/diy/ - Do It Yourself

this thread's digits brought to you in part by the MAX9791 Windows Vista-compliant class D speaker amps with headphone amps
(oddly specific, that)

Thankfully at my previous job everything was C and high structured. Would've been nice to get to work on the 4-processor board though.

>>1489795
>it's real

hi /ohm/ ,
anyone know if i can power a op amp with an arduino ? i know op amp need +x and -x v , how do i get the minus voltage with the 5v of 3,3v of the arduino ?
pic related is the amp im trying to interface
https://www.mouser.fr/datasheet/2/609/AD8221-748861.pdf

>>1489917
Well you could just throw a square wave into a capacitor and use a diode or two to reference it to ground, which will give you a -3V-ish rail. But when it comes to low-voltage DC op-amps I use something that can run off ±5V (or ±3.3V I guess) with the ground rail at halfway (2.5V or 1.65V respectively). At those low voltages it's best to use a rail-to-rail amplifier. Looks like the OPA2350 can get you those specs and it's on ali. SMT tho. But searching "rail to rail amplifier" got me nothing with more than 5 orders, so I don't know what I'm doing wrong.

I use a voltage divider between 5V and 0V to set my ground rail, usually with a buffer amplifier on it to keep it stable but sometimes just a couple of capacitors instead. Depends on how many op-amps I have free.

TL;DR, decide what op-amps you need to buy before buying them.
t. not an expert

>>1489939
wait I meant to say ±2.5V/±1.65V ∆V = 5V/3.3V. Anyhow your AD8221 can take ∆V=5V but since it's not rail-to-rail you'll be looking at a voltage range between 1.9V and 3.9V. That's fine for low-voltage stuff, but if you want to amplify it before it goes into an ADC for maximum resolution or send it to some earphones I'd reconsider your choice of op-amp.

pic related

>>1489917
If you want to run it all from 3.3V, rail splitting is not an option but you can generate a negative voltage with a standard charge pump circuit powered by a GPIO pin. You can then connect the Vref pin of the amp to the common GND and the negative supply does not need to be that stable. D1 and D2 should be Schottky diodes like the (cheap) 1N5817 because you lose 2Vf. The supply current of the AD8221 is only about 1mA.

>>1489939
>>1489942
>>1489973
Many thanks for this well detailled explaination, but i ask myslef about the real purpose of the ± supply, can't i just put the minus pin at gnd ?

>>1490015
if you want to multiply two numbers, you need a zero to multiply against, right? preferably one shared with the rest of the system.
instrumentation amplifier ≠ op amp btw

>>1490016
i don't get it, what i want to do is amplify the signal coming out of a photodiode, and put it into the input range of the arduino (i use a due so its 0-3.3v), what difference should i acknowledge between instrumentation amplifier and op amp ?

>>1490015
>>1490017
you do not need a negative rail for this circuit, just a rail-to-rail op amp. none of the terminals of this op amp will ever be below 0V, specifically because the op amp's feedback will produce a positive output to cancel what would otherwise be a tiny negative voltage generated at the inverting terminal by the photodiode.

>>1490017
an instrumentation amp is a block that's designed to reckon the difference between two voltages and multiply them by some relatively low gain factor, all with respect to some given zero point
an op amp is just a high-gain differential amplifier that multiplies the difference between its inputs by a very high but imprecise gain factor
a plain op amp with a low input bias current and a common-mode range slightly below 0V is good enough for a photodiode detector. an LM358A might do in a pinch, with a typical input bias current of -15nA (but max of -100nA), which may be adequate for your application if you don't need much dynamic range

>>1490017
There's an unconventional way to use an instrumentation amp to measure the photocurrent of a diode. In case of the AD8221 (± supply) you get Vo=49.4kΩ*Iph or about 0.5V at 100lx for a BPW34. The conversion gain is fixed because there is no access to Rf.
Source: https://www.radiolocman.com/shem/schematics.html?di=184339

>>1490028

Yeah ive seen thoose LM358 boards, but ive already had the pink instrumentation amp so it would be nice to use them, also i need good stability are repetability, and high frequency range so a dedicated board made for instrumentation sound nice (i guess)

>>1490050

dont understand your shematic, is the blue region the AD8221 itself ? if so this is nice, less wiring is better for me, that is also why i bought a ready to use amp board

>>1490072
>blue region the AD8221 itself ?
Yes, simplified internal circuitry.

if i use voltage regulator in constant current mode is it safe to short the output wires?

>>1490111
if it's a boost converter in constant current configuration, you might want to think twice about doing so with your hands
if there isn't a load on it already, there is some risk of blowing the regulator going directly from open to short
otherwise, should be pretty safe

>>1490050
bretty clever

How do I increase the maximum unclipped output of a two stage common emitter to emitter follower circuit?

>>1490111
I believe they can have a minimum voltage that isn't 0V. But it's surely specific to topology.

>>1490191
Further apart voltage rails?

>>1490191
Increase the supply voltage and recalculate the circuit for max Vout.

>>1490191
get some dc bias into Q2, by shorting C5 or adding a resistive divider to the rails

Does it make a difference in energy use of a circuit if I use a potentiometer as a resistor or as a voltage divider?
I'd guess not because using a pot. as a resistor in series basically is a voltage divider right? So if I use it as a resistor or connect all three terminals doesn't change efficiency right? Obv. to do the same thing within the circuit.

>>1490191
Having a 10k on that emitter output is atrocious. Not to mention that circuit would have 50% loss of signal due to the lack of biasing. It'd be better off reducing the resistors for the first stage and coupling that. I saw that the circuit was used for a mic pre-amp, but even then, it doesn't make sense.

>>1490273
I was talking about in a general sense I just got the circuit off google

>>1490251
a voltage divider pot has a constant power consumption and a linear voltage to turns relation. a pot as a resistor has neither, but it will consume less power simply because you've got another resistor in series with it. since you can just get a higher resistance pot, there's never a reason not to use it as a voltage divider when you're just trying to get a reference voltage.

>>1490375
Unless you want the nonlinear behaviour of using a rheostat in series with a fixed resistor, which might be helpful for audio (pseudo logarithmic) stuff.

I'm getting a new (used) car but it won't have an aux jack to plug my phone into and it doesn't have a tape player either. It seems the only option to play music from my phone is an FM transmitter or gut the console and replace the stereo which I did with my old car and while it worked it was messy and some stuff got damaged while getting the old unit out. So I'm probably looking at an FM transmitter.

I'm looking for a circuit that will accept stereo input. I do not want to mix R and L to get mono and then modulate that. I want true stereo. On top of that I need to be able to add a power amplifier after the output to drive the antenna. Yes I know this is probably illegal as it'll exceed the low maximum output power that FM transmitters are legally required to have but I think it's unlikely I get caught in a constantly moving car and I really don't want interference fucking with my day so I'm willing to risk it. Even then I'm probably not gonna put more than a few watts of power into that antenna so the worst thing is maybe a car in front of or behind me hears a bit of whatever I'm listening to if their radio is dialed into the same station. It's not like I'm gonna interfere with some radio station outputting at 10kW.

>>1490388
for FM stereo just >buy it

>>1490394
If I get a commercial one can I at least mod an amplifier on the output of the mixer but before the antenna?

>>1490388
And why would a DIY circuit be better than existing solutions? They automatically seat themselves in unused frequencies and I imagine they transmit stereo. Plus some of their enclosures are handy for docking a phone directly to.

>>1490396
Yes you can, but their output amplitude is typically more than sufficient. Regardless, I think they already have an RF amplifier there, so you'd just need to change the gain resistors. Also only AM radios use mixers, FMs use some other wacky shit like FCOs or PLLs or something.

You guys know how all the test equipment out there with those nice push button switches and/or keypads? where the fuck do the companies get those made? I want to have a keypad in a potential future project but I hate the *, 0, # on the bottom. of the 3x4 keypads. Or in the case of the 4x4 keypads, what I mentioned above, and the A B C D keys/labels. Where would i get custom ones made (just out of curiosity) and short of that, how would I change the labels on those buttons to be ones I want?

Hi I’m new. Are there any brick and mortar stores that sell stuff to start making circuits?

Any recommended ICs for lithium ion boosting to 5V? Looking at fairly low-power, no smaller than 1/20" / 1.27mm pitch, and possibly room for overdischarge protection. Though the latter should be simple to implement with an op-amp and a FET or so. I'm already going the TP-4056 route for charge protection, but if there's an IC that encompasses both and doesn't shut down at low current I'd also be happy to switch to that.

>>1490421
Depends on where you are, but in most places there are at least online stores with fairly low shipping from local warehouses. Physical stores typically have huge markups that you'd do well to avoid. But I just go from china with free shipping for very cheap prices. 3c each resistor isn't even very good on AliExpress, provided we're talking SMD.

>>1490422
"low-power" is not a mA or W rating. I'm sure you could find plenty of single-chip solutions (or chip + cap) but you're not actually speaking your problem.

>>1490439
I've no idea what kind of power expenditure I'm after since it's a hodgepodge of different levels of a lot of different types of ICs and modules. All I know is it should be less than 1W. Perhaps 500mW at the most, 50mW at the least. The thing has multiple settings, so at the lowest power setting I could probably get below 50mW. Thanks CMOS.

I'm looking at the FP6298 (from "Feeling Technology") right now and it covers my specs (input 2.6-5.5V, low shutdown current of 0.1µA) but it has a quiescent current of 6.5mA while switching, and I think I might be able to do better with something that isn't designed for a maximum of 4.5A. The MT3608 looks a bit better at 1.6mA when switching and doesn't have a heat-sink pad underneath it (no reflow), though it does have a <1mm pitch. I just feel there's a lot of options out there and I'm not optimising very well.

Ok so i am proceeding with my simple smart fuse.

Basically it uses a hall effect monitor to shut off the power it the current breaches set limit.
So, what if i upon the breach of current, instead of turning the power off, it would rerouted it through a lm317 set into constant current mode, set to that max current limit?

Now i know this would work.. BUT... what happens if the device only wants to draw more than maximum allowed current for say 10 seconds, and then wants to go back to drawing say only 50% of the max allowe current, wont the lm317 just keep showing same into it forever?

Pic related

>>1490463
Or just power it through a current-mirror/regulator-based current sink in the first place. When you have a large enough load resistor (it wants to pull little current) there the mirror will simply send in as much voltage as it can, so you'll get roughly Vcc to it and hence less than the maximum current. Increase the current/decrease the resistance enough and the voltage that the current mirror wants to put through the resistor is such that the voltage across the load is less than Vcc, at which point it's regulating current. Problem with this system is that you can really produce a lot of heat in your regulator if it's set too high, but for 100mA or so you should be fine with a single LM317 and no other circuitry.

>>1490466
>Or just power it through a current-mirror/regulator-based current sink in the first place
i have no idea what that is, but it sounds pretty complicated, i want to keep it really simple. heat is not an issue i have massive heat sinks.

What i want to know is if with the setup i described the load will be able to actually go back to using less current after the relay triggers on, of if the lm317 will just keep the relay on forever since it will just keep forcing the current through the load and the hall effect sensor will keep detecting that and keep the relay on.

>>1490474
lm317 is a linear regulator. In the constant current mode lm317 will not output more voltage than provided by your power supply. It means that even though it's called "constant current mode", it will not be able to keep this current up if the load is not able to sink it while keeping the voltage below Vcc-dropout.
If you remove the load completely, you'll get zero current (obviously) and about Vcc-dropout voltage in the output. With a true constant current source, the voltage will go to infinity, trying to sustain the current constant.

>>1490474
> if the lm317 will just keep the relay on forever

nobody can answer that but you because we dont know at what current the hall-effect thing will trigger on. if its higher than the LM317 current then it'll reset to direct operation, otherwise it wont.

