/ohm/ - MOS 6581 SID edition

[Anonymous No.1474893]

thread with dry caps: >>>>>>1471126→→

>RULES (draft, comments welcome):
0. Mains wiring goes to /qtddtot/. PC assembly to >>>/g/.
1. Know the law. You are assumed to understand Ohm's Law and Kirchhoff's circuit laws. If not, try /sqt/ or /qtddtot/
2. RTFD. Re-read all documentation/datasheets pertinent to your components/circuits before you ask.
3. Pics > 1000 words. Post schematic/picture/sketch/9001.5 hours in MS Paint with all part numbers/values when asking for help.
4. Read the problem statement before replying.

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

>Problem ideas:
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/

>Principles (by increasing skill level):
Mims III, Getting Started in Electronics
Platt, Make: Electronics
Geier, How to Diagnose & Fix Everything Electronic
Kybett & Boysen, All New Electronics Self-Teaching Guide
Scherz & Monk, Practical Electronics for Inventors
Horowitz and Hill, The Art of Electronics

>Design/verification tools:
LTSpice
falstad.com/circuit/circuitjs.html
NI Multisim
CircuitLab
iCircuit for Macs
KiCAD (pcb layout software, v5+ recommended)

>Components/equipment:
Mouser, Digi-Key, Arrow, Newark (global)
RS Components (Europe)
eBay/AliExpress sellers, especially good for component assortments/sample kits (caveat emptor)
Your local independent retail electronics distributors
ladyada.net/library/procure/hobbyist.html

>Related YouTube channels:
mjlorton
paceworldwide
eevblog
EcProjects
greatscottlab
AfroTechMods
Photonvids
sdgelectronics
TheSignalPathBlog
BigClive

>Li+/LiPo batteries
Read this exemplary resource first: https://www.robotshop.com/media/files/pdf/hyperion-g5-50c-3s-1100mah-lipo-battery-User-Guide.pdf
>I have junk, what do?
Take it to the recycler.

[Anonymous No.1474897]

does anyone know of any textbooks or resources for synthesizer design? I can poke around with shitty square wave schmitt oscillators but at the end of the day they're just novelties

surely there must be some textbook which describes synth design from the ground up?

[Anonymous No.1474901]

This thread support bought to you by the trusty Keystone 4893 standoff

Quite neat how it's a SMD part, comes in a reel and is also machine pickable ;^)

[Anonymous No.1474910]

if I'm going to be majoring in computer engineering and I'm still doing my Gen Eds, should I focus on learning/improving my programming skills or my circuit fundamentals?

[Anonymous No.1474937]

>>1474842
yeah, Li+ chargers take note of the state of charge before they start charging. over-discharged and fresh new cells need to be charged at a very low rate until they're within the normal service voltage range, otherwise they become fire

>>1474891
a better example would be a kit of (1000) 50V ceramic caps for $4. try to beat that price for an equivalent kit from a domestic seller

>>1474893
>>1474901
son, am proud
t.the usual OP

>>1474897
analog synths grew out of the technology and design of analog computers. there should still be some good literature to be found on that
TI has a nice collection of op-amp circuits with some tutorial http://www.ti.com/lit/an/sboa092b/sboa092b.pdf
Nuts and Volts had some good material on "understanding and using op-amps", including OTAs
also there's nothing wrong with looking at/aping existing module designs

[Anonymous No.1474938]

>>1474910

the world is a varied place, but in general, computer engineering is 97.5% software. coz anything your chip can do, i can do in 100 lines of code.

[Anonymous No.1474944]

>>1474897
>does anyone know of any textbooks or resources for synthesizer design?

Make: Analog Synthesizers

website is http://musicfromouterspace.com/ which is also a learning resource.

