/diy/ - Do It Yourself

this thread's digits brought to you in part by the Exar MP7636 multiplying DAC

An amusing dumpster fire
>>>/sci/10217956

>>1517640
wow

I saw a fascinating video inside PCBway as some guy films the process all the way from drilling to thru-hole plating, etching, solder masking, silkscreening, quality control, etc. It's done by a pajeet though.

I bought the red soldermask instead of the green, am I an edgelord? Also I figured out that since it's best to use a very flat surface to press the stencil and the PCB together (both for etching and masking), you can easily get that with a cheap photo-frame's small square of glass.

>>1517653
m8 soldermask is more expensive than sending your gerbers to the chinese

>>1517657
$1.50 for a tube? I don't think so.

>>1517639
now there's a handy jellybean for something.

>>1517731
>using general purpose external ADCs
From what I've seen, you get one tailor fit for your purpose if it matters, else you just use the one in your ESP32.

>>1517753
16 bit DAC that has its output four quadrant multiplied with a separate input, ain't gonna get that built into a micro.

What's the difference between box and mylar film? Is it simply size and form factor? Does the box add shielding?

>>1517793
As in, cap like in your pic vs. a round-shaped one?
The plastic case gives some mechanical support and makes automated handling easier.

>>1517795
Yes sorry I wasn't more specific, thanks that was exactly what I needed to know

>>1517734
>you'd still need an opamp that gets near positive rail inputs
No, not at all. The opamp output is decoupled from the 40V supply voltage by the npn low side transistor and only sees its base. The opamp could run on a 5V supply without problems.

Hope this goes here:

For reasons that would waste space to explain, I needed to make a very quiet custom fan arrangement for someone's PC that I'm fixing up. The fan is 12v 0.45a, but at full power, is insanely loud. I tried running it off of a 5v source, but that wasn't enough to make it spin up. What ended up working was running it off of 12v with one of those in-line "noise reducer" adapters. Got everything put together, and it works just fine. Only possible concern is how hot that resistor is getting.

Now I know that resistors can get hot, and this is an resistor that's meant to be on a 12v source for a PC fan (no idea what the actual ratings of the part are). The part originally came sealed up pretty tight inside sleeving - when I reassembled it, I popped a couple holes in the sleeving just to allow for any minor bit of heat dissipation/airflow. This thing feels like it's getting really hot, though.

The heatshrink on the original part has 125C marked on it, and the stuff I used says it's rated up to 135C for continuous operation. The resistor is held on both sides with my heatshrink, inside it's original sleeve with pinhole vents added, and will be running continuously whenever the PC is in use. I've had this running for about an hour, and it gets too hot to hold. I know 100C is boiling, and the lowest number here is 125C, but still... The wiring is 28 gauge, and says it's rated for 300v 80C, so I'm kinda wondering about that as well.

With the information provided in this short novel, is there any cause for concern?

>>1517813
60C is about the touch limit. if you can touch it firmly for a second or two then it's fine.

i was able to find a page for an "na-src7" which is 51 ohms and has a 2.7W rating. the na-rc8 is 150 ohms so i would imagine it's intended to drive an even smaller fan. as such you should consider whether using it on a (supposedly) 5W fan will leave you with enough airflow.

>>1517813
don't BIOSes these days have custom fan PWM settings, or is that just the good ones?

>>1517826
Here's what's going on:
I'm trying to install a small GPU into this PC, but the fan on it is insanely loud and is incapable of being adjusted by any means. What I'm doing to try solving this is putting in a larger fan with a small gap between it and the heatsink, then run it at a lower RPM to cut down on noise, but still move more air than the original fan did. The only place I really have to connect the fan is directly from the PSU, as the fan header on the card does jack shit, and there aren't any other headers left open on the board.

Thus far, this seems to be what I need, but I'm just concerned I'm gonna start an electrical fire in the wires running to the fan.

>>1517820
>Touch temp
I was actually looking that up. I can hold it firmly for like 5 seconds before it starts getting really uncomfortable, and 10 seconds is when I have to let go. At some point, I should get an IR thermometer, because right now I feel like I'm doing about as much proper science as balancing biles is doing proper medicine.

These adapters were leftover from installing a Noctua heatsink for someone, which used 120mm fans. I just looked up 3 different models of 120mm Noctuas, and their amperage ratings go from 0.05 to 0.12. I know fuck all about electrical engineering, so I have no idea what the amperage pull through a resistor might do to it. Should I be looking for a fan with a lower amperage requirement, because I double-checked, and this fan really is 12v 0.45a - pic related.

I plan on doing tests, but I wanted to make sure this part was even safe in the first place - I can check PC component temps to make sure they're being properly cooled, but I don't have any means for checking the temp on an individual resistor.

>>1517837
>Should I be looking for a fan with a lower amperage requirement
you still need significant airflow over the gpu so i wouldn't do that.

>I have no idea what the amperage pull through a resistor might do to it
making the current go through that resistor before it gets to the fan means the resistor burns off some voltage that the fan would otherwise be getting. less voltage on the fan means it'll draw less current. so an equilibrium gets reached where the fan is drawing less than 0.45A because it's not seeing a full 12V. drawing less current = spinning slower = quieter. so using a 150 ohm resistor probably meant for low power fans will be limiting your big fan's current so much that it may not be spinning fast enough to cool the GPU. there's no free lunch. it's a tradeoff between cooling and noise, and you get to pick what's acceptable to you. just be aware that 150 ohms is a large limiting value for that fan so it might be spinning too slowly. if you don't like the tradeoff you'd just have to get a noctua with better bearings and airfoil design.

>>1517843
The fan that's on the GPU right now is 60mm, but it is also in a very stupid arrangement:

Whereas large GPUs use blowers that pull in air and push it down a shrouded heatsink because that is logical and efficient, this thing is just a normal fan attached directly to the base of the heatsink, not even attached over fins. Because of that, it's blowing air directly into a wall, and just kinda hoping it moves air over the surrounding fins. The motor is also placed almost directly over the actual card's processor as well, which means it's getting as little airflow over that area as possible.

The plan is to remove the old fan entirely, then install this new setup at ~1" distance. Currently, the air being moved by the fan, which has been measured at fuck-if-I-actually-know CFM, feels pretty impressive for the noise this thing is generating. The card functioned with a failing fan that would randomly start/stop for a bit before I found the source of the noise, so I'm sure it'll be able to handle booting into the OS and opening HWMonitor to check if it's on fire or not while I initially test this setup. If temps hold steady at idle, then I plan on running some stress tests.

I'll let you know if I die, though.

>>1517846
i'm getting into things i don't know much about now, but as i understand it there are two types of fans. some are designed to create a pressure differential, and some are designed to move air with a low pressure differential. case fans are the latter so i think the result of your setup will be that very little air will be forced through the fins because that's a high friction path whereas simply going around the gpu will be low friction. i'm interested to see the outcome though. also this has nothing to do with your original question. the resistor dropper is at a safe temperature, but it probably is running outside spec.

>>1517848
If you can tell me the proper resistor that I should be using, I can replace it no problem, and I'd much rather be using the right parts than hoping to get away with the wrong ones.

I'll be doing temp tests later today, though - I'll post the results.

>>1517851
Hook the + lead of your fan to 12V, and the - lead to 5V. This gives your fan 7V, which should be enough to make it spin without having to burn off voltage/make heat with a resistor.

>>1517857
I'll delete System32 and make some cool crystals that /b/ told me about while I'm at it.

Does this solder wire look good for through hole?
Mostly veroboards.

It's dirt cheap and I wanna buy as much as I can

>>1517874
> 60/40-ish
Check
> Rosin core
Check. Jump on it.

>>1517858
use a multimeter to measure the voltage between the 12v rail and the 5v rail
it would only be a problem is the black/minus/negative lead on the fan was grounded in some way - it isn't

>>1517636
I have a nice amp from around 80s (Diora WS 4421), I would like to use it but there's a problem. When I was checking it (with some little chink 8ohm speakers), I noticed that right channel is perfectly fine, but left is very quiet, barely can hear the music. It's not the fault of speaker, cable or input because I checked. What can be the problem and how can I fix that?

>>1517874
0.8mm 63/37 2.2% noclean flux?
eutectic alloy, lowest melting point (183°C)
I would buy a kg if it's less than 30€/kg
will last forever and a day

>>1517901
get schematic, compare channels, find fault, repair

What circuit would I use to generate a tiny negative voltage (2v) and positive 20v from a single transformer coil? I`m thinking about making some sort of voltage doubler+the rectfier but it isn`t working.. I`m a brainlet that started a couple months ago pls help

>>1517916
stupid image. post your circuit diagram instead.

>>1517927

>>1517933
Take AC voltage from the transformer via a capacitor, half-wave rectify it with a diode and limit/regulate the voltage with some suitable way.
Another option would be to put some diodes in series with the existing bridge's negative output. That way the point before the diodes would be negative in comparison to their other side, which you would then ground.

>>1517773
what do you even use it for?

>>1517858
dumb fuck

Reposting this nice dumpster fire
>>>>>10217956

>>1517950
Yup. Limiting resistors and transistor as simulated Z-diode. Leftmost cap only as large as necessary for -2V current.

why did someone make this thread if the last one's at 400replies 105images?
isnt the image limit 150?

also how do i stop pic related from sending bullshit to a or b if they're low
do i need a much higher resistor in front of a and b (relative to the ones to ground)

forgot pic

>>1517979
Because the bump limit is 300 posts.
Put diodes in series with the inputs.

>>1517950
thanks

>>1517985
don't image posts still bump after a certain point?

also is this a good xor?
adding the diodes and boosting the previous one is 8 components while this is 7

>>1518000
No. Bump limit is bump limit.

Depends entirely on your idea of "good", but discrete RTL is pretty bad in general.

>>1518008
can you suggest an alternative? i've got 100npn and 100pnp bjt's and some resistors and want to make some gates and eventually an adder
what is even the advantage of TTL when using discretes?

>>1518000
Oh, and since you apparently are planning on omitting the base resistors: doing so makes the circuit very unreliable and you have to limit the input current some way in any case.

>>1518020
Discrete transistor logic is all about compromises. You can make it suck less at the cost of complexity.
The main problems with the simplest form of RTL are the low noise margin and low speed. TTL tries to improve both. Alternatively, you can use additional negative supply and more resistors to improve the noise margin.

>>1518033
is rtl slow because of resistor capacitances?
if so, why? the ttl circuits also have resistors

>>1518045
The output of an RTL gate is active in one direction only (transistor pulling output low), while a typical TTL output is active in both directions. This alone makes RTL's rising edges slower with practical pull-up resistors. RTL's base resistors also limit the available drive current (particularly turn-off current) to the transistors.
Typical TTL gates are built so that in addition to having faster output stages, the internal transistors are driven harder than in an RTL gate of comparable current consumption.

Maybe you should check some antique calculator's or computer's schematic for ideas.

All that said, RTL is probably the best choice if you just want to play a little with discrete transistor logic and don't really care about the practicality or performance.

>>1517639
I am still looking for LaTeX sty files for such sheets. Any help?