>>1490484
>if its higher than the LM317 current then it'll reset to direct operation, otherwise it wont.
When the load hits the hall effect limit then it will engage the realy with lm317 and instantly turn off since it's now bellow, so the relay will go into rapid fire mode and blow up

>>1490474
current mirror
make R_load larger than R and the thing will act like a normal voltage source
make R_load smaller than R and the thing will act like a current source

>>1490399
>>1490398
>>1490394
Well fine you sold me on a commercial unit I guess. But I didn't get one of those cheap ass 10uW ERP FM transmitters. Nah, I got one of these 0.5W transmitters. Takes 12VDC in so I just need to get the right adapter and it should be fine. I'll probably pick a band up around 107MHz which is relatively empty around where I am. I don't anticipate having interference issues unless I go down to NYC or something.

>>1490422
>no smaller than 1.27mm pitch
I hope you'll make an exception here for SOT23-6, otherwise your options are very limited. if only you could do 0.5mm QFNs you would have a shit-ton of options, some of which do everything in one chip (except battery protection) and even allow you to run while charging (which is a bad idea with stand-alone Li+ chargers)
>protection
the DW01 (SOT23-6) and two external FETs (often TSSOP-8 but obviously use what you like) is the cheap, available game in this town

>>1490447
anything bigger than SOT23 is probably going to be designed specifically for larger output currents. this would be a fine opportunity to stretch your wings a bit
if you have that small of a load, switched-capacitor converters might be worth looking into

>light #2
Ok so this project I’m using this red lighr as a base. Got 3 of those little ~2.5” LED strips, wired them in parallel, attached them to the face for extra brightness. Got a dimmer switch, but that cunt gets really hot so I think that plan is out and I have to find a PWM deal or use the regular on-off switch.

Ripped out the AAA batteries, installed Li-Ion. Cheap power bank provided the USB charging port and protection for the battery.

My issue: Same shit happened when I ripped apart the other power bank- when I had the power bank all together, it powered the LEDs fine. I rip it apart, get the board off, wire from the board straight to the light, and it will power a small light fine, but then it will flash and shut off like an overcurrent protection when hooked up to the bright light. What changed? Could a shitty soldering job or connection make the light pull more amps?

There is also some sort of switch on that board that puts out 2.2V when nothing is plugged in, then the full 4V+ when a USB is plugged in. But since I ripped that USB connector off, I need to figure out how to wire it so it puts out full voltage when the light is on. Might put a 3-way switch so it goes from 1:off to 2:2.2V—>4V to 3:lights on.

>update
Just found some power banks with a 2.1A output so I’ll tear that apart when I get home and see if it holds up.

>>1490495
Seems interesting i will look into it to see if it will be a better option for creating a current limiter than my relay setup

>>1490517
One more question- what devices use PWM? Is there any cheap gadget I could buy at Walmart to rip apart for the switch that could run ~4V?

I had a busted DeWalt drill and took apart that trigger but can’t figure out the wiper mechanism now that the pieces went flying everywhere. It worked when I still had the trigger in one piece.

Suppose I want to use a 3.3v microcontroller as a CNC controller.

Exactly what should I do to keep this puppy from making the mcu misbehave?

The MCU already is inside an old flex-atx case, but I could make a folded steel sheet case later on for superior noise immunity. What should I ground to what? Should I ground the motor shell? The atx case to PSU gnd? Should I use a small can just for the controller unit?

>>1490521
spoiler: it's not
it eats > 50% of your power unless you add emitter resistors to unbalance it
a single current sensing series resistor that draws drive away from the pass transistor will almost certainly be more efficient

>>1490527
what is your reason to expect that it would make the mcu misbehave?
>What should I ground
everything to a common point (aka star grounding)

>>1490521
It's not a limiter. It is a constant current source (PNP) sink (NPN). It delivers a constant current to your load.

>>1490530
so when are you going to show him the current multiplier instead of wasting 50%+ of his power as heat?

>>1490529
>what is your reason to expect that it would make the mcu misbehave?
When people use spindles like that with Arduino they complain of the arduino disconnecting from the PC and false-triggering the endstop sensors. If the noise spikes can interfere with a 5v controller, they're probably even worse for a 3.3v one

After I ground everything to the same point, what do I ground it to? To the circuit's ground?

LOLI COLLECTION :


urlcut.ru/PORNO


==========
rgtykiomkevrtbhynj

>>1490532
I don't think he's looking for a transconductance multiplier. Maybe if he was trying to make a mixer or log amp or something but that's really not suitable for this application.

>>1490536
you need to draw a diagram of signal and power connections in your setup. there are too many variables
>what do I ground it to?
earth

>>1490463
next time use real dimensions (voltages/currents) on your diagram

I`m hacking a webcam to turn it into a microscope. If I invert the lens I can see individual grains of dust and bubbles in the diode glass, I`l need to make a casing with adjustable focus for it to work tho

>>1490574
These are the pics without inverting the lens, the inverted ones are crappy without a proper enclosure and slices

What value and type of fuse (roughly) for 230V should I use to protect a circuit with two SMPS in it, one 12V/500mA and the other 3V/1000mA? I don't expect either of the supplies to operate under more than 10% load at any given time, and I want the fuse to blow if they ever start nearing their limits, because I'm kinda paranoid about fires (I put a 60°C thermostat in there, too...)
Now I understand the maths (accounting for the meh efficiency can I expect out of these chink supplies) but I have no clue about how the power draw develops when I plug one of these in. Is there high inrush current that will blow my fuse even though the real load will drop off to a fraction of that? Will using a medium or slow fuse help? What is a good value to start?
Sorry for the retarded questions, really, but I figured it might be a bit too specific for /sqt/.

how do capacitors affect components if/when the only connection common to both is the ground plane?

in a lot of circuits, I see capacitors placed in parallel to the output of an IC or something, with the capacitor connected to ground. I don't understand what effect this has as it's not filtering anything to the next IC in the chain (unless I'm mistaken). See pic related.

>>1490588
>forgot pic
I know it's an RC oscillator, but I don't just see this done with oscillators

would this technically form one voltage divider, since the two in series on the left could be considered one resistor of 2k ohms?

>>1490589
With an oscillator or supply it's an active component. For normal circuits it's for high-frequency filtering. Look at the AC/DC models of capacitors and inductors. This is a very basic thing, so do understand their uses.

>>1490576
well. this shots are from the diode metal wire, I`m shotting them with my hand, it still has no box/base to speak of

>>1490623
my hair

>>1490595
>one voltage divider
one amperage divider

>>1490626
maybe we'll be able to see your penis next

>>1490634
wavelengths of light aren't small enough for that

>>1490588
Uh, the usually put caps on the power pins of an IC to ground to minimize high frequency noise by shunting it to ground. In that case it IS a low pass filter. Typically caps on input and output are simply coupling caps to block any DC from passing through but they wouldn't go to ground. There are some circuits with caps on their output that do go to ground though, most notably active peak detectors which us the capacitor to store the peak voltage. The diode preceding it and the buffer following it are what prevents it discharging. Aside from that there are some RC oscillators that probably use a config like that and so do most SMPS though you won't see the cap right on the output of the IC unless its a low power low voltage SMPS with all the switches and the coil on chip

anyone here knows about GPU repair? i have an old gtx460 card which just stopped working one day which im trying to fix now
6 mosfets looked burnt and there was a short between the gate and drain when measuring in-circuit
however, the short is somewhere else, because when i removed them, it didnt disappear
pic related shows the part of the board where the mosfets were, the shorted positions (blue lines) and a shorted ceramic cap between the inductors (green circle, not a direct line of sight but looks good). can it be that the gpu itself is shorted somehow, and is there a way to diagnose it?

>>1490634
>>1490640
it`s a microscope anons not a radio telescope. If you are really interested just ask ur mum ;)

>>1490515
SOT23-6 is close enough to 1mm, and it's not like 50 pins. QFNs will be impossible with my current setup, both for my photoresist etching and soldering technique. It's a right hackjob and will likely need a fair few mods as I build it, so I'm not going for professional PCB manufacturing just yet. DIY vias with copper wire will be interesting since I've got myself some 2-sided FR4. I might want to try peening them over before I solder them in place.

As far as protection goes, is the internal little PCB on the lipos I'm planning on getting sufficient? They're made for RC purposes (so probably not the best choice as far as current vs longevity) but they're common and cheap. Two 200mAh's if you're curious.

>>1490623
>>1490626
Getting a fair bit of chromatic aberration there, might want to mess about with the different lenses to minimise that.

what's the TL072 or NE5532 of the comparator world? the LM393 is trash but at a glance it seems like if you want better than ~1us response time you're stuck with less ubiquitous, more modern parts.

>>1490772
The lens was dirty and kinda fucked, I was holding the assembly with my hands so all images are affected by my shaking, I only lenses that came with the 5 $ webcam. I`ll make a base with adjustable focus where I can put petri dishes and stuff and it`ll work better.

>>1490772
>SOT23-6 is close enough
then you have any number of boost (AP36xx or equiv) or buck (AP34xx or equiv) regulators at some > 1A current ratings in the jellybean jar. how much you wanna pull?
>is the internal little PCB on the lipos I'm planning on getting sufficient
those will generally provide overcharge protection, overdischarge protection, and possibly overcurrent protection
>two
if you're really lucky it adds balancing but maybe not
goncern

>>1490774
>broadly available medium-speed comparator
LM311 maybe

>>1490799
They'll be in parallel with low current draw so the voltages at the terminals will be equal, I shouldn't have to worry about balancing at all. Roughly 150mA max, 15mA min at the boost input. I'll look into the AP36xx and AP34xx lot thanks.

>>1490806
if you're interested in a SOT-89 boost converter with a 500mA switch that claims low 无负载状态下输入电流 (12µA), and you don't mind a datasheet written in chink runes, consider the QX2303L50E (there are fixed voltage variants)

>mfw trying to design a milliohm meter that will measure individual milliohms with single digit percent of accuracy
WHEW what a journey. The journey continues

>>1490854
Just feed like 10A through it from a MHz constant-current switched-mode power supply.

>>1490869
Nah too noisy, and that wouldn't present enough of a challenge

>>1490879
>noisy
MHz switching with enough filtration should be quite possible to get into the <1% range. Plus DSP helps.

>>1490881
Nah.
I'm not buying a 10A power supply just for that purpose. Can get by with 1A and still have 0.1% accuracy worst case

>>1490885
That's true, though I'd actually suggest you make the supply instead of buying one. What are you using to amplify the signal before the ADC anyhow? An instrumentation amplifier?

>>1490886
i want a milliohm meter because i want it to measure resistances (and its a fun side project) for an electronic load to test a supply that i want to make. its a cycle of i want this project because i want to use it to do this project etc etc kek

>>1490886
chopper amplifier, with precision resistors in differential amplifier config. if i use a chopper i can get away with 0.1A too with 1% precision or thereabouts. with instrumentation amplifier, the voltage offset is like 50uV and that can give like 50% error if measuring 1milliohm

A thing i want to know and learn more about is ADC noise floor and ADC ENOB, because i want to use a 16 bit ADC, but I dunno what my ENOB is or how to find it. but even if its 14 bits that gives me 1/8 mV resolution, or if its 13 bits 1/4 mV precision
I'm probably going to have to use a 2.048V reference because 1.024 seems to be rare

Will current flow back through a solar panel if I put a capacitor in parallel with it?

>>1490966
Yes, slowly. Keeping PV cells with a bias on them can actually be bad for them, so I think they often have a diode in series. At least when one cell in parallel with another is shaded, and its output voltage drops below what the unshaded cell is producing, the reverse voltage on it can cause permanent damage. Though I think this is only with some types.

Is it true that you can connect 2 aluminium electrolytic caps back to back and effectively make a non polarized cap?
I get that if I put 2 identical caps in series like that I'd have half the capacitance.
If I don't have a 1uF bipolar cap, could I get 4 1uF polarized ones and put 2 series pairs in parallel to effectively get a 1uF bipolar?
Only asking because I'm having trouble finding 1uF bipolar caps on Farnell(Newark).
Maybe there's just a filter that I'm not seeing

>>1490977
Yes, but it's a really shitty method. Basically an Al-electrolytic cap short-circuits when in reverse, a current flows through the electrolyte. This current will make some heat and possibly even electrolyse some gases if there's enough of it, though typically this doesn't happen since it's blocked by the other cap. So don't use it in power applications where you'd have significant current flowing through the capacitor, and don't use it in audio applications because you'll get shit quality.

