slave microcontroller.

>>1524232
I would use some transistors to create a counter. Hit DATA1 with a high voltage and the PCB returns DATA2 high. Pull DATA1 low. Setting DATA1 high for the second time, if the type of module you're polling has an ID of 1, it will return DATA2 low, because this is the second pulse. if the ID is set to 2 or higher, then it will return DATA2 high. If DATA2 returns low, then you've reached the end of the ID and can begin normal function. Tge module then permanently latches the +5v line high to the rest of the module circuitry until it's powered off. As long as DATA2 keeps returning high, you increment the counter on the AVR and keep going. A discrete counter shouldn't be hard. 8 bits should give you plenty of IDs.

>>1524240
>a counter
Isn't that a lot of transistors?

>>1524240
>The module then permanently latches the transistors to the +5v line as well as DATA1 and DATA2 open to the rest of the circuitry of the module
Fixed

>the data lines are used by the modules so i can't just use those two wires to replace most of the module's logic with an I2C slave microcontroller.
Transistors fix this.

>>1524241
Not a terrible amount, but I'm p sure there are ICs for this.

>>1524240
>>1524242
This is pretty clever ! Thank you !

>>1524190
If the supply isn't loaded there is only going to be basic over-volt protection. Make a dummy load at the current rating and see how it looks.

P.s. you don't have a normal supply to compare. Since old console were based on wall-warts, they would've had over-volt by design. It is literally impossible to do a supply without overvolt when you're doing a transformer+rectifier+capacitor, unless you're also okay with undervolting, which can be more dangerous.

>>1524242
4 NANDs per JK FF, perhaps you can drop that down to 3 for a T FF, and assuming shitty RTL it's only one transistor per gate, so you've got a minimum of 3*8 = 24 transistors and at least twice that in resistors (48), that's basically an entire 64-row breadboard of components, half if you're really cramming. A complete waste of time if you ask me.

>>1524258
okay so do it with a 4c PIC then fuck it ¯\_(ツ)_/¯

>>1524258
Couldn't he have fewer using a ripple-carry adder?

>>1524259
this anon gets it. that larping faggot and his NAND gates, I swear he needs to join the 21st century already

>>1524232
anyway, many of the smaller MCUs even have an internal, vaguely precise RC system clock on the order of 8MHz. synchronized clocks shouldn't be necessary, just use one line as conf clock and one as conf data, and bit-bang slowly. use the slave micro to control a bus switch e.g. 74LVC2G66, if you're not using the micro to control the peripheral itself
Pic related for example: c and d are both open-drain, pulled up by host. times are mostly determined by the code on the slave MCU and how fast it can run. at least t1 after power comes up (a few µs), the master starts pulling c low for at least t2a and less than t3, and the slave puts a new bit on d within t2. the master raises c and samples the bit, saving it somewhere and waiting at least t2a. after however many bytes the master wants, if the master decides to ignore the peripheral, it just leaves c high and forgets about it. if it wants to activate the peripheral, it pulls c low for more than t3, waits until t4, then sets up lines for the peripheral. the slave sees the low for more than t3, so gets off the lines within t4 and enables the peripheral function before t5, then goes to sleep until reset
... shit, reset. what are you doing about warm boots?

>>1524262
>this anon gets it. that larping faggot and his NAND gates, I swear he needs to join the 21st century already
Same guy btw

>>1524260
To be honest I'd prefer a simple 8-bit counter IC, like a 74269, 74461, etc.

>>1524260
I think you'd need 2 of these.

>>1524263
one half of you needs to kys, the other can stay

>>1524262
>pic related
>doesn't pic related
stop throwing errors, 4channel

>>1524084
I already know how to use KiCAD, I just don't want to be assed to make my own benchmark board

>>1524262
i was fooling around on EasyEDA and i have to admit your solution is a lot more compact and less expensive overall, since those little NAND ICs do add up quickly.

i won't do any live module connection because during operation, the host PCB will be hidden away. Thanks !

>>1524269
is there any condition under which the host might reset (watchdog etc.) but the power supply won't drop to reset the peripheral's ID chips?

>>1524271
no, in the I/O section of the main board there will be the possibility to disable each connectors so that only the ones recognized by the main microcontroller get power. I can easily turn on and off the supply during the reboot procedure of the host.

How do I stabilize (to a certain degree) Germanium transistors against ambient temperature changes? (I.E. stabilize the bias point).

>>1524292
>Germanium transistors
Why? Not to say that I don't approve of your use of obscure components, that is. But they had better be in an obscure tin-can package like a TO-99. It will depend on your use case. If it's for an amplifier, I believe there are somewhat common topologies for amplifiers that use two transistors (possibly one NPN and one PNP) to compensate for thermal drift. I don't pretend to know what "bias point" means, but I think that's what you're looking for.

>>1524294
I know about those, also the reverse-biased diode trick, but they're not that good for my use case.
I need it to build distortion/fuzz effects for musical instruments, and all that clipping just doesn't go that well with those solution.
Hence why I'm looking for some form of more sophisticated bias compensation and/or just thermally stabilizing the transistors somehow.

>>1524295
Put them on a thermal mass with a heater and feedback circuitry to keep it at around 50°C, increase the voltage rails and/or decrease the amplitude of the signal, that's all I can think of. If clipping is the result of the problem, then the latter suggestion could prevent the clipping itself, but it won't prevent your gain from drifting. The former suggestion would be similar to an oven oscillator (OCXO) I saw in an EEVblog video recently. Of course, if you did put the junction in an oven, you'd need to change the circuitry since I imagine the hFE would be much higher than it was at room temperature.

Got a datasheet? I'd quite like to see how much thermal drift it typically has.

>>1524295
You just have to find a topology that doesn`t relly on hfe that much ;)

>>1524299
I might have not explained it correctly, I want it to clip, just at a certain bias point (which I want to be stable).

>>1524301
Hmmm, that might be an interesting pathway to explore, as it is quite different from the "usual approach" in effect design.

BTW, how bad of an idea is it to have a microcontroller monitor the temperature and current then automatically adjust the bias voltage - thus effectively stabilizing the bias point? Could it be done with (reasonably) cheep-ish parts?