Hey /ohm/ I bought some buck converters for couple projects I had for over the weekend.

>pic related.

I am running 18v using AA batteries. I am powering two devices off of this battery pack, one is a raspberry pi, and another device (RFID) and it seems after some tests that buck converter is causing a lot of Radio intereference, since my rfid reader can barely read when I have RPi powered on.

I did some googling, and it seems people like to use lm417 voltage regulator, and I am wondering if I use voltage regulator with capacitors and resistors will that eliminate the Radio interference like I am having with a buck converter?

>>1518113
oops I meant lm317 voltage regulator.

>>1518115
Yes, every ham radio operator knows that. Anything that switches or does PWM pollutes your place with EMI/RFI to an extent that you cannot use indoor antennas. The only solution is a linear regulator. From my experience there's no way to silence a switch mode PSU.

>>1518130
Have another right next to it that's 180 degrees out of phase?

>>1518130

Im new with electronics, but I am assuming that means lm317 will still give me EMI/RFI, correct?

>>1518138
LM317 doesn't switch, so very low noise, but also inefficient, all the excess voltage is turned into heat.

Why are you running 18v?

>>1518130
what about a faraday cage/wrapping it in aluminum foil?

>>1518138
The LM317 is clean, no EMI/RFI.

>>1518113
>>1518138

Adding a linear regulator may not help much with the EMI, because that just comes with the nature of switching converters. It may get it working for you if the lines coming out of the regulator to the load are long, since they will act like antennas if directly connected to the output of the converter, but no guarantees.

Best you can really do is add more/better filtering to the output (which may well include a linear regulator; use a LDO regulator if efficiency matters to you), and put it in a small, grounded metal enclosure.

>>1518144
rfid reader.

>>1518148
true, but after doing some googling, it produces a lot of heat, like anon said here. >>1518144

>>1518150
I was thinking of adding 100uF capacitor to the output to see if that helps. You think it will or no?

>>1518153
you can solve the electrical infetterence by using two separate regulators. the linear for your rfid and the buck for your pi. regulators aren't perfect at filtering input noise so you can filter higher frequencies even further by adding a choke on just the rfid's power line.

>I was thinking of adding 100uF capacitor
that won't help much. larger capacitors are bad at filtering high frequencies. at a bare minimum a choke and a ceramic cap adjacent to the rfid might work well enough.

>>1518164
Even having the buck converter in the circuit with a linear? I did buy a 10 pack of these buck converters for other projects.. Wouldnt using two buck converters, one for rpi, and another for rfid help with RFI?

>>1518169
>one for rpi, and another for rfid
the raspi is contributing some noise for sure but the buck is almost certainly the bigger problem. you can try.

>>1518150
what happens if you don't ground the enclosure?

>>1518172
So adding a second buck converter doesn't affect the range anymore than just one buck converter did. Adding rpi, produced same results if I didn't have the second buck converter. Which means you have to practically put id card right up to the reader and push on the enclosure a little bit.

So, to eliminate the buck converters and minimise RFI, it seems I can try to use a arduino nano (on pin 30 it can handle up to 20v it seems) and then add esp8266 to the arduino nano so I can still achieve wifi capabilities.

But I can't help but feel that I will still get RFI as if I have buck converters..

Hey guys I was wondering, is it possible to use one power switch to operate two power sources that power two different devices, and turn them both on?

>>1518224
yes. DPDT or DPST switch.

>>1518248
oh thanks anon, just curious two different power sources will have different voltage (one is 5v the other is 12v) it wont interfere with those two?

>>1518256
Those switches are isolated, there shouldn't be any problem.

>>1518287
Thanks anon!

>>1518132
that would be fairly directional

>>1518153
wait, does the RFID reader take 18V? if so you need filtering on the *input* of the switcher and probably some shielding around it

>>1518169
Why don't you just add a shield?

>>1518113
>18V
>AA batteries

>>1518294
>wait, does the RFID reader take 18V? if so you need filtering on the *input* of the switcher and probably some shielding around it

Yes, and Im probably gonna go arduino route, and eliminate the buck converters from it.
For the filtering of the input would it be a ceramic capacitor on + and - is what you mean?

>>1518333
Another anon suggested foil around the buck converter, is that what you're suggesting?

and yeah its most like right under 18, when Im using 12 AA batteries.

I designed a board recently that connected a sensor and motor driver to a Raspberry Pi. The board drives a DC motor off a 12V floating supply with PWM and sources a few amps of current. The ground of the 12V supply is shared with the digital ground. The Pi is powered off its own power supply. I noticed switching noise manifesting as quick spikes on the sensor's digital output. The output and supply of the sensor ran underneath the motor driver circuit (stupid, I know). In trying to fix the noise I cut the trace with a fairly wide gouge between Pi GPIO and the motor driver, and attached the sensor before the cut. But the noise is still there.

I have two theories:
1. The noise is still present because the ground plane is bouncing when the driver switches
2. The driver circuit is inductively coupled to the cut, floating trace underneath it, which is capacitively coupled to the part of the cut trace the sensor and GPIO are on.

Which do you think is more likely? My next revision of the board is going to have the Pi powered off 12V through a linear regulator. I'm trying to consider how to solve this problem beyond moving signal traces away from the switching power supply. I'd prefer to stay two layer so ground plane isn't the best solution. My plan right now is to have copper pours for the digital side and the power side with the power input on the power side, and have the two connected by a relatively high impedance connection (thinnish trace). I'm hoping that this will keep transients isolated to the power side.

Thoughts?

>>1517846
>>1517848
Okay, so temp tests concluded:

Stock fan
idle - 36C
stress - 51C

Modified fan @ 1"
idle - 36C
stress - 51C

Modified fan @ 2"
idle - 37C
stress - 53C

And just for the sake of curiosity:
No fan
idle - 54C after 10 minutes
stress - haha no.

Overall, I'd say this is pretty successful. Noise is way down and temps are pretty much identical, which is also well within spec. I have some other things to try, but Attempt #1 seems to be completely usable if I run out of time. I found an actual voltage controller among my computer junk, which I may end up using and just installing entirely inside the computer so the person I'm setting it up for can't find it and fuck with the setting. It'd also make fan replacement easier if I didn't have to have it entirely customized, should that become a thing.

I would, however, be interested in knowing if there was a more appropriate resistor that I could be using in a setup like this. As previously mentioned, I can't tell my ass from a hole in the ground when it comes to that level of working with electronics, so I would have no idea where to even start looking or what to start asking.

>>1518356
This
>>1517857
is the right solution

>>1518359
And here I thought you guys were fucking with me. Works perfectly.

I threw in a little grounding just to be safe, though.

>>1518373
kek

Give it to me straight doc.
How bad am I if this is my first board?
I'm only motivated by the fact I've seen boards far worse work fine in commercial products, so in far or less confident this will work.

>>1518391
Also I'm thinking of cleaning my next board before soldering and changing my tip

>>1518392
>>1518391
Apart from the circled solder joints it looks pretty solid to me, some of them look balled which indicates the solder may not have flowed onto the board, only onto the component legs. Wiping the copper side down with isopropyl before you start never hurts and will help prevent this balling, the rest look like they could use a touch more solder. Rework these joints and you should be golden. For the good joints you've used a good amount of solder, drill outs are consistent and you've dressed the resistors/centered them well. It seems your tip is an ok size as evidenced by the fact I'm not seeing any evidence of cleaned up bridging between the tracks. Not bad for a first attempt anon.

>>1518392
>>1518391
Use more flux for those dry joints (either apply and remove some solder or buy standalone flux), perhaps use a dedicated IC socket so you can replace the IC if it breaks/solder it in place without overheating it, maybe add some epoxy/hot glue over the output wires since they'll likely be prone to work-hardening and snapping off if you're not too careful. Of course, putting strain relief in the form of chassis/panel mount sockets through a project box would be a preferable idea.

What iron is that anyhow? Chances are you just need to clean it a bit with some copper wool.

>>1518409
I've seen manufacturers use IC sockets but I thought it was just more messy and cumbersome but I'll look in to it, do ICs go bad quick?
I'll buy flux paste since I'm heading out now and give it another shot, thanks, I'll also have to look into epoxy since I think hotels will just get messy.
My soldering iron is an old JBC 40s.

>>1518405
I'll definitely wipe my board down next time since I was a retard and cleaned it with windex.
Thanks anon that's exactly the kind of input I needed!
Also should I flux and solder those balls till they go cone shaped? Liquid or paste?

Also what's a good way to keep components snug on the board while I flip it around and solder? I use alligator clips on the terminals

>>1518415
ICs are more robust than your TO-92 transistors, the main worry is whether you blow them with heat while soldering. I've got a tub of jelly-like rosin and it works a treat, but ensure you have some isopropyl alcohol and an old toothbrush to wipe your boards down with since it makes a bit of a sticky mess afterwards. No-clean flux looks to be pretty good, but a flux-pen doesn't apply it liberally enough for me, so take your pick. Add a tiny bit of flux on those joints and reflow them, and they should go nice and cone-shaped or "filleted". If you screw the circuit down in a box (which you should do) then epoxy isn't really needed, but it's a handy thing to have lying about in any case.

Keeping components on a board is a bit of a tricky one, I just tend to solder my components in order such that I can always just have the PCB resting on the component I'm trying to pin down, or try to squeeze them to the board with a finger that I try not to burn, but bending the leads to either side always works if you don't mind the slightly offset leads. I've seen jigs that hold the PCB and push a thick block of open-cell foam against the components, which looks to hold them in fairly well.

>>1518415
Take some iron wool and clean the copper, then put flux where you`ll solder and on the component legs (clean them the same way). It`ll be much better

>>1518441
I've never understood this flux meme, and I solder a lot. I don't even have any. My flux is in the solder wire, which is more than enough. I use 0.7mm Sn60/Pb38/Cu2 2.5% flux and solder at 250°C, works like a charm.

anyone ever had an AMS1117 just go *pop* with no load? I'm pretty sure I just didn't and that the problem was poor etching in my kitchen sink board fab, but wanted to make sure (and maybe sub a regulator more tolerant of low load operation in version 2)

>>1518373
10/10 conscientious workmanship

>>1518422
no-clean flux paste ftw

>>1518405
not the anon you replied to, but I actually bought this flux gel, and I put that on surface/wire of where I will be soldering. Shit works amazingly, but question is, am I doing it right or am I doing it wrong, but somehow it seems correct since results are not negative?

>>1518455

>power supply oscilates
>no matter what I do my bjt oscilators do not oscilate
T-t-thanks

>>1518452
It shouldn't break without load, but it won't work properly, either. Usually the voltage setting resistors are a sufficient load. It also needs an output capacitor to stop oscillation.
Your other stuff after the regulator might go *pop* if you don't take care of these. The regulator itself is most easily killed by momentary overvoltage.