Also it's only 1µF, you can get ceramics (all SMD) or film caps (thru-hole) at that value easily.

>>1490982
Never mind, I just found a pack of them in my component press.
I really need to catalogue my shit

>>1490984
>catalogue my shit
I should do this too, but then I'd need to actually organise it instead of having it lying on and around my desk in little paper and plastic bags. I guess a spreadsheet is the way to go, lest you want to dabble in SQL.

>>1490894
>its a cycle of i want this project because i want to use it to do this project etc etc kek

so much wasted effort. to make a load, i'd just wire a bunch of light bulbs in parallel, varying the number by screwing wires (or not) into a big terminal block. incandescent and halogen bulbs are essentially free, as everyone is switching to LED.

>>1490985
>I guess a spreadsheet is the way to go, lest you want to dabble in SQL.

spreadsheets are fine if your data-set is small and you can find things by sight, or with control-F, but for large data-sets, you need something like microsoft access. you can create searches visually, no need to learn SQL. you can also do ranged or fuzzy searches, like all caps from 5-10uF, which you cant do easily on a spreadsheet unless you can program macros.

>>1490987
SQL isn't hard to learn and doesn't limit you to the M$ botnet, if that matters. To find "caps from 5-10µF in a spreadsheet I'd just make a conditional highlighting rule and sort by capacitance.

But as far as a variable load goes, if a FET package (or three) will handle the heat you're making you can make a very simple constant-current load with a FET, op-amp, and passives. Otherwise a whopping great rheostat from ebay or whatever might do the trick.

>>1490799
>LM311
that's it, thanks. i didn't see it because i was using digi's propagation delay category and it wasn't populated for that chip.

>>1489791
are richmeters anygood?

>>1490693
bump, i have several cards with similar simptoms
they arent that important to me, im just using them as practice

hi, maybe you guys can help me understand what is happening and if it is a bad thing:
my car's radio has no headphone jack, so i opened the radio, found 2 pins which said "right out" and "left out" and soldered them (and earth) directly to a headphone jack i had laying around.

now, when i connect my mp3 player to it (using a male to male connector), the audio goes out through the car's speakers. but if i disconnect the mp3 and connect my phone instead, no audio comes out, and instead the car's speakers keep doing a muffled rhytmic beat. anyone know the sound i'm talking about? what does it mean? should i put a diode inbetween when i soldered it?

however, if what i do is start playing music on the phone, then connect it to the radio with the radio off, and then i turn on the radio, the sound comes out perfectly fine.

>>1491162
What you described is a totally normal. You connected your jack to a point in the circuit that is both 'out' and 'in' at the same time but is not 'strong' enough to drive a headphone, which almost shorts the signal. The same happens when you connect the player. It shorts the radio and feeds in its own signal instead. When you disconnect it you hear the radio again. Sometimes it is sufficient to switch the player off while leaving it connected.

>>1490693
All I can guess is replace the mosfets. If you look online the problem is likely common enough. I should take a heatgun to a 970 I got from a friend, because it had burnt mosfets, but I wasn't able to solder them right I think. He went through 2-3 (warranty!) of those cards because they were just made wrong.

>>1490774
LM6172

You pay for em though.

>>1490977
Ceramic caps are nonpolar/bipolar. I just checked and they have plenty of 1uf caps. If you need AC voltage, you could put a few nf caps in parallel.

>>1491176
this guy's a bit too pricey for the sort of chip i was after, but i'll note that i also found the lm319 as a dual comparator option similar to the lm311. too bad they insisted on breaking out the output emitters so it doesn't come in a nice dip8.

Would you trust these aliexpress latching switches for stompboxes?
If not what do you recommend

>>1491245
They're probably perfectly fine.

>>1490894
a lot of those ancient op amps have offset null pins. connect trimmer to V- and adjust. then all you have to worry about is temperature drift

>>1490987
NoSQL seems ideal for this sort of thing, actually. Solr actually wouldn't be bad for this at all

>>1491240
the LMV393 etc. are good and fast too, if the low voltage works in your application

>>1491245
none of those will be click-less, in case that matters to you. I think the bottom two with a flip-flop and a couple of FETs would be best. it would be better if they were latching so as to eliminate the flip flop
ideally, you want something with at least 1kg of actuation force. so the rest of the switch is more likely to be designed to withstand stomping

>>1491245
What happens if you are white and try to use the negro switch?

holy shit i didn't know /diy/ had threads like this, i would think /g/ would but there aren't ever things on these topics

>>1491249
>negro is considered an offensive term in the US
>it literally means black in spanish
I don't get this.

>>1490015
Go look at what your gnd is connected to: it's the battery's - terminal.
I don't know what the rest of these guys are going on about.

>>1491248
Not using ancient op amps

>>1490987
I've done that to load a powe rsupply's output. Didn't like it

>>1490577
>>1489791
Pls?

microscope guy, made a cardboard base for it. here is some moss, I don`t think I`ll get to see microbes withour a better lens, or a extension tube

>>1491318

>>1490973
I thought it is not a good idea to connect PVs in parallel at all. One is always going to produce slightly higher voltage than the other. Connecting them in parallel with a diodes will make one of them mostly useless. Also, dealing with higher voltage is easier than with higher current.

>>1491331
When all panels are approximately evenly illuminated (same angle to the sun) the voltage difference should be negligible such that the internal resistances of the panels is enough for them to share the load. It's not like their voltages won't sag as you pull a current from them.

>wire up 6 LEDs on a breadboard
>power from computer's USB
>USB supply voltage with LEDs disconnected: 5.2V
>current drawn by LEDs: 40 mA
>supply voltage with LEDs connected: 4.2V

Which is more likely
>my computer's USB voltage actually drops by >20% under only 40mA load
OR
>my piece of shit $20 multimeter is not measuring the voltage/current correctly

>>1491350
If it's just meant to be a data port that can't charge cellphones then it's entirely possible its sagging that much. $20 DMMs are usually fine for measuring that kind of voltage. But a circuit diagram could help. You used dropper resistors, right?

>>1491346
Yeah, you're right. I still don't understand the benefit of connecting them in parallel. Except probably for resilience - if the solder work fail on a single cell, the whole plant is dead.

>>1491351
LEDs are wired up like this.

>it's just meant to be a data port that can't charge cellphones
Funny you should mention that - I also tried a bunch of random USB wall warts, including 2 that came with phones. All were similarly bad. Even my good phone charger, which is supposedly good for up to 2.85A, drops from 5.05 to 4.55 with the LEDs connected

>>1491354
charge your phone and measure the drop.

>>1491352
Well often you'll have thousands of individual cells and putting them all in series will result in a voltage that's a pain to deal with. Same reason cells in car batteries are in series+parallel arrays.

>>1491354
That does sound like an illegal operation (pic related) so I imagine it's not meant to output much power at all.

>>1491359
I'd have to cut up a USB cable for that, which I don't really want to do right this minute.

>>1491362
>That does sound like an illegal operation (pic related) so I imagine it's not meant to output much power at all.
I have no idea what you're trying to say here, but since I'm sitting here with 2 phone chargers and 3 tablet chargers and every single one of them is (supposedly) going far into that red range the moment I hook up a few LEDs, I'm inclined to assume the problem is with the multimeter.

>>1491364
In that case I'd believe that the DMM is the case. I'd get a USB tester, they're pretty reliable for a sub-$5 product, plus they give current and voltage.

>>1491365
>USB tester
Hey that's pretty cool, I didn't know such a thing existed. Thanks anon.

>>1491316
yes, there is inrush current. no, it is not easy to find a value of fuse on the primary which will withstand that inrush current yet blow when the power supplies are overstressed (need a scope and a current probe)
to protect against the power supplies taking a catastrophic shit, put a 3A slo-blo on the primary
to protect against the powered devices taking a catastrophic shit, depend on the power supplies' overcurrent protection or add fuses on the secondary side of each power supply

>>1491354

sounds to me like you're using long thin wires, and they're causing the voltage loss. to check, measure voltage near the xformer, and then near the LEDs.

>>1491449
Oh shit, thanks anon, that was very helpful. The wires are fine but the contacts in this breadboard seem to be kind of shitty. I run a jumper wire from A to B, measure resistance from one of the contacts adjacent to A to one that's adjacent to B, and it comes out as >20 ohms (when the jumper wire on its own reads as 0).

Is this normal for breadboards, or just what I get for grabbing the cheapest one I could find?

Would a dim power led on an ESP32 be from, undercurrent, over voltage or undervoltage?

>>1491471
I don't think that's normal. Try taking the back adhesive tape off, removing the metal contacts, and polishing them up. But it might just be a single bad connection or corroded pin (which can happen with tin-plated steel pins) so just unplug and replug everything for starters.

>>1491486
Probably not overvoltage, at least if everything else is working. But over time if you overdrive LEDs they will dim. If its resistor is set to let too much current through this could well be the issue if you've had it on for a few dozen hours, but I doubt it. Undercurrent and undervoltage are not independant issues, one causes the other. Either your power supply can't supply enough current so the voltage is dropping (more likely) or your power supply has too low a voltage so not enough current ends up flowing (less likely). Now's the time you bust out your $10 DMM and try not to short the SMD pins with the massive probe tips. And fail, naturally.

>>1491493
It was undervoltage

>>1491494
Another mystery solved!

>>1491172
thanks, i will try today and see if it boots

>I'm new to electronics, where to get started?
It is an art/science of applying principles to requirements. Find problem, learn principles, design and verify solution, build, test, post results, repeat.

Opie is pretentious as fuck

>>1491171
but should i leave it as it is? might it not cause problems in the future, for the car's radio or whatever i connect to the jack?

i've turned that poin in the circuit to be "out" and "in" at the same time, but originally it was only "out". i'm worried it might blow a fuse or something

also sorry i'm dumb, what does it mean that it "shorts the radio"? isn't that a bad thing?

>>1491520
HALP!!!

I still have this damn thing on my workbench. It’s the same shit that happened with the other one where I used a different power bank.

Still trying to power these LEDs off some Li-Ion batteries, want to use the chip out of a power bank so I can charge it via Micro USB input and have the battery protected.

Made a USB cable that goes to + and - alligator clips. The power bank was supposedly good for 1.0A, the light draws ~0.6A at full power. When I had the USB out plug to gator clips, it powered it fine. But I want to wire the output straight from the board so I have protection from draining the battery too much.

Used the regular switch, it flashes and shuts off, overcurrent protection I think. Used the dimmer switch and somewhere between 0.4A and 0.5A, it shuts off and I can’t run the lights at the full 0.6A. Two different power banks, both rated at 1.0A output, they both powered the light fine through the USB port. But when I rip the USB port off and try to solder straight to the board, I get the over current protection BS at around 0.5A.

What am I missing here?

>>1491522
I would not expect any damage because the radio->-T->-amplifier connection (T is your added jack in the signal path) is at a very low power point in the circuit. A short connected to the T simply mutes the radio.

>>1491530
>when I rip the USB port off
At least desolder it, jesus what a mess.

>>1491546
I mean kinda. Problem is it’s small, that SMD stuff you goys are talking about that’s really difficult to solder cleanly. I fucked up by not leaving those 4 tabs on there for the USB out. I think I only need the two outer tabs for this project though.

Also another questions for you fegs... any brick & mortar chain stores that sell small spools of thin gauge wire? Like 24AWG and smaller? Home Depot doesn’t have shit smaller than 18AWG, thinnest I could find at the usual places was 24AWG x 2 speaker wire.

>>1491550
Wire made for punchdowns inside cat5 is pretty good, fits breadboards nicely.
I'd have snipped/ sawn/chiseled off each solder pad from the USB connector and simply reflowed the solder on each so I could take them off individually without needing a 1cm-long iron tip.
Sure your bodge job didn't end up shorting the output to ground or something similarly whacky?