>>1524305
Which _specific_ circuit are you talking about?

>>1524305
>Hmmm, that might be an interesting pathway to explore
m8, that`s literally the first thing you should do before thinking about diodes and shit when dealing with bjt, a design that relies on Hfe, is a bad one.See for example when you add a degeneration resistor to the emitter of your usual amplifier, it diminishes the dependance on Beta.

If I have some 24v pilot lights I want to have patterns show on them (blinking/lighting in order/etc) via arduino, is a mosfet or relay the best way to go about doing this?

>>1524314
mosfet/relay not mosfet or relay*

>>1524314
open them up, replace bulb with led.

>>1524314
open them up, replace led inside with regular led or remove the voltage/current reducing components connected to led inside..

>>1524317

Anyone know what this might be?

>>1524308
>>1524311
Thing is I'm dealing with common-emitter amplifiers for the most part (such as a fuzz face and the like).
They do have a certain tonal character that I'm after, but the problem is the (somewhat notorious) instability.

>>1524332
AMD programmable array logic
did you even google

>>1524333
>not using JFETs for your musical distortion
y tho

>>1524342
>AMD programmable array logic

thanks. is there a link to the pdf?

>>1524342
Because I want a the harsher sound of BJTs, also why would I want to emulate tubes with JFETs if I can get my hands on some proper vacuum tubes?

>>1524350
so, no, you didn't google

>>1524362
fair enough. try setting the operating point with a current source maybe?

>>1524317
Could not be further away from giving a useful answer...

>>1524365
One of the things I thought about (but hadn't much chance to test) was using a vactrol to try to control the bias point, maybe by using it in a voltage divider topology?
Because that way when the transistor starts leaking more it'll move the bias voltage.

>>1524314
mosfets (or bjts) are much better for this. you can get ULN2004s/3s that are just 7 bjts in one package that should make your life easier. they're dirt cheap.

A tiny part of my nixie stays lit even when my transistor is not supposed to be conducting.

It happens on every digit, always in some corner of the nixie wire.
Any idea why this is?

Anode is connected to some ~150V and cathode is connected to the ground through an NPN bjt transistor (BC547B), controlled by gpio from my atmega2560. Tried different supply voltages for the nixie too.

I've also tried different values of anode resistors (currently using 10k), the same thing happens

Is it due to the bjt leaking current or something other i can fix?

4.8mm between prongs, each prong ~0.7mm thick. Does this shit use some wacky current or what ? Is got my first meter, it claims that the bulb when connected gets current with frequency of ~30k Hz and is unable to tell me the voltage or amperage.
The meter is a clamp meter (Uni-T UT213C) , but for both AC and DC. I presume bulb uses AC but I tried DC too. Meter is even supposed to be true RMS I just can't fathom why it doesn't tell 12V as is written on the lamp and some of the bulbs. Meter isn't fully broken, reads proper voltage from the wall, proper amperage when I'm charging the phone, purpose voltage from batteries. I do know that for clamp reading wires need to be separated so I'm not fucking this up.

>>1524414
>BC547
Data sheet, max Vce?

>>1524414
your transistor isn't rated for 150V so it's basically acting as a 50V zener diode when off.

>>1524333
The traditional fuzz face always had a dc feedback and was stable. I've seen other circuits though where the 'fuzz' transistor was exclusively biased by its own Icb leakage. This is not kosher, no matter what odd Ge-diode you use.

>>1524421
...please read on how to properly use a multimeter. And how a light bulb should read.

>>1524470
you're a basic fuzz face

>>1524526
I'm certainly using the clamp properly (clamp closed all the way, only hot in the very middle of the clamp, neutral outside the clamp, yeah I know it could be the other way round too), I'm still not getting any amperage reading.

>>1524370
>Vactrols
I thought you were worried about instability...
Vactrols vary so widely from one to the next that you're going to have even worse process control than these original distortion boxes.

>>1524382
ty senpai

>>1524305
Look into how temperature compensated oscillators work if you don't want to build an oven, though I doubt their workings will be too different to those compensated topologies you were dismissing earlier. To simply shift the bias voltage as a function of temperature you could probably do that just fine with an analog temperature sensor, some op-amps, and a handful of passives, if not a second germanium transistor with a current source instead of the temperature sensor.

>>1524414
>BC547
Kek, use a properly rated transistor next time.

>>1524421
Just use the resistance mode bro. Once you know its resistance you can calculate how much power it draws at different voltages, and hence figure out that yes, it will blow up when you put mains across it.

Can someone tell me if this highside current monitor works? I will only be able to build something next year, so I want to have something already in mind to make when it gets there. There is a voltage divider because VCC is like 50V, and I need to step it down so the opamp can handle it.
>inb4 use a monitor IC
Can`t find in local stores, their IC inventories are barely organized by name, no search by function and guys working there don`t know electronics so if I ask for a highside current monitor they`ll look at me like I`m a alien. And on other online stores I can only find arduino shields to monitor current and they are too big.
>inb4 diference amplifier using 3 opamps or one with resistances
Rsense is 0.1 Ohm or less, the current will go from mA to 5 A, the voltage drop across it in much of the range will be the same size as the error I`ll get from most opamps I have here, and I can`t find instrumentation ones.

>>1524621
What's that BJT doing? Isn't that positive feedback?

>>1524621
Can't read your scribble but I think the eqn for Vo has a Vs term in it. That would be bad because Vs depends on the load and on the line voltage.

>>1522826
Slightly out-of-band for this thread, and I apologize in advance for that, but I'm rebuiling my resume, and this time I would like to seed it with as many keywords and key-phrases as I can (and still be truthful) so that when employers use the search engines all the job websites have, I'll get more hits and hopefully better choices. So does someone have a source of keywords and phrases I could use as a resource? Thanks in advance, it's much appreciated. :-)

>>1524621
You really need an IA with gain for the common-mode input voltage range. If accuracy isn't much of an issue then a single op-amp IA will work, but then the op-amp itself has to have a high Vcm, so it's more practical to use a 3 op-amp IA design.

>>1524688

>>1524693
???????????

>>1524689
>single op-amp IA
That's new, how does the circuit look like?