>>1518468
oh, it was a fixed AM1117-5.0. there was nothing connected after the reg, not even the output caps (bad etching). I presume, then, there was just some short between in and out

>>1518452
The AMS1117 is a low voltage regulator that resembles the LM117/317 (i.e. floating, no GND connection). It has a minimum load current of 13mA at most and 2mA at least. Typical value is 5mA for the fixed version and 10mA is specified for the adjustable version. This is because the quiescent current has to go somewhere. With no load at all there is no regulation and the output approaches the input voltage. If that is a problem better use a grounded regulator.

>>1518467
normal lack of skill

>>1518467
>call power supply an oscillator
>call oscillator a power supply
Problem solved.

>>1518450
>I've never understood this flux meme
Newfrens often leave the solder on the tip untill there is no flux left.

>>1518415
>>1518441
>>1518409
>>1518455
Like >>1518450 (that's some fucking tidy work by the way), I have never in 20 years of hobbyist and professional electronics, used flux, and the shops and labs I have worked in haven't either. I did try it a few times in my early years but it just seemed like an unnecessary extra step with no positive result and extra mess to clean up. It seems like a very American thing based on forum and industry observation, but then again, Americans can't even pronounce the word "solder" correctly, so I guess they need all the aides they can get ;)

We have rosin core solder and an isowipe will clean the board plenty, message ends as far as my experience goes. And I say this as a greasy wog who has very oily hands.

Also, I would avoid the steel wool, it's going to scratch the shit out of the finish and make it look ghetto as fuck. Can also leave metal strands all over your board/workspace which may or may not introduce shorts, and will get stuck to any tool which has even the slightest hint of magnetism to it.

> Also what's a good way to keep components snug on the board while I flip it around and solder?
For resistors, caps etc, I give the legs a little pull out sideways to hold them in, for DIP packages (chips or sockets) I bend out diagonally opposite pins. for things that cant sit flush against the board (LEDs that need the height to get to a panel for example) your alligator clip method will do the trick nicely.

>>1518491
>can't even pronounce the word "solder" correctly,

solder: "sodder"

salmon: "sammon"

>>1518496
Holder = "Holder"
Colder = "Colder"
Older = "Older"
Folder = "Folder"
Bolder = "Bolder"
..
Solder = "Sodder"

WHY.

I cannot watch burgerclap instructional videos for just this reason. No matter how hard I try to ignore it, the red mist inevitably descends.

>>1518500

I live in the usa south, and I can't explain this:

hog = hog, kinda like "hahg"
frog = frog
smog = smog
log = log

etc.

but...

dog = dawg

I'd feel like a retard saying hawg, frawg, lawg, etc., but dawg sounds just fine

maybe we are retarded

>>1518502
perfectly normal and expected phenomenon
https://en.wikipedia.org/wiki/Vowel_shift

>>1518502
Thats just a southern drawl kind of thing, I can get behind that, but ignoring a single letter because.. why? FFFF..

Anyway, sage because off topic. As you were /ohm/.

>>1518467
>oscilate
Ok, managed to fix that piece of shit of psu, apparently it`s good at full load and has a ~1V dropout at 25V and 2A. Tested it with a 1100w universal motor, as expected it got really toasty at 2A with the ANCIENT pnp power transistor I got from a tube tv (I`m using it because I`m retarded and blew 2 of the good TIPs I had bought). This is my first project so far I`ve learned:
>plan ahead, deciding to make the power stage on a separate PCB was a good decision, because I knew I`d screw up in the controls
>osciloscope probes are shorted, sparky
>making over complicated transistor and opamp equivalent circuits was a waste of time, to fix the oscilations I had to try diferent capacitors to see which one worked better
>always test ICs before starting other shit to avoid wasting 5 hours troubleshooting
>don`t go crazy with PCB complexity because I have to hand-draw them or make in a perfboard
>always check if the .1F capacitor bank is discharged before handling

>>1518578
Why make your PCBs by hand when you can get them from China in a week for $20?

>>1518579
because I`m not american, mail services here are criminally expensive and slow, my currency is 1/4 of a dollar so chinkshit is expensive, taxes..
>inb4 free shipping
no free shipping, I have to pay for all foreign packages regardless of weight or size

Does it matter whether the resistor is on the right or on the left here for the thing to be active low?

>>1518646

no, because they are connected to the same wire which is assumed to have negligible resistance compared to 4.7k.

as far as what "active low" means to you and your circuit, that's another issue, and I don't understand the difference between the L and the three Hs. they all do the exact same thing.

This question is probably unrelated, but I don't know a better place to ask. I want to print datasheets and manuals of what I'm working with, because you can have all needed info right before your eyes, removing the need to scroll between pages and Alt-Tab between documents, and can make notes and mark important bits with color liners. So, what is the most inexpensive way to do so? I suppose that dot matrix printers still give the cheapest price per page and are particularly undemanding for quality of paper, but I'm not sure they have resolution that is high enough to print (semi)modern datasheets' PDFs. So, any thoughts or experience?

>>1518657

why not keep everything as-is on the computer or web, and just print a page or five here and there on any laser printer.

I had to google to see if they still actually sell dot-matrix printers.

>>1518581
brazil?

>>1518690
yes

I just slightly touched a film capacitor with my soldering iron on accident and a small corner melted away.
Do I have to replace it?

>>1518579
I might invest in this, where should I start?

Common jellybean shunt current monitors?

>>1518660
Plenty of used ones are sold for next to nothing. The ribbon cartridges are still being produced.
>>1518708
Check if it's not shorted and you'll probably be fine.

>>1518729
nevermind, I`ll make something with transistors.

>>1518450
>>1518491
I think it's partially >>1518479
>Newfrens often leave the solder on the tip untill there is no flux left.
But also when I'm using an underpowered iron on a heavy ground plane, I find the time it takes for the solder to wet to the pad is long enough such that the flux on the drop of solder on my tip has burned off at that point. I solder with a substantial blob on my tip because it gives better thermal contact than the shitty conical tip. Not sure how common this technique is, but it certainly helps heat shit up fast.

Can any electrician's give me advice for an apprenticeship? I have an exam coming up and have no clue what to study for
>In preparation for this exam, you should review the areas of:
>Electronic Symbols
>Electronic Theory
>Electronic Logic
This seems pretty broad and there's no study guide or anything I can find for the test.

>>1518746
I'd say that means:
>know how to read a circuit diagram by its symbols
>know Ohm's law, Kirchoff's laws, and perhaps the characteristic equations of diodes and BJTs
>digital logic gates
really not sure about that last one

>>1518746
>I have an exam coming up and it's like I was born yesterday, I'm so fucking clueless.

where do people like you come from? and why aren't you flipping burgers or bagging groceries?

>>1518748
Thanks m8

>>1518749
maybe it`s not for a electronics related area but they are in the test anyway. You sound like a cunt desu

>>1518748
>Ohm's law, Kirchoff's laws
If Kirchoff then Om.

>>1518500
>I cannot watch burgerclap instructional videos for just this reason.
Imagine being me and having to listen to Big Clive talk aboot sollldering.

>>1518077
I'm also interested in this.

>>1518813
Imagine being a rational human being who can watch both without any issues.

From what I've seen, there is some traditional reason for the pronunciation to be "soder" instead of "solder", but I think that might have died off before the American-English "soder" came to be. Either way, America is the only significant English colony and has a higher population than the homeland, so there's nothing wrong with having a highly influential dialect of your own, but I'd prefer it if you at least made the distinction that it isn't "the original English language".

Don`t worry americans, soon you will be speaking spanish and the Solda will be different enough from the Anglo-Hindi equivalent across the pond.

>>1518711
Get KiCAD, design your board, export the gerbers, send them to JLCPCB or some other cheap Chinese fab house

>>1518350
Pls respond

>>1518581
you might be just the sort of person that "we have too many components to list, please write us" is made for
anyway, next time you do shell out for chinkshit, consider some dry film photoresist, or toner transfer paper

>>1518729
depends on what kind of output you want ofc

>>1518711
pcbshopper dot com

>>1518739
surprisingly, this is a very good way to do very fine-pitch soldering
(combined with a lot of flux :^) )

>>1518786
/ohm/ RULE 0

>>1518952
to what? anyone knows what a fookin' spike looks like. post diagram of your power distribution setup

>>1518952
Firstly I'd add some ceramic capacitors across the rails on the board side, plus everywhere else that makes sense. Check the frequency of each power supply to figure out which one is causing the problem (microphone + Audacity's FFT is a good method for diagnosing audio-range noise).

>>1518963
>very fine-pitch soldering
Oh yeah that's how people just drag a bead along a QFP and they all just turn up real nice. I guess no-flux is for perfboardlets.

>>1518963
+12V

>>1518963
GND

Both +12 and ground come in from the connector on the bottom of the picture. The Pi connects via the connector on top and has its own 5V supply which is regulated to 3.3V. All my digital stuff is 3.3V

>>1518969
DC motor driver (the switching supply) is the 4 FETs on the right side of the pic. The sensor is the bottom right connector, and the output runs right through the DC motor driver

>>1518969
>no proper ground plane
Not necessarily your problem, but I wouldn't leave my PCB like that.

>>1518974
Yeah, for the next revision I'm going to keep all the power and all the digital stuff isolated and try to give them each a nice gorund pour. I'm hoping that, or some other tricks, is enough because going to 4 layers would raise the price by a lot

>>1518982
I should say separated, not isolated

>>1518969
I presume the 3-pin connector at lower right is your sensor? if so, then you are connecting the sensor's ground to the one place on the board where ground bounce respective to the pi's ground will be worst. instead, connect it directly to the point in the grounding system which is closest in potential to the pi itself. how about pin 39?

>>1518991
That's right. But I cut the sensor output trace and soldered wires to the Pi connector to get it away from the switching supply. So your suggestion was basically already implemented, and the picture of the noise I posted above is after that fix

>>1518997
oh
the only other thing I see that's potentially problematic is the pi board probably being a lower impedance connection from BLS to ALS and C6 GND than your GND-bus trace across the pi connector. maybe cut between pi connector 25 and 30, and 39 and C6, and add a jumper wire between HIP pins 14 and 17
also, for the same reason, consider adding some wire between A1 pin 7 and the motor GND input, and cutting the equalizing traces, to keep ground current out of the pi
just for fun, try grounding the sensor to other ground pins on the pi connector and see if there's one that bounces less

>>1519011
I'm impressed you either know that IC or took the time to look it up. Why would the grounding of the gate driver cause switching noise?

I'm going home for Christmas and won't have access to the board, so I'm planning on just inferring what the problem is and designing rev B from home. Are we in agreement that ground path issues are the most likely culprit?

>>1519017
looked it up
>Why would the grounding of the gate driver cause switching noise?
something's got to pull the gate down to somewhere. in this case you're discharging the A half's low-side FET into the pi. if you haven't left yet, you could test this theory by isolating the spikes to one or the other half of the HIP driver
>Are we in agreement
yes, absolutely, within a subtext of power distribution generally
just my own personal suggestions, but, since the HIP is basically symmetrical from side to side, I'd turn it around and put the power side closer to the power side if the circuit, and move the bootstraps out to the side of the IC. just to keep the distinction between "current ground" and "voltage ground" more clear and to ease routing somewhat. I might also look into using DPAK or D2PAK FETs and using the board as a heat sink. I would also suggest using horizontal-side/vertical-side routing strategies so that you could keep a ground plane maximally intact, especially if vias are "free"
if you don't mind a minor cost adder, you could add some fast isolation between the two halves and not have to ever worry about ground bounce

redpill me on norton opamps or current feedback amps, because I`m not understanding them

>>1517646
I would rather not learn anything then.

