/diy/ - Do It Yourself

>>1494960
diodes with the pullup. unless you're running <3.3V you should have plenty of headroom for the next CMOS device to see a solid logic 0

>>1494969
yep

>>1495036
this thread's digits brought to you in part by the MBI5036 constant-current LED sink driver with blah blah blah

My smart fuse is finally finished.
All i need to do now is to make some pretty 3d printed case for it.
The knob sets the maximum allowed amps of current, up to 5000ma, and if it crosses that limit, the fuse will shut off the power, as can be seen with the led on the power supply current read out.
No more smelly burnt shit and shorts

>>1495061
Sweet!

I finally finished this light. Has a 20650 instead of the 3x AAAs, USB rechargeable, magnetic base, and the switch I used had a high and low. Not sure how long this will last.

Only thing I woulda liked to do different is mount that circuit board in a way where I could view the battery level lights. Maybe when I get gudder at this. A 3D printer would be a blast to design and build a proper case.

>>1494929
I replaced the part and it still doesn't work so I assume it's my computer. Is there like an update I need or something? I've installed every driver I could find and it still won't recognise it or let me interact with it with libusb. It's an owi edge if anyone is familiar.

In a linear power supply, how is the output supposed to be handled such that if you dial it down from say 10V down to 1V, the output will also go from 10V to 1V when there's no load? the output caps won't just discharge on their own. How do commercial supplies handle this?

Why isn't there continuity between my phone's USB ground and AUX ground? I measured the impedance between the two at about 6MΩ. This is causing noise issues when playing audio in some circumstances.

>>1495270
Just put a large value resistor to ground on your output. This will lightly load it at all times allowing your output cap to discharge. You will be wasting some power in that resistor but it'll be very small assuming you use an appropriately sized resistor. 1MΩ is probably a good fit for most circumstances. Larger resistors decrease the power but increases the discharge time. Smaller resistors do just the opposite.

If this was a switched mode power supply the resistance of the feedback network would usually be enough to provide the required loading. For that matter linear regulators also have a feedback network which should also discharge a capacitor internally through the IC. How are you accomplishing variable regulation without any kind of feedback network?

Is it ok if my Li-Ion battery is leaking liquid out of it?

>>1495307
Feedback network draws less than a milliamp at 30V. It'll be ages before the output got from 30V to say 1V
There's no way to cause that to happen sooner without a legitimate load?

>>1495310
Simplest solution would be a MOSFET to ground on your output and a push button and when you change voltages you close the switch which turns the FET on a shorts your output cap. A better solution would be a series P-FET and shunt N-FET. P-FET is normally on N-FET is normally off. When switched is closed P-FET turns off preventing you shorting your supply. N-FET turns on and only serves to short out the charge on the cap. Done.

To be honest though both solutions are crude since it always relies on a manual reset. You really want circuitry that will detect a change in set point and enable something like a current source to drain the cap. I'm thinking current sense resistor in the feedback path with a differential amplifier across it which drives an N-channel MOSFET on the output acting like a current sink. It'd be operating in it's linear region though, not in saturation. I dunno if this is a GOOD design or not but it's a rough one I came up with after 2 minutes of thinking and no refinement.

>>1495314
forgot my 9000 hours in ms paint pic

>>1495308
jesus no

>>1495308
Depends, you're gonna need to do a taste test. Brackish and metallic taste is good. Sweet taste means something has gone horribly wrong. Sample it multiple times to be sure.

>>1495318
Meh, 3/10, would not mix with tonic again.

And what will happen if I try and charge these? The one is 1.6V and that’s the only I have any sliver of hope for.

>>1495315
>>1495314
Cool thanks for the idea and genuinely helpful advice bro
I've got another question for you since you seem to have a greater breadth of knowledge - https://www.eevblog.com/forum/projects/general-purpose-power-supply-design-7488/
Why does this guy have the op amp in the "regulator" section wired up as a comparator?

>>1495323
I think the thing to do with them is to throw a very low constant current through them and see if they get to 3.2V. Maybe around 20mA or even less.
https://youtu.be/aEVxVVET0nE
https://www.eevblog.com/forum/beginners/reviving-a-0v-lithium-ion-battery/

>>1495298
there's no continuity because the audio output ground is driven
you need isolated power and/or isolated audio. isolated power is probably easier to provide

>>1495330
what, U1? you mean, with + and - interchanged? the base drive for the pass transistor is provided by Q3. when the ref voltage is lower than the feedback voltage, U1 turns on Q2 more, which steals current from Q1's base and slows current flow to the output

>>1495270
a smaller output cap, just big enough to prevent major load spikes until the regulator can supply them, is probably a good way to do it

>>1495346
Oh, if it's a driven ground can I used 3.5mm to RCA shielded coaxial cable and then just whack this on the end for some common mode rejection?

I sort of outlined the problem I'm having last thread. New car. No aux input. Got 0.5W FM xmitter. Plug phone to transmitter with currently unshielded 3.5mm to RCA cable though I've also used unshielded 3.5mm male to 3.5mm male cable the transmitter has inputs for both. Transmitter is plugged into car accessory outlet. I have both plugged my phone into usb for charging when playing music and have left usb disconnected. This makes a big difference. With USB disconnected I have major issues with random pops and cracks that are nearly loud enough to blow a speaker, even at relatively low volume levels the spikes are that high. I absolutely cannot play music with the USB disconnected if I do not want to kill my speakers. With USB connected it eliminates those large pops and cracks but I still have problems with audio frequency whine that increase in frequency as I accelerate. I originally thought it to be magnetic coupling from the alternator but I now believe it to be a ground loop.

I have two accessory outlets. One has a USB thing that plugs in to charge the phone. The other powers the FM xmitter. Both of those should come from the alternator and should be at roughly the same potential. The sticking point I think is this psuedo-floating audio ground on the phone's aux jack. I think this is causing the ground loop. This is NOT at the same potential as the transmitter/USB. I'm not sure how to solve this. Maybe isolation transformer but I'm not sure and I'm worried that they won't have a flat frequency response over the audio range without buying something ridiculously expensive. Plus transformers have shit CMRR that degrades rapidly with increasing frequency if they aren't driven and I don't know how much drive capability the phone's output power amplifier has. It does a pretty poor job driving 56Ω cans on its own

Pic related, a loop antenna amplifier and power injector. this is my second attempt at "Russian style" boardmaking combined with the copper adhesive-backed tape. the trick, apparently, is to cut a V-shaped notch between pads on the laminate with two strokes, one stroke leaning to each side, then finish up by shaving off any burrs with a third stroke flat to the board
this is a fun way to build

>>1495346
Yeah U1
afaik that's a comparator topology. so all it would do is go to a voltage rail or to ground only, no? and create noisy output while it's at it
Also I can't quite figure out how he's got the current limit laid out - does the second stage of the current limit function as an integrator?

>>1495361
>>1495346
Also forgot to ask, in my falstad simulated circuit, even a 10uF would take a while to discharge with no load

>>1495358
You should go for Manhattan style construction instead to avoid cutting up your ground plane like that.

This is too cheap to be true, right?
90% efficiency 6kW inverter for just 40 bucks?

https://www.aliexpress.com/item/ping-pongs-chinkverter/32869436148.html

>>1495363
ehh maybe. I wonder if this copper tape will stick to Kapton tape very well, still seems kinda rough for SMD.
I could also just use a double-sided blank and drill a through-hole if/when I need a really good path to ground

>>1495364
its a modified sinewave anyway. Those tend to be cheaper, plus aliexpress which knocks down price a bit more

>>1495369
I am not sure what are you trying to say, so it it good to go? I want to use it to connect things like computers, blenders, mini fridge etc

>>1495369
Also, how do you tell if the invertor makes a pure sine wave?
When searching them most of them have no mention of this at all, like this one, i have no idea which one it is
https://www.aliexpress.com/item/chinkchongpong/32852734769.html

>>1495367
It generally works better with THT parts

>>1495373
99.9999999% of all electronic devices couldn't give less of a shit about what kind of abomination of a sine wave you shove in them since they rectify it to dc anyway, so as long as the average power is more than the device needs it will be fine
the only exception is shit like ac motors and some medical devices, so don't try to run things like MRI from that inverter

>>1495373
Yes you are good to go for most things

>>1495377
Usually one that is pure sine wave will go out of its way to say its pure sine wave.

>>1495381
This. Only certain applicances like things involving motors - heavy inductive loads.

>>1495362
add an integral load to the supply. some switchers have a minimum stable load so they do this too.

>>1495381
>couldn't give less of a shit about what kind of abomination of a sine wave you shove in them since they rectify it to dc anyway,

true, but pic related UPS worked fine with my old PC power supply, but the new one clicks like crazy so much that I bought a sine wave type UPS. (the voltage scale is weird; I must have used a divider or something).

and yeah, this is hardly a sine wave, to be sure.

>>1495399
that is the most square sine i have ever seen

>>1495404
they just added in a little extra frequency content for you. more is better.

>get idea for a fun new project
>order chinkshit from ali to make it
>by the time it arrives in a month be already burned out on the idea and don't feel like doing it
>toss the chinkshit into a box
>week later
>new idea
>rinse and repeat
>then again and again
>the box is full now
>it doesn't even show the really expensive shit i got and never use like the clamp ncv multimeter, or the scope and solar panel setup with agm deep cycle batteries and other shit
>break down in despair because even though i want to stop more than anything i just can't
>have currently 12 packages from ali in transit awaiting delivery
>running out of space, money and sanity
WAKE ME UP INSIDE WHY CAN'T I WAKE UP
please help

>>1495410
Well where are you and what gear do you have exactly? May not be useful to you but you could always sell it off online for half price, at least get something back from a fucking atrocious habit.