>>1524621
there will be an error in Vo inversely proportional to the beta of the transistor
consider running the single op amp with V+ = Vsupply and V- supplied by a zener, not necessarily related to anything else. you'll probably have to use a TL081 or similar whose Vcm range includes V+. Pic related, simulated V+/V- = +50/+30, Ioutput=1mA/A

terrible Vid related: me measuring a 1.0µH ±1% inductor with a homebuilt LC meter

>>1524743
I don't get why people put so much time into making circuits on protoboards and homemade boards like this. If you're not from a third world country, you can have China make a dozen PCBs for you in a week for $20. Designing it takes less time than figuring out how to put together some hack like this, the end product looks and works better, and you get experience that's actually useful in the industry. What's the appeal of doing it like this?

>>1524744
impatience mainly
respecting the homebrew and self-educational culture of amateur radio
>a generic array of 2x3 pogo pins at 0.100" spacing + "Dupont" connectors = shit's cash
>VSSOPs + no soldermask = tears
>polyimide tape alone is not enough to heat-shield components
>1mm spacing between component body edges is comfy for old farts
>as constructed, this thing drifts like your father
also
>implying I didn't lay that out in KiCAD
>implying a week is anything close to a guarantee especially during the "harvest season for merchants", Chinese New Year, or tariff uncertainty
>implying the USA is not a third world country
>implying you are going to be any goddamn use during the zombie apocalypse

>>1524535
Anon. I mean are you actually measuring the light bulb in a circuit. A bare light bulb does not generate electricity, and should read a dead short or very low resistance.

>>1524743
How'd you do it? I wanted to make one using just a simple resonator for the L, though I'd have to do calculations. If I could measure analog fast enough (or use a large capacitor+slow measure) I could digitize it.

Another useful circuit for inductors would be a saturation tester.

>>1524713
>>1524689
>>1524683
>>1524628
thanks guys

>>1524744
Most perf board circuits I do are layed out in kicad.. They always turn up diferent but idk, it gives me an idea of how the circuit blocks wil be.
>boss, the supplier for our stand says it`ll be delayed, ETA is until the con is over
>can`t you just solder one?
>no, it takes 2 weeks for it to be mailed from china, sorry boss

>>1524767
Personally I'd rather etch a board with permanent marker than try to make something work with perfboard. And since etching in itself is a time consuming and problem-riddled process, I basically never get around to doing anything beyond breadboard prototyping. But hopefully that will change when I get my photoresist in the mail. Chinkshit is my lifeline.

>>1524767
Eh, that argument gets weaker if you need to order anything else. I've had an application I needed specific kinds of mosfets or op-amps. If I worked in a real setting I'd probably have a lot of useful shit from the get-go but what can you do.

>>1524769
>Personally I'd rather etch a board with permanent marker
Toner transfer never worked for me, so I do that too.. ICs make everything hard kek.
>>1524770
>you get experience that's actually useful in the industry.
It was against this, in both my internships (EE student fag here) making shit that worked, in the middle of nowhere/delayed supplies kinda made bosses happy.

>>1524771
Now that I use nail-polish, the hard part is actually drilling things straight, since I've got a 1mm drill bit in a drill-press chuck that's only rated from 1.5mm to 12mm or so. Being able to print out boards with holes in the copper for the bit to seat itself in should help greatly.

As a poor student getting a basically useless degree, are there jobs you can get by having a thorough understanding of electronics and PCB manufacture?

>>1524761
here's the oscillator I used, based on a circuit by Phil Rice VK3BHR, but with a slower comparator (I didn't have the original LM311 on hand)
with SW1 in shown position, an external inductor Lx is connected in series with (and therefore adds to) L2-L4, forming a tank circuit with C6 (a 3300pF 2% film cap I had laying around from another project). U3B provides positive feedback, coupled bidirectionally through C5+C12. U3A is a voltage shifter feeding the oscillations to the micro. the micro counts the frequency using its 16-bit timer and a busy delay loop, then calculates
>L=1/(4*%pi^2*C*f^2)
which is then formatted and printed to the LCD. the Zero button saves the reading and simply subtracts that saved reading from future readings. normally you short the test leads together for a 0-valued Lx then Zero, then connect to the inductor to test
capacitance measurement works similarly, with Cx added in parallel to C6 and the internal inductor, and basically the same math with L and C exchanged, and you leave leads open to zero
>though I'd have to do calculations
computer algebra systems are helpful. e.g. wxMaxima was told that
>f_LC(l,c):=1/(2*%pi*sqrt(l*c));
then told to
>solve(f_LC(L,C) = f,L);
and did so instantly
>saturation tester
hmm intredasting... my main purpose for it is checking hand-wound RF toroids for correct value. once this LC meter achieves its final form (a factory-made proto board and possibly even a box) I'm torn between building a small toroid winder or a scalar network analyzer

>>1524771
that yellow press'n'peel paper works sort of well for me. QFNs are naga happen but SOICs and 0603 (inch) are fine. I sure do miss the solder mask though

>>1524770
parts stocks are built item by item, I've found. ofc you aren't gonna stock shit you don't (yet) know you need. I will say those resistor/capacitor/inductor cut tape assortments on ali are nice and pretty convenient for the impulse build

>>1524775
>As a poor student getting a basically useless degree, are there jobs you can get by having a thorough understanding of electronics and PCB manufacture?
I got internships at a place that made electric motors, most emergency pcbs I did where THICC test setups for big IGBTs.

>>1524781
>made
>tfw very little manufacturing of anything is done in my country

That IGBT stuff sounds fascinating though, were they HV BLDC drivers or VFD-tier inverters for induction motors?

>>1524783
>>tfw very little manufacturing of anything is done in my country
M8, I`m in brazil you don`t get worse than that I think.

Uhm.. I think one of them was for a SMPS they were going to use somewhere else and the other was for motor drivers, one of them was for a homopolar generator that gave very HIGH current (It was pretty cool, never thought I`d see one of those), and the last one before I got out was for controling highcurrent in something they put in a vacumn chamber, no idea what it was just did what they said.