[Anonymous No.1474951]

>>1474938
I figured.
The more I learn about electronics the more I truly grasp the meaning of microprocessors and microcontrollers being in everything

why design a circuit when you can drop a MC in with some glue logic and get the same result

Assuming you're in the field, what languages do you work with most?
I really want to do kernel and operating system programming along with embedded programming and circuit design.
I figured the best/most universal language for that stuff would be C++ so I've been focusing on that

[Anonymous No.1474964]

So arduino has several 5V pins. Is it safe to short them together?

[Anonymous No.1474965]

>>1474964

they're already shorted.

[Anonymous No.1474971]

>>1474951
yes, software has eaten the world
>why
speed and latency, of course
>glue logic
which, if you're savvy, is written in Verilog

[Anonymous No.1474975]

Asking again, because the ham radio general seems to be dead.

New to RF and I want to explore it more as I can't take the courses that I would like to right now. What are some software I should get familiar with for RF and Microwaves?

[Anonymous No.1474984]

>>1474971
whoops I meant that why question rhetorically, though your points still stand

[Anonymous No.1474985]

>>1474965
Hmm then why isn't the arduino working?
Could it be because i connected 12.5V to the raw pin? But specs say it should handle it and when i try to power it with the FTDI prorgammer it works fine. And i doubt i could melt the voltage regulator without damaging the rest of the board

[Anonymous No.1474986]

>>1474975

not sure the hams can help you anyway, coz they're not very academically-minded. what i'd do is get the course syllabus from the school's website, then visit the bookstore to read thru the textbook a bit. take a picture of the index and table of contents. see if you can download a copy from that website that has all the text books.

i took a course in microwaves, and had fun playing with the pipes and antennas, and shit, but found RF in general to be a disagreeable thing that made very little sense.

[Anonymous No.1474987]

>>1474985
>i doubt i could melt the voltage regulator without damaging the rest of the board

you doubt wrong. that scenario can certainly happen.

check your drawings to see if there's a fuse you might have blown.

[Anonymous No.1474991]

>>1474987
i don't think so. in order to burn it it has to get at least hot first so it would mean after connecting the 12.5V it would turn on for at least a little bit, there is no way a voltage the board is basically rated for would burn the regulator isntantly and not touch rest of the board
Maybe i will try soldering the pins into the holes instead of just sticking them in, in case it's just a bad contact

[Anonymous No.1475080]

Continuing the constant current lamp driver from the last thread.

>>1474640
I tried this circuit with a 1 Ohm shunt. It works like a charm. Based on my measurements its input draws 104 mA at 13,4 V supply and outputs a constant current of 244 mA. At that current the lamp voltage settled around 4,3 V. Based on these figures the efficiency is about 75 %, not bad at all and might still improve with a Schottky diode instead of BYW 29.
Now would need to add some resistors & transistors to enable current adjustment.

>>1474879
This looks good too, especially because it would save one op-amp from the circuit. Have to try it out, thanks for the idea.

[Anonymous No.1475088]

>>1475080
Wait, can't a positive-temperature-coefficient bulb experience thermal runaway when running at constant current?

[Anonymous No.1475102]

>>1475088

question makes no sense. a constant current is the opposite of a runaway current.

[Anonymous No.1475119]

>he doesn't disguise his uc as a balun

[Anonymous No.1475121]

I fell for the ridiculous >german engineering meme and bought Sennheiser momentum in ear earphones.
After a year, the same thing that happens to the cheap chinese earphones, happened to mine: one earphone stopped working.
If I pinch the cable near the jack, it starts working again for a milisecond, so probably there is the problem, some wire just cut off.
Can someone give me a quick rundown on how to repair these properly?.

[Anonymous No.1475125]

>>1475102
not constant voltage tho. conceivably, even with I^2 remaining fixed, R *could* increase if its tempco were positive and high enough

[Anonymous No.1475139]

>>1475121
the gauge of wire used in IEMs is so small that they're practically impossible to satisfactorily repair

[Anonymous No.1475143]

>>1475102
For semiconductors you're right. When the temperature increases, the resistance decreases, so if you have a constant voltage across it the current will just keep increasing by way of I = V/R. Constant voltage + increasing current = increasing power = increasing temperature. Put a constant current through it and the decreasing resistance will simply decrease the voltage, keeping it in a steady state.