>>1495410

You could form a support group with Bepis for people who just buy shit but never actually do anything.

>>1495410
you should choose your projects better, buy stuff that you can use for a bunch of different things.
I have a box like yours too, and when I get an idea I can dig around in it and find what I need to put it together right away.

>>1495410
>agm deep cycle batteries
what batteries, where'd you order them from, and how much? I'm looking for cheap batteries for a battery backup system.

>your boards will NEVER get delivered

I tried putting the fans together, but only one spins. Can anyone tell me what I may have done wrong? I believe they have enough power

How do i avoid reading datasheets for components? I fucking hate it so much

They never just put the relevant information on top, it's always buried in 10 pages of useless shit that only .1% of people who use that thing actually care about

>>1495476
You don't. Reading datasheets is a skill you have to practice if you want to be good at electronics.

>>1495476
git gud with your ctrl+f and learning where to look.

how do you call that thing in opamp that describes how close to a rail can its output be?

>>1495482
output swing

>>1495364
Too good to be true. No way they can sell a 6kw inverter for so cheap when they are hiring the best marketing and advertising people in the business.

>>1495490
>outdoor

>>1495490
I ordered it anyway. My guess is that it works, but it's probably actually rated for less, if it can do 1kw it will be enough for me

I want to try to measure voltage (actually current) with the arduino analog pin, since the hall sensor seems to be really shitty, but the voltage can be 0 - 30V which is too large for arduino. So i was thinking of using an opamp with a resistor between output and - leg (amplifier mode) to sort of reduce and clamp the voltage to a safe 0 - 5V.
And then connecting the opamp output to the analog pin of the arduino
Do you think it will work?
See pic related, max current can be 10A, which the op amp multiplies into 4.2 something V, and if for some reason the current jumps to more, the op amp will be limited by its 5V rail and won't put out more than that, so the arduino should be always safe.
Seems to work fine in the simulation.

The only problem i can think if is what happens when i try to power the arduino from a wall adapter and the measured circuit from another adapted? will it shit itself? do i have to connect grounds of those two, is it safe to do so?

>>1495511
you don't need an op amp at all. 10A*0.1 ohms=1V FSR, and the atmega328p has a 1.1V reference onboard. you'll want a zener just in case you short the 30V supply.

>do i have to connect grounds of those two, is it safe to do so?
this depends on the adapter. measure the resistance from both V+ and V- to L and N of the adapter. if there's no continuity on at least one of the adapters then they can share a ground.

>>1495473

dont be a baby asking for help. so some basic troubleshooting: (1) remove heat-shrink, test if connections are good. (they wont be if you used solder without flux). (2) if you see nothing wrong, proceed to take it all apart and test each part individually. (3) take out multi-meter and measure impedance and current.

two most likely causes are bad solder connection, and broken wire.

>>1495517
forgot to mention a capacitor in parallel with the zener can be used to average the shunt sense voltage which can help if you're measuring something digital

>>1495473

dont be a baby asking for help. do some basic troubleshooting: (1) remove heat-shrink, test if connections are good. (they wont be if you used solder without flux). (2) if you see nothing wrong, proceed to take it all apart and test each part individually. (3) take out multi-meter and measure impedance and current. alone and in parallel.

two most likely causes are bad solder connection, and broken wire.

>>1495517
If you use the 1v1 reference can the pin still handle 5V input?
The main reason for the op amp was to protect the arduino from overvotlage.
I am afraid that the zener won't be fast enough, since shorts are very real scenario here.
Not sure about the cap, from my basic understanding the cap will just charge up and then basically act as "open" circuit with massive resistance.
The project will be used to measure power consumed by various devices and the resolution i need it at least 20ma

Where can I see the schematics for these things?
Is it even possible to mod them without destroying them?

>>1495546
>schematics
>for anal light up stick
It's literally battery, LED, maybe resistor for current limit and a switch. Fancy switch, but still switch.

>>1495542
>If you use the 1v1 reference can the pin still handle 5V input?
yes
>I am afraid that the zener won't be fast enough, since shorts are very real scenario here.
i'm not sure why this concern is so widespread since despite some research i can't find any data to back it up. zeners are used frequently in crowbar circuits and smps snubbers. the ~10pF pin capacitance will give you a 100ns time constant on the input pin which should be adequate, but if you're worried you can stick a 1nF cap on there which will eliminate any risk.
>Not sure about the cap, from my basic understanding the cap will just charge up and then basically act as "open" circuit with massive resistance.
this is true and it's basically what you want for DC. if whatever you're measuring has an AC component to its current consumption then the cap will average that out. you can do the same thing digitally though.
>the resolution i need it at least 20ma
you're kind of pushing your effective bits here. the internal reference will be more accurate than the 5V supply rail at any rate.

>>1495552
okay thanks, i'll try doing it your way. i do have a 5.1V zenner on hand too so that's nice.
i had no idea it would be this easy.

Can you get seven-segment displays with integrated BCD decoders or similar? I hate when I do a nice job laying out a circuit and then there's just a forest of cables going to the basic displays I need. This is a thing the world deserves and if it doesn't exist I'll burn society to the ground.

>>1495551
I'm new, man.
All I know is that there are these blueprints for every electronic device ever and that all theoretically should have one laying around somewhere.

>>1495571
You can get multi digit 7 segment modules with I2C driver chip onboard,

>>1495571
Not that I know of. There are modules like >>1495605 mentioned or there are BCD modules as well. Something built-in would not have much of a market share so it would be quite expensive.

I suggest you build adapter yourself if you need so many of them that wiring them becomes pain in the ass.


>>1495584
here you go. Well, quite literally it is just led/battery/switch connection in series, just nicely packaged.

>>1495607
Thanks, fren.

you those little screwdrivers you can stick into an outlet and they light up if the wire is live?
How do they do that without closing the circuit?

>>1495552
>the internal reference will be more accurate than the 5V supply rail at any rate.
I don't know about Arduino in particular, but ATMega's internal reference sucks some serious ass. Power rail has a good chance of being more accurate.

>>1495571
Sure you can, but you probably don't want to.
TIL311 is pretty well-known single-digit display with built-in decoder, but it isn't strictly 7-segment.

>>1495613
>those digits
eurgh jesus christ who'd make something like this

>>1495613
yeah you're right the datasheet says it's 10%. the regulator is at least 5%. sad when your rail is better than your reference. he can still measure the shunt directly though without losing resolution if he first measures the reference with the rail, and then measures the shunt with the reference.

>>1495614
Oh, the smart guys at Texas Instruments. And Hewlett-Packard/Avago/Broadcom too.

HP's displays look nicer, though.

>>1495571
>>1495605
Furthermore, why don’t you go with an LCD? More pixels = more better. And they typically come with a built in driver and/or I2C backpack so you can run the whole motherfucker on 4 wires.

>>1495609
capacitive coupling

>>1495609
Non-contact voltage probes work on coupled-capacitive sensing. As the magnetic fields change around energized conductors, the measured capacitance at the tip of the probe changes.

>>1495270
A regulated supply doesn't need an output capacitor. The "tank" capacitor which converts rectified AC to DC goes before the regulator.

>>1495620
>>1495622
i see, that sounds complicated
i want to make a make simple tester to check if an outlet is on or not.
i guess it will be easier to use two wires, to actually close the circuit between live and neutral. I was thinking i could basically do this, in series, metal stick, 11500 ohm resistor, a LED, another metal stick
you stick the sticks into the two holes in the 230V outlet and it lights up the led with 0.02A if there is power. but one problem i don't know what to do with is what if the voltage is slightly different? since the mains can jump around anywhere from 220V to 250V it will just burn the shit out of that led

>>1495622
>magnetic fields change
>capacitance changes
interesting meme

>>1495633

http://hackedgadgets.com/2014/05/24/non-contact-voltage-test-pen-teardown/

>>1495633
if you go the route of shoving something into the socket, put a 1/2W 100k resistor in series with your LED and a standard diode like a 1n4007. the current through the led will be around 1.5mA +-20%ish whether the mains is 220 or 250V. LEDs can't handle much reverse voltage so you need another diode for the negative half cycle.

>>1495633

you must have a case of the Dumbs today. just calculate your resistor based on the high voltage. so use 12.5K instead of 11.5K

for the screwdrivers you stick in the slot, they use small neon bulbs in series with a resistor, around 100K. the bulb needs such little current, that the few electrons that flow through your thumb and into the surrounding air is enough to light it.

>>1495410
holy fuck this hits home so hard

>>1495378
ok, gramps

>>1495410
this anon
>>1495469
gets it

>>1495476
the absolute state of amerifat education

>>1495473
>soldering gun
found the problem, gramps

>>1495653
I'd have hoped binary clochmed would have given up by now

>>1495631
A linear regulator should have an output capacitor.
>>1495270
And linear regulators also have a minimum rated load current, so there should not truly be no load.

>>1495702
>minimum load current
floating 317 has, but not grounded 78XX.

How are big LED matrixes like https://www.adafruit.com/product/2278 put together?