>>1519039
I saw another one about a 4-layer PCB manufacturing that was more interesting anyway, instead of positive photoresist for the etching they used a negative photoresist on the outer layer, then electroplated more copper onto the outside traces (and vias/thru-holes I think), then they plated tin over all the traces. Then they stripped off the photoresist, and etched them such that only the un-tinned copper was removed. Then they stripped off the tin.
https://youtu.be/sIV0icM_Ujo

I can`t change the 0.1 resistor value, but I need to change the current which trips the current limiting. Currently it it turns on when the current across the resistor is about 7 A, I need it to be close to 4 A.
A voltage divider on the base would give me a ratio very very close to one, which is useless even with 1% resistors. How could I achieve this? Or in some way implement foldback (I have no idea on how to do it with pnp and with Rsense before the pass transistor.)
>ib4 use op amps
Can`t because local stores have no opamps with common input including VCC when VCC=25V like it is here, and even on the web is hard to find and china takes too long. (I live in a shithole)

>>1519039
if you can learn anything through a thick Bengali accent, more power to you

>>1519063
why can't you change the resistor?
>no opamps
how about the TL081? that's basically jellybean status

>>1519063
>Can`t because local stores have no opamps with common input including VCC when VCC=25V

They don't have antiquated shit like TL07x? Some old JFET amplifiers are kinda-sorta capable of handling positive rail inputs. As in, they aren't guaranteed to be able to do it, but even manufacturers say it is usually ok.
Why is it impossible to change the resistor or add another one (or at least a piece of wire) in series with it?
If the input voltage is stable, a voltage divider on the base should work. As in, upper leg to Q2 emitter, lower leg to ground and the mid point to Q1 base. For 12V input the lower resistor would be like 10k and the upper 220 ohms. No super precise resistors needed.

>>1519066>>1519072

Because increasing the resistance will increase voltage drop at full load.. Ow shit I had forgotten abou the tl072! Thanks m8!

I've been making guitar pedals using veroboard diagrams from http://tagboardeffects.blogspot.com/ for a few years now.
I want to get into PCB territory and I'm ready to learn how to use KiCad but I just wanted to know if there was a similar website to the one I'm using but with gerber files to every effect pedal so I can just get away with ordering chinese printed boards that weren't designed by me.

My other idea was drawing up pre-drilled vero board designs just to save time and to get my work looking neater.

>>1519063
>common input including VCC
I thought this meme had already been laid to rest at the end of the previous incarnation of /ohm/.
>>1517982
Also, it is not difficult to make an adjustable and stable high side current limiter even with your 0.1Ω series resistor.

>>1519135
>I thought this meme had already been laid to rest at the end of the previous incarnation of /ohm/.
what

>>1519025
Good points. Thank you

>>1519133
PCBs aren't suitable for true audio. a golden ear can tell the difference between pA and fA of leakage. i'd advise you switch to point-to-point wiring on nylon spacers. circuitboards are a poor imitation.

>>1519197
yeah, I'm gonna have to ask for sauce

>>1519200
sauce

>>1519133
>using someone else's gerbers
Where's the fun in that?

>>1519197
Nylon? Jesus, you clearly don't care what your music sounds like
https://en.wikipedia.org/wiki/Superinsulator

>>1519200

a "golden ear" isn't just a meme, it's real, and only occurs in about 1 of a million humans; the type who can hear the benefits of pic-related.

>>1519218
>This configuration provides incredible imaging and detail across the entire audio spectrum, deep, tight, powerful bass, sweet transparent mid range smooth and linear high frequency response. This cable provides incredibly low resistance.
this is what true quality looks like, kids

>>1519224
I just noticed it's going to a wall outlet. That's hilarious

>>1519218
>the type who can hear the benefits of pic-related.

There is no benefit to pic-related, ergo those people to not exist. QED.

>>1519241
enjoy your muddy tin ear, pedestrian.

Is it safe to set a buck converter to 5V and then connect it directly to my phone via usb cable?
I am tired of all the shitty chink chargers that do like 1A it takes forever to charge that way

>>1519225

>>1519252
yes, but without your device knowing it can draw more than 500mA from the cable, you'll still be waiting a while. see e.g.
https://electronics.stackexchange.com/questions/177788/how-do-i-design-a-2a-or-more-power-supply-for-my-consumer-usb-devices/177792

>>1519260
just short the data wires

>>1519197
Audiophiles are not people

>>1519252
The phone's battery controller limits the current that the device draws from the charger, so you will need to tell the phone that your makeshift power supply can provide more than 500 milliamps guaranteed by USB specification by putting voltage dividers on data lines. Google the specific resistor values.

Crossthreadposting from /ham/, just in case someone knows this better
Will a coax notch filter work well as an FM trap? Trying to get some decent signals out of my chinkshit RTLSDR, and I've never built filters before, so I'd like to learn

Hi good people, I'm trying to make a power supply for a nixie tube.

I'm fairly sure I got all the pieces in the right places, the output voltage shows around 100 volts but my coil (green resistor lookalike, bottom left in picture) makes a rising high-pitched noise and dies smoking after ~10 sec of connecting a 9V battery.

Is this just due to the coil being too weak or something else?

Also instead of 100 pF C3 i used a 2.2 nF capacitor. I don't think that's the issue as the coil dies also if I remove the cap altogether and just connect the resistor to the ground, but maybe worth mentioning

>>1519304
what is the current through it? from the diagram when q2 is conducting the coil is shorted to ground.

>>1519304
Coil is too weak, should be 100µH ≥1A ≤0.3Ω

>>1519310
>coil is shorted to ground
of course it is, it's a step-up converter

>>1519304
You know it's not that difficult to make a converter with variable duty cycle, it just requires 3-4 op-amps and a comparator or two. But yeah your inductor is passing too much current for its rating, turn up the frequency to >15kHz or get a different inductor. Also I hope that electrolytic is rated to 170V.

Plot the thing in Spice to see how much current is flowing, that always helps.

>>1519320
>>1519331
Thanks, I'll get a higher rated inductor.
Is <0.3 ohm crucial for any reason?

Don't have any more caps lying around for turning up the frequency, but also I don't see how that would help, the power on the inductor would remain the same with the same duty cycle?

The output electrolytic is rated for 350V

>>1519345
Less resistance means it drops less voltage and heats up less. Couldn't say for sure about that value in particular, but it sounds pretty good.

>>1519345
Not crucial but reduces I^2R loss. My flat 13x10x5 mm 100µH 1.2A ferrite SMD coils have 0.28Ω. 1A is specified where you probably got your diagram from, supply for 6 nixies.

>>1519358
so if i pump 100 amps through a 1milliohm resistor it should be ok right

>>1519382
>so if i pump 100 amps through a 1milliohm resistor it should be ok right
p = i^2 r

100 x 100 x .001 = 10 watts.

if the resistor is rated for 10 or higher then it's ok

>>1519383
>p = i^2 r

how do people who post on /ohm/ not know this

i blame arduino

Looking for a "helping hand" set, but there seems to be a big price gap between the 2-clip models and the 3+ clip models. In fact it's probably cheaper to buy 2 of the 2-clip models, but that's unnecessary clutter I'd like to avoid. Any thoughts?

>>1519403
I made me a 1-clip alligator model.

>>1519263
true, they're gods

>>1519421
they're autists who listen to equipment instead of music

>>1519387
me too. I almost have some sympathy for the /leftypol/ people on cripplechan who screech continuously about theory

>>1519421
Audiopathy is not a treatable mental disorder, it's a chronic state of delusion. Vendors know that and exploit it adequately.

How do i limit maximum current without changing voltage or resistance?

>>1519436
fuse

>>1519382
The component in question is still primarily an inductor. As long as its resistance is reasonably low, it has no significant effect on current. Thus, higher resistance means more heat.

>>1519403
Probably because the 2-clip versions are made by boatloads, while the other models are much less common.

>>1519447
on all the 3+ clip helping hands I could find with an exhausting 10-second aliexpress search, the arms are higher quality (1 clip per gooseneck or coolant flex pipe) than the cheap 2-clip models (basically 2 clips on a bar)
fwiw, 2 clips on a bar is usually just enough for wire splicing and connector souldering

>>1519436
google bjt current limiter
>>1519435
what is their excuse for shielding mains wiring? And only shielding it for 8 feet???

>>1519452
I'm imagining there's a bunch of ferrites inside the loom. maybe some creative braiding

>>1519436
RULE 2

>>1519452
>what is their excuse for shielding mains wiring?

I don't know abou the shielding, but /g/ says the prop the cables up off the floor in this pic >>1519218 because floor vibrations affect the mains power and you can hear it in the audio.

In other words, these people are insane, but wealthy, and are easily separated from their money, so it's a win-win, and we get some good laughs as well.

>>1519458
I suspect would be cheaper to install a floating floor, thus eliminating all conducted external microphonics. you know, like they do in actual working recording studios to keep noise off the recording

>>1519458
>>1519457
>>1519459
why not use coaxial cables?

Ohhh shhheeeiiit look what the girl ordered for me! Can I be part of your sekrit club now?

>>1519472
>girl
>sekrit club
no

>>1519472
Rocco the family parrot uses Amazon Alexa to shop while owner is away.

>>1519479
What if I make the green light blink?

For real tho this is going to be a dope xmas. My brother got me the DC power supply and a set of 3/8” impact sockets, the girl got me that Arduino kit, the Ridgid drill, impact driver, and sawzall, and I think she got me a Knipex Pliers Wrench too.

I think next step I want to get one of those soldering stations with hot air for around $100 from Amazon. Gonna wait for the holiday paychecks to start rolling in and I’ll order it.

>>1519464
because those arent propietary sockets with monofilament gold cables selling @ 6000 a cm

>>1519484

what kind of deviant family tells you what you got a week before christmas. and at worst you can only open one on christmas eve.

>>1519484
okay but you gotta post nudes on /b/ kek
score, bro. OT but what did you get her?
sra-solder dot com is selling this guy https://www.sra-solder.com/aoyue-888a-2-in-1-digital-hot-air-rework-and-soldering-station for $90 shipped. it has been good to me for a year or two

>>1517636
That image has a lot of potential.

I've been reading up about synthesizers and therefore oscillators and came across a picture of a crystal oscillator and thought it looked familiar, I've seen them on a motherboard before. When I look at one, it says "25.02FL" which doesn't seem like a hz measurement. What does it mean? There's another one labelled "SKC24.576H2"

Or is it some arbitrary bullshit specific to that/a motherboard and it wouldn't be useful to me?

>>1518500
It comes from archaic pronunciations of the current version.

>>1519492
They're typically MHz values, but not always.
SKC24.576H2 is clearly 24.576MHz (common value), but 25.02FL probably isn't 25.02MHz, but 25MHz (common value) or something entirely different.