>>1524784
I bet it's worse in Ireland than in Brazil. Highly unionised = high wages = outsourcing manufacture to places where a dollar goes a lot further. At least labour in Brazil is cheap enough to make manufacturing viable. Not that I'm in Ireland or anything, but either way there's a shortage of jobs.

>>1524775
you need the bits with the larger shanks
https://www.aliexpress.com/item/35-10pcs-1-0mm-Carbide-Micro-Benis-Drill-Benis-CNC-Jewelry-Rotary-Tool-For-PC-Boards/32820312970.html
since industrial board makers' CNC drills can't tolerate much wear, there's a fairly steady supply of used/refurbished bits still quite suitable for manual usage. I hope your drill press has really tight runout

>>1524786
>>>/pol/

>>1524784
and they were to be used for very little time, just until the real deal arrived or got made properly. No way a diy board is going to take 500A
>>1524786
>At least labour in Brazil is cheap enough to make manufacturing viable.
``no``
>. Highly unionised = high wages
my man, until last year there was a national fund that simply took 1 day worth of wages of EVERYONE every year, and distributed to unions, regardless if you liked them, joined them or suported them. Also even today making deals with your boss is kinda legally worthless because ``the unions and government protect you from yourself```. Unions here is cancer and gay and they are thousands.

I worked in a WEG subsidiary, WEG and embraer and companies like them fir a niche brazilians companies manage survive that encompass cheap-ish high quality rugged stuff.

>>1524787
Got some on the way, don't worry. But I wish they just had 5 0.8mm and 5 0.6mm or so, having one of each size from 0.3mm to 1.2mm is a bit stupid.

>>1524789
I take that back, I am sorry for you being in Brazil.

>>1524780
For calculations, they're simple, since the oscillation is just an LC(R) expression, I just haven't done it. If you can effectively measure the signal, you can just provide a brief pulse and watch the ringing (or charge a cap and watch the ringing). I didn't think too much about R and C, with a carefully chosen value, or switchable value, you could set the ringing to occur in an audible range, which would be much easier than explicit peak detector/analog sensor. R would just be a drain but could also slow the ringing.

A saturation tester might be very useful, since you can't normally observe the current limit of an inductor before it, you know, saturates.

Related, there are some circuits for voltage-controlled capacitors in research papers.

On parts, there are things I occasionally see that I wish I knew about before. One set of op-amps I had an opportunity to get, they had a very high bandwidth, though they were in the tiny-ass surface mount format.

>>1524414
>BC547
>Data sheet, max Vce?
>>1524453
>your transistor isn't rated for 150V so it's basically acting as a 50V zener diode when off.
>>1524585
>Kek, use a properly rated transistor next time.
But nixie driver circuits (namely SN74141) is rated for 55V, same as this transistor. What gives?

Anyone know what kind of connector this power cable is, furthest pins are 13mm apart

>>1524821
need more to go on than that.
https://www.jst.co.uk/products.php?cat=31&nm=

>>1524830
>https://www.jst.co.uk/products.php?cat=31&nm=

it is a tein EDFC controller i bought secondhand, got a lot of motors and cables but did not include power cable, have found out online which pins are which but cant find anywhere where i can buy a new power cable. as tein is a japanese company its very difficult to order through them if it even is possible, no suppliers anywhere in the euro region as far as i can tell. so i decided to just buy a universal connector which would fit

>>1524761
Yeah it's producing light. I guess maybe it gets AC and I still didn't check out if the AC part of the clamp meter works properly, but how come that if I touch the meter to the metal (conductive metal, resistance of bulb when removed is same and very low doesn't matter what part of the prong I touch) prongs of almost fully inserted, brightly shining bulb I get no voltage reading, but meter appears to read both AC and DC vintage properly when checking it with batteries and outlet.

I want to make a batch of products that rely on a 12au7s tube.
Will these be scarce or hard to source for a reasonable price in the future?

>>1524806
see dave jones' nixie design videos. with an underrated transistor like that you're basically modulating the voltage across the tube from 150V (vh) to 100V (vl). you can calibrate it such that the current at vl is low enough that no part of the display illuminates. this is done by trimming the supply voltage and limiting resistor. you don't seem to have read up on what you're doing though so for christ's sake just get a 150v rated bjt.

>>1523249
ey, I think i bought the same kit. I've been having fun blinking leds for now, I'll try using the 8x8 matrix next.

I got a 4x4 pad too, I think I'm going to make a game of whack a mole.

>>1524884
define "future"
the tales of Curtis Electromusic are instructive. after Mr. Curtis died, one synth manufacturer made a deal with the widow to get a special run or two of some of the chips for his own production. the only other place to get those chips for service etc. was on the eBay aftermarket for up to $80 each. a few years later, a Latvian company cloned the more sought-after chips for $15 retail, and the widow did another run for public sale
the point: manufacturers for boutique items like triodes come and go. if you want to guarantee availability, best is to stock them yourself

>>1524840
You could always check whatever extension wires you're using. Could be something dumb like a line break.

OK a 400nm LED looks slightly purple to human eyes, but that's not because it's a mix of blue and red light, correct? If that's the case, then presumably they only trigger the blue sensors in a digital camera, so why do pictures of these LEDs appear purple? This can only be because my monitor is showing me both blue and red light, correct? Does the color processing of the camera or computer decide to make it purple instead of blue?

>>1525174
Might be there's an all-spectrum "brightness" pixel, along with individual colors. If something was low "blue" but high brightness, it would come up purple.

I might just be flat wrong, for a surprising reason:

https://www.quora.com/Sensors-Why-does-infrared-appear-purple-in-digital-cameras

According to that chart, there's red photosensors that pick up at ~400nm, and maybe at like 380, are the same sensitivity as blue is.

>>1525207
Ah, yes it looks like both camera sensors and human eye receptors have a spike in the red sensor higher up. Interesting how that results in purple appearing to have a higher frequency than blue, and the mythical color wheel.

>>1522826
why isn't there element14 on the YouTube channel list

>>1525238
They don't look that bad, about Make tier.

Can you use a manicure sterilization oven thingy for SMD reflow soldering? They're really cheap and go up to 200ºC, if I use an arduino + thermocouple... could I use it to solder PCB's?