The interface for this thing is:
> shift some bits into a register
> set address lines to the row you want
> selected row lights up, showing colors based on what's in the register

I assume on the low side they've got some kind of LED driver like https://cdn-shop.adafruit.com/datasheets/tlc5947.pdf (but less fancy, since there's no PWM), but what's on the high side? I guess there's some kind of demultiplexer involved, but how did they rig it up to so it's capable of sourcing >1A?

>>1495721
>tlc5947
they say it's just a 16-bit 'HC595 or something like that, plus a 16:1 demux
>but how did they rig it up to so it's capable of sourcing >1A?
MOSFETs on the demux outputs of course

>>1495721
if you take adafruits fucking cock out of your mouth for two seconds and pick up a book so you can actually learn something instead of buying price hiked shit from china and following a retard guide for gluing the lego pieces together and copy their arduino code libraries already written for you.
but if that's too much work then just forget it.

the normal learning curve follows quite nicely from voltage dividers into driving transistors. did you miss some steps? get yourself a nice book and work through it, if you jump around online courses you are going to have holes. transistors are pretty important.

>>1495726
>MOSFETs on the demux outputs of course
Thanks, anon. I didn't see anything on the back of the board that looked like a transistor to me, so I just assumed they had something fancier going on. I guess some of these chips must be MOSFET arrays?

Sorry for being a brainlet but I can't find anything explaining this.
How do these things work? Are the red/green pieces LEDs that get plugged into the battery?

>>1495799
looks like it

I'm looking for temperature controlled soldering stations on aliexpress and there's a bunch for less than $40. Anyone got recommendations for a decent one?

>>1495870
Go to /csg, and then get redirected here. T12 stations look pretty good.

>>1495870
depends, what are you going to be soldering? a decent assortment of tips is a necessity if you plan to work with varying levels of technology or lead-free solders. the 1.6mm screwdriver tip is good for through-hole board work and minor wire joinery but won't do you any favors with SMD
a lot of /ohm/budsmen don't think much of the 900 series but I don't mind them. tips are dirt cheap too. see for example
https://www.aliexpress.com/store/product/Celsius-Fahrenheit-Temperature-Adjustable-Soldering-Benis-YIHUA-908D-Soldering-Benis-Station-For-Soldering-Benis-In-Bagina-XDD/401349_32789194324.html

Does this type of power connector (marked A B) have a name?

It's two ~1 inch prongs which interface with slots in the PSU which goes on top

>>1495894
Forgot pic

>>1495895
This is what it looks like by itself, but on the site I found this it's just branded as a PS4 power supply connector, is it a custom part or can I buy them somewhere for use in projects?

Anyone know what this was for. Its on my motherboard. it is labled PETS.

>>1495906
Probably test pads
"Performance Evaluation and Testing System", or something similar.
Ask /g/

I'm making a digital circuit simulator for fun, basically a Logisim clone.
Currently working on the interface. There's a lot of little things to decide on (whether to allow diagonal wires or not, stuff like that).
So if you have any particular gripes with UI/UX of Logisim, I'd like to hear them.

>>1495918
That it requires X version of java, and that there aren't handy shortcuts for each tool. Ideally you'd just press "1", "2", "3", etc. for wire, gate (select specific gate with scroll wheel), drag, delete, a bit like on LTspice. Making sure it has a nice interface for making truth tables and using them to generate a Karnaugh map and simplified expression is also a feature I quite like. I know it's a bit abstract, but some way to generate circuits from timing expressions of some sort with both synchronous or asynchronous logic might be an idea.

>>1495928
>That it requires X version of java
Done, it's in C++.

>handy shortcuts
Sure, easy.

As for the rest, it's out of the scope for now. My current goal is to make working with CPU-sized circuits easier than in Logisim: better editor and a lot faster simulation. After that I'll think about the rest.

>>1495702
The minimum load requirement, as in the case of the common LM317, is still only ~3 mA, and that is still pretty slow

>>1495517
>this depends on the adapter. measure the resistance from both V+ and V- to L and N of the adapter. if there's no continuity on at least one of the adapters then they can share a ground.
So what happens if there is continuity on both and therefore i cannot connect their grounds? Does that mean i am shit out of luck and cannot use the GPIO pins to receive signals from the circuit?

>>1495517
I don't know where he will find a ~1V source capable of 10A. If he can, its the more practical solution for sure though

But if he has the will, he'd probably learn more fashioning his own set up with op amps and such

I'm another anon who is making a milliohm meter. stimulating exercise for sure. And looks like costly. But i'm enjoying it so far. It's all just in my mind and on paper thus far. going through datasheets and learning shit and such

>>1495945
it's not very common for wall adapter outputs to be connected back to mains through anything other than a small capacitor. are you saying both of your adapters aren't isolated? if so then you are SOL but you can easily find another adapter.

>>1495953
he's measuring across an 0.1 ohm shunt resistor, so the voltage across that resistor will be 1 volt at 10 amps. that's the full scale voltage i'm referring to. it'll only be higher for fault conditions (>10A) that the device isn't intended to measure, only withstand.

>>1495959
>he's measuring across an 0.1 ohm shunt resistor, so the voltage across that resistor will be 1 volt at 10 amps. that's the full scale voltage i'm referring to. it'll only be higher for fault conditions (>10A) that the device isn't intended to measure, only withstand.
That's contingent on the 1V being able to source 10A to begin with

>>1495963
he's making an ammeter, not an ohmmeter. it's a 5-30V source capable of sourcing 10A into an arbitrary load (so a car battery i guess). the shunt is just put in series to measure that current.

>>1495964
I see what you mean, that should work great then

>>1495036

Robomower is kill. Rest in pieces shitlord. 8 wheel drive igloo cooler incoming by summer.

>>1495968

>>1495969
buy a drone and mount a weed eater on it

>>1495036
Can I connect fuses in series in order to break higher voltages.
Or it is better to order proper ones

Hey lads, not much experience here.
I have an old washing machine electric engine and some circuit boards salvaged from the same machine, and I want to spin the motor up to around 1300 rpm for a rotational coffee machine project (1300 gives you around 15 bar with 100ml water and a 20cm radius for the rotation).
Since the machine could easily do 1200 with a full drum that weighed way more than that Im pretty confident that the engine can handle it, but I have no idea how to wire the circuit boards in a way that they control the amount of voltage going to the engine.
I had two semesters of analogue electronics, so Im fairly sure I wont electrocute myself out of sheer stupidity while tinkering, but this stuff is slightly beyond what I learned as a physics undergrad...
Ideally I want a variable resistance somewhere, so I can calibrate the machine to make the best possible coffee.
The motor is an AC one. Will check how many rounds of wire in the coils when I get to the basement.

>>1495959
So i measured the power source (it's an old computer source), between the 230V DC contacts an a DC 5V output and two of the metal contacts for the 230v cable had shown no continuity, but the third one beeped. it was the one that is not in line with the two remaining ones, i think this is probably the green earth wire. hopefully that means i am good to go

is electronics a good career/job/trade for a loner?

>>1495887
I'm thinking largely through hole for ICs with a lot of pads like MCUs and SMD for passives and maybe transistors.

They were clearing out some parts from old customers in my company today and I claimed these.
Niggers pull 1.4A at 52V.
Is there any cheap way to run of these from a battery driven portable supply?

>>1496140
Current-regulated boost converter, perhaps?

So I just had a chinese electrolytic capacitor blow and it vented a pretty sizeable amount of gas/smoke before I was able to shut the power off. I probably got about a lung full of the smoke accidentally, i feel fine, but should I be concerned about having inhaled this?

>>1496237
As long as you're not breathing that shit in day in and day out
But I wouldn't go out of my way to inhale it either kek

>>1496240
Alright, as long as I didn't scorch the insides of my lungs im okay. Thanks.

>>1496237

>>1496237
Nah shit's fine, could happen 100x and it wouldnt do shit. Soldering fumes on the other hand, are way worse, becasue of how they stagnate, repetitive exposition and w/e.

>>1496237
It's highly carcinogenic.
I am so sorry son, you have a year left at most.

What's a good multimeter to go for when it comes to electronics?
I have this really old one, but it has an analogue display, leads thinner than my twink arms and I don't think it has ever been calibrated.

>>1496302
How much are you trying to spend? Fluke is what almost every pro uses, but you can get auto ranging ones for $25 that work well.

>>1496305
200 is pushing it. 300 if I get a blowjob with my purchase. So thereabouts.

>>1495702
switching regulators do, but I have yet to see a linear regulator that does

>>1495943
quiescent current is the consumption of the regulator itself, not a minimum load

>>1495987
nope. the first fuse to blow will stop most of the fault current, and how will the rest reliably get enough current to blow?