>>1491552
I don’t think anything is shorted. I didn’t really solder much on there. Plus I tested it with some smaller LEDs and it worked, and then hooked up that dimmer switch and I can get close to 0.5A (full power is ~0.6A) before it goes into the protection mode.

And as I mentioned, it did the exact same thing with the first LED monstrosity I tried to create. When the powerbanks were still together, it powered the lights fine at 0.6A using pic related. I gave up on the first one and just wired the board up so I could charge the battery properly with it, but then had to wire the battery straight to the switch/lights because it wanted to go into protection mode. So that first one there is no protection from over-draining the Li-Ion battery, but upside is I have the little battery monitoring lights so I can press the button as I use it and when it gets to 1 out of 4 little blue LEDs, I plug the micro usb in to charge it.

I’m baffled though. There was nothing inside the case of either that was taken out, both were just batteries and little boards with USB in-out and a little button to see battery life.

I grabbed a power bank rated to 2.1A so maybe I can rip the board out of that next.

>>1491552
>>1491558
>using pic related
My bad. This thing powered them fine, but I can’t go full power hardwiring the lights straight to the board.

And yeah, two different power banks did it. Both rated at 1.0A. I think the first light was slightly higher powered, like 0.75A while this one is closer to 0.6A.

Time for more trial and error...

If the small one powers it through that USB cable fine, there needs to be something more. Like maybe it needs to be connected to something more than just the two pins.

>>1491532
i still dont understand what a short is but ok, ill leave as it is. thanks

>>1491626
Do you know what a circuit is? In this context a short (or short circuit) is a highly conductive connection at a place made for a lightly conductive connection. In your case the headphone and the output of the player act as a short to the audio signal of the radio which normally goes to the (lightly conductive) input of the amplifier. I'm trying to avoid technical terms like source, impedance, load and attenuation.

>>1491641
ok, i understand now thanks. less resistance in the path of my input so current flows that way instead of out of the radio. so this is the proper way to do it, and safe? shouldnt i put a fuse somewhere just in case?

flipped the breaker to a light socket im replacing, light went off. turned the lightswitch off and checked with my ncv, getting a weak reading. popped it off and checked the terminals with a multimeter, getting 2vac. is it actually off or is there a fuckup somewhere in the wiring? house is literally brand new first owner.

flipped the breaker to a light socket im replacing, light went off. turned the light switch off and checked the socket with my ncv, getting a weak reading. popped the socket off and checked the terminals with a multimeter, getting 2vac. is it actually off and just picking up some inductive current from other wiring or is there a fuckup somewhere? house is literally brand new first owner.

newfriend here, just getting started and semiconductors are cool shit
if i wanted to light pic related without killing it(under 20mA), i'd need the voltage across it to be between 3 and 3.6v right
also say i wanted exactly 15mA with a series resistor, but have a 5v source - i can treat it as a 200ohm resistor at 3.5v(3.5v/0.015A)
then calculate total resistance 333ohm(5v/0.015A), 200 of which is from the led, meaning i need a 133ohm resistor

did i fuck anything up

>>1491700
V=IR doesn't work for nonlinear components. solve it this way: (5V-3.5V)/15mA = 100 ohms.

Is there anything fun I could do with 50+ induction heating coils from a few super expensive cooktops? Cleaning house and I might just toss all but like 2 of them.

>>1491799
railgun

>>1491799
not sure what frequency these run off or if the circuitry to generate that frequency exists inside the coil assembly, but you could mount one under a table to do devious things to peoples' keys (or phones)

>>1491804
These are just the coils. I do have the controller board from one of the cooktops somewhere I think.

>>1491803
Good idea, always wanted to make one.

>>1491520
to get started, it's useful to understand what it is and what it isn't you are doing
pretension helps keep the junkposters down... sometimes

>>1491663
/ohm/ RULE 0

daily reminder to try a new method of fabrication from time to time
Pic related: Peltz oscillator board made "Russian style" (cut and peel copper from a laminated board with a pen knife), about to be soldered

>>1491799
they usually are made out of litz wire, which is expensive

>>1491276
Anything can be offensive to an american as long as you say it a loud and angry way.

Are Mammoth brand potentiometers good quality?
There even cheaper than aliexpress

>>1491799

Pawn them off on eBay or CL. As >>1491845 says, litz wire isn't cheap. More accurately, litz wire isn't cheap in small quantities, since it's a fairly niche type of wire.

>>1491868
>cheaper than aliexpress

>>1491879
Me too, that's why I came here to ask.
https://www.mammothelectronics.com/products/mammoth-16mm-single-gang-solder-lug-potentiometers-linear-b?variant=35159430855

They have tons of cheap stuff wich I don't know if I should trust

another example: 4th-order Butterworth high-pass filter, copper tape on unetched single-sided copper-clad FR4 as ground plane

>>1491804
>keys
most keys are brass - won't work
> (or phones)
I wouldn't consider it a harmless prank if you fried my phone and I found out it was you.

>>1491810
>These are just the coils.
nice, heavy gauge magnet wire
if I didn't sell them I'd clip off the leads and connectors and save them for the wire

>>1491584
I don’t fuckin get it.

The 0.6A must be too close to the peak for that 1.0A power bank. If I plug it straight into the USB of the new bank, it flashes like protection mode sometimes, but then seems to work if I run it through a switch. That double cell bank is rated to 2.4A, so I think I will rip that apart and see if it holds up.

>>1491881
it should be cheaper to send back if it's not

>>1491804

>>1491931
kek wtf mate

>>1491442
Welp, 100mA blew, 500mA didn't (both fast), so I'll just find the sweet spot in between. Thanks anyway.

>>1491990
the purpose of a fuse isn't to protect the circuit, it's to prevent catastrophic damage like the fires you mentioned after the circuit has already failed. so the 500mA fuse will be sufficient unless you have a bizarre failure mode.

>>1491993
I know, I'm just being autistic about it (realistically the thermostat is useless, too, since any current that would heat the supply that much would have blown the fuse long before that could happen.)

>>1491901
So with the new one, the protection must be coming from a bad connection between the battery and the board. Once I cut the nickel strips and tried to connect them with alligator clips, it gave me the flash. Connected the + with alligator clips but held the tabs for the - together and it worked fine.

So whatever protection must be gauging the current coming from the battery to the boars, not the output to the device, right?

>>1490419
Keypad customization / changing / ordering. Anybody?

>>1492023
which ones? can you be more specific?
prototype membrane keypads: $1300 http://www.melrose-nl.com/membrane-switch-prototyping.html
prototype silicone keypads: https://almaxeis.com/rubber-keypads/silicone-rubber-keypad-prototype/
it looks they can do semi-custom keypads, using existing molds/layouts with custom legends
if you want a REALLY nice keypad, screw the membrane shit and install tactile keyboard switches. custom keycaps are comparatively cheap

>>1492017
The current sense mechanism on those things sometimes uses a thin, snaky trace instead of an extra SMD resistor, so by messing about with the components you may well have changed its resistance and caused a problem like that.

>>1490536
Optoisolators.

>>1492026
>that company
yeah see I never would have found some place like that, unless I knew of them before hand. Membrane is a possibility but just need ideas
>silicone
>request quote
goddamn it that means its going to cost a kidney too kek
>tactile keyboard switches
I'm not making a keyboard. I want to make a numeric pad with ok/enter/backspace/custom functions/whateverelsekeys for projects I want to do
>pic related
These companies must get those things made somewhere
Momentary on tactile switches also feel really cheap. Not to mention I don't think I have the means to machine a panel, or to make / find key caps, or to label those keycaps etc

>>1491990
Usually slow fuses are used in cases like this. Schurter (and others) have guides for dimensioning fuses, but like the other anon said, it's kinda difficult to do properly when you have to take the inrush current into account.

>>1491990
Bro just get some NTCs

How do i take voltage that can be from -10v to 10v a turn it into 0v to 5v, where 2.5v = 0v (middle between 10 and -10)

>>1492159
here you go. the math to calculate the resistor values is tedious but fortunately your case was easy to figure out by just setting the input and pulldown equal and tweaking the pullup value.

>>1492176
Thanks. Is it some sort of voltage divider?

>>1492190
yes. you can think of it as creating a static reference voltage from the supply rails like you'd normally do with two resistors, and then using a third resistor to pull that reference further up or down with your input signal.

>>1492193
It should be mentioned that the effective resistance of the ±10 V source is part of the game. An open input produces an output voltage of 10/3 V, independent of the real values chosen for the R--R--R/2 network. If that matters, some input buffering would be required.

>>1492066
just saying the Cherry MX + generic keycap + paint stamp route will get you any key you could ever want
>must get them made somewhere
see "silicone keypads", above. couldn't hurt to request a quote. semi-custom might only cost you a couple of teeth. it might even cost less than a grand

can this also measure DC or only AC?

>>1492296
only AC. for DC you'll need a hall effect current sensor.

>>1492299
Fuck. Another complication. I wish i could go back in time and kill tesla so we wouldn't have to deal with the annoying dc everywhere today.

>>1492296
Only AC. A current transformer (or any kind of transformer) works by way of magnetic flux from mutual inductance. When a current flows through a wire a magnetic field is generated around that wire, and the opposite also happens where a magnetic field generates a current. But this can only happen when the magnetic field is changing, because the voltage across the secondary coil (the output) of a transformer is proportional to the change in magnetic flux with respect to time. Faster changing flux means more voltage. But a faster-changing current through a transformer's primary coil means the magnetic field amplitude cannot increase to as high of a peak, such that you get a constant voltage ratio between the input and output. In the case of a current transformer however, the thing is acting in a saturated mode since the primary coil is just a single wire, the result being you get a constant current through the output that is proportional to the current going through the primary. This is why it's a terrible idea to unplug a current transformer while it's on an active wire, since the constant current will cause an increase in voltage until it reaches the breakdown voltage of either the wire insulation or the air between the wire ends. Basically it will spark a bunch.

See Faraday's and Lenz's law.

>>1492301
Just use a sense/shunt resistor bro.

I'm interested in traffic control devices and CD4000 discrete logic. I've been playing around with this traffic light controller circuit. Current version uses two CD4017 counters, a single CD4027 JK flip-flop + some OR & NOT gates.

The basic works, but does /ohm/ have any ideas how to implement different time delays for different light phases? I've seen a lot of 4017 circuits and most/all of them are using a lot of diodes to connect 4017 outputs in parallel.

>>1492324
Anything wrong with a normal monostable multivibrator? Have a series of FETs on resistors to customise the delay if you want. Also is it normal for your traffic lights to go orange/yellow before they go green again?

>>1492324
4000 series logic is amazing for autism projects. i whipped this up just now. the combinational logic blocks i was too lazy to draw out are either nothing (just a straight feed from Qn to the following gate) or one or more AND gates that are used to select a specific time delay such that the delay is the sum of the periods on each AND'd pin. so your overall delay for each light period is Fcrystal / (2^Qn1 * (2^Qn2+2Qn3+...)) or something like that.

>>1492303
Should be fine with this one, it has zeners built in to prevent retards from electrocuting themselves.

For all you homos wanting to get into EE, Big Clive is a big gay. Wish he did the follow up video he promised.

https://youtu.be/tSePiQuei8M?t=1989

>>1492338
>zeners built in
aw, that's no fun

>>1492303
>a current transformer .. is acting in a saturated mode
No, it isn't. Core saturation is a fault condition. Can you imagine why that is? It has to do with the relation of H and B.
>since the primary coil is just a single wire .. you get a constant current
No, you don't. np=1 does not mean constant current. Why would it?
>proportional to the current going through the primary.
Now what, either constant OR proportional to the primary current.
You need to work on your misconceptions.
>See Faraday's and Lenz's law.
Yes, do it.

>>1492324
If you want to keep the design synchronous, then yeah, you'll use varying number of clocks per light phase. Diodes are just a cheap way to OR a bunch of outputs.
Variable speed main clock would be another way, so that your light state machine also selects the clock speed for the next phase.

>>1492340
>Big Clive is a big gay.
This isn't exactly news.