>>1525174
>so why do pictures of these LEDs appear purple?
Because cameras don't filter out UV and that's why they look purple. This is also why LED lighting is bad.

>>1524175
PIC18F4550 has USB, so you can use it as a HID and not use any driver at all

>>1523576
>I've got some dry-film photoresist on the way from china
Chinese dry film is usually garbage. The best one for DIY usage is Kolon Accuimage from South Korea. It works even when laminated cold.

>>1525248
But purple on a computer screen is a combination of red light and blue light. Cameras and screens and the human eye all work by having a seperate sensor for red, blue, and green light. Therefore, unless the camera automatically processes a strong blue light with very little red or green as a purple and so adds its own red, or its red light sensors are sensitive to light around 400nm. Turns out, it's the latter, and the same applies to human eyes. And as far as producing light in a single wavelength goes, LEDs are the king. Perhaps behind dispersed laser light.

>>1525261
Where do you recommend buying that from?

>>1525245
>200°C
kek

>>1525238
>>1525241
whom, if any, should they replace? AfroTechMods might not make the cut

>>1525255
how garbage is it, tho?

>>1525261
>Cameras and screens and the human eye all work by having a seperate sensor for red, blue, and green light.

Human eye has different proteins that are bleached by different wave lengths, cameras work differently. They have just an array of the same kind of sensor with a color filter array called Bayer Filter. The image captured is a mosaic, and then it's averaged to give RGB values.

There are UV filters just to filter out the UV light from the sky, which causes a blueish haze on the pictures. The bayer filter doesn't filter out UV, and it's within the CMOS detection curve for both blue and red light. That's why they look purple on pictures.

>>1525265
>kek
It has a 70w mica heater so I think it can get a lot higher than that, but... how fast? Reflow solder takes about 4 minutes in total

>>1525265
>how garbage is it, tho?
Wavy, spots, comes with areas already exposed and so on. Use it for very rough circuits only, with thick traces, or else you're gonna have a bad time.

>>1525268
All the PCBs I've seen Clive make with the stuff look perfectly fine, sure you didn't just get a bad batch?

For the oven I'd want to set it up with a temperature controller such that it follows those recommended soldering heat curves you see on datasheets.

>>1525270
Clive only ever showed boards with 32 mil and above traces, go below 16 and it becomes really difficult to get anything decent. I stopped home etching just because it's a major waste of time, btw.

>>1525270
>For the oven I'd want to set it up with a temperature controller such that it follows those recommended soldering heat curves you see on datasheets.

From my original post:
>if I use an arduino + thermocouple... could I use it to solder PCB's?

Also, since JLCPCB is going to do pcb assembly starting next year I really doubt it's worth trying. Maybe it is if it turns out to be very expensive per board

>>1525272
>I really doubt it's worth trying

I think I may try reflow soldering using a sandwich and some botched boards I have

>>1525272
it might be good for vapor phase soldering, if you can find a way to keep the heat transfer fluid in and to control the board lowering/lifting

>>1525277
now that's what I call breadboarding

>>1525272
I won't be doing much, if any, SMT work, so it shouldn't prove an issue. Conversely, missing out on the plated through-holes will hurt a little. I may try to solder/rivet a piece of 1mm copper wire in a 1mm through-hole, then drill it out to 0.6mm if I can get a good enough clamping solution. Chemical means aren't really terribly available to me.

Anyone got any good resources for wiring up a cnc machine?

How do I into electronics/PCB design? I need to design a microcontroller-enabled peripheral with some onboard power electronics for switching cells in a battery pack on and off depending on load.

t. naive mechanical engineer

>>1525272
>JLCPCB is going to do pcb assembly starting next year

That's big. Got a source?

Do Hantek scopes have any actual problems that would interfere with a novice hobbyist? $230 for a two channel 100MHz scope or $280 for that + integrated signal generator is really tempting.

>>1525310
Simulator, Multisim is business standard, I haven't tried KiCAD for board layout but I've been meaning to. Controlling a motor is tarduino-tier.

>>1525310
PCB design: watch an introductory video on KiCAD, it's good shit.
Electronics: uuh, get a breadboard and a DMM and start messing about? Being able to use a simulator is also handy, but I wouldn't use it as a replacement for a breadboard. Solid-core wire > jumper wires.

>>1525321
>I haven't tried KiCAD for board layout but I've been meaning to
Go for it man, it's an incredible tool.

>>1525313
you read chinkrunes? they already offer it on their China-facing site
https://www.sz-jlc.com/portal/newV2/smt.jsp
highlights courtesy of google translate:
>Patch mode innovation
>This kind of hundred-piece welding has always been a chicken rib in the industry.
>Many patch factories are unwilling or refusing to provide such services for various reasons.
>Jiali Chuang has taken a different approach and invested heavily in creative use without changing the feeder or
>Only change the individual Feida mode to solve the SMT sample production problem

Hi! I'm making a MIDI controller using arduino that is going to have all the components (knobs, buttons) connected to a PCB on the front panel. The arduino is on the bottom panel. What's the best way to connect all these components to pins on the arduino? I would ideally love to have a single clicky cable that just snaps in the arduino. Can you reccomend an arduino shield or a different way to do this?

>>1525361
>Hi!

Hello there! There's an arduino thread in the catalog. /ohm/ is for actual electronic work.

>>1525365
my bad

Hello there, planning to make a simple first person shooter in real life. The gun won't shoot bullets but send a signal and if a person gets "hit" a loud alarm beeps.
I'm sure these already exist. My questions are
What are they called?
Can I use door magnetic sensors to implement this?
Thanks

>>1525368
To be more clear I want to make the "gun" that sends the signals and put sensors in important parts of the players' body

I've got a small LCD display from a kit but it has some kind of controller soldered on, it has 4 pins (gnd, vcc, sda, scl). Any clue of what it is and where I can find some info?

>>1525310
i can recommend altium designer. i think its interface is the most developed. electrical engineers are used to working with winforms trash that hasn't been rewritten since 1996. i'm also a meche though and when you're used to modern UIs from cad software like NX it's hard to regress.

>>1525369
laser tag?