For an incandescent, when the temperature increases, the resistance increases also, so if you have a constant current across it the voltage will just keep increasing by way of V = I*R. Constant current + increasing voltage = increasing power = increasing temperature. In this case, using a constant voltage supply will lead to negative feedback, which is how most dumb soldering irons work. And since those bulbs get up to a good ~2500C I imagine their resistance increases by quite a lot too.
>the resistivity of tungsten becomes 15 times greater when it is glowing hot
Says once source.

>>1475121
>>1475139
I'd recommend soldering the wires to a new jack, but instead of using whatever shroud comes with the jack just fill the bastard up with epoxy. Should work great to stop anything in there from breaking. Putting a ~1" length of heat-shrink on the insulation before the stripping could help for strain relief, provided it sticks there properly. Other strain reliefs like a spring or more elastically-curing polymer could also suffice.

[Anonymous No.1475162]

>>1474746
>Not him but I'd build shit if i had money to buy a ton of extra components to tinker with and learn
oh please. when I was a kid I used to take home vcrs and shit from the dump and strip the components off of them. my biggest investment was a soldering iron.
>get shit for free
>take shit apart
>use the shit to build other shit
this is basic stuff, come on.

[Anonymous No.1475165]

>>1474901
>surface mount
>tin plated
are these really meant to be soldered in? I don't think I've ever seen soldered standoffs like that. I like the idea.

[Anonymous No.1475169]

>>1475139
>practically impossible to satisfactorily repair
my ass. I repair these lots. it's tedious, but not impossible at all.
Separate the wire from any plastic threads that might be twisted with it
Burn the enamel carefully with a flame
Scrape off the residue with cloth
Then solder just like normal wire. sometimes use extra flux, it can be hard to get it to take solder.

[Anonymous No.1475191]

>>1475162
Not the shit I want to make, sorry.

[Anonymous No.1475239]

>>1475165
Yuppo. You are probably used to the idea of standoffs being press fitted in, well nowadays you can get them in SMD varieties

[Anonymous No.1475249]

>>1475165
>Mounting Hole (M.H.): .213 [5.4]
>Solder Pad Dia. (S.D): .306 [7.8]
apparently they are. fancy that

>>1474746
srsly anon you can get a good amount of stock for $100 including a few ICs of whatever interests you

[Anonymous No.1475270]

>>1474893
so what does this thread have to do w/ the 6581?

[Anonymous No.1475272]

>>1475270
whatever anyone brings to the table
so usually nothing
I was going to make it a "get to the point Dave" edition but OP beat me to it

[Anonymous No.1475273]

>>1475249
>srsly anon you can get a good amount of stock for $100 including a few ICs of whatever interests you
That's the thing - I don't have a job yet kek

[Anonymous No.1475276]

>>1475272
so why call it the 6581 thread if it has nothing to do with sid

[Anonymous No.1475277]

>>1475273
ohhhh, well then
if you can scrounge up $20-$25, you could get a less complete but still useful kit of resistors, ceramic caps, a modest grab-bag of transistors, some op-amps, a few 555s, and maybe some elcaps
the trouble with pulls is that you may or may not know what you're getting and you probably won't know if they work. this can lead to lots of frustration. also, some components take a bit more than a pencil iron to pull
hot tip: back before lead-free, a buddy of mine used to float boards on an electric skillet full of peanut oil and just pop components off with appropriately sized pliers. I don't know what he'd use for a heating media today that would survive 450°F without unpleasant decomposition

>>1475276
to confuse newfriends
here, have some porn

[Anonymous No.1475295]