>>1492406
Ok I'll admit I don't understand how current transformers work, I was assuming the current relationship instead of the voltage relationship was caused by some unideal mechanic of transformers. The main one of these I know of is core saturation. Relations of H and B is not something I've studied, I've only done E&M in insulating dielectrics and in a vacuum; I only just learned D and E fields. If there's a neat place to learn about B and H fields in a transformer core that would be neat. Am I correct in assuming a current transformer wouldn't work with an air core?

I wasn't saying that np=1 was what directly causes the current proportionality, but rather that it causes the extreme turns-ratio disparity which causes the unideal effects which causes the current proportionality. You could just as well have 2 or 3 or a dozen turns. But a few hundred might not work. "np=1" was just a qualitative description/observation, not a quantitative one.

And by constant current I was trying to describe a current source, which in hindsight isn't a spectacular use of terminology. You know, a PSU operates in constant current or constant voltage mode, but it's not like you can't change the current or voltage on it.

>>1492407
He's only been gay for about a year. Thought it was interesting.

OK so I'm salvaging parts from a Dell PSU and stumbled with a transformer with only two terminals. What are this suppossed to do?

>>1492450
post pics. if it's on a ferrite E core it's probably the PFC inductor.

>>1492454

Hello /ohm this totally isn't my homework I promise. If I were to do this with two amplifiers would it be easier with two common emitter amps or 1 common emitter amp and 1 emitter follower amp? I was thinking of using 1 common emitter with the emitter bypass capacitor and a second common emitter amp without the capacitor.

>>1492450
>a transformer with only two terminals
a transformer with two terminals is a choke

>>1492496
Common emitter + emitter follower. If you don't use an emitter follower for the output you're gonna have to use a really small collector resistor to be within the output impedance specs and an even smaller emitter resistor to get the right bias and that means quiescent current will be really high which will kill a 9V battery really quick. Voltage gain of 50 should be easily achievable by either full bypassing or partially bypassing the common emitter amp. To get the input impedance above 15k you just have to choose your bias resistors somewhat carefully. R1||R2 needs to be about 1/10th of beta*RE or smaller to satisfy the loading criterion but also needs to be above 15k per your design specs. Another way to increase your input impedance is bootstrapping so you may want to consider that as well.

>>1492407
>not wanting to keep the design synchronous
RST is async on the 4060
the rest of the design, well, I can't think of any good reason not to keep it synchronous

>>1492324
>>1492333
if you're feeling particularly autistic and want to add cross-traffic, just add an OR gate to the CD4060 RSTs coming from the cross traffic green and yellow on each cross path
also, you can add a switch to sense traffic, then add a time delay (1/2CD4015 + 1/2CD4072?) for an output that stays high for some period after the last car has passed over. integrate with the square AND blocks on this or the cross traffic ways however you like

>>1492496
Just use an op-amp circuit, that's what they're made for.

>>1492496
I know I effectively did your homework for you but here's a two transistor amplifier that should meet all the design requirements. It attempts to keep distortion low with negative feedback but it's still 3.1% so not great. Not a terribly well refined design but a good starting point.

>>1492534
Not a very helpful answer and not a very good one. op-amps may dominate low frequency electronics but transistor amplifiers are still commonly used and the knowledge of how to properly design them is still worthwhile. In RF electronics you never see op-amps. Microwave amplifiers are built with simple common emitter amplifiers, of course using BJTs with GaAs junctions and other special materials to allow them to operate at 10s or 100s of GHz. op-amps are just too slow and transistors are not. Hell even for VHF band stuff transistor amplifiers beat out op-amps. A basic 2N3904 jellybean op-amp has a gain bandwidth product of 300MHz which is plenty high enough to allow it to be used in the FM band. Most op-amps could only dream of reaching 100MHz. Many of them barely have any gain at 100kHz open loop. There are some high speed op-amps on the market that are pushing into the 100s of MHz range but they are expensive and offer no major advantages to transistor amplifiers. I won't dispute the usefulness of op-amps under 100kHz but transistor amplifiers are still useful and worth learning about.

>>1492549
>2N3904 jellybean op-amp
Obviously meant 2N3904 jellybean BJT

>>1492549
While you're right about that jellybean part thing, RF amplifier ICs are definitely available these days. I'm looking at ones on mouser that go up to 89GHz, though for $400 each. And there are ones that function at 10GHz for less than $20. (plus some oddballs that say they operate at 140GHz for $2, not buying it)

>>1492553
RF amplifier ICs don't contain op-amps though. This $140 40GHz amplifier uses a transistor based cascode amplifier configuration. If you look at the schematic architecture page (if available) for most of those ICs you won't find the typical op-amp differential amplifiers, class A gain stage, level shifter, compensation, and output push-pull amplifier.

>>1492553
>>1492557
Fucking forgot pic, man I can't post right today

>>1492557
You're right about that I suppose. I more just meant jellybean ICs that don't require you to model them with every other component in the circuit to check that their output and input impedances are messing everything up, but I guess you can hardly apply the "ideal op-amp" laws to an RF amplifier.

>>1492549

no way that thing has 100 ohm output impedance given the positive peaks have to go thru a 4.7K resistor. design me something better, faggit.

>>1492562
It has a 210 ohm output impedance, Rout = R6||(beta/R4). Close enough. You figure it out yourself then you fucking ingrate. I did more than enough work for you.

>>1492235
>just saying the Cherry MX + generic keycap + paint stamp route will get you any key you could ever want
I'm not looking for keyboard keys nor the feel of keyboard keys

>>1492496
>>1492549
It doesn't say anything about the stages needing to be directly coupled... Normally you'd just do a class A with the bypass capacitor on the emitter and AC couple it to the emitter follower.

>>1492557
Radio POWER amplifiers are a pain in the ass because they're usually designed for a uniform output resistance (50 ohm). I had a project which was in the AM, but called for up to 1000v pk-pk, line and neutral (i.e. one wire had the + and - polarity), and the output resistance was known to change from 200 ohm to up to 7 kohm within 5-10ms. Ass of a project that was.

>>1492586
Why would you not directly couple it? Why are you going to AC couple it which now forces you to have another two resistors to bias the emitter follower and on top of that you now introduce additional phase shift through you capacitor which while not a problem in this circuit could lead to instability and oscillation as part of a larger circuit?

>>1492571
Yes yes, you want a custom silicone keypad, but you don't want to pay the normal price for a custom keypad, you don't want to make a front panel fitting for your custom keypad and probably would like to avoid using a gold-plated PCB for its contacts, too. Pic related is totally unacceptable, because it does not look and feel like a silicone keypad. Stock silicone keypads are unacceptable, because one with exactly the correct labels does not exist. Did I get it right?

>>1492595
>but you don't want to pay the normal price for a custom keypad
Wrong. But 1000 dollars for one is absolutely unrealistic by any stretch of the imagination
>you don't want to make a front panel fitting for your custom keypad and probably would like to avoid using a gold-plated PCB for its contacts
Wrong
>Pic related is totally unacceptable, because it does not look and feel like a silicone keypad
Wrong, and there's no way i can change the label for it
>Stock silicone keypads are unacceptable, because one with exactly the correct labels does not exist.
Wrong, only because I can't change the labelling on some that already exist

>Did I get it right?
Mostly didn't

>>1492599
>Wrong. But 1000 dollars for one is absolutely unrealistic
In other words, it's correct because you're unwilling to pay the normal price.
>Not to mention I don't think I have the means to machine a panel
But somehow this does not mean that you don't want to make the front panel?
>Wrong, and there's no way i can change the label for it
Those (or at least very similar switches) have replaceable labels. In the most simple case, it's just a printed piece of paper or transparency film.
>Wrong, only because I can't change the labelling on some that already exist
So it would be wrong if you could actually do it? But since you can't, it's correct.

Still sounds like I got it mainly right.
The bottom line is: use something else than silicone keys. Or pay.

>>1492591
It's just a lab/homework assignment man. Usually they use 1 cent transistors that aren't very consistent, so the simulations are often ass. When you're doing it all out on the breadboard it's a lot easier to do only AC coupling. Two stage doesn't really matter, but we were suggested to do a 200x(400x?) 8ohm three stage design (jfet for input impedance, bjt/fet for higher gain, emitter follow for low impedance out).

>>1492599
just buy an existing keypad and get rid of the lettering with a whiff of acetone or brake cleaner. Add new writing to it with some epoxy paint and a tiny brush or stencil it or something. As the other anon said getting mechanical keyboard switches is a fairly easy thing to do, as is just getting an ordinary keyboard numpad, chopping off the parts of the membrane+layers you don't need, and replacing those keycaps and throwing it in some sort of bodged case.

I unplug all my desk shit when not using it, recently the PC started to do a weird thing:
When I plug it in again, the PSU takes like 5 min until it can be turned on.
I have a spare PSU, but I want to repair this one for the keks. Anyone has any idea where to look?
Caps seem to be OK.

>>1492605
>In other words, it's correct because you're unwilling to pay the normal price.
wrong, because otherwise anything with any sort of custom keypad available commercially would cost at least a thousand dollars each
>But somehow this does not mean that you don't want to make the front panel?
Don't put words in my mouth - that was your assumption
>Those (or at least very similar switches) have replaceable labels. In the most simple case, it's just a printed piece of paper or transparency film.
Okay. Can you show me?
>So it would be wrong if you could actually do it? But since you can't, it's correct.
Wrong again - There are quite a few keypads that would potentially suit my needs - but you offered no solution to them unlike >>1492607

>Still sounds like I got it mainly right.
>The bottom line is: use something else than silicone keys. Or pay.
Sounds like you got it mostly wrong still and are just finding unfeasible options that i will shoot down just so you can try to make me out like some bad dude.
You seem to be going out of your way to antagonize me instead of helping so bottom line is: give something realistic, actually helpful, and don't misconstrue what i'm saying. Or piss off.

>>1492607
>Acetone, epoxy paint, and stencil
I suppose a stencil would work. Do they have any with full words? Or how do i make words like "triangle" at a size that will fit on a tiny keypad button? I have never done stenciling or anything of that nature
But anyway keyboard numpad and keyboard keys is far too big for my application, even if we pretend I went for the keyboard numpad option. I want a numpad that's the size of the one on your tv remote or something, which is why I used the example image of the arbitrary waveform generator front panel

>>1492617
Check voltages. and don't shock yourself. Some heatsinks are at mains potential. I had a feeling they were, but i accidentally shorted something fiddling around one time in one while it was plugged in. There was a big flash kek
There might be schematics online for it if you can find the make and model of the pcb

>>1492618

>>1492607
So far I can only seem to find stencils that are for "normal" size writing ie too big for small keypads and such. And stencils for words seem even bigger

>>1492605
NKK makes the switch you linked, but even making a keypad "manually," I run into the same problem above - that of producing labels

Keypads are a goddamn pain in the ass

>>1492606
Don't get lazy and design your circuity shitty just because it's only homework. Bad practice you learn in school are bad practice that will carry over into your professional life. Nobody is going to hire you when you're throwing away money with parts you don't need on your BOM.

3 weeks and a lot of hack soldering later...

I fucking hate electronics. Everything is too small for my manbear sized hands.

>>1492648
Magnifiers and precision tweezers are a thing

What are some reasons why an Arduino Uno would only be putting out ~3.6V as a digital high on an analog pin?

>>1492656
*Nano

>>1492651
I actually got a magnifying glass with those helping hands but I haven’t really needed it on this project.

I think maybe a soldering station with some balls and fine tips will be next on my list. I underestimated how much I would use that soldering iron when I got it.

I got that cheap HF iron and then grabbed the medium duty Weller, and I haven’t even touched the Weller yet because I didn’t want to ruin it as I was just learning.

As for the tweezers... I got that covered but I got pic related last week and I love them. The pliers are thinner than your average “precision pliers” but the grip is way better than tweezers.

>my old ass analog osciloscope got here, vendor said it was working
>pluges in
>fire and sparks
>dies
>2 transistors caught fire inside and god knows what else before the fuse blew.
Thanks obama. I don`t think I`m smart enough to pull this weird ass dusty PCB that is wrapped arround the vaccum tube and find whats broken. Lets hope the vendor doesn`t fuck me over and I don`t end up with a lot of money in old parts.