>>1525381
Ah didn't know it was that popular

Can I use electrolytic capacitors for pic related oscillator? Would polarity of capacitor be an issue here?

>>1525400
(+) to collector side

>>1525426
Will that work for the whole cycle though?

>>1525374
> modern UIs from cad software like NX
That used to be modern in 2006

>>1525313
>That's big. Got a source?
I asked on their facebook page

When the chinese say they're preparing to launch something, it means it's already done. They already offer it in China.

>>1525455
UI design peaked at ribbons

>>1525371
Look for a serial number and google it.

SDA/SCL is I2C. If you have a common dev board like an Arduino then there's libraries to drive I2C devices.

>>1525434
Yes, it will. If you want to know why, use Tesla's method: Close your eyes, put yourself in the position of one capacitor and observe what happens to you during a cycle in slow motion. You will find that most of the time you're charged correctly and the reversal is very short-lived and of a very low voltage. No problem for electrolytics.

Dear /ohm/, I am stuck: please help.

I am making a remote controlled VTOL aircraft, but I am running into real issues with brushless motor efficiency at low duty cycles, and I'm not sure what to do about it. The problem is that I need about 40x the thrust during takeoff that I do in cruise, but as it would turn out, running the motors at about 2% duty during normal flight is horrible for efficiency and battery life. The whole point of this aircraft was to take off vertically and then fly very efficiently in order to achieve long flight times, but that's turning out to be somewhat difficult.

How do I throttle down my motors that much while still maintaining at least 80% electrical efficiency? I'm looking at buck converters and also possibly putting a bunch of FETs in the battery pack to dynamically rewire it in series/parallel, but I still haven't come up with a satisfactory solution. What do you recommend, /ohm/?

I would ask /rcg/, but they don't know anything since they just buy prebuilt drones.

>>1525539
Use two motors

>>1525371

also there are sketches that scan for connected I2C devices. and they are all (almost) based on the HD44780 driver.

What's the best way to learn how to use an oscilloscope? I pulled the trigger a few months ago when there was a sale, and now I have a Rigol DS1054Z laying around. I wanted to get a scope thinking I'd find a use for it if I had one, but problem is that I don't really know how to get started.

>>1525539
PULSE THE MOTORS AT 100%
YES THE BLADES WILL GENERATE DRAG, GET OVER IT

>>1525567
eevblog.com/forum/beginners/ has an Oscilloscope training class

>>1525567

there's a button on the lower left that is labeled GOOGLE.

>>1524470
It does drift with temperature, unless you *extremely carefully* pick the transistors...
The old Dunlop reissues even went as far as using thermistors (which are an inconvenient solution for me).

>>1524537
Even if I carefully match the LED and photocell?

>>1524585
Analog temperature sensors and opamps sound like a cool idea, I'll just have to check the costs.
Thanks for the suggestion!

>>1525310
>>1525374
Altium is nice but costs thousands of dollars. If you pirate it you'll have a hard time with the libraries, and with explaining in interviews how you have the software for personal use. I second KiCAD.

I've been learning theory and something I don't quite get is how to use a higher-voltage voltage source to power something. I know intuitively I would need to add a resistor to drop the voltage down, but how would one calculate it?

For example let's say I want to power a flashlight that uses 4.5 Volts batteries as source, but I have a 9v battery rated at the same current.(theoretically) I would measure the resistance of the whole flashlight and then make a voltage divider to make sure there is a 1.0v drop across the resistance I will add.
I feel there's a quicker way of doing so or alternate way of doing so instead of measuring the equivalent resistance. I know what Thevenin's theorem is and I feel it kinda applies here, but I'm not sure how to use it in this scenario.

>>1525628
oops meant to say 4.5V drop not 1.0V

I know that with this method I would be wasting power, but that's besides the point.

>>1525628
You should only use a resistor to lower your voltage if your load is purely resistive. E.g. in your flashlight example it's fine if it's an incandescent bulb. But if it's LEDs, there's probably some constant current drive circuitry that your resistor isn't going to play nicely with. You want a linear regulator if you prioritize low noise over efficiency, or a switching regulator otherwise.

>>1525539
>he doesn't use constant-speed props

>>1525628
measure the load current at the desired voltage. then choose your dropper resistor to drop Vsupply-Vload volts at that current

>>1525630
he knows, he's learning theory

>>1525371
usually these are shitty I2C I/O expanders like the PCF8564. the general idea is that you bit-bang the expander's outputs as if it were a whole GPIO port on your MCU, in order to supply the 4-bit+control signals the HD is looking for. shitty

>>1525539
perhaps you need to modulate the frequency based on a pulse width known to be efficient, sort of like >>1525573 suggested but a bit faster

>>1525630
A cheap LED flashlight will just use resistors for current limiting, and an expensive one with a CC switching regulator wouldn't mind having a resistor there at all, though you'd probably cook the resistor. For one with a linear CC regulator, adding a resistor there would probably still work perfectly fine, as the heat load would be split between the resistor and regulator and prevent the regulator from overheating from the overvoltage, since the resistor calculation is still perfectly valid.

Does anyone have/know of a general list of active components that are just useful to have around for hacking and prototyping? I wanna just make a big Mouser order to beef up my part stash but besides a few things like 74xx ICs and various rated BJTs/MOSFETs I'm not too sure what to grab.

>>1525761
1. alibay component assortments are great for beefing up your part stash
2. Digi-Key is usually a little cheaper in my experience
3. what you should buy depends a lot on what you'll be doing and what you have already
but anyway, as generic advice
>op-amps of various supply voltage ranges, input types, and speeds (esp grab some TL082s which are handy for high-side current sensing etc.)
>comparators, likewise
>small-package 74xx#G logic is really handy sometimes, esp 1G66, 1G3157, 1G19, 1G57/1G58/1G97/1G98, 1G14/2G14, 1G17/2G17, 1G79/1G80
>crystals/oscillators? Si5351 clock generators are handy, too, see the datasheet
>switch-mode regulators?
>power inductors?
>Schottky/fast-recovery diodes
>small MCUs (<=20 pin)
>dragon dildos <- very important for the dedicated experimenter
>a big bag of 100nF caps to go with that 74xx stuff
>LEDs. everyone loves LEDs

>>1525761
tl074
ne5532
lm393
lm358
tl494/ka7500
uln200x
555/556
*4053/2/1
*4011

>>1525588
Here you have it all: Germanium, a 'thermistor', an op-amp and your 'Vactrol®'.

low-volt boost converters don't get any easier than Pic related

>>1525761
oh, I almost forgot
>voltage references e.g. TL431
>generic switch-mode controllers e.g. MC34063
>op-amps, in particular, MCP601/MCP602 are good low-volt jellybean devices, MCP6002 good to have for low-volt RRIO

Anyone ever mod a cellphone to run off power from the outlet? I got a Nexus 6 I wana work with but the li-on battery is swollen and now a fire hazard.