>>1496068
>through-hole DIPs
the 1.6mm screwdriver will handle those
>SMD
how small? the tip itself should generally be about the size of the joint you're trying to make. if you get a 900 series iron, you can get tip assortments on ali for $5-$10. make sure you get one with 1.2-D (screwdriver) and I (point) tips. if it comes with some conical tips, more's the better for soldering e.g. DPAK transistor slugs to the board or wire tinning
general guide to tips: https://www.hakko.com/english/tip_selection/series_900m_standard.html

>>1496237
could be GBL. don't drink alcohol for the next day or so and you'll be fine. do drink alcohol and you might wake up in a bath tub full of ice and a kidney missing

>>1496140
>t.not an executioner
obviously you don't have that nigger hooked up right
>72W
first, divide by 0.8 to account for losses in the step-up converter, then decide on a power source that can deliver 90W without crying for the desired period of time and meets your definition of portable
>52V
if your power supply provides more than 60V or so you won't need to worry about the 0.8 discount. in any case the power supply is likely to be the biggest expense
once you have the supply, if greater than 60V or so you can use the HV9910 and a few external components to build a buck-mode constant current driver. if less, you may have to cobble together a current-regulated boost converter that can handle 90W, which is not necessarily trivial
tl;dr: no

>>1496302
>analogue display
jesus
http://www.eevblog.com/forum/testgear/multimeter-spreadsheet/

>>1496307
Sheeet, you should have money for a Fluke. I think the 117 is kinda the base model, doesn’t have all the bells and whistles but should do everything you would expect a multimeter to do and you won’t ever doubt it because it’s yellow.

And then the Fluke 87 is the real pro go-to but that’s $400. They have tons of offerings too so it depends what you want.

Otherwise for the same price as a 117 (a little over $100) you could get a nice Klein or Extech with more features and you wouldn’t be disappointed. Lots of decent Amprobe meters in the $100-$200 range.

>tfw got a bunch of my dad’s tools when he died a couple years ago but no idea what happened to all the Flukes he used to have

>>1496302
Any Fluke or Agilent bench multimeter. Forget the handheld ones. That shit is for electricians.

>>1496313
>nope. the first fuse to blow will stop most of the fault current, and how will the rest reliably get enough current to blow?
So, I will end up arc-welding everything around?
Shit, I have to buy some 15A 50VDC fuses then...

Any recommendations on where to get my PCB fabbed for cheap? Preferably without costing an arm and a leg on shipping.

>>1496389
pcbshopper dot com

>>1496389
I got 10 for under $20 including shipping from JLCPCB.

>>1496366
i can second this. i'd rather have a decent bench meter and a trash handheld than a nice handheld.

>>1495943
correct me if I'm wrong:
At 10 volts you'd need a 3.3k resistor to ensure a minimum load of 3mA. If you have a 1uf filter capacitor that's only a time constant of 3.3 milliseconds. so it's going to discharge pretty quickly.
and that's assuming the regulator won't sink any current. will it? I never tried that with a 317.

>>1496313
my bad, the 317 does have a minimum load requirement (*typically* 3mA, max 10mA). wew

>>1496417
sounds right. but the '317 itself will start feeding current to its output well within 3.3ms
the output cap is mainly there to stabilize the regulator and provide support during load transients. stability during input dropouts is the business of the input cap. note also, some 3-terminal regulators are prone to self-destruct when there is a large capacitance on the output and the regulator input is shorted to ground. so put your big cap on the input side and/or place a reverse diode across the in and out pins of the regulator
>sink any current
the 317 itself sinks about 50-100µA to the adjust pin

>>1495987
If you're really playing with exciting power you are going to want a sand impregnated fuse so it doesn't destroy everything in the process of blowing.

>>1496313
>quiescent current is the consumption of the regulator itself, not a minimum load
That's correct, but the LM317 *does* have a minimum load

>>1496422
>>1496417
>LM317
Sorry I should have specified. Yes 3.3 is fast by our standards, but I suppose I was just hoping for that nice near-step response (double digit microseconds time at most)
Although to be fair, I've only heard step response (in the context of power supplies) refer to rise time only. You guys are probably right and 3 ms is plenty fast for no external load

>>1496434
>my reply
Sorry only just saw your >>1496422 response

>>1496434
The minimum load is normally covered by the divider R1,R2. The total idle current of the circuit is still smaller than the 5mA typically drawn by a 78xx regulator.

>>1496528
Yes that's correct

Can anybody help double check my shit
The comparator detects negative output voltage. It turns on the transistors on the right to shunt some current and bring the voltage between the 10k resistor and the diode (where the voltage is important) to somewhere above 0V

The output RC anti-aliasing filter should be slow enough (roll off point in the 10kHz - 100kHz range maybe tune it closer to 10kHz) for the comparator to react before the output voltage changes to negative

The protection is to prevent a negative output voltage from being present on the input of the ADC

Not sure if I should protect the op amp inputs too. The protection would be for if I were ever to measure resistance of inductors.

>>1496424
Not to exciting, but automotive fuses won't do it, since they are only 32v, and I have 42

>>1496563
Also
the O is just just an LED to show a fault condition when the transistors are turned on. It'd flash on and off rapidly as the comparator switched on and off rapidly in response to potentially medium frequency negative voltage crossover so not sure how obvious it'd be. Could always just fashion a circuit to fill a capacitor to make the light last like 1 sec at least or something

Hello. I`m trying to make a psu, with only CV it werks, but the CC part is manbling my jambos. What I`m doing wrong? Why the differential amplifier isn`t working? Cropped part is CC U2B is a comparator that should check if the max current has been reached.

yeesss my deadly lazor is finally here!

>>1496661
I put Q2 as a switch to ground the main transistor base when max current is reached.

>>1496663
3d printers use pwm to control their fans right?
i really hope the laser will be able to turn on and off fast enough when suing the pwm signal on the til port, so i can simulate beam strength since it can't be regulated linearly, it can only turn on and off

Are there any cheap microscopes good enough for looking at dice of old ICs?

I decapped one. But even with an extra magnifying glass between it, this is the best my shitty microscope can do.

>>1496661
is positive supply to the opamps a good amount above 10V?

>>1496719
12-0V. For some reason the diff amplifier is not working, which makes the transistor saturate even though max current wasn`t reached.

Working on my traffic light logic circuit... I managed to get it working somehow. The problem is that the binary counter that counts the time delays doesn't reset every time when it should.
I used an NE555 for generating a clock pulse and some CD4000 & 74HC logic ICs.

>>1496736
Main circuit:

NANDs: CD4011 & CD4049
NORs: CD4001 & diode-ORs
NOTs: 74HC14

"clk_div" in this circuit is an asynchronous binary counter made of 3*74HC74 D flip-flops.

>>1496661
How is Q1 being driven? It's only got the connection to ground or is floating that I can see

>>1496787
without the current control it works fine.

>>1496567
why not just a Schottky diode?

>>1496661
>>1496664
uh, is there a resistor between Q2 collector and U2A output?

>>1496739
try synchronizing your resets with a separate DFF

>>1496805
>uh, is there a resistor between Q2 collector and U2A output?
Nope. But the output of the diference amplifier is fucked up for some reason. It should give the voltage drop across a 1 ohm resistor, AKA current

>>1496661
Um. How's this PSU? Where's energy storage inductance?

>>1496818
He's making a linear power supply obviously not a switched mode so there's no energy storage inductance and even with switched mode power supplies you only find energy storage inductors in forward converters.

>>1496818
He's making a linear power supply obviously not a switched mode so there's no energy storage inductance and even with switched mode power supplies you only find energy storage inductors in forward converters.

>>1496661
>>1496790
Your differential amplifier works just fine
If there's more current flowing than you have it set for, U2B will shut off Q2 completely. Like if your diff amp sees 1A and you have it set for 0.1A
If there's less current flowing than you have it set for U2B will switch on Q2 completely trying to sink current from U2A, since U2A's output is right on the collector of Q2

>>1496813
you're asking the op amp for the difference between the bottom of the current sense resistor and (20/2)V. which is a bit over 9V (accounting for probable common mode input range violation). you're not actually dividing the input by 2. make R5 20k and add a 20k between ground and the - input of U2C

>>1496842
How would I make it saturate ate the desired current instead of just fucking everything up and turning itself off?

>>1496805
>why not just a schottky diode?
Most professional articles say to use a schottky diode
I considered it but if i have a schottky diode in reverse right on the output of the op amp, the output (at the capacitor) will end up having a negative voltage across it anyway esp if the negative voltage duration lasts anything more than a like a few tens of microseconds it seems.
If I had a function generator and DSO i could probably test this in real world
falstad simulator seems to suggest that a schottky wouldn't be enough

>>1496844
The difference amplifier is working as intended dude. Don't listen to this guy

Shift registers are literally magic, like holy shit. Does a more useful discreet logic chip even exist?

>>1496844
Not him but his difference amplifier is working just fine. 10k resistors aren't ideal (100k gives a more accurate result) but its fine as you can see

>>1496863
>>1496874
>working just fine
with ideal op amps, which the LM348 decidedly is not. it would work with an OPAx196 which has RRIO, or if he generates a few volts of V- headroom below the output ground
otherwise he'll need to translate to a higher reference voltage here and elsewhere (and yeah I got the resistors wrong originally, corrected here)

>>1496890

>>1496874 here
>RRIO
Does that stand for rail to rail input output?
>few volts V- headroom
This is so you can get to zero volts right? Because most op amps only get really close to zero and not precisely zero (even ones described as rail to rail)
Finally, what's the reasoning for the 10k resistor connecting the inverting input to ground?

>>1496908
Do the math my man. It`s the ''general'' op-amp configuration with 4 impedances.