>>1492331
>normal monostable multivibrator
I'm planning to use a 555 + ½ of CD4027 connected as Toggle flip-flop to produce the clock signal. Is this what you meant?
>is it normal for your traffic lights to go orange/yellow before they go green again?
Yes, this is the ordinary light sequence in my country. It just shows the difference between red->green and green->red transitions.

>>1492333
>>1492512
Your circuit would be great for controlling multiple lights but however I only need one light for cars + one for pedestrians. Also I don't have any cd4060 at hand, but have a huge stash of CD4017 & not afraid to use them!

>>1492407
Variable speed clock sounds like an interesting idea. I thought about dividing the main clock signal down with some cd4027 flip-flops connected as T-flip-flops (or yet another 4017?). Then a logic circuit would select an appropriate clock frequency for the cd4017s controlling the lamps.

Here's the current iteration of my nearly-all-4017-controller. It uses two extra 4017s in both lamp counters to inhibit the clock signal, keeping the counter in "green" state for a longer time. I also tried to re-organize the clock distribution to keep positive-going transitions from occurring at cd4027 J, K and clock inputs at the same time.

>>1492781
;_; fuck man, I was really happy with this.

>>1492786
Honestly I have no clue on how to trouble shoot this thing. If anyone has experience with old ass chinese analog scopes please help. I can`t even measure the capacitors because I have no clue on how to take this pcb out of the chasis and it`s hugging the tube very tigthly.

>>1492618
the pros use pad printing, basically dipping a hard rubber piece with a mirrored positive of the desired image in paint, then rolling it along the key
talk to your local rubber stamp company, they may be able to help

>>1492662
>I'm using stone knives for precision work and I'm a tripfag
ok

>>1492656
how much load are you putting on that pin?

>>1492788
Checked the transformer it makes A LOT of noise, but it`s outputing from 30 to 40 V on it`s secondary taps, seems about right to me. How do I test the capacitors? A bad cap could cause a short in the rectfiers and fuck up the transistors

>>1492786
>>1492788
I've had a blown resistor that still read correctly somehow. If not, drop in another transistor and use the same R value as the one on the right. The transistors *may* have been matched or precision or this or that, but a lot of design is done with redundancy or predictability over autism, so component values don't matter as much as you might think.

>>1492803
>how to test caps
Desolder, test, and solder them back. idgaf

If you don't have a RLC tester you can use a constant current source and oscope to test.

>>1492804
the blown resistors is something like 1ohm. But the problem is what caused it to blow up.

>>1492805
> oscope
>need a scope to fix first scope
Fug, the board where therectifier is is attached to the main panel and two other boards and all of them are making a square arround the tube. fucking old stuff and being a poorfag

>>1492806
I'd guess either a bad transistor or a bad voltage regulator (if it's driven from a regulator). I've had a bad voltage regulator once, which caused an inductor to overheat, though it didn't break anything in itself.

>>1492808
Oh, right. You can use a voltmeter. A constant source charges a capacitor at a constant voltage rate, so if you give it 10-30s you can see if the capacitor is charging consistent with the rating on it's label.

>>1492808
one of those $30 chinkshit memescopes might be enough, or not

>>1492810
Do transistors get bad over time? They can go ded on your ass? Pic related
>>1492811
yeah I`ll do that. The caps don`t seem stuffed or leaking tho. I still have to figure how to remove the board so I can get to the solder.
>>1492812
This is a chinese memescope, but from the 70s.

>>1492814
The thin squares on top of the caps and on the right of the rectfiers are also rectfiers, so it seems.

>>1492783
No I meant for timing delays. Using a monostable variant means you give it a clock pulse and it releases a high output for a certain amount of time before it goes back to low and stays that way.

>>1492815
>>1492814
One of the `big power transistors` has a short between Emitter and collector. Dunno if it caused the fault or was the fault. rly makes ur think

>>1492831
If it's a FET then that could just be the body diode. Though I imagine you've looked at its name and found it to be a BJT. In any case, once you replace those ancient trannies with modern transistors it should behave a little better.

>>1492865
Yeah, I`ll look into that. They are PNP power Transistors. (NEC A1009), I`ll look into similars (I`m taking sugestions) and find replacements for the other bjts that went arabic.

>>1492790
>stone knives
Like those ceramic things? Those are all the rage these days.

Ok goys so after 3 weeks, maybe I have a theory on this. I don’t think it’s an overcurrent protection that shut the thing off. I think maybe the voltage drop caused by the light going full blast is tripping that board from the power bank into thinking the battery is super low or dead so it wants to shut off or limit the current. Could this be the case?

What are the best soldering iron tips for high thermal conductivity and medium/low heat capacity? I'm guessing it would be better to get tips with the element inside them as opposed to ones where you only replace the tip in order to get these specs. The T12s look fairly good, though perhaps a little long.

It's about time I got myself a proper soldering station.

>>1492895
Also anybody have any simple ideas to get around this? The little board has 4LEDs to tell battery level and when I start turning the light up, it goes from 3-4 lights down to 1 and then the one dims as I went full power and it cuts out. I’m going to try and charge to a full 4.2V and see how she goes since I was at 3.8V on the battery before.

>>1492914
This too... does anybody have a copypasta for decent soldering stations or irons? I know about that Hakko for around $100 but since I’m still just fucking around with this, I would be down with a $50 Chinesium copy if it works long enough for me to learn what the hell I’m doing.

>>1492917
There's a lot of variety when it comes to soldering stations (especially on alibay) and there isn't a nice spreadsheet like there is for DMMs (that one is on the eevblog forum). I've been wanting to make such a spreadsheet for a while (since I'm a bit of a spreadsheet autist like that) but as someone without any such iron it would be hard to do it with my own knowledge (unless there are datasheets for the tips somewhere), and in either case I'd need probably some way of collating data from a bunch of users on this general. If anyone wants to help with this project or otherwise has ideas I'm all ears.

>>1492914
Big Clive recommends Yahua stations. Think you can get those with the pump too. Recently someone I know got a pen-type soldering iron, but it uses a 12-24V DC input, and has temp sensing built in. Something like 30-60W.

Can someone explain something with transistors and heatsinks to me?
I can't seem to find any answers on google worth a damn. Usually just one-off answers, whereas I'd expect a mountain of answers for something so commonplace.

Why are transistors electrically isolated from the heatsink? Even if its the only transistor on that heatsink. Multiple transistors, obviously isolation for in case their collectors are at different potentials. But even so, all of the tabs are electrically isolated. Why is this done?

>>1492914
>It's about time I got myself a proper soldering station.
I feel you. Chinese ones will do for most applications. But if you want to burn some money, get a JBC station. Even their lowest model costs like 500 USD kek (i fucking want one)
Currently using a chinese one. Can't supply enough heat if I'm working on a really chunky, heavily soldered section of PCB

Oh just remembered, TS100 soldering iron is good too

>>1492962
What about something like pic related?

Why is this inductor/transformer present like this in this power supply circuit? I believe this came out of a laptop power supply, but I procured it in this state (all I've done is replace a blown cap)
The circuit symbol for it is LF1, what does the F stand for?

>>1492959
having that voltage just hanging around on a big piece of metal is usually a terrible idea and limits your options wrt physical reinforcement to the case etc.

>>1492967
common-mode EMI choke, keeps high-frequency switching noise off the mains

>>1492966
Yeah that's a chinese one and it'll work great. I have one similar, less the hot air bit

>>1492968
>having that voltage just hanging around on a big piece of metal is usually a terrible idea and limits your options wrt physical reinforcement to the case etc.
but it doesn't have to touch the case, like in some ATX power supplies. Yet in those, the heat sinks are mains voltage anyway. So what was the point of electrically insulating the transistor? IT just doesn't make sense

>>1492970
When you're poking about with a screwdriver or have a loose wire, it's just one more thing to go wrong.

>>1492974
There has to be a better reason for than "some random person might open up our product and accidentally short something out"
Besides, in my example the heat sink was at mains potential anyway, even though the transistors were electrically isolated with thermal pads so what was the point

>>1492980
Why were the heat sinks at mains? Because often the thermal pad on the back of a BJT is referenced to the collector, so depending on whether the transistor was on or off you'd have a vastly different voltage at it. Heat sinks are often soldered down into the PCB, and grounded for the same purpose that we ground the chassis.

In general I think it's a good idea to stop current from being able to flow through the heat sink because you might get bad conductivity and heating up in spots.

>>1492980
>thermal pads
ah, so they didn't have the rest of the insulation setup (bushing and insulating washer) on the screw? might be gap fillers, used for thermal conductivity instead of electrical insulation

>>1492987
They're used for both, but there are electrically conducting and insulating versions of thermal grease. AFAIK the conducting ones conduct heat better as well. I don't know if you can get electrically conducting thermal pads.

>>1492968
So not a transfomer at all, and the F stand for filter?

>>1492995
F is *usually* fuse, insofar as such things are standardized
that's still a perfectly reasonable deduction

>>1489791
I want to increase chooch factor of this 8" electrical ride.
I have bought a proper controller (without any speed limits and zero-start), it even came with LCD screen, which is nice.
What type of wire should I use for signal stuff (throttle, brake, +36V for LCD, tx, rx...)?
Can automotive fuse deal with 36V Li-ion battery (42V fully charged)? And how many amps it should have, considering that controller is limiting current to 15A.
Can I connect two different batteries (with BMS and shit) in parallel via fuses (lets say 7.5A ones, since even chinkest 10S2P battery will blow this up easily)?

>>1493028
The controller may have an internal fuse.

>>1493031
It doesn't. Positive and neg connect directly to mosfet bridge and cap.
Btw, it is updated KU63 (it is possible to find a schematic), that does have SMPS for mosfet driving, unlike original, where it all was done via LM317. Chip from XCM MCU...

>>1493028
you could probably get away with 24AWG for the signal wiring as long as they're not sending appreciable power through it
>fuse
ATC fuses are only rated to 32V. a 125V fuse and holder rated for the appropriate current would work fine
>different batteries in parallel
NEVER connect *different* batteries in parallel

>>1493049
What DC voltage are those good to?

>>1493052
Not him and i hope i'm not speaking out of turn, but afaik, wires are limited by current, generally speaking, and not by voltage

>>1493061
I meant the fuse. I know that DC fuses have to be able to block the arc while AC fuses don't have to worry as much since the voltage passes through 0. It's why AC fuses typically have a lower voltage rating at DC.

>>1493062
Oh derp sorry bout that anon

>>1493052
oof, that's a good point. 125Vac ~= 30Vdc
maybe use 250Vac rated fuses and holders

>>1493072
I'm surprised it's that low. Some boats/trucks are bound to use 48V systems, might find cheap fuses for those, though 240VAC fuses aren't particularly expensive. Cheap DMM fuses might be the way to go since those things are good to at least 100VDC.

>>1493082
meh it was a guess based on ratings I've seen. always check the datasheets
there is also telecom. also there are these
https://www.littelfuse.com/products/fuses/automotive-passenger-car/high-current-fuses/157_5701.aspx

What is the easiest way to shift voltage range? For example you want to turn 10 - 20v to 0 - 5v to make a sensor output compatible with arduino.
Does it have some special name?

>>1493108
voltage divider + op amp

>>1493108
Voltage divider, but that scales it by a factor rather than subtracting a fixed amount
A level shifter deals with voltage like that but a level shifter only deals with 0 and 1 logic levels and nothing in between

>>1493108
Logic level shifter, or something similar, at least for a digital one. There's a method for this that uses a single FET and maybe a couple of resistors. For an analog one whatever >>1492176 is looks promising. calculate its DC "transfer function" and you should be able to apply that circuit to what you're making.

>>1493111
Level shifter based on a current mirror can deal with analog signals quite well. But you need a level to shift *from*. In your case it should be either 10 or 20 volts dc.