>>1525862
Replacing the battery would be pretty cheap. The phone itself will run off some voltage rail(s) that it creates itself with voltage regulators/converters running off the battery, but intercepting past this circuitry to feed it a direct 5V/3.3V will be next to impossible, especially with a modern phone. So I'd just get a 4.2V isolated flyback converter and solder it directly into the battery slot. There should be another contact there for a thermal sensor of some sort, some googling should find you what sort they use so you can spoof it with a resistor of the right size (or use a proper replacement next to the transformer body so it thermally throttles if the converter gets too hot). A simpler method would simply be to use a (low-dropout) linear regulator on an existing 5V usb supply to feed the phone a nicely smoothed 4.2V. It's a little wasteful, but if you're running it off mains it shouldn't matter too much.

>>1525787
>Digi-Key is usually a little cheaper
Try LCSC

>>1525883
>spoof it with a resistor of the right size
usually a 10k thermistor, but replace with a 10k resistor to pretend it's always 25°C and sunny
>linear regulator
meh, one diode drop should be enough

>>1525886
I did
>can't sort by value

>>1525883
That second thing might be more useful. What i've done is taken the old swollen battery and taken the circuitry that connected it to the phone out. That way I can still plug whatever circuit I have in the battery slot and for all the cellphone knows it's just a regular fully charged battery.

I'll looking into the liner regulator idea. That sounds like something I can do. I've directly soldered the power lines of an old USB cable to this battery circuitry and simply connected it to a 5V USB power supply but I get nothing.

This is also super ghetto and niche so I can't find a fucking soul on earth who's actually done this before. At least through google I can't.

>>1525887
>one diode drop
If the converter output is smooth enough then probably, but depending on the converter topology you may just be able to feed it the 5V. But if not, I don't like the idea of using a diode drop that might turn into only 0.3V or so when at standby current. A 3.7V zener circuit would be a better choice.

>>1525891
>I've directly soldered the power lines of an old USB cable to this battery circuitry
>I get nothing
What battery circuitry? The stuff inside the phone itself? I hope you didn't cook it.

People also use arrays of phones for spam/crypto purposes, so you might want to look into that, but I think they just keep the batteries in and plug a USB lead into them all.

>>1525897
>What battery circuitry? The stuff inside the phone itself? I hope you didn't cook it.

Essentially the little plug that connects the LI-ON battery to the phone. It's entirely possible i've just been an idiot and cooked it. I think a fully charge LI-ON battery does 4.2V and the USB power sources does 5V so maybe it just broke over time? It's fine I have spares because I knew this would happen.

>>1525897
>I don't like the idea of using a diode drop that might turn into only 0.3V or so when at standby current
hm, good point. I can't be sure whether or not the radios are expecting a boosted voltage below 5V but above 4.2V. linear or buck it is

>>1525891
overvoltage protection circuitry, or the missing temperature sensor, could have caused the battery management IC to cut the battery off. or you just cooked it

>>1525905
Using a buck for 0.8V seems a bit silly, a linear may even be more efficient. Not to mention much smoother. I'm guessing the phone's battery circuitry isn't terribly focused on noise rejection as batteries tend to be pretty clean voltage sources, so the phone may not play nice with a buck converter's ripple, meaning you may need a larger capacitor than what it comes with. Especially if you get a cheap chinese one. Bulkiness and line/audio noise is also a reason I'd steer away from a buck.

>>1525891
>What i've done is taken the old swollen battery and taken the circuitry that connected it to the phone out. That way I can still plug whatever circuit I have in the battery slot and for all the cellphone knows it's just a regular fully charged battery.
This doesn't always work. Case in point:
>https://hackaday.com/2018/11/21/battery-swap-keeps-sansa-clip-chugging/

What can I do to increase the current sourcing/sinking capacity of a MCU pin? I mean, what kind of external component can I use? Suppose my MCU can only source/sink 3mA on each pin. Is there such a thing as a bidirectional line driver that will work with up to 200 khz signals?

>>1525914
That looks to be an issue with the thermistor or possibly inline protection circuit. Messing around with resistor values should get that to work.

I need a sanity check. I'm trying to find complex impedance of a network by applying a sinusoidal signal at frequency f and FFTing the current and voltage.

V = IZ, where all V, I, and Z are all complex.
Z = V/I

So dividing the FFT of V at f by the FFT of I at f should give Z. But it seems like the imaginary part of the resultant Z is the negative of what it should be. I mathematically tested it with a 1F cap and 10Hz signal.
V = sin(2*pi*t*10)
I = cos(2*pi*t*10). Verified these on LTSpice so I'm pretty sure I didn't screw that up. Doing Vfft/Ifft gives .0159j, but it should be 1/(j*w*C) = -0.159. Anyone see the problem?

>>1525928
What are you trying to drive? 200kHz isn't particularly fast but if you're trying to drive big capacitances you might run into issues.

>>1525947
>1/(j*w*C) = -0.159
Typo, 1/(j*w*C) = -0.159j

To make it more clear, the impedance I calculate with this method is the complex conjugate of what it should be. Is this some weird Fourier transform property?

>>1525910
>may
>guessing
yet they run from shitty/dirty af USB chargers, with a maximum permissible decoupling capacitance per the USB spec, and sometimes both battery and USB simultaneously, all flowing into the same system voltage bus from which point-of-load regs hang, all of which generally run at > 1MHz frequencies, and don't have any particular problems with noise. methinks you don't give mobile system designers enough credit
>bulkiness
really now

>>1525947
it seems to me like a matter of perspective. any reason not to consider current flowing into the DUT as negative?