>>1496909
>the ''general'' op-amp configuration with 4 impedances
is not a topology i'm familiar with
Not to mention the output is not reflective of the current passing through the 1 ohm resistor as is the case with the left, correct differential amplifier topology

>>1496908
>rail to rail input output
that is correct
>Because most op amps only get really close to zero and not precisely zero (even ones described as rail to rail)
also correct, and usual op amps are especially poor near V+, especially on their inputs
>connecting the inverting input to ground?
offsetting the zero level of the amp to V+/2 to keep it out of the op amp's no-no zones

>>1496911
the output *with respect to V+/2* is indeed proportional to the current passing through the resistor

>>1496921
>RRIO
Okay thanks for confirming
>poorly near V+
I've begun to understand this and the reason ive started to use negative rails if I want to have a true zero volts output. Annoying though unless you use a charge pump voltage inverting IC though

>offsetting the zero level of the amp to V+/2 to keep it out of the op amp's no-no zones
>the output *with respect to V+/2* is indeed proportional to the current passing through the resistor
I see what you mean, I think. Now he'll need to subtract off the 10V though, no?

>>1496436
If you really wanted that, maybe you would put a shunt regulator across the capacitor and have it track the series regulator. When the cap voltage is higher than the adjusted voltage, the shunt regulator would conduct and drain the capacitor down very quickly.
idk if that would actually work.

>>1496872
>Does a more useful discreet logic chip even exist?
demultiplexer?

>>1496930
yes, but not if he offsets whatever he's comparing by the same amount, which is pretty easy to arrange in case of a pot

>>1496872
some of the 74xx1G logic is highly useful, on a functionality per unit size measure
otherwise I'll confirm other anon's answer, mux/demuxes

>>1496947
>yes, but not if he offsets whatever he's comparing by the same amount, which is pretty easy to arrange in case of a pot
Can you show me what you meant? Because it honestly sounds very unconventional. He may as well use or fashion up an instrumentation amplifier at that point

>>1495333
Power supply is still on my wishlist, so I’m attempting to see if the solar panel with it’s real low current will charge it and hold it. Has gone up a little bit but I haven’t had it charging too much the past two days because I don’t want to leave these batteries unattended and start a fire or anything.

Also 3 out of 4 of these batteries are kinda bloated. The one that wasn’t bloated had the lowest voltage at .006V while two of them were .150V-.200V and the last was higher but I rubbed off the marker, maybe like .500V.

>>1496951
Just use a huge resistor in series with any old voltage, 5V USB is enough. Just keep an eye on it.

>>1496951
don't use bloated lipo batteries you walnut

>>1496953
Ehh half the fun is learning that lesson the hard way.

But I’m curious as to why the bloated ones actually have a little charge, and the flat one is almost completely at zero.

Alright so my buddy got a broken air compressor for free and he wanted me to look at it. pic related. I have more pictures too.

When I plugged it in it didn't do anything, so I tore it apart and started checking components with a multi meter. The power cord looked fine and it checked good with the multi meter. The on/off switch worked like it should. The pressure switch was closed like it should be at ambient pressure. There was a decent amount of rust on the steel components, so I cleaned up some of the connectors and re-seated them and that managed to get it going, but it kicked off at ~60 psi. (It's rated up to 110psi) I checked the pressure switch thinking it was weak and opening too early, but it still had continuity. After that i removed the power switch and pressure switch to rule them out and used a jumper lead in their place so it was a straight shot from the plug to the motor, but it still wouldn't kick on again. After tearing into it some more I found a thermal resistor on the motor and I think that might be the problem since it would've got hot building up pressure, then prevented current flow until it cooled off again.

What do you guys think? Shitty thermal resistor that comes on too early or corroded wiring with high resistance causing the motor to draw too much current? I'm thinking about ditching the crusty spade connectors and soldering everything back together and using heavier gauge wire.

>>1496977

He's an amazing wiring diagram of the compressor. I'm going to apologize to your eyeballs in advance.

>>1496965
Different modes of failure. My guess is the ones at zero have failed via an internal short circuit, any chemical voltage is discharged quick enough through an internal resistance that's lower than it should be. The bloated ones will just have been overdischarged, producing some chemical byproducts (gasses) since they've followed a non-reversible reaction path. Of course it's impossible to recharge a battery and expect that gas to go back to what it used to be, meaning the battery will have a permanently lower capacity.

Has anyone here programmed an STM32 microcontroller using CubeMX and TrueStudio? How the fuck do I read a voltage value across a resistor through an analog pin? Like I have all the physical components setup correctly, but I cant get the simple code correct and nothing online has helped me. First time programming one or using a microcontroller that isnt an Arduino

>>1496950
the current-set voltage VDC3 in
>>1496661
is relative to whatever the zero-voltage of U2C is, which is ground. instead, in my design, the set voltage would be relative to virtual ground. if he is using a pot to set that voltage, all is well and easy. if he's setting it digitally, my idea isn't so great after all

>>1497006
I would use the (poorly documented) ADC read input function from their handy library
show/link us your code fren

>>1496980
You sure it's not a thermal fuse? Those are one shot.

>>1496825
>a linear power supply obviously
>comparator output controls power transistor
o rly?
Also,
>>1496661
>Hello. I`m trying to make a psu

>>1496825
>you only find energy storage inductors in forward converters
Looks like you have no idea what you're talking about.

>>1496663
okay so this is a stupid question, but i will use a computer power source to power this laser, and then a pin in my 3D printer's mother board (the 3d printer has it's own power source) to send a signal to the laser to turn it on and of (on the TIL pin on the laser's board)
Do i need to like connect grounds or something?

>>1497048
>comparator
It looks like you're the idiot here
LM348 is not a comparator, nor are any of the op amps there in comparator configuration

I have a relay blown in a bench power supply.
I have a relay with the same pin configuration and current rating but with a 12v coil rather than the 24v one that's blown.
The coil resistance is 360 ohms so can I just stick a 360 ohm resistor in series with the coil and use it?

>>1496661
>>1497053
U2B is a comparator that checks the current through the one ohm resistor and the set point.
But that isn`t a switched mode power supply, I don`t even know how to make a linear one yet.
t. guy that made the thing.

Guys, how would I improve the design and make the thing saturate at the current setpoint instead of shitting itselft?

>>1497061
Yes, half the voltage * twice the current = same power consumption of the 12V relay. The other half goes to the resistor. Twice the power consumption of the old 24V relay in total. Be aware of that.

>>1496981
Ahh ok that makes sense.

So after that happens.(assuming it’s not 100% kill) can you still charge the thing up to 4V but with a lower Ah capacity? Or is it only gonna hold a charge at <3V now?

>>1496736
Man, what a nightmare that wire jumble, yo.

I have quite a sizeable amount of high power N-Mosfets. I have absolutely no use whatsoever for them. Should I order a bunch of PCB's from JLCPCB and sell them with a bunch other components as a DIY kit?

>>1497085
>Man, what a nightmare that wire jumble, yo.

that's why god invented arduinos

>I have quite a sizeable amount of high power N-Mosfets. I have absolutely no use whatsoever for them. Should I order a bunch of PCB's from JLCPCB and sell them with a bunch other components as a DIY kit?

no. just no.

>>1497048
Its a fucking pass transistor you moron. Not unlike one you'd find in a LINEAR REGULATOR. Its being operated in its LINEAR REGION. Its not being switched at high frequency.

>Looks like you have no idea what you're talking about.
Looks like I'm the only one that has any idea what I'm talking about. Forward converters are unique in using an output energy storage inductor. I did forget to include the push-pull topology as well though.

>>1497075
Power consumption should be fine, I've bigger problems now though.
Someone had hooked something way to powerful to the supply. The relays that feeds the transformer's secondaries into the bridge rectifier had blown. Turns out that the protection diodes also blew and the bridge rectifier itself not to mention a few traces burning out. I've managed to get all them replaced from stuff to hand and the power supply comes on but outputs 100v which can't be adjusted.
I've now found a giant 230v 15a npn transistor with a short between all pins. I'm fairly certain I've nothing close to that to hand

transformer question:

pic related comes with an 80 ohm resistor that goes across the two pins from the current transformer. the top trace is with about 6 amps and the 80 ohm resistor, and is what I would expect.

The middle trace is with about 600 ohms, and the lower one is about 1000 ohms. Note that the output voltage magnitude is a lot higher as you would expect. The part I don't understand is the distortion. If it's a simple transformer I would expect the sine wave to be more sinusoidal (assuming the input is a sine) as the load decreases (increasing resistance). I tried melting the black potting but it seems to be epoxy that doesn't melt easily.

I can live with it using the 80 ohm resistor, but I'd like to understand what is happening. I don't see why they'd have any other components inside the device, and even so, what would cause the voltage to collapse to zero in the lower plot?

>>1497117

presumably the magnetic field is saturating and causing chaos. the 80-ohm is presumably equal to the coil's internal impedance thus providing the maximum energy transfer.

>>1496980

a thermal resistor (thermistor) wouldn't act as you describe. i'm sure it's a thermostat, instead. and yes, they definitely can go bad, especially if they're open to the elements. time to dig for documentation. online stores that sell parts for appliances might have the info you need.

>>1497125
>presumably the magnetic field is saturating and causing chaos

hmm. so the input current of 6 amps is enough to saturate the core, but when there is a sufficient load on the secondary, that current opposes the current in the primary, thus reducing the flux below saturation?

If that makes sense, I still don't see why the lower trace goes up in a more or less normal manner and then collapses to zero. Any theories about that?

>>1496980

i was right. it's a thermostat.

>>1497091
Why not? People need to control motors all the time in university classes

>>1497151
>Why not? People need to control motors all the time in university classes

that makes sense. His plan sounded more like a grab-bag when I read it. A proper motor driver at a good price is fine.

i have a couple questions regarding the full 3 bit adder
i'm trying to make it out of transistors, is there a better/more compact way than this?
also can just connect the wires of an OR gate instead of using transistors?