>>1519488
I have this one on my wishlist. Was debating between that and a couple others between $70-$90.

>>1519487
I learned a couple years ago that if I don’t tell them specifically what I want, they will spend a bunch of money on shit I never use. Like a Flip video camera years ago that I literally used once. Amazon wish list is great, the only problem is I know what they ordered because it disappears from the list and I can see it when I’m tracking my own stuff.

But it will all be wrapped and under the tree on Xmas. I know what I’m getting but I didn’t open any of it yet. The Ridgid stuff has been sitting in my garage since Thanksgiving and it is so tempting to open it up and shear off a few screws with that impact driver. And I really want to play with the power supply too. 6 more days.

>>1519493
Oh bullshit. "beginning" has its roots on the Old English word "beginnan", but we pronounce it with the "ing" on the end, because that's how we spell it now. To justify it as a call back to the root origin of the word is retarded for two reasons, number one, it's current usage and spelling is "solder"; How the fuck would you justify ignoring the "ing" in beginning like you do the "L" in solder, and number two, "sodder" is not even the correct pronunciation of those words anyway.

https://www.google.com/search?client=firefox-b-ab&q=souder - "sodeh"
https://www.google.com/search?client=firefox-b-ab&q=soudure - "sodyer"

SORT. YOUR. SHIT. OUT. BURGERSTAN.

>>1519500
You've never heard someone speak in a different dialect have you?

>>1519502
Plenty of times, but an entire country, from Bostonians to Pensylvanifags to Jew Yorkers, all pronouncing "colder" correctly, but mispronouncing "solder" in exactly the same method, does not imply an regional dialect variation. It just implies a bunch of motherfuckers who for whatever reason ignore the "L" in one word, but not in a handful of other very similar words.

"I need to put my drink in a cup hodder so it stays codder for longer while I sodder this odder circuit board"

Gimmie a fucking break.

>>1519503
You're a moron who doesn't even understand the image in >>1519493 I bet you rage over, "fillet" too.

>>1519505
>"fillet"

it's weird that americans pronounce that like the french version (or wherever it came from), but we don't pronounce Amarillo like the spanish version.

>>1519505
The image indicates the historic roots of the current word solder. You have tried to justify pronouncing it as "sodder" by claiming this pronunciation stems from one of those archaic meanings, none of which are pronounced "sodder".

What am I failing to understand here?

>>1519510
Is that beer gluten free?

>>1519510
that Americans have nothing but retcon

>>1519510
>What am I failing to understand here?

How language works, kid.

>be E&I apprentice
>do industrial install, troubleshooting, and calibration
>take home is $32,000 ($28,800usd) before tax
>apply for Uni for EE
> wifes friends husband did the program and was hired upon graduation at $60,000 salary with a $10,000 raise after a year
>I'd be 29 when I finish the program
Is it worth it /ohm/? I've always been interested in the upstream side of instrumentation

>>1519484
as someone pointed out in another thread, buyfaggotry is worth,less in electronics. Just get a book nigger and build shit

>>1519527
7th semester monkey doing EE here. The good thing about it is that if you dedicate and want; you can work in ANY area. Really, the more you learn more, the more things become EE problems.

I haven`t applied for a internship yet because I live in a cancer hipster city where technology = phone apps and there aren`t good ones. The country is leaving a recession now (hopefully) but there was always a spot for EE fags. If everything fails you can always make a living doing residential/small electric projects kek.

>>1519527
I'd say stick with the trade. it's harder to outsource
not that you shouldn't learn about the upstream side of instrumentation anyway, but you don't need to go back to school for that, necessarily

>>1519508
they may be pronouncing fillet wrong

>>1519527
You'd be only 8 years older than a typical grad and making more than twice as much as you were before. It's very worth it. Probably. Are you in the US? I doubt anyone's going to be exporting all our EE jobs any time soon, but maybe that's a problem in other countries.

What's the reason for R1 AND R3?
I thought they were termination resistors, but I think there's more to it, since R3 shouldn't be there according to that theory.

>>1519563
provides a dc path to ground, prevents charge from C1/C2 from coupling into the prev/next stage, causing pops or damage

>>1519563
They give charging/discharging route for the capacitors when the pedal isn't connected to anything. If you remove them, connecting the pedal can produce a very loud pop when the capacitors charge through your amplifier's input or the guitar mic.

>>1519533
>The country is leaving a recession now (hopefully)
kek, ok

how does one git gud at electronics? My knowledge is rather basic but I'm not sure how to go beyond.

I've fixed headphones, mouses, toasters, mp3 players, a yoghurt machine, mostly by soldering one or two things back on. But what about diagnosing an amp? I just fell upon an early 80s Marantz, it worked for ten minutes and then died. Not certain where to look for a problem. Need to build up my security confidence to manage high voltage without shitting up anything.

How do you even unsolder a 10000µF cap without shorting it and make big sparks?

>>1519634
practice, practice, practice
diagnosing an amp, as it was taught to me: get or trace out the schematic (elektrotanya is an excellent resource). then, after checking obvious shit like power rails and fuses, you start taking voltage measurements and plotting them on a copy of the schematic. then you look at your plot and step your way from input to output and around the feedback loop, mentally converting between voltages and currents as necessary, looking for the place where the reading doesn't make any sense. (for advanced cases, you can do the analogous thing with ac, using a signal generator and a scope) also as taught to me: you can informally think of a BJT as a current-controlled current source. 1 electron goes from base to emitter beta electrons go from collector to emitter
of course, know shit like Kirchhoff's and Ohm's Laws, how diodes and transistors work, etc. there are books in the OP, maybe Geier would be a good start?
>security confidence to manage high voltage
be more worried about shitting up yourself. if you suspect there may be high voltage around, keep one hand in your pocket (literally) while probing around, to avoid killing yourself via fibrillation. the late, great Bob Pease used to solder neon lamps with resistors into a circuit where unsafely high voltage was present, as a dancing orange warning to not put fingers on components to take temperature
>10000µF cap
drain it with a resistor. got 1/4W resistors? need to discharge 50V? P=(E^2)/R and R=(E^2)/P. attach a 10k resistor with clip leads and come back in five minutes (three 10k * 10mF time constants), after which the voltage will have dropped to 50 * (1-(1/(e^3))) ~= 2.5V.

>>1518951
Or, if you have any sense of nationalism and don't want to sabotage your own economy, buy from local pcb fabs. In the US, pcbs.io is the cheapest.

>>1518966
I have tried this, but I have better luck dispensing solder paste with a microscope and reflow. I wish someone had told me a few years ago how important a microscope is for small pitch components.

>>1519678
And by reflow I mean, a toaster oven. No controller or anything else.

>>1519674
>practice, practice, practice

right, and that applies to everything, but I think some minds just absorb electronic theory better than others. I'm an EE and I just can't handle complicated non-digital circuits like pic related. In a radio or audio amp, something comes in here and goes out there. But in pic related, while that is still true, everything seems to affect everything else, so when my charger died, I probed around a bit and gave up.

>>1519678
>don't want to sabotage your own economy, buy from local pcb fabs
Yeah, I did that. The local factories cost ten times more and the quality isn't really better. Well, unless you pay 20-30 times more. They aren't even much faster than the Chink fabs, unless you give them even more money.
I'm not a Burger, though. Maybe the Burger fabs aren't that expensive.

>>1519680
ah, circuits with a magic chip and magnetics in them are a bit different from a mostly or entirely discrete audio amp
for these kinds of circuits, I'd test the easily testable stuff first: back-drive voltage into the output and see if the reference and opto-LED work, does the opto-transistor move if I vary back-drive voltage up and down, test the FET and diodes, look for obvious shorts/opens in the transformer and primary side caps, ensure continuity through the EMI filter... when you have eliminated all other options, it's probably the magic chip
>some minds just absorb electronic theory better than others
possibly true, early training in critical thinking, cause-and-effect, and basic arithmetic under time pressure (see Alexa's "Railroad Math" game) seemed to help me in practice

>>1519694
>burger fabs aren't that expensive
kek. Advanced Circuits has okay prices for protos on larger boards with mildly restrictive design rules, but aside from the panel pools, USA protos are expensive

>>1519707
>back-drive voltage into the output and see if the reference and opto-LED work,

that's a good example of where my "practice, practice, practice" is lacking.

thx, maybe I'll look at it again.

>>1519387
Ardufag here. I'm beginning to realize I don't actually know that much. I can use Sparkfun and Adafruit modules, and I can use data sheets ok....ish.......

So, how do I become a man so I don't die an Arduino onions boy bitch? Also, any tips on coming up with projects that can't be done by throwing Sparkfun/Adafruit modules into a big pile?

>>1519716
>that can't be done by throwing Sparkfun/Adafruit modules into a big pile?

don't fall for the meme. what you can do is cool. if you can learn some more advanced electronics, then that's cool too.

>>1519731
Thanks. I totally get what you're saying, and I don't mind the premade stuff, but I'd like to do some stuff that I have to build from scratch. Sometimes I get tired of thinking "Hey, I'd like to make this! Wait.... wait, yeah, I could just get a couple cheap boards, tweak some example code, and be done. The hardest part would be soldering (if I even need to) and enclosing the project." Sometimes that's fine. I've certainly done it a few times. But I like HAVING to make a new board from scratch becuase a commercial solution doesn't exist. Maybe I just need to get over myself, ignore existing solutions, and do it from scratch anyways.

>>1519716
meant to say onions boy bitch. IDK how TF I typed onions.

>>1519748
>not knowing about word filters...

>>1519748

lol are you seriously that new? welcome. have a soda

hi guys I have a huge doubt about ip and FPGAs. with the recent price drop on low power FPGAs i find it more appealing to use them for my projects instead of microcontrolers, i'm using several peripherals including several CMOS cameras, sd card, temperature, humidity, light, gas sensors,as well as an rf transceiver module and a display.
Most of these components come along with their own ICs working as interfaces which I find redundant since the fpga itself can be programmed to function as such. Shit is, would i be able o sniff around these ICs long enough in order to find the schematics and implement it on verilog or are we just not quiet yet wen it comes to open source hardware?

>>1519448
2 clips on a bar is what I would be using to hold a PCB, but one extra.

>>1519679
How about reflow in an oven with a thermocouple inside the thing and a little feedback controller hooked up to a triac/relay dimmer? I've been meaning to design such a device for a saucepan anyhow.

How bad of an idea is to parallel two bjts of diferent models?

>>1519767
Most FPGA softwares are living hell to use. Literally a jumbled mess of thousands of modules with a UI out of the 70s that manufacturers shat in our face. Don`t mess with FPGAs, they`ll bring only pain. Also VHDL and VERILOG are C A N C E R

>>1519777
Not that bad, if they're in the linear region both will act as individual current sinks/sources. Provided you're using individual base resistors that is. Using them as switches however and you could get a large discrepancy in their currents thanks to ones forward voltage being slightly lower, so watch out for that. If you're using a BJT as a switch then you should probably switch to a MOSFET anyhow.