>>1493111
It doesn't have to be exactly 0 - 5, it just has to fit entirely within that range, so if the op amp can magically divide the voltage by say 0.2V then it would move from 10 - 20 to 2 - 4 range with is perfectly fine, since the arduino can read it now with its analog pin
Can you provide a simple example circuit? I do have some op amps at home to try it with.

>>1493115
If you can afford to loos a one bit of your ADC precision, just use a simple voltage divider. It will bring 10 - 20 v range to a 2.5 - 5 v range.

>>1493115
If loosing 1 bit is not acceptable, then it's going to be the same divider plus an op amp multiplying by two.

>>1493114
I'm assuming he has an analogue signal that can be anywhere from 10V to 20V and wants to turn it into one from 0V to 5V. Ignoring the fact that I don't think arduino ADCs work up to 5V, I think that voltage divider with an input resistor will work just fine since the transfer function gives a "V_out = A+B*V_in" format. Though certainly much uglier.

>>1493123
Ok so it's impossible to linearly shift 10V to 0V with resistors and a 0V rail alone, but with a -5V rail it's a piece of cake.

>>1493049
>you could probably get away with 24AWG for the signal wiring as long as they're not sending appreciable power through it
I think even thinner might work. The problem is that some wires don't like vibration and they just break. I tried using stranded ethernet cable before, it ended up breaking.
>NEVER connect *different* batteries in parallel
Why? They all have BMS, so they shouldn't do spectacular shit. Also I would a fuse between them, just in case chink BMS fails and doesn't open circuit in case of excessive current.
Dunno, you can't use different batteries in series for sure, but I see no problems with parallel connection, as long as they are charged to same voltage.
>ATC fuses are only rated to 32V. a 125V fuse and holder rated for the appropriate current would work fine
What will happen to automotive fuse at 42V? Gap seems big enough to prevent arc.
Good thing about automotive stuff that it is doesn't break with vibration, not sure if glass fuse will survive this.

>>1493129
If they're just MOSFETs and a little IC on them then I think you're right, they will auto-balance since they're in parallel (provided it's same number of cells in series with similar enough chemistry). But if they're in anything with voltage feedback, like with a linear regulator (god forbid) or a buck/boost regulator then you can't put them together. You'll still have to have them on seperate charging circuits of course. If you want to share the current unevenly in a way that the cells wouldn't do naturally I'm not sure how you'd do so. If you trip one BMS's overcurrent protection then you'll trip the other's in short order.

Glass fuses are pretty damn tough, tougher than your windshield or mirrors in any case. There are also ceramic fuses which are even tougher. At 42V I would not be surprised if you could get a sustained arc within a piddly little blade fuse for a few seconds. Just look at the kind of distance a tig torch works with, you can stretch that arc at least a few inches with enough current (which a big battery can provide).

>>1493108
Here you go, mate. It's inverting, but I don't think it matters much since you're doing ADC anyway.

>>1493130
Yes, they are basically MOSFETs with ic, that monitors total current and monitors voltage on each cell. One (stock) battery has balancer circuit, while second battery (that I have to buy) is a hoverboard cheap battery, it will have only protection, no balancer, but in think they won't go out of balance with low currents (3.25 A per cell isn't much)
In case of overcurrent and two fuses blow, I can simply install spare 15A/16A fuse for stock battery and continue driving until battery is empty.
Also I will check for proper fuses, since I don't want jet engine here.

>>1492790
>how much load are you putting on that pin?
Ha ha, oh man, I forgot to limit the current on the base of the transistor. It's sitting there on my desk instead of in the circuit. FML.

>>1492873
Managed to find the same transistor on a store here, got the last 4 of them. The smaller ones that went arabic can be replaced by a bc557 I think

>>1493167
Anyone know how to take this pot apart? It`s the only thing keeping me from acessing the board

>>1493172
Pull the rubber knob off the front.
If you don't have the strength ask your boyfriend to do it.

>>1493174
Ah, I thought they were screwed or glued somehow. Thanks anon. I was just afraid of breaking something, strenght isn`t a issue ;)

>>1493172

that big blue cap is looking definitely bulgy.

Holy shit this was almost impossible to solder.

That stupid heatsink was stealing all my soldering iron heat and i was only BARELY able to half melt the solder

>>1493262
yay it's working! those crafty chink pongs have done it again. a tiny little led that takes 230v directly like nothing
god that light hurts my eyes so much

>>1493224
>IND. BRAS. IND.
Huehuecap?

>>1493283
almost finished.
this is great, no longer will i be enslaved to the grocery jew

>>1493262
>>1493283
>>1493302
So this is for growing vegetables?

>>1493302
>cardboard
shiggy diggy doo

>>1493319
yes.... vegetables....

Scope anon here. The seller `forgot` to tell me it`s a 110v device and the backplate told me it was 220/110v but I later discovered it was either a 220v or a 110v device. (I`m not that fucking retarded it`s literaly written 220v use fuse 0.5A, 110V use fuse 0.something). So after plugging in 220v mains it went arabic, I replaced the assploded PSU parts, it turns on now, get a trace (a fucking ball) but it doesn`t move and after a while smokes comes out of the HV side. Into the trash it goes. Dumbest mistake and purchase of my life. I think I`ll off myself because someone studying EE doing this deserves no mercy
>>1493284
>>1493284
Yes,but all Minipa stuff is copied from china. This is a kiksui something 20Mhz shitter.
>>1493224
Checked, it`s working. It`s shaped like that.

>>1493262
I have these, the best way to solder that shit is to put them on a clothing iron and heat up while you solder.
>>1493324
I think I`ll just pussy out and shell 500$ on a digital one before our next president makes our currency zimbabwe level and I can`t buy my chinkshit anymore.

>>1493326
I`ve managed to find the service manual on a chinese site but it`s unreadable mess, as a A0 sheet split in 8 A4 scans all on diferent orientations.

>>1493324
I just got a cheapo USB chinkscope for $50 and couldn't be happier. it works great and i can read it on my huge ass lcd monitor. it even has isolation so you won't fry your call of doody machine if you fuck up

>>1493329
M8 I think I`ll do that. The guy at least sent me 2 new 20Mhz probes which are like 1/2 of what I paid for the scope. The bad part is that I can`t even use the transformer for a power supply because it`s 110v and makes a really angry noise when on 220v. If I contain it`s anger with Kapton tape it might be doable but still.

>>1493302
Fuck. It looks like i might not be able to get away with completely silent growing and might have to slap a fan on it. It had it running for 1 hours and the led was at 80°C and the heatsink was too hot to touch at around 60°C

>>1493342
I made a grow box with a 50w led, what I did was to take a sheet of aluminum and use it as a base for the led (going from end to end of the box). It dissipates most of the heat, and I`ve added a ventilator that turns on if it gets toasty

>>1493343
this. you do NOT want the lighting waste heat INSIDE your box

>>1493350
>>1493343

I did this. It turns off if the fan is on and it is still toasty too and only turns on when you press the on button again.

>>1493134
Now, non-inverting and with 75% less bs.

>>1493352
but you're just drawing hot air from the LED through the plants on one side. mount the fan closer to the light and you'll get good airflow on the plants
if you're bringing a PIC into this anyway, may as well use a PWM variable-speed fan and just run it proportional to temperature

>>1493364
I did that, I was just retarded and made the wrong drawing.

>>1493262
wait, why does it have L pole? with 220V AC it doesn't mater which way you connect the wires does it?

>>1493329
Does yours use open-source software? It's not necessarily a requirement for me, but you can find $50-$150 portable scopes with limited range or shitty interfaces that don't work well with live logging, $200-500 desk versions with LCDs, or $50-100 for a USB box. I think I've seen Windows only versions and cross-system versions but I'm not sure.

>>1493368
I'm assuming it's because the aluminium PCB is grounded to the N line and hence so is the heat sink (if you're using conductive grease).

>>1493397
Well.. i am using a cable with 2 prongs so every time i plug it in i enter a lottery where i could be making the heatsink be 230V. I hope my luck won't run out soon. But then again it's not like i plan on touching it when its on, i am scared of the whole thing coming crashing down as is since it's literally just loosely placed cardboard boxes on each other

>>1493415
>cardboard boxes on each other
shamefur_dispray.jpg
imagine what happens when you bring plants and water into the equation
you really, really, really need to fix that first. rig up some picture wire and a wooden hutch, geez

>>1493415
Don't you have a 3-prong outlet? I'd test if there is continuity between the prongs and the heat sink and the PCB just in case.

>>1493323
More like weeds.
Don't forget about ventilation, or police might arrive on party

How do I remove alternator noise coupled into my aux cords in the car? Creates an awful noise that increases in frequency with RPMs. Is it common mode noise? Do I use a common mode choke? Is there some kind of common mode attenuator with op-amps I can make? Where do I put the circuit? Output of phone or input of stereo?

On a related note but not quite are 3.5mm male-male TRS connectors shielded? I know they have a ground but is it just a ground wire or does enclose the two signal wires? If I converted the 3.5mm to RCA and then drove the ground lines with a buffer amplifier whose input was the signal applied to the center conductor would this create a guard ring that reduces noise even more than if it were grounded? Should I just drive the guard ring with a DC voltage instead?

>ISL238533 instrumentation amplifier with programmable gain
>2 bit tri-state logic
>reduces component count
Fucking hell. I'm runnign this on split +/- 2.5V supply and trying to get it to interface with regular 0 - 5V logic is a pain in the ass. reducing component count my ass
Anybody have any ideas?
it's on split supply, and the gain setting pins are referred to V-
the gain setting pins are G0 and G1
in this case with my op amp's supplies:
"0" is -2.5V
"1" is +2.5V
"Z" is 0V
rest of my circuit is 0 - 5V stuff
I need:
-2.5 -2.5 (0, 0) to set gain of 1
-2.5 +2.5 (0, 1) to set gain of 10
gnd gnd (Z, Z) to set gain of 100
+2.5 +2.5 (1, 1) to set gain of 1000
It wil be controlled by an arduino
All circuits i've come up with require multiple transistors already and are inefficient and inelegant
pls halp

>>1493550
Try a new battery. If it will make any difference - it is alternator, if stays the same - it is probably ignition pickup.

>>1493584
set the pin to an input to get the 0V level?

>>1493262
This is where high end soldering stations come in
alternatively TS100. videos seem to show that thing able to supply enough heat fast enough to solder right to things like yours

>>1493592
gnd and 2.5V is easy. -2.5V less so. getting them all to work without having to give each value a different circuit is hard

I just bought this beauty because my fluke 17b+(chink) wasn't accurate enough for my current project. What digital multimeters you use and why?

>>1493597
i use the eevblog BM235
I really want one of those big boy benchtop 6.5 digit DMMs

>>1493597
Nice choice. I'm rocking a Hioki DT-4252. It's not that suitable for small-signal stuff, has no mA/µA range and only goes down to 0.000µF on the capacitance range, but it's bloody bulletproof for mains testing. When you remember to flip the leads back from current mode, that is.

Hi, I'm a new to EEPROM and IC component in general and I was wondering, in the datasheet of my 28c16 (pic related). They talk about operating Voltage, but no operating Amperage. Does that mean I can run it on 5v 2A or will I burn it ?

>>1493643
>new to EEPROM and IC component
Sounds like you're new to the whole concept of electricity.
ICs operate on constant voltage. They consume as much current as they need, provided the source is powerful enough.
You cannot force 28c16 to consume 2A without raising the voltage above 5v.

>>1493644
Aaah i see now thx

>>1493643
>no operating Amperage
it's called current, here
look for the Icc or supply current specification in the datasheet, it varies by mode and operating speed

Hi there,

I am currently "upgrading" an old PSU. Basically designing a completly new PCB and changing the displays. While I'm at it, I thought I would need some more power, so now I want to change the power transistors to TIP142 and use 4 of them simultaneous.
Now to my problem, I don't know about the power dissipation of the cooler. The easiest would be to calculate the heating of the old Transistor and use the maximum rating of the PSU to set as a maximum of produced heat of the new PSU. As you might have read, it was only one Transistor until now. I'm not quiet sure, how the inner Transistors on the cooler would behave with the new setup. Is there a possibility to calculate the tc (thermoconductivity) without any graph, only with the Volume and the surface? The Formular doesn't have to be exact, I just want to calculate the tc of the cooler for the inner transistors, to be sure they won't reach 70°C.