>>1497157
>also can just connect the wires of an OR gate instead of using transistors?

not directly, you need diodes to create a wired-or. (otherwise, your ones and zeroes will collide to create ½'s.)

>>1497161
but why? for example in my pic, if one of the AND's return 1v and the other two return 0, i'm not gonna get 1/3v, am i?
as far as i know there shouldnt be a problem as long as the other 2 AND's dont take the voltage somewhere else(where)

as for the diodes, they would actually drop the voltage so how would that be better than just wires

>>1497125
>>1497148

Awesome. thank you.

top kek

>>1497108
pls halp fremd
>>1497064
>Guys, how would I improve the design and make the thing saturate at the current setpoint instead of shitting itselft?

>>1497165
>as for the diodes, they would actually drop the voltage so how would that be better than just wires

i was joking about creating ½'s. there's no such thing. when you short two or more gate outputs, you get an unknown/unpredictable state. the reason you use diodes is that it isolates the outputs. a high on any diode will produce a high after the gates (that's the OR function), but no current will flow backwards into an LOW output coz the diode blocks that, so you dont have unknown states.

https://en.wikipedia.org/wiki/Wired_logic_connection

>>1497165
>as for the diodes, they would actually drop the voltage so how would that be better than just wires

i was joking about creating ½'s. there's no such thing. when you short two or more gate outputs, you get an unknown/unpredictable state. the reason you use diodes is that it isolates the outputs. a high on any one diode will produce a high after the diode (that's the OR function), but no current will flow backwards into an LOW output coz the diode blocks that, so you dont have unknown states.

https://en.wikipedia.org/wiki/Wired_logic_connection

>>1497165
>as for the diodes, they would actually drop the voltage so how would that be better than just wires

i was joking about creating ½'s. there's no such thing. when you short two or more gate outputs, you get an unknown/unpredictable state. the reason you use diodes is that it isolates the outputs. a high on any one diode will produce a high after the diodes (that's the OR function), but no current will flow backwards into an LOW output coz the diode blocks that, so you dont have unknown states.

https://en.wikipedia.org/wiki/Wired_logic_connection

at 5V, a TTL high is anything above 2.2V, so losing 0.7V is no big deal. it's different but similar for CMOS.

A mosfet with a variable gate voltage becomes a variable current source?

>>1497211

and a nice heat source too!

>>1497064
did you simulate it with better op amps yet? to me, it looks like it will limit current to the set point by stealing *some* base drive from the pass transistor as long as current is over the set point. maybe you would need to reduce the gain of that "comparator" stage for stability, but maybe not

>>1497244
Managed to get it working.

>>1497245
Basically the large PNP power guy has it`s base current limited by Q1. Q1`s current is controlled by both op-amps, but Q3 gives priority to current limiting.

>>1497246
NPN power guy*

hello friendos
i have a white wheel which has a 0.5mm dot on it. i need to count how many time that wheel spins by counting the passing dot
what would be the easiest day to do this with arduino?

>>1497249
*black dot

>>1497157
>is there a better/more compact way than this?
Dunno about better, but you can produce the carry by summing the bits "analogically" and comparing that to a threshold value.
Transistor level design also allows fusing the two xors partially together so that you don't need two full XORs.
>can just connect the wires of an OR gate instead of using transistors?
Depends entirely on what kind of logic you're using. Nothing prevents you from designing AND gates which allow wired-OR.

>>1497249
>the easiest day to do this with arduino?

1. make the dot a lot larger. 1/2 mm is TINY. use something that reflects light or infrared and a sensor to detect the reflected

2. drill a hole and let the light shine through it

>>1497255
If i had my choice of medium i wouldn't use a retarded tiny damn dot for sure. But it is not up to me, i am reading an energy meter and that is how it is set up unfortunately
I was thinking some focused beam laser ir diode but i have no idea if something like that even exists

>>1497249

a reflective optical sensor. but 0.5 mm is kinda iffy. if you can make it 2mm, should be easy.

if you cant then you'd need a interrupt sensor that's similar but goes on either side of the wheel. (like they have in computer mice).

>>1497259
>>1497255
it has to be possible because their own overpriced piece of shit reader does it somehow as well

pic related it's the meter

>>1497259
>reflective optical sensor
I actually have several of these on hand
https://www.aliexpress.com/item/10pcs-TCRT5000L-TCRT5000-Reflective-Infrared-Optical-Sensor-Photoelectric-Switches/32818041584.html
But i am not sure how to focus the light to get feedback on something so small. maybe the dot has a diameter a little bigger, could be 1mm, i am not completely sure

>>1497249
tcrt5000

>>1497261

are you sure the "dot" isn't a tiny neodymium magnet?

>>1497264
>tcrt5000
Well that is the same one i linked above, i just don't know how to recognize such a small black dot on a white background with it

>>1497263

if the dot is big, then the sensor will give you a nice big logic signal that you can feed directly into an arduino, or whatever. but if it's small, then you'll just get a small blip, which will require amplification, or the use of a comparator to turn a that small blip into a valid logic signal. so it's doable, but more fiddly.

>>1497265
I think so. The wheel has actually numbers printed on it and next to the numbers are the dots. There is no ir diode there, so i know the data transfer is not optical. The number wheel is also sealed in plastic is i know there is no physical contact reading going on.
Do you think there could be a magnet hidden under the number wheel? Intriguing idea, i will definitely check tomorrow.

>>1497265
in the manual the call it the "pulse output"

>>1497268
>>1497269

If it turns out that it is a small magnet, I had a similar pulse meter situation and I had to get a really sensitive hall switch, A1125LUA, to sense it. what they had in the original was a tiny reed switch but it also was a lot closer to the magnet than I was able to place my hall.

if you use it, it directly interfaces with arduino GPIO.

>>1497277
I have reeds switches in my supply box as well. If it turns to be a magnet that would be pure awesome because nothing is easier than detecting a passage of a magned, there are so many simple solutions to that.
But my fear is that as per usual, the meter will contain some beyond retarded proprietary solution thing that will be impossible to DIY a reliable sensor for

>>1497266
it`s an analog device, when the albedo changes you can amplify the diference (AC) and detect it.

>>1497299
btw, I used the same sensor so sample my blood volume on the finger. The signal was amplified 280.000 times IIRC.

>>1497352
moving average and some filtering.

>>1497248
>>1497246
>>1497245
G R A P H S

>>1497356
Current state of the project, I put ideal opamps there because I was trying to troubleshoot some stuff but it works with the lm348 as planned.

>>1496805
>try synchronizing your resets with a separate DFF
Do you mean adding a D flippu-floppu between the output inverter & clk_out pin? Should it be triggered with the same clock in as "clk_div"?

>>1497362
you should probably make sure it works with all LM348s, and calculate power dissipation of pretty much every component in there and make sure it's sane. I'm also not loving Q3 in that transmission gate type of arrangement, I'd rather see it rob Q1 of base drive, but not gonna argue with it if it works

>>1497391
no, on clk_div's reset, replacing the NOR, and without the clock phase inversion
it may take a cycle longer for reset to happen but I get the feeling that's not super critical in this design

>>1497082
Should be the same voltage (4.2V) since the voltage of a cell is dependant only on the cell's chemistry.

>>1497362
Questions:
Why are you using 1-ohm for a current sense resistor instead of 0.1 or even 0.01 ohm? You can get some gain from U1 no problem.

You look like you're using U3 to invert the output of U1? Shouldn't you have a resistor between output of U1 and inverting input of U3?

Also I'd use a bandgap reference for D1, much more accurate and stable, and they're easy to get and cheap these days.

what do you think my chances are of replacing an 0603 capacitor with a through-hole film cap with bent leads? for reference, I've done it with 0805 before easily
pic unrelated

>>1497571
doesn't sound impossible

>>1497571
for me, going from 0805 to 0603 is the difference between easy peasy and starts to get fiddly
how big is the new cap and its leads, what's your DEX stat, and what's your saving throw against soldering fumes?

>>1497049
anyone? please?

>>1497640
yes, you need to connect grounds

This might sound like a stupid question, but would it be a bad idea to use an analog comparator in the design of a digital magnitude comparator?

I'm working on designing some discrete component designs for fun, and the amount of gates needed for an 8 bit magnitude comparator is pretty huge, so I was thinking why not just use two DACs and an OpAmp configured as a comparator?
In that case, I only need to do digital logic to check for equivalence, which is piss easy.

In all seriousness, I'll probably just wind up buying a 74xx series comparator, but I'm curious how feasible using an OpAmp would be.

So arduino mini eats about 10ma when normally on.
I need to power it from a couple of AA batteries.
The problem is this nets me about 14 days of function on 2 batteries.
This is way less than i need. I need at least a year
So how would i do this?
What the Arduino does: there is a reed switch connected to it and arduino counts +1 every time it clicks on. when it counts say 50 pulses it uses an attached esp wifi module to send the number to a server.
The problem is that the reed switch triggers constantly, at a rate of about 1 pulse per second. Therefore, i am unable to put the arduino to sleep mode, since it would need to wake up every second to count the pulse.

So what do i do?