>>1519783
MOSFET for linear psus are ok? People tell they are shit in the linear region for power stuff. Mine is 5A0-25v

>>1519786
With feedback I don't think there's anything wrong with them.

>>1519488
Oh, and
>what did you get her
-Opal & Diamond earrings
-Topaz bracelet
-Topaz ring I broke down and gave her early when she had a long day
-Vera Bradley messenger bag for her new laptop she just got
-Ceramic measuring cups/spoons with flowers painted on because she said she wants those in her kitchen when we buy a house together
-Nice stainless measuring cups-spoons because the ceramic ones are too nice to actually use and she complained the size markings on the plastic ones get washed off
-Pic frame with pic of us on our last weekend out of town with seashells we collected that weekend

And then I fucked up with her family because I was going to get them a $100 gift card to Olive Garden, but I really wanted to buy tools for her dad and ended up getting sterling-diamond necklaces for her mom and grandmother and a PS4 game for her one sister and I’m on the fence about whether I get her vegan asshole sister some film for her new camera, a gallon of ice cream, or nothing at all.

>noodz
Bruh, they’re good. I’m still amazed because I don’t have the patience for dime basic bitch white girls, but somehow these dime Latinas come at me. And this one is a keeper so I can’t post em.

>>1519528
I don’t need anymore flashlights. I’m thinking that will be a fun next step to get an idea of what I can actually build in my garage. I think a DIY power supply and soldering iron will be coming up on my list, and would be fun to build some sort of meter.

Could someone please recommend me a book on control theory? Thanks.

>>1519770
You don't want to reflow in things you'll be putting food in. And a toaster gets the whole board up to the right temperature at once (basically), not just one side

>>1518373
That water pan underneath the pot will fuck up your grounding, the pot needs to contact the table

>>1518500
anon, you might be altruistic

>>1519197
Your opinion should be shot

>>1519218
This had better be a meme, because even theoretically there is no benefit to pic related.

>>1519484
>soldering stations with hot air for around $100 from Amazon.
You should be able to get one for $50-80, chink quality really doesn't matter until you get up to $250 or so. If you want a GOOD soldering iron alone (which you definitely don't need at your level, and really don't ever need), then get one with the all-in-one tip+heater attachments, they have better thermal characteristics.

>>1519535
>>fil-it
As a burger, disgusting

>>1519760
kek

>>1519809
I dunno, pirate one

>>1519809
>recommend me a book on control theory? Thanks.

You're welcome.

>>1519758
>>1519760
I......um..... wtf is going on?

>>1519760
Also why do I get an 11 year old?

tard here
if i where to add a motor to use as a fan how would i change my circuit to fit it in since im using a pot to control the duty cycle of the led i would like for the motor to activate by a simple switch and not the pot without lowering the current to the led

tard here
if i were to add a motor to use as a fan how would i change my circuit to fit it in since im using a pot to control the duty cycle of the led i would like for the motor to activate by a simple switch and not the pot without lowering the current to the led

>>1519935
If you want to PWM dim it, add another transistor (and appropriately sized base resistor) to the output of the 555, pic related. Choose the base resistor by dividing the fan's intended current by about 50 or so (a low estimate of the hFE of the transistor). If you want it on all the time, just put the fan in parallel with the 5V voltage source. If you want to switch it on and off manually, put a switch in series with the fan (works for either arrangement).

Also since the LED pictured must be pretty high-power (3V voltage drop means 2V are being dropped by the 1Ω resistor hence the thing is letting 2A of current through) you'll be dissipating 1.4W in the transistor at maximum duty-cycle if the thing is switching fully saturated. It almost certainly isn't switching saturated though, since you have a 7kΩ base resistor, which would limit the collector current to somewhere between 30mA and maybe around 150mA. i.e. you're probably overdriving that LED to shit, since a normal 5mm LED can only handle 20mA. If you do happen to have a high-power LED, use a MOSFET to switch it.

And since a fan is an inductive load, give "fan protection diode" a search and add one in.

>>1519941
Thanks a lot I'll try to get this working
It's also a high power nichia LED, max current is 1.8A

>>1519941
>>1519948
It'd be better to put a pot in line with the transistor, and then tune the pot to match the current through the LED. Or since a 5V in leads to 3.3-3.5V out, just do a (3.5.-.7)/R = 1 to 2 A.

1.8A*2V = 3.6W, doable with a decent heatsink. Another option is to use a pot and a logic-level mosfet (start with 0 gate voltage!). I've seen more mosfets with tabs for mounting to a heatsink than bjts, but either can work.

Realistically he should be using a current buck regulator.

>>1519634
Sedra Smith Microelectronic Circuits was the book we used. Covered a lot of fine detail on bjt/fet amplifiers. There's probably "cruder" books designed for audio amplifiers and circuits themselves. Ultimately, without a more formal background you'll have a lot of trouble designing "your own" circuits, but most circuits are of integrated designs and are off the shelf these days. Not a bad thing, circuits have gotten way better than they used to be.

>>1519527
I made 50k starting and got laid off before a year. It depends on who you can get in with. I have been unemployed since February, despite pushing out tons of applications. I feel that CompE would be more lucrative, but I went EE and this is who I am now. I like Physics and maths, so I'm not mad about my choice, but for money it's not that great without connections.

>>1518450
I think it depends on iron temp.
>>1518739
I've found the opposite. I have a hot iron, and it burns the flux like crazy. Shitty 60W Home Depot Weller. I put a grounded socket + dimmer switch into an outlet box just so I could turn down my soldering irons. The best iron I use is the USB soldering iron, though there's been a few computer stuffs that've used some new high-temp solder that hates melting. It's not that it's lead free either, I have some old lead-free stuff I use all the time and it doesn't have an issue.

>>1519680
Majority of those components are just filters/fly-back preventers. The electrolytic caps (the polarized/curved-side caps) are the main tanks for stabilizing the voltages. The zener circuit in the lower right sets the output voltage, most likely. I looked it up, since the circuit clear isn't used for filtering/flyback prevention. The component is a programmable shunt. It looks like a zener so I had to search to figure out wtf it was. By changing the shunt voltage, they can control when the optocoupler is active, controlling or not the chip is dumping enough current into the transformer.

>>1519963
>iron temp
Oh, I (>>1518739) do use lead-free solder, and typically run my iron at 325-350°C because of it. Not that I can tell that's accurate since the numerated sticker on my iron's temp-control dial looks a bit off.

If you've got a non-temperature-controlled iron, then it's no wonder that you're burning flux. If you're not constantly hitting up a ground plane then your tip's temp will just keep increasing until it hits the self-regulation of the PTC element, which could well be over 500°C. I was going to compare it to those 8W USB irons that get up to like 550°C, then saw that you use one seriously. You should probably get a $10 iron with a temp-control trimpot on the side if you don't want to invest in a $30 T-12 station or whatever.

Anyone know if it's possible to use a tweezer tip on an Aoyue 888a station?

Hi frens.
I have this Black and Decker ri cooker that just stopped working out of the blue. I took the bottom off to look at the guts and it seems pretty simple. How should I go about diagnosing and repairing it? The light doesn't come on at all when plugged in. Nothing appears damaged at first glance, but I wanted to know what you guys think before I start taking it apart.

>>1519984
Do you have a multimeter? If not, the first step is acquiring one.

>>1519941
I was about to post that this circuit isn't working because most of the current is going to the motor and about 50mA is going to the LED but I changed the value of the 7k resistor to 1k and the LED is receiving close to 600mA. Very beginner in circuitry, I understand the reason for some of the resistors and know the capacitors in a 555 circuit change the frequency of the pulses but I'm kind of lost at what the resistors that connect to the transistors do

>>1519991
Yeah I have a multimeter, albeit a cheap one, it does the job. I was wondering what/where I should test.

>>1519984
>ri cooker
>unsure if typo or das raycis
Check if it has any thermal fuses that may have died out, test continuity of the microswitch. Does it have an electronic control panel or a thermostat?

>>1519984
First thing's first, look for a fuse, then jiggle all the connectors to ensure good contact, then check the resistance of the element (R = V^2/P = 120^2/30 = 480Ω), then check that the switches are working as intended. The fuse may be a thermal fuse bundled alongside a critical component. If the fuse/thermal fuse did blow, there was probably a reason for it and it might be within the cooker itself (i.e. wasn't a power surge). So make sure you check that the element hasn't short-circuited, and that the wires have a low resistance and aren't hanging loose before replacing the fuse.

>>1519967
The USB iron works very well, and the portability and rate it gets hot is very convenient. The replacement tips are more miss than hit I think, it's better to buy a new one (though I modified mine with a switch).

I had a temp-setting iron buy almost immediately the dial stopped working.

With a dimmer+socket I can change the power on any iron I want.

>>1519992
(3.5V-.7) / 7k = .4 milliamp, *beta ~= 100 = 40mA. Your circuit is working as designed. At 1k you get ~2.8mA * 100 = 480mA. Since estimating beta as 100 and I was 25% lower, same with 600/480, your transistor has a beta about 125. We can also just go directly, you need to halve/third your out resistor.

>>1519716
I think you need some pressure in your art. e.g. size limitations, performance requirements
for size pressure, you could start by putting some finished application on a single board, with all the outboard components included, to give you a better sense of board design, and get you thinking "now I could..."

>>1519767
you probably won't find schematics for existing complex ICs, no. and if you did, they would probably have significant analog portions that you wouldn't be able to merely digitize
you might be looking for the term "sensor fusion" in which you connect digital temp/humidity/etc. peripherals to an FPGA, have the FPGA querying the sensors, and deliver "fused" readings, possibly with some precalculation, to a host micro (which could be a simple core right on the FPGA, if there's room)
fwiw there are a couple of free IP core consortia; one of the more famous is OpenCores (creators of the Wishbone bus)
what FPGAs just got cheap, out of curiosity?

>>1519782
so? just wear your hazmat suit and stop licking your fingers

>>1519956
you don't need tabs if your FET has a low enough Rds(on). AO3400 in SOT23 can handle 5A continuous

>>1519977
that's a really good question. it looks like the 866 and 888A use the same handles, so the hot tweezers for the 866 might just plug right in

>>1520010
so you got a dud. millions of users didn't
>dimmer
you're only regulating power, not temperature. good if this hack works for you, but it is nonetheless still a hack

>>1520010
Er, 280, not 480. Beta is not a good measure if we're not sure of the output voltage from the 555.

>>1520011
You need heatsinking if you're going to use a mosfet in the linear region. He could dispense the 1ohm entirely and lose all the heat in a mosfet if calibrated. Or a transistor.

>iron temp
You think your iron temp is anywhere near accurate? If it's a cheap unit all it'll do is turn down a mosfet, it won't actually do sensing.