>>1525974
>consider current flowing into the DUT as negative

Then the imaginary part has the right sign, but the real part is negative

>>1525974
>shitty/dirty af USB chargers
Those don't go into the electronics in the same path as the battery, moot point.

>>1525532
Thanks

>>1525983
hardly. in all PMICs I've seen, the input lines all wind up in the same place, with nothing but a FET (and maybe a buck) between them and the system common voltage bus, which Allwinner/X-Powers call IPSOUT, benchmarq/TI call VSYS. that loosely- or non-regulated bus then feeds other regulators, each with their own decoupling/filtering caps, some of which are switching and some of which are linear, the bulk of which are contained within the system PMIC. where the system designer needed extra ripple control, he already had enough dV/dt motivation to deploy it where it would do the most good

https://www.aliexpress.com/item/LCD1602A-16x2-White-Character-Dot-Matrix-LCD-Display-Module-Black-Background/32470535629.html

how am I supposed to drive one of these generic dot-matrix displays?

>>1526069
download the arduino library lmao

>>1526069
find the HD44780 datasheet
this module datasheet contains a brief summary https://www.openhacks.com/uploadsproductos/eone-1602a1.pdf

Not /ohm/ per se, but do you have recommendations for a room air conditioning unit that's easy to control with a raspberry pi / arduino?
I currently have a cheapo fujitsu I wish to be able to control on a schedule but it only has a remote IR controller. (I can set the reference temperature, mode (cooling / heating / drying) and fan intensity)
My first idea was to use a transducer and record the signals, then play them back to control it via IR but this seems convoluted and the unit has coil whining so a new one is due anyway.

>>1526105
I found this useful to learn how to generate IR codes. It goes on to describe Sony and NEC codes
>http://www.remotecentral.com/features/irdisp1.htm

>>1526112
Thanks. I found something similar too https://nicegear.nz/blog/hacking-a-mitsubishi-heat-pump-air-conditioner/
I was hoping to find something like an API for AC units but I guess the only choices are either to record and playback the IR signals, to reverse engineer the signals like https://cheesecakelabs.com/blog/hacked-air-conditioner-controller/ , or to buy a unit with a wifi API and look from there https://medium.com/@sushilks/hacking-mitsubishi-ac-controller-part-1-c31ab8364ef1 But I don't feel comfortable having to go through a company's server to turn on / off my AC

OK boyos I'm electrically retarded and need help mating a driver with LED array. The array will have 40 LEDs in total, 20 of which have a forward voltage of 2.0-2.2 volts and forward current of 400-500mA. The other 20 LEDs 1.4-1.6 volts and current of 400-500mA. All LEDs are 3w units. I was hoping I could get what I need from AliExpress store Chanzon because I picked my other parts from there also. I was about to post a link but the site thinks ali links are spam.
And to clarify, I am retarded. Explain your reasoning pls.

>>1526155
20x2.2 = 44
20x1.6 = 32
32+44=76v
Buy a constant current driver between 400 and 500 mA capable of output at least 76V, more is fine.
Connect all leds in one series string.
That is just one solution of many, depends on what your goal is I suppose, one series string means if one led dies open circuit they all fuck off, you can split them into several smaller circuits you don't have to worry about the voltage if you stick with current driven supply

Sidenote: What are you doing about heat dissipation?

>>1526155
>the site thinks ali links are spam.

screen cap the array(s) and description

>>1526159
Okay that sounds reasonable. But isn't that like 38 watts? Are the '3w' LEDs actually less than 1w or what's fucky in this situation? I'm shooting for 120w output. For heat dissipation, i'm going to mill some fins on the back of a aluminum plate I'll mount the LEDs on and add a few computer fans if necessary.

What kind of circuitry do I need to make a power-bank from scratch? Besides the li-ion, of course. It needs to be USB-chargeable, give 6V and shut down when the cell voltage is below a certain threshold.

>>1526186
You need a special controller for it
https://www.youtube.com/watch?v=Fj0XuYiE7HU

>>1526069
>>1526078
https://www.youtube.com/watch?v=hZRL8luuPb8
Those are character displays and you really shouldn't need a driver to learn how to use one. Read the fucking datasheets.

Is Mims the Loomis of electronics or are his books actually readable and helpful?

>>1526240
>you really shouldn't need a driver to learn how to use one. Read the fucking datasheets.

well which is it, seymour? if he goes by the data sheet he will sure as hell have to write a driver, but if he does what the other guy said, use the arduino library then all the hard work has already been done and you do not have to "read the datasheet".

Have you ever looked at the datasheet that you suggest he read? yeah, he really needs to use pic-related to drive an led display.

>>1526279
>>1526279
the link he posted was a character display, not a fully addressable dot matrix/LED display

using a driver for the latter is completely understandable, but you can drive a character display with a shift register and then it's basically CLI I/O

>>1526314
>the link he posted was a character display, not a fully addressable dot matrix/LED display

the pic I posted is for the same character display that this is about >>1526093 and we are all talking about the same generic lcd 2x16 or 4x20 or whatever, and while you can do custom characters, it is in no way fully addressable dot display.

I think you and I agree except that you are confused as to what my point was here >>1526279

we need a new thread, so I'll go make it

new thread >>1526326
new thread >>1526326
new thread >>1526326

>>1522826
I'm trying to add a drive cam to my Jeep - already successfully did it to our Hyundai.
Running into an issue with the 'add a fuse' devise. It bottoms out when the black casing widens out, almost as if I need a device with longer prongs.
Anyone run into that problem before? Suggestions?

>>1526329
>new thread
But this thread will remain for like a 3 days more.

How do I get boost converters to work? I seem to have the circuit correctly done but I cant figure out the values. I have a 3.3v output from a 555 but only get ~3.8v with the converter. Im using a 100uH coil and 20uF capacitor with a motor load that works at 3-30v. I tried 800uH and of course the voltage was higher but thats way too high im seeing circuits that end up with 30v and dont use such high values for the inductor.