>>1497666
I can of course use a bigger battery, but it cannot be much bigger or too expensive since i need to make lots of these things

So the main advice i'm after is current draw reduction while maintaining the desired functionality

>>1497662
if you are wililng to have <= or >=, and your DACs are nice and linear,
>it would need to wake up every second to count the pulse
then fine
otherwise, you may as well use something like a shift register

>>1497666
>since it would need to wake up every second to count the pulse.
and?
for how long does it need to wake up?
show entire power distribution system

>>1497666
have you sanity-checked this project?
3Ah means your wi-fi, assuming 200mA current draw, can only be on for 15h for the whole year.
>since it would need to wake up every second to count the pulse
and?
for how long does it need to wake up? like milliseconds, then it can go right back to sleep again. it shouldn't need to be awake more than 0.1% of the time
>while maintaining the desired functionality
use pin change interrupts and see
http://home-automation-community.com/arduino-low-power-how-to-run-atmega328p-for-a-year-on-coin-cell-battery/

>>1497676
wifi is not a problem, it can send the data like once every 30 mins if needed and be off for the rest of the time
does it really take so little time to wake up? i thought it takes about 3 seconds to boot like when i checked it on my serial monitor

>>1497677
>i thought it takes about 3 seconds to boot like when i checked it on my serial monitor

I believe that has to do with arduino trying to figure out if it's being flashed or if it is in normal AVR mode. There may be "arduino" ways to fix that, but going to non-arduino mode will get you the fast wakeup time.

>>1497677
it shouldn't need to come out of a full reset when waking up. it should resume exactly where it left off in your program. at most it takes some time for the crystal oscillator to stabilize (a few thousand cycles)
if you use this sleep mode you should consider using an external time base that's reasonably power efficient and fairly accurate, like one of those 50 cent DS1307 modules. hook its interrupt pin to an AVR pin with pin change interrupt, set it to alarm at a specific time, then just sleep until something happens. let your interrupt routines handle the counting and time-to-call-the-ESP32-flag settings

I need a surface mount sized solution to very short range proximity sensing. This is for a hacked together rotary encoder type application.
The best I could find was tyhe EE-SY1200 phototransistor sensor, though its not a perfect solution for this.
Anyone got any suggestions?
Note - I would have like to use a hall effect sensor as that would be perfect, but I can't because it will be adjacent to other fluctuating electromagnetic fields so would read totally inaccurately.
Thanks

I got one of these T12 ali stations recently, but it doesnt seems to go over 370C (when checking with an external iron temp tester). Anyone know what can be wrong? I've set the cold junction/ambient temp, and tried the calibration (which obv. doesn't work because the last step is 400C+). The PSU gives stable 24v, so it seems like software, but who knows.

>>1497563
>1 ohm res
I just did that to simplify the math, this is not the "real" project. I`ll probably go with 0.1.
I`ll use a proper voltage reference for Vset and Iset until I can be bothered to go ucontrol.
>>1497563
Because I`m stupid. After long thinks I got to this.
>>1497410
power dissipation is ok, control loop gets little power (it`ll run on lower tension than on the pics), I just need to wait my usb chinkscope to get here so I can get to build it.

>>1497740
Top left
>CC mode step response
Top right
>CV mode step response
Lower left
>CV mode with 4A limit

>>1497740
The 15mA ripple in the CC mode is not a problem because it`s drawing 4A anyways and that`s near the limit for the BJT, and if it`s a problem I`m too stupid to fix it because extra caps will make the transient time take forever and I don`t think that is good.

>>1497722
After some googling around it seems some people have fixed similar problems by replacing the PSU DC wires with bigger gauge ones. I guess whatever cheap ones they use can't supply 3A, thus limiting the output? I kinda wonder why this is not picked up and software and the screen keep showing 400C even if its waaaay off though...

fuck... i was trying to check one of those really tiny resistors and i had it on my finger and was looking at it under a magnifying glass and then i tried to smell it to see if it smelled burnt and it fucking flew straight into my nose and it's gone..
i mean it's really tiny so it shouldn't cause any real problems right?

>>1497771
rip anon

>>1497771
SMD claims yet another life... this is why I tell people to stick to through hole...

>>1497771
>i mean it's really tiny so it shouldn't cause any real problems right?
Was it previously soldered to a PCB? With lead solder? Just something to consider when your organs starts shutting down

>>1497410
Thanks for your suggestion. I tried it and it seems to work the same way as with the NOR. So I think there's a logic design mistake in this circuit.

The input pins on the left change their states at exactly the same time when the D-FF (or NOR) should reset the binary counter. That seems to be the reason why the counter doesn't always get reset back to 0 after each counting period.

I tried to remedy this just by making a D flip-flop input data register. The register holds the input data bits long enough for the reset circuit to do its job. Although it works in simulation, I'm not sure if I could add any more ICs to my proto board...

i just managed to repair my first broken icb
feels pretty great
it was some rtc module i got from chinks and the battery wasn't working, so i troubleshooted that shit with my multimeter and turns out, that one tiny 1mm zener diode wasn't soldered in on one side. so i ejaculated a bit of solder onto it and now it's as good as new
yay me

is gold from old circuit boards just a meme? I was fucking around with junk boards and noticed that my soldering iron tip had yellowish metal on it. Is it economically feasible to try it out if I can get a steady supply of older circuits?

>>1497743
>10µF on the TIP120 base
found your terrible transient response

>>1497771
F

>>1497972
yep. it costs almost as much to extract the gold as it's worth

>>1497951
I was trying to avoid the propagation delays from the wired-OR setup by adding that pipeline stage. what if you take your clk_out from the DFF at top?

>>1497972
only if you are chinese

Found a couple of old matched JFETs ICs and found out I can turn triangle waves in breddy gud sine waves using a diferential pair. It seems my bench supply project will also generate some signals. Nice. Wish me and the other psu anon luck

>>1497722
Why would you ever need a temperature over 320C

>>1497992
Not him but I crank it up when soldering large tabs, just to get more punch by storing as much energy in the tip as possible.

>>1497662
> would it be a bad idea to use an analog comparator in the design of a digital magnitude comparator?
Yes.

> the amount of gates needed for an 8 bit magnitude comparator is pretty huge
If you're minimising the number of 74xx/40xx chips, it's an 8-bit adder, 8 inverters, and an 8-input NOR gate. Given inputs A and B, invert B, add A+(~B)+1 (8-bit adder, with carry-in set to 1), if all 8 output bits are zero => A=B, otherwise if carry-out is 1 => A>B, if it's 0 => A<B. That's assuming there isn't a dedicated 4-bit magnitude comparator which can be cascaded (there may well

Either way, that's going to be cheaper than 2 DACs and an op-amp.

>>1498027
The optimal solution for this is to get a station with higher maximum output power, perhaps up to 200W. Getting a tip-type with the thermocouple really close to the element and just overvolting it is what I'm thinking of doing.

Where do you radio nerds get your coax cable? I'm just starting out and want to listen to airband and do ads-b tracking.

>>1498039
R/2R ladders are pretty cheap, if one's time hand-matching and assembling resistors is of little value. just sayin'
74x85 is a cascadable 4-bit mag comparator, 11ns per 4 bit slice

>>1498056
the radio nerds are probably on the /ham/ radio general, over here
>>1484287
that said, last time I stocked up on shielded cable it was RG316 from ali the chink

>>1498039
So, just perform subtraction in 2's complement.
Makes sense.

>>1498117
I thought you could get 8 bit R-2R DAC ICs for like 50¢.

>>1498039
If he only needs equivalence (rather than <, >), then XOR gates is far more cheaper option.

>>1498056
>want to listen to airband
But why?.. It's utterly boring.

>>1497722
How do you like that station? I'm in the market for a cheap temp-controlled station.

In general, how much power should I expect to need for normal-sized SMT soldering?

Unrelated to that, but if I start doing the solder paste, stencil, toaster oven technique, does that restrict me to only having components on one side of the board?

so floating power supply means the dc output - is not connected to AC input N ?

>>1498209
>How do you like that station? I'm in the market for a cheap temp-controlled station.
For the price I really doubt you can beat it. Apart from the temp thing I mentioned (not even reaching 400C, when it should work up to 480 or something), there is nothing wrong with it. It really heats up instantly and solders everything I throw at it. Build and PCB-quality on the KSGER is damn solid too, probably one of the better thing I've ever gotten from Ali.


I'm also getting some quality 0.5mm2 silicon wires to replace the dinky PSU to controlboard connection today. So we'll see if that fixes the temps.

Can you recommend me a Binary/BCD -> 7-segment decoder/driver IC? My poor DIP-8 pic can`t handle all those pins.

>>1498235
They're all BCD to 7-seg, I just search on the wikipedia pages for "list of all 7400 series ICs" or "list of all 4000 series ICs" and usually get a few results. Then I narrow it down to ones with the right combinations of I/O latches, active high/low I/Os, whatever the fuck tristate means, price on AliExpress, and whatever else I want.

>>1498237
thanks!

>>1495036

I'm building a chinkshit stabilized power supply (the 30V 3A one, you probably know it.)
It asks for a 24VAC input with maximum current of 3A.
I found a killer deal on a 100VA transformer that I assume can output about 4A at the maximum, maybe less but most likely more than 3A.

My question is: can I use it? I'm aware that inrush current isn't being transferred down the secondary (what with it being spent to magnetize the core in the first place), but will the circuit - barring failures - possibly allow a higher current to flow than it can handle? My guess is no (a SMPS doesn't, after all, unless it shits the bed), but dropping a big ass transformer in the equation is making me question my judgment.
Anyway, the Chinks are saying that for the full 30V-3A output, a >90W transformer would be needed (though I bet that would fry the transistor in no time even with good cooling), and since 90VA essentially translates to ~90VA, I suppose I'm not very far off the mark?