>>1520014
obviously linear dimming is poor counsel
>my temp
yes, cheapfag, my iron is digitally regulated and holds temp quite well

yay, my pixel led strip finally arrived from china.
it's so cool you can program each led separately
too bad it's going into THE box as yet another cool chink shit i ordered and have no idea what to actually do with it

>>1520022
post pic

>>1520024
of what? it's a regular 5m roll of pixel leds

>>1520025
i don't know what those look like

>>1519992
A bipolar junction transistor (BJT) is basically a really shitty current amplifier. It has a gain that varies wildly from ~50 to ~300 from piece to piece and as a function of temperature, and isn't really a constant you can count on for building a circuit around. The current from the collector to the emitter is equal to the gain (or hFE or beta) multiplied by the current from the base to the emitter. The base-emitter part of the circuit acts like a diode, so will basically have ~0.7V across it whatever you do. So without a base resistor, you'd basically be shorting a power source to ground through a diode.

The time it takes a capacitor to charge is proportional to C*R, so doubling the resistance or the capacitance will double the charge rate. The 555 timer charges the 100nF through 1 diode and through the 1kΩ resistor, and discharges it through the other diode and the "dis" pin, hence making a different time constant each up and down cycle.

>>1520010
If you wanted you could run your wall-iron off rectified mains, use an MCU to measure the resistivity, and PWM the thing to keep a constant resistance.

>>1520026
just type led strip into google

>tfw no way to work metal
>tfw no metal sheets
>tfw psu board is ready
>tfw I`ll have to make the case out of cardboard, plastic or fibre board.
Pls kill me if my house fire doesn`t

>>1520037
it wouldn't be completely unreasonable to mount the xfmr on a plastic box, to bring back some of that old tube gear aesthetic, minus the tubes

>>1520037
I’m tellin ya, you gotta get in good with some HVAC people. My dad used to take me to his shop after hours sometimes and it was so quick to make little boxes and cases like that. Grab some scrap metal, bend, cut, spot weld, and you’re done in 5min.

Or just break into an HVAC shop and do the same shit.

>>1520042
>spot weld
holy shit nigger you just gave me an idea, I have some BIG caps 65V caps here, I can use the power supply to weld some aluminum scrap. Thanks!

>>1520064
You probably can't, but go ahead and try anyway, I'm sure you'd find some retarded way to blow up the power supply eventually.

>>1520064
There was a video some anon posted awhile back and I tried to make something similar. Car battery (video recommended ~100CCA for spot welding nickel strips on his battery packs), relay, momentary switch, and copper nails.

There are a ton of DIY methods, but this was the most simple I saw.

I still need to fuck with mine more. I tested it with the battery to wire straight to nails and it blew holes in the nickel strips. Then I wired it all up and I obviously have a shitty connection somewhere because it doesn’t nearly have the same balls. It will weld the stuff, but it takes like 3 seconds of glowing before it melts and welds. Maybe I need a capacitor like you, and an oldfag recommended graphite rods instead of the nails. Video used larger copper nails too and all I could find at Home Depot were small copper plated trim nails.

>>1520065
>imblyign the real Bepis would ever tell somebody to not do something that would destroy electronics

>>1520068
the wire on those clamps looks pretty thin too. i might be wrong but i think every milliohm matters here so you might want to get something thicker.

>>1520071
I believe it was 10ga wire for the actual power, then like 18ga to the switch/relay. The same wire and clamps worked fine when I was going straight to the battery, but I guess it’s DC so it loses power pretty fast once you start adding more length to the wire, right?

I think the relay was rated for 40A, all of the universal relays at the auto parts store were about the same, so could be that too. Or just a weak connection somewhere. I shelved the thing shortly after finishing it so I need to go back and play with it more when I actually have a use for it.

>>1520069
> protecting your reputation this hard
> on an anonymous basket weaving forum
you're as cancerous as the nigger sticks you smoke you faggot

>>1520078
Jesus. I’m simply saying the real Beppy wouldn’t be caught dead posting such a logical comment.

>>1517636
What kind of plug is this one?

>>1520082

>>1520082
did you google the product?

>>1520085
I figured out the it is a 2-20p plug. But I don't even know where it is mainly used and where the fuck I can buy an adapter

>>1520082
That's a Kensington plug. Lower part N, upper part L.

>>1520093
too short to be NEMA prongs. more likely, it's some custom attachment for the manufacturer's custom national plug adapters
did you google the PRODUCT?

>>1520065
Stupid shitposter, with mosfets and capacitors.
>>1520068
thanks

>>1519786
>>1519787
Brainlet here, why would they be shit for power? Why would you need feedback?

Newbie here, why does pic related fuse not blow?

>>1520129
Because it is rated for a higher continuous current than that battery can provide.

>>1520119
Take a look at a power mosfet, like the irf510 or anyother, they have a safe operating area graph (SOA), SOAs are only defined for switching aplications.

>>1520134
Won't current be inf if resistance is zero? Or is this because of battery ressistance?

>>1520129
This guy sounds like he knows what he is talking about. Cheap 9V batteries like that don’t seem to be able to put out more than 0.2A. Remember it’s like tiny AAAs in there set up in series, so it’s only going go put out the current that those can provide. Not like a big lead acid battery that can put out hundreds of amps.

>>1520129
>measure short circuit current
>buy fuse for that current

>>1520011
Thanks, that's some good advice. One of the projects I've been focusing on is a data logger. I don't have a 3d printer (yet) so I've had to use what I have for enclosures. I decided I want to make a data logger that can fit in a pill bottle. I want it to work with a variety of sensors that can be swapped out, and I want it to last for months, even years if possible, so it has to be very low power. I want it to be very tough too. A non-techy user should be able to bury it in a field, strap it to a car, etc and come back months later and download the data from the SD card or EEPROM (via RS323 to USB adapter and a computer application). This project has held my interest for a long time, and I think it's because it has such serious restrictions.

>>1520149
Thanks, how does that fit into the v=ir equation?
>>1520159
Last time I measured short circuit current(with mains admittedly) I had to get a new meter.

>>1520167
take ten 1 ohm resistors, pu tthem in parallel, measure the voltage ;). It fits in ohms law because batteries have a internal resistance which limits the output current. Capacitors have less, which is why they are used when you need a big surge of current like in welds and stuff. That navy gauss cannon used a hugefuck capacitor bank to power the coils.

>>1520182
There was a Youtube video I watched, I can’t remember what dude was making, but he got some fuckhueg capacitors from a hybrid bus regen braking system. Maybe it was another one of the spot welders but I think he said he got a few of em used for <$100. Now that looks fun to play with.

well lads I bought an SMPS for my stereo amplifier, and am trying to figure out how to hook it up to mains AC. I am familiar with the screw terminals for +/- but not sure how to address the earth pad thing around the standoff screw hole, in the top left corner.

Do I solder the earth wire from my mains cable (UK) to one of the tiny earth pads?

>>1520198

Doing something with nixie tubes, forgot to order decoders and can't wait for them;

Can I just use my uC GPIO to drive the nixies? I'm confused as to how to get it to work - putting the pin to low or high just changes the voltage from 0 to 5 volts, which would still be enough for nixies to have ~150V drop

Does simply configuring the pin as output prevent the current from being drawn?

Configure it as input to light up the tube, configure it as output to turn off the tube?

>>1520136
plenty of power BJTs have SOA graphs as well

>>1520257
did you even read the datasheet and look at the output arrangement on the pins where you'll see a big fuckin' diode to Vcc
did you even google

>>1520262
I know they have, but mosfets usually show their SOAs ONLY with switch times. Not continuous running, because they are designed for that.

>>1520218
I don't have a frame/project box to screw it to rn

>>1520262
>did you even read the datasheet and look at the output arrangement on the pins where you'll see a big fuckin' diode to Vcc
>did you even google
So you're saying it won't work?
Because Vcc is 5V and nixie voltage is 150V, the diode would get raped?

>>1520264
yet, they still work, and one can derive the SOA for any pulse length up to dc. amazing educational resource follows
https://www.infineon.com/dgdl/Infineon-ApplicationNote_Linear_Mode_Operation_Safe_Operation_Diagram_MOSFETs-AN-v01_00-EN.pdf?fileId=db3a30433e30e4bf013e3646e9381200

>>1520119
according to the above resource, one reason you need feedback is thermal runaway at low Vgs

>>1520283
at best, the clamp diode will simply hold all segments on continuously, for a little while. can't speak for the health & safety of anything else on the 5V supply tho
so use an external transistor. the ULN2003 array, commonly used in this application, is rated for 50V. you might (or might not) be able to get away with a 2N2222 or 2N3904 or 2N4401, all rated for 40V, or you might need something a bit stronger like a 2SC1815, or a proper high voltage transistor like an MPSA42 or KSC1009

>>1520305
>you might (or might not) be able to get away with a 2N2222 or 2N3904 or 2N4401, all rated for 40V
I'll try it with BC547B, have tons lying around, thanks!

>>1520273
then use a nut and a bolt

>>1520328
fug thads smart arigatou

>>1520317
>>1520305
hey it works what do you know

>>1520337

>>1519984
>How should I go about diagnosing and repairing it?
start at the electrical plug
measure continuity from the prongs to the inside of the cooker
keep following continuity until you get to the heating element
the heating element should have somewhat low resistance
you'll have to trick it into thinking it has water in it before you have a complete circuit from one prong to the other of the plug

>>1520305
>yet, they still work, and one can derive the SOA for any pulse length up to dc. amazing educational resource follows
Thanks m8.
>>1520037
So I found some sheet aluminum and those aviation pliers or
whatever they are called and made pic related. Now instead of a housefire I`ll be received with a nice shock everytime I handle the PSU because of the lack of grounding in my building.
I literally cut it with those pliers and bent them with my hands. Feeling very Pajeet. Atleast it fits and they slide over eachother without much obstruction.

>>1520375
I`ll buy screws tomorrow now all I need is a reliable way to measure current, a current limiting circuit, and to program a pic or a atmega328p to measure voltage and current and display them on some 7 segment displays on the panel that`ll make.

>>1520375
>aviation pliers or
>whatever they are called
snips, aviation snips

>>1520375
unironically pretty impressive for a first time with less than basic tools
>measure current
the easiest way would be a power resistor in the - leg of the supply, where V is directly proportional to I. you could use this for both readout and limiting
just make sure your voltage reference for regulation is relative to the load side of that resistor

>>1520386
I`ve already put a equivalent 0.1Ohm resistor in the input (after rectification, before controls and pass transistor). I want to make a foldback current limit thingy but can`t think of a pnp version. (See pic).Both the control board and power transistor board have a opamp with a unused side (lm358) so maybe I could use them somehow but I think the maximun input is vcc-2v or something. I could just chug a pnp with the emitter on the input, base at the emitter of the pass transistor and colector on the base, but that would consume more power than foldback if it reaches the current limit.
>unironically pretty impressive for a first time with less than basic tools
thanks m9

>>1520375
>because of the lack of grounding in my building.
god why

>>1520386
Putting a resistor on the - leg makes me feel icky. + is much nicer.

Can I use a 12v relay with a 9v battery?

>>1520617
probably

>>1520618
What about 30v relay?

>>1520620
probably not

but try it, nothing's going to break

>>1520397
that's why in the - leg of the supply: your 358 should work fine at and near ground

>>1520504
y tho, it's how it's usually done, esp if your pass device is npn. so you variably level-shift the secondary side of the power transformer a bit in a load-related fashion, whoop de doo

>>1520617
coil rating ≠ contact rating. which one are you on about?