Anyone have a good guide on making a bench psu? I have an old 5A 110/12v transformer here laying around. Most circuits I find on the web is just
>lmao just put a lm317 there man

>>1493875
why not empirically? since heat sinks are generally specified in °C/W, just apply watts and measure delta°C under actual airflow conditions
also, if there is no fan on it, simply adding a fan will massively lower the thermal resistance of any heat sink vs. still air
you could also increase efficiency in the design phase with a tracking buck pre-regulator, as another anon a few threads ago considered, so that you don't really have to worry about it

>>1493882
lmao just put an lm358 there man
>don't actually do this, use a proper rail-to-rail I/O op amp

>>1493886
>why not empirically?
Thats a method that needs time (The PSU is broken atm and I would have to mount a heat source), I hoped for a formula for fast reverse engineering. Empirically is the method that will be used this weekend, if I have no other choice.
>no fan?
It's a PS2304D, so no fans. Also I want to keep the noise as low as possible and I suck at mechanical work.
> tracking buck pre-regulator
Great for efficiency, I will keep that in mind. Before I add more regulators, I want to calc if the efficiency is high enought for the heat sink. The basic Idea right now is, to redesign the PSU, but keep it as analogous as the original.
Meanwhile I am keeping placeholders, to mount a µC and replace the regulating opamps. I even play with the thought to let this µC get an ethernet/USB connection, to display a signal curve defined on my laptop. Would be really cool, but also very time consuming.

>>1493875
I'd advise getting/making a new power supply, one that doesn't use darlingtons but FETs instead. Upgrading that sort of thing probably isn't the best use of your money.

>>1493903
I understand this point very much, but this time it isn't about money.
I want to have fun with my hobby and have some "antique" analogous PSU to show off.

>>1493886
>lmao just put an lm358 there man
>don't actually do this, use a proper rail-to-rail I/O op amp
pls I want something decent my man. It`s going to be my first big boi project

>>1493906
In that case go for it. Better make your new PCB with nice and curvy traces.

>>1493911
that's legit the structure of a decent linear power supply, my dude. it will work nicely for relatively well-behaved loads and reach down to millivolt output. current limiting, component selection, etc. is left as an exercise for the engineer
if you don't like this design, you could also consider one of the old linear regulator controllers e.g. µA723, UCC2837 and follow their detailed design procedures in their application notes

>>1493937
So what you are suggesting is make a linear regulator but discrete? The dude was asking for a bench power supply.

>>1493954
what do you think a bench power supply is?
a proper bench supply isn't going to connect a buck directly to the output because output ripple will be embarrassing, especially as the set output voltage approaches zero (which, additionally, very few if any monolithic regulators will do)
maybe that dude should reply with actual specs that he needs or at least some suggestion as to what he'll be hooking up to it

>>1493960
Well having an output rail a volt or two above the circuit's 0V will solve the problem of not reaching very small voltages, and these days buck converters are probably good enough to give you sub-0.1% ripple, though having a ~1V drop linear regulator on the end would definitely make it easier.

>>1493028
Shit is good even with one battery. But yeah, I need to hide wires and controller somehow

>> - (-- +
So pic related is a GBS8220. I want to power an LM1881n using the connector on P6 so it's removable, however from what I can tell, P6 is a power input, not an output. Would reversing the positive and negative make P6 a power output? I'm also going to wire an L705CV 5v regulator with the LM1881n.

Anyone got recommendations for cheap fab houses? I need <5 pieces of a small board, 90x60mm, preferably 4 layer but I can probably make do with 2. No advanced features. I'd like to spend less than $30.

>>1494016
jlcpcb has good deals for first time customers, looks like your board in 4 layers will cost $29 for 5 copies shipped to burgerland. if you're not a first time customer, it's more like $42
see pcbshopper dot com and play with the numbers as you like

>>1494022
Thanks.

I'm coming to KiCAD from Altium. One feature I can't seem to find is a way to toggle the way wires and tracks bend, like in this picture. I was thinking it'd be "Switch Track Posture" (/) but it doesn't seem to do anything. Anyone know if there's a way to do what I want?

>>1493937

>>1493960 Unironically.
Big boi commercial PSUs are basically elaborate discrete regulators designed to meet certain output specifications - voltage range, current range, power, ripple, efficiency, safety features
You can buy a regulator and make that your power supply. for sure. But a "big boi" one is going to be made to specs desired as above - mostly because regulator most ICs can't cover all those bases.

>>1494014
>Would reversing the positive and negative make P6 a power output?
No. It's just in parallel with the barrel jack by the looks of things, so leave the polarity as-is. Check the continuity just to be safe though, could be a diode or something in there.

>>1494029
/ has been the key to do that for quite a while. check under Route → Interactive Router Settings... and set Highlight collisions. also check your design rules

>>1493967
as much as I love me a good switcher, they alone can't provide solid output with negligible ripple over sub-mA to multi-A loads from 0 to tens of volts and anywhere in between

>>1494052
So do the common high-power power supplies use a switcher for the majority of the work and a linear regulator setup of some sort for smoothing?

>>1494040
>1/2
So does this mean I can just wire them with the intended polarities?

>>1494064
>>1494040
2/2

>>1494059
I can't speak to anything recent, but older bench supplies of small to medium size used a discrete linear regulator circuit
newer gear might go the pre-reg route, unless some fugger's got a patent on it

>>1494052
Is there no way to stay on walk around or shove mode and still have the ability to change posture?

>>1494077
on shove mode, / will change posture but it might not if it sees something in the way that it can't shove. see if "Allow DRC violations" makes it work for you, if not, post a screenshot of what you're trying to make it do
I haven't had an occasion to use walk around mode yet

>>1494059
I might be wrong but i think modern bench power supplies are fully switched mode outputs, unless you buy one specifically that is linear

https://www.aliexpress.com/store/product/high-voltage-cap/2349218_32832256637.html
How sketchy would it be to use some of these caps in a tube amp? I need two 16uf caps rated at 450v and one 30uf one rated at 500v for the power supply section. I was thinking of just shunting the positive leads on one of them together to get a 32uf.

>>1494022
pcbshopper looks interesting, but it missed the fact that JLC is doing a 10 boards for $2 deal. As soon as I'm able to confirm my email and make an account I'm going to be buying some.

>>1494095
I wouldn't work that close to the edge of the spec, especially with the chinkshit factor. why not a film cap?

>>1494098
yeah pcbshopper only knows about regular prices. that's why I made a point to mention JLC specifically

>>1489791
Hey fellas. I built this preamplifier. I works very well and the amplified sound out of it is pretty clean and clear for a high gain amp. The issue is that tube V3 makes a sort of hissing noise. The noise isn't electrical, it doesn't come through the speakers (I only hear it when the speaker output is switched off) it is some sort of mechanical sound. Any one got any idea what's up with that?

>>1494064
Yes. If you draw out a rudimentary circuit diagram (pic related) it should make sense, provided you have a qualitative understanding of Ohm's law and that voltages add in series and currents add in parallel.

>>1494083
That's what I was thinking.

>>1494102
looks a bit like a Soldano
how are your supply voltages when the speaker output is switched off? if C and/or heater voltage is too high, any number of exciting and damaging processes could be going on inside that tube

>>1494109
It is, good eye. It is a SLO100 preamp. I run it into my Twin with a looper for a sort of best-of-both-worlds sort of deal. I love this thing.

The heater is a little high at ~6.7VAC. The B+ for that 12ax is a little high at 400VDC. My twin runs plates at 410, though.

>>1494118
ah yes, that cathode follower third stage is pretty distinctive. noice
hmm, yeah, I'd drop half a volt to each heater if you can and see if that helps the hissing problem

>>1494118
but before you do that, can you swap the tube out with another?

>>1494107
Thank you for your input anon.

>>1494125
>can you swap the tube out with another?
I did that first thing. Even bought a new JJ because the guru geniuses on the guitar/amp message boards seemed to agree it was really quiet. One weird thing is that adjusting the volume knob effects the mechanical hiss. I think maybe it is just the amount of current flowing through the cathode. As well, lifting the ground from the tone stack seems to make it act up a bit to.

Between this and my last post I went through the three major B+ dropping resistors and found they were somewhat far off from their nominal values. One of the 15k ones was at 13k or so and the one that is supposed to be 10k was almost 13k as well so I went through my collection of those resistor and found the ones which were closest to the nominal values. The B+ on V3 is now about 393VDC which is marginally better.

Chop sticking the Volume pot also seemed to have some effect on it. I am reluctant to think that the hiss is coming from that, but removing the tube would also remove current flow from that pot so I could have been mistaken in my diagnosis.

Pic related is a gut shot for all you sinners. Please no bully, it was my first point to point build.

>>1494142
>what are PCBs

>>1494142
jesus fucking christ

>>1494147
Whoa shit, PCBs? Thank you for pointing out the existence of these, I'd never heard of them before, thank you citizen.

I'd never done a point to point build and this was low parts count so I went for it. It doesn't look great but it is easy to work on and it works quite well. Normally I build amps with eyelet board because they are simple enough to do so with. The time and expense of designing a PCB, especially when you have to take into account routing heater and B+ so you don't get any noise is often not worth it.

I did use a PCB in part of my twin, the bias vary trem to minimize footprint but normally I stay away from them for the things I do.

>>1494149
Eat a dick and be glad you've never worked on any electronics from the days when point to point was the standard. I could show you rats nest nightmares beyond your wildest dreams.

Pic related is a tube phonograph and radio unit my neighbor brought me. This is a piece of consumer electronics from 1948, a Bendix Model 676B.

>>1494152
just buy a perf board you brainlet, it`s basically point to point minus the fire hazard. That shit from 1948 is better than your shit because atleast the wires are isulated.

One of the bare wires is a ground bus, the other is the 6.3V heater bus. All B+ carrying wire is covered.

I have perfboard but no need to use it, point to point worked quite well. I also have strip board, tag board, turret board etc. I use what I think will be most prudent. I don't feel unsafe using this device.

>>1494154
This desu
Perf board and also use wire and route the wire neatly where needed. It works great and looks decent too

>>1494154
>>1494157
Perf board is for shitters. Just get some copper-clad FR4 and go to town with the magic marker. For that kind of through-hole PCB it's more than accurate enough and it's not a bitch to solder. Nail polish works a treat if you can't get a marker to stick in the etchant, though it's a pain to clean off again.

Is there such a think as Y dupont wires or Y pin adapters?

I often need to connect several things to a ground pin or voltage pin but boards usually have only one or two, so i am force to cut several dupont wires and make my own Y wires which is incredibly tedious and annoying.
Yet i can't find any Y pin adapters or Y dupont wires on aliexpress. Don't fucking tell me nobody has thought of making anything like this yet.

In development it's fine since i can use bredboard, but i don't want to fucking have to use breadboard in production, and deal with wasted space and fallen out wires all the time

>>1494162
Make a bus-bar out of a pin header with a wire soldered across the short side.

Hello guys,
I am thinking about building a """big""" 7-segment-display myself.
I thought about driving it with rgb leds because muh colors. Now I ordered smd rgb leds, but i didn't think about them being common anode. That means that i am unable to put them inseries, obviously.
Now i thought about putting them all in paralel. BUT, if i do it like pic related, i'd have (if i use 10 leds for each segments) to put fucking 30 resistors on each segment.
Can't I just put one resistor at the vcc? Wouldn't this just increase the drawn power on that one resistor or am I a brainlet?

Any other idea to design it? each segment should have around 10-12 leds.

>>1494171
here's a pinout of the type of leds i have

What could happen if I use the transformer of a 110v appliance to make something that runs on 220v? It`s a 80W transformer. From my limited knowledge the only problem I imagine happening is the coil isulation failing, but when I pluged it in all I got was the angry noise of a metal tab vibrating. On long term operation what else could happen?