Thanks for any answers.
The manual is here:
https://img.banggood.com/file/products/201505080459530-30Vinstall.pdf

>>1498277
W = VA * power factor, for that design I'd assume power factor around 0.6, so for continuous maximum load use you'd want more like 170VA transformer rating.

Higher current rated transformer is fine. A bit higher voltage might be nice too, with a 24V nominal one there's not much room for transformer sag or a lower than standard mains voltage before you start dropping below 30V out max.

That transistor heatsink looks waaay too small to use at 3A when the output voltage is set low.

WHY THIS FUCKING DIAC DOES NOT TRIGGER REEEEEEEEEEEEE

>>1498314
Thanks.
>That transistor heatsink looks waaay too small to use at 3A when the output voltage is set low.
The heatsink is not included, I bought a 40x40x20mm one with a fan and I'll try to aluweld a bunch of smaller fins to the sides. Regardless, I don't plan on using it at 3A for prolonged periods of time, I just like to have some headway.

>>1498316
You know the green multisim parts are the ones that have no spice model, only footprints, right? Of course your DIAC won't trigger because as far as your circuit is concerned the TRIAC isn't there. I'm a bit surprised its not giving you a convergence error honestly.

>>1498318
100VA transformer is probably alright for your uses then.

I've been looking into getting a raspberry pi to connect to a monitor, external HDD, mouse/keyboard, and wifi/cat5 internet, for the purposes of watching torrented movies/shows, playing audio files, and watching youtube video.

I do some breadboard stuff and make simple circuits, but all this raspberry pi shit just looks like plug-n-play legos/adafruit modules and OS downloads or preloaded OS on SD cards. Am I missing something or is it really that simple? I'd rather not purchase one only to find out that I need to purchase another model or whatever.

Seems the "Raspberry Pi 3 - Model B+" looks like it has what I need if I get a heatsink, PoE Hat, and other small things. Though, I have a box of heatsinks already.

>>1498326
>You know the green multisim parts are the ones that have no spice model,
I had no idea. thanks m8

Just wanted to share this pic of 74HC595 shift register

>>1498342
>Seems the "Raspberry Pi 3 - Model B+" looks like it has what I need if I get a heatsink, PoE Hat, and other small things

any Pi will play 1080p video. you dont need a super duper model for that, tho a v.3 is needed if you wanna use wifi and bluetooth. and a v.2 is the minimum if you wanna have a nicely responsive user interface. you dont need any heatsinks if your box has ventilation holes. other than the Pi, you'll need
- a 5V 2.5A xformer with a micro USB plug
- a tiny pushbutton switch to turn the Pi on. it connects to 2 pins on the header. to turn the Pi off, you use the menu, or set it to shut-off after, say, 40 mins of idling.
- a wireless mouse. or, better yet, a windows-media-player-edition compatible infrared remote and receiver. see pic.
- a 2+ gig micro SD card with the latest openelec or libreelec. use win32diskimager to put the image on the card.

(no idea what you're gonna do with a PoE hat. if you're using this near a TV, there should be a bunch of power sockets available.)

>>1498412
Thanks very much for the tips.

>>1497977
>what if you take your clk_out from the DFF at top?
I never came to think about this. It really made a difference. Now the circuit finally works like it should, without any extra D-FFs. Many thanks to Anon.

Is there any way to read an id from a rfid chip from say 2 meteres?

Chinks have stuff like this, but it has only 5cm range.

Basically i want to have something small i can put say into my wallet (and it doesn't need batteries) and then i have some reader thing that can dettect that this rfid tag or soomething similar is within its range somehwere around
And i need the range to be about 2 meters.
I don't care about exact position in space, just want to know it is in range

>>1498445
UHF RFID can go over a couple meters

>>1498450
>UHF RFID
goodness gracious, that is really expensive!
the regular readers cost like $3 and these uhf ones start at like $80 on ali.. that is crazy, i expected them to not cost more than twenty bucks.. what is a poor ghetto slav boy like myself to do now?

>>1498342
People will tell you pi can play hd video which is bullshit, it has hardware to render x264 directly to the video meaning you get an overlay, it literally draws on top of whatever the computer is doing, you can't move the window or see what's happening behind it. But you can get kodi or whatever that makes it bearable but it has to be built for pi to use the hardware decoder, should be in the repo if course.
Just treat it like a really shit computer tower that runs on 5v. I run a few wall mounted 19" monitors, 12v dell soundbar output via an ebay power supply made for quadcopters works great and kills the pi when you turn the screen off. No sound though...
They use the same bus for ethernet and usb so torrenting or streaming out should be slow and shit but I don't know enough to say but its a common theme to look into.
But easy to set up and get going, if you can install Linux you can do this.
The small media Bluetooth keyboard with tiny touchpad that fits in pocket it absolute necessity but have a spare keyboard or ssh if your plan on lots of setup typing...

>>1498389
based IC.

>>1498451
Regular readers also have shit range. If you want to read short range RFID tags at longer distances, you need to have reader that is outputing shitloads of power AND is sensitive enough to pick up RFID tag signals. At which point UHF RFID becomes cheaper and easier to implement.

Or use fucking bluetooth, it can survive on single CR2032 for quite some time if used properly.

>>1498458
>>Or use fucking bluetooth, it can survive on single CR2032 for quite some time if used properly.
i need at least a year on no more than two coin batteries in this case
from what would this "bluetooth tag" be made though?
The functionality i need is very simple, i just need the tag to have some secret password string saved on it and the reader will read this string.
I would like to make this work with arduino or raspberry. What parts would i need?

>>1498461
Do you not have a phone?

>>1498467
you would have to install an app and phones do not by default broadcast location for security, so there would have to be pairing and stuff as well and that is too complicated at this point

>>1498412
>any Pi will play 1080p video. you dont need a super duper model for that
as is already pointed out, it will only play x264 smoothly because it has hardware acceleration for it, Pi 2 and 3 might cope with other codecs at 720p and possibly 1080p, but Pi 1 wont

>>1498496
There pretty much isn't a reason to play any other format, though. Everything is available in x264, and if by some miracle it isn't, you can just let Handbrake chew it through the default x264 preset, should take an hour tops even if your CPU sucks ass.

>>1498461
>i need at least a year on no more than two coin batteries in this case
Energy harvesting?

>>1498235

Retarded question to end the dead thread with-

If I’m trying to charge a battery with whatever power supply, does the power supply voltage need to be higher than the battery? Like if the battery is at 3.7V and I have a solar panel reading 3.0V hooked directly to the battery, will the battery be charging?

>>1498751
Basically, the battery is in parallel with the voltage source. Two things in parallel have to have the same voltage. So you slowly ramp up the voltage on the PSU (or just give it a constant current) and the battery's voltage will equally slowly ramp up, charging it. If you're just using a resistor to limit current, then by knowing the kind of current you're trying to put into the battery you simply work out the voltage that should be across the resistor to get the correct current, and set your power supply to be that much above the initial cell voltage.

To be perfectly safe, to charge up a cell with a resistor to trickle-charge it, you'd have a resistor value such that you get a safe current flowing at the start and no current once it's up to full-charge, i.e. the PSU voltage should be set to 4.2V for a lithium ion. By extension, (V_max - V_initial)/I_max = R_charge, where V_max = 4.2V, V_initial is 1.6V or whatever your shitty cell was, and I is 40mA or whatever.

But this method will be slow as shit, especially at the end since the current will be so low, a problem you wouldn't get if you used a 12V source and just kept an eye on it. Better to throw a multiturn trimpot on an LM317 and use it as a constant-current regulator and follow a decent curve while constantly monitoring current and voltage. Could probably write an MCU code for that if you have an Arduino and a digital pot/DAC lying about, which you probably don't. PWM and a low-pass filter could cut it also, though I'm not sure how much feedback you'd need.

>>1498760
Umm ok sweet, I understood some of it but this is mostly above my simpleton mind.

I ripped apart those old solar lights, then tried hooking up the little solar panel straight to the battery to see if it would slowly charge up those flat batteries at all. Panel voltage was around 1.0V with the light, battery was like 1.8V. Then I poked and probed and clamped and it was reading around -0.002V, so the current was going from the battery to solar panel? I thought I was doing something wrong and then remembered those Li-Ion chargers are always 5.0V out going to the ~4.0V battery, so I came here to find out how retarded I am.

Yes the current was going backwards. But clamping at such low DC currents isn't very accurate anyhow.

>>1498776
Oh yeah clamping that low was definitely worthless. I just threw it on there because it was already on my workbench. I had the regular meter but the negative current was so low that I wasn’t getting shit until I put the lead in the <500mA plug. First time ever using the microamp range on that meter.

So where does that current go? It ends up as a very slight amount of heat in the solar panel or something?

>>1498783
>So where does that current go? It ends up as a very slight amount of heat

nah, dude.
sunlight in = electricity out
electricity in = sunlight out

Get your visas ready, it's time to migrate

>>1498817
>>1498817
>>1498817
>>1498817

>>1495906
connect that bit to your dog

>>1498767
panel voltage open circuit will need to be be above battery open circuit voltage (unless you have a boost converter). Li-ion chargers do ~3-4.2V, 5 is way too much.