>>1520629
Oh ok thanks
Says coil 12v and contact 30v. Is 30v the max?
>>1520626
>>1520618

For some reason no stores here have 9v ones. Just tried, works, thanks.

>>1520645
30VDC is the maximum that the contacts could seperate without making a nasty arc. 12V is the recommended for switching the relay on, but there's some leeway between not enough voltage to open the relay and so much current that the coil burns. You can always buy a 5V relay and put a resistor in series with the coil to get the intended current.

>>1520420
Because brazilians are stupid, building codes are strict but not enforced, and even if you make a correct project the ``````````contractors`````````` just wing it at the building site.
>>1520629
do you mean put the 0.1 resistor in the - leg?

>>1520650
correct

>>1520648
That cleared it, thanks

>>1518450
>I've never understood this flux meme, and I solder a lot.
Maybe if you're only soldering through-hole components and largish SMDs, but you can't really do something like pic related without applying flux liberally.

>>1520683
OP is diamonds

I want to measure a couple of 10 to 50V values using a uC, it has a 4 channel adc. If I put those values directly in the uC it`d fry. I`m thinking of using resistive dividers, is that a good aproach or there are better things I could do?

>>1520707
a voltage divider is fine and probably the most accurate thing handy

Using chinkshit makes me feel like the macgyver
I had to improvise here because of course the battery they use doesn't exist in the civilized world

>>1520719
you sure?

>>1520721
yes i am, no size of regular a batttery fits there

>>1520719
i will probably have to buy another of those fatter batteries, since when the thin one runs out of juice the fatties will be stuck at half capacity left and going nowhere since the thing one won't be coducting

>>1520722
really, it looks like an A23 battery slot to me, which is a pretty regular battery for things like garage door opener remotes and the like
post dimensions

>>1520730
I already fell for that trap and now i have an useless overpriced energizer A23 battery in my drawer

>>1520740
measure, and compare to a list of battery dimensions. I'm actually kind of curious now

>>1520829
Nah i'm too lazy to do that, it works perfectly with my solution so that is good enough for me, i think the inteded battery is supposed to be several lr coin cells

>>1520683
How do you do this? Flux the shit out of the component and pads, put a blob on the tip, and drag it across?

I've got a mainboard from a Sony blueray player that won't turn on. I've been mapping it out the best I can to figure out what's wrong with it, although at this point I'm 75% sure it's an internal failure in the CPU because there isn't much in the way of components on this board and I've ruled out the obvious issues AFAIK.


The confusing thing is that I keep finding "traces to nowhere". Some go to a single SMD pad that's been tinned but not connected, and the pad isn't part of a larger footprint that just wasn't populated either. It's just a single, lone pad. Maybe they're test points, but they don't look like traditional test points. I'm also seeing some traces that end in vias. They just hit a dead end, go to a via, and the via doesn't connect to anything on the other side of the board, not even the ground plane. Is this how multilayer boards connect outer layers to inner layers? The vias could also be test points I guess, but I've never heard of that either.

>>1520848
The vias are probably going to inner layers. The single pads are likely test points. Can you post a pic?

>>1520846
yup.

>>1520846
in short, yes, that's drag soldering. there are hoof tips with a depression to hold extra solder but you can use a plain bevel tip with slightly impaired efficiency

>>1520846
>put flux on the pads
>insert the component
>put some more flux on top of the pins
>put a tiny blob of solder on the iron
>run the iron across the pins

These fucks are also a pain in the ass to solder unless you apply flux on the pads or the contacts. There's practically no surface area you can apply heat to.

>>1520850
Well I took some pictures but apparently my phone takes pics that are over 4mb. I'll try to think of something. Any ideas?

>>1520850
OK I know it's stupid but I just downloaded it and cropped it to get it under 4mb. I know there's a better way. But for now, here it is.

>>1520830
>several lr coin cells
Yes, four of them in series. Battery is called 4LR44, 6V.

>>1520875
It looks like those are non-populated passives

>>1520875
The example I'm getting at is in the middle here. There are three test points on the this trace in the middle and the two vias on the end go nowhere on the other side. This is a 5v rail if it helps. I've been trying to find the 5v regulator that generates it, but the only IC attached to it has 5v on two pins, 1 ground pin, and two pins that connect to the cpu. The external PSU is 12v, so I doubt the cpu is sending 12v to this chip. I'm too nervous about measuring the voltage on the two pins connected to the cpu though, because I don't want to short them out if.

>>1520879
But what would the upper two be populated with? They connect to the same trace, you'd just be shorting the component out.

>>1520884
Hard to say. The trace is a little thinner between them. Maybe they put that in in development with the idea that if they needed a resistor there, they'd cut the trace, then left it in for the final design.

>>1520887
Any idea what this is? It's the chip I described above that seems like a 5v regulator. The top two pins are both at 5v, and one powers a USB port and the other powers the power LED (which is controlled by a transistor connected to the CPU). The middle pin is ground, and the bottom two connect to the CPU. Can I test the bottom two pins without shorting the CPU?

>>1520908
Keep your multimeter in VDC and you can test whatever you want. I don't know what that chip is but maybe someone else does.

>>1520912
I did that once when I measured the voltage across a voltage-limiting resistor for an LED. The LED blew out and I always assumed it was because the multimeter shorted the resistor. I did the same for the transistor for the power LED on this board to turn the LED on. Although now I'm realizing that this makes so sense, because you can measure the voltage on a battery and it doesn't melt your leads so...... now I'm totally fucking lost.

>>1520940
Either your multimeter is fucked up or you had it in current mode. Voltage sense is high impedance, so the circuit doesn't see the meter

>>1520944
Soooo I just tested this. My main meter, the one that I've seen this issue with, is Sparkfun's generic 830 model. I put it in 20vdc mode and touched it's probes to the collector and emitter of the transistor and the LED turned on. Then I got a Healthkit 2372 I inherited and did the same thing. The LED didn't light up. How is this possible? I always thought this was weird but it never fully occurred to me that this was not how it was supposed to work. I guess I'll be using the Heathkit from now on. I've been wanting to invest in a nice Fluke (or the EEVblog meter) for a long time now, and I think this just pushed me over the edge. Any recommendations for a model? I'm willing to drop $200 for a meter that I can get reasonably rough with, that will do pretty much everything I could want, and that will last a long time. I'd like it to have capacitance, inductance, etc functions because I currently have nothing that can do that.

>>1520944
Your meters have infinite resistance? That is pretty cool, if you sell this technology you can make billions.

>>1520963
I said high, not infinite

>>1520961
I don't understand the fluke worship. Half of the fluke meters I see don't even have basic stuff like capacitance sensing. I use something like these:
https://www.amazon.com/Professional-Counts-Multifunctional-Digital-Multimeter/dp/B01J2J6QVG/
https://www.amazon.com/Multimeter-Auto-Ranging-Electrical-Resistance-Capacitance/dp/B075CYZ4N5
and my meter works great for electronics.
I guess if you're doing mains work and trusting your life to the meter then get a fluke for the name-brand QA. But otherwise I don't think they're at all worth the money

>>1520944
>>1520963
This is why I keep even trivial instruction manuals. the manual for my meter says it should be 1MOhm. Is that too low?

>>1520993
My Heathkit meter just confirmed 1MOhm on the meter.

You know,now that I think about it, I may have used ohms mode when I blew out that LED years ago. So I would only look the transistor issue for this. I'm going to try to find the input impedence on the Heathkit and then short the transistor with a wire to see if the LED is even brighter (if it is, that means the input resistor on the shitty meter was at least doing enough to dim the LED).

>>1520999
lol so the Heathkit has 10MOhms input impedance. Does this explain my problem?

>>1520856
a small bevel tip is good for them. put solder on the tip, slide up to the tip of each lead, and pull back once it flows. still needs bathed in flux

>>1521008
it explains the fact that you're seeing light on the LED when you put the old meter on the circuit. a few microamps * a few hundred = 1mA, enough to visibly light that shit up
>>1520999
>short the transistor with a wire
if you don't know how the transistor is connected, you risk blowing out whatever is driving the transistor
tbqh you need to understand that you are working on a computer. you're unlikely to successfully debug a computer with a multimeter unless it's something basic like a power supply problem. in all likelihood you will need the service manual to know right from left in this thing

>>1521033
>a small bevel tip is good for them. put solder on the tip, slide up to the tip of each lead, and pull back once it flows. still needs bathed in flux
Exactly my point - it needs to have flux. All the retards ITT who think flux is useless seem to have no idea just how much it's necessary when you solder something more than an LED to an Arduino.

>>1520963
don't vacuums have infinite resistance

>>1521059
You need to have SOME conductivity to measure the voltage.

>>1521079
Not true, you can measure the voltage through purely electrostatic means. Even modern ADCs don't work like a galvanometer-based voltmeter of old.

>pull back once it flows
>needs bathed in flux
Flux is the lube of the unskilled.

>>1521116
stop trolling

>>1521091
huh. Can you elaborate? I guess I'm not entirely sure how ADCs actually do their thing either.

>>1521033
I was going to short the emitter to the collector. That should be fine, right? The base, and really the whole transistor, shouldn't really be effected at all.... I think.

And yeah, I figured I'd only be able to figure out so much with a DMM. But the blueray drive shows NO signs of life. The power light won't even come on when I hit the power button. So I'm starting by looking for power issues. Do you have any advice on looking for other issues? There are very few chips on here, and if the CPU is dead, IDK how to tell.

>>1521121
An ADC can be made with a comparator and a DAC that searches through a selection of output voltages until the comparator finds the values on either side of the measured voltage. If your comparator is constructed with MOSFETs, then there's effectively no net current into the FET gate, it's just a capacitor.

In a similar way, voltages can be measured by introducing a capacitor to a system and measuring the charge that flows into the capacitor. Use a very small capacitor and this charge can be insignificant compared to that stored in the stray capacitance of whatever it is you are measuring, which is similar to how an electroscope functions.

>>1521122
yeah, that should be fine. probably won't tell you much tho
I'd probably look for a logic signal from the mainboard that enables some of the power supplies on the power board. I'd probably be googling for a service manual right about now, too. did you check elektrotanya?

>>1521121
> I guess I'm not entirely sure how ADCs actually do their thing either.
There are a bunch of different techniques with different trade-offs between speed, accuracy and cost.

But regardless of the technique used, putting a FET op-amp buffer in front of the ADC mechanism results in negligible input current. Unlike BJTs, FETs don't need to draw current to operate. Any current which they do draw is a consequence of the fact that nothing is a perfect insulator.

>>1521129
That signal has been on my list of things to look for and I'm about there. From what I can tell so far though most of the rails seem to be on all the time. The power button goes right to the CPU so that's not very helpful. I haven't finished tracing +12V yet though. I haven't heard of elektrotanya before, I'll give it a shot. Getting model number for this might be tricky though. I can't seem to find the manual (it was my parent's) and there's no model number on the case.