>>1475277
>you could get a less complete but still useful kit of resistors
I suppose I have a lot of that from stuff I've got or bought in the past. Right now i'm trying to recap razer mako speakers because its being a fucking piece of shit and i largely like it aside from cheap chinese shit, heat, and the odd failure mode its in
(tweeters make hiss for a while when powering up, goes away eventually; subwoofer makessome low level noise too now)
I'm also trying to design and build a lab bench power supply, and an electronic load. The latter came up because I realized i wanted a way to test my power supply that wasn't just having a shitton of high wattage resistors (though that certainly works) plus its a good learning experience
also trying to learn control theory and how to tune op amps properly so they don't go into oscillation and shit

[Anonymous No.1475312]

>>1475295
Op amps shouldn't go into oscillation unless they've got a big phase difference from the input to the output. Are you variable load anon?

[Anonymous No.1475316]

>>1475312
>Phase difference
well obviously, but yeah
>variable load
I think so?
Might also have to rethink the whole circuit too so there's that

[Anonymous No.1475319]

>>1475162
Not so easy these days with the proliferation of tiny ass SMD components and Chinese chips with no datasheets.

I'll add that while you should be able to SMD solder these days it's hard to prototype and breadboard with SMD parts and there are many parts these days without exposed leads like QFN packages that you pretty much can't solder/desolder without a reflow oven or at least a hot air gun. Modern electronics are just not friendly for salvaging. Also since almost everything is done digitally these days it's harder to find analog components at all. You open up most stuff these days and you'll just find resistors, caps, and a micro you can't do anything with because you have no way to program it.

Unfortunate reality is today's world is not the same as when you were a kid and the things you were able to do then simply cannot be done now.

[Anonymous No.1475321]

>>1475319
Agreed
But i've already explained to him what i'm trying to do / make
The other thing is I want to learn how to program microcontrollers (that aren't arduinos)
So that's another world I need to get acquainted with too. With JLCPCB, I can easily have them make my finalized project (with microcontroller), but first I'd need to learn how to program and use one properly
I think that's harder to do since a lot of shit involves arduino these days

[Anonymous No.1475324]

>>1475321
Also to add, i was looking on digikey and common smd components are dirt cheap. 100 resistors for a few bucks or something
Unless you need something specific like high wattage shit
>>1475319

[Anonymous No.1475325]

>be installing electrical boiler with boss
>boss tells me to turn of the boiler breaker
>do it
>boss takes wire cutters intending to go berserk on the old boild cable
>ask him "ummm.. shouldn't we like multimeter those cunts first just to be on the safe side?"
>he turns to me all condescendingly
>i'm paraphrasing "Kekkity kek, are you retarded and autistic son? You turned the breaker off so therefore by definition, there can't be any tasty electron juice left in them copper snakes"
>he proceeds to snip the wires
>nothing happens
>he shakes his head at me for even wasting his time with my retarded suggestions before
>grabs the wires
>ZAAAAAAAAAAAAAAAAAAAP
>house goes dark
>stands there for couple of seconds with blank expression on his face
>try very very hard not to say something that would get me fired, that i REALLY wanted to say at that moment
>dumb tyrones living in that house had their breakers misaligned
the takeaway is, always check your wires when working with high voltage your dumb cunts

[Anonymous No.1475330]

>>1475325
HAHAHAHA
what a stupid fucking faggot
I hope you immediately walked out just so you could spare him from hearing actual laughter from you

[Anonymous No.1475336]

>>1475321
Okay, I'm really failing to see your problem though. Just get a PIC micro with the little dev board, and you're set, preferably the debugger too whenever you have the money for it. I'm not sure exactly how this relates to being too poor to buy components? If anything microcontrollers are better for poorfags since you can usually assimilate the functions of multiple discrete analog and digital circuit blocks into the micro thus saving on external components. It's not an arduino and PIC MCU's are well documented and are very widely used though more on an industry level rather than a hobbyist level. Plus the IDE MPLAB X is free but you'll need a debugger like a PICkit 3 or 4 or whatever they're up to these days to make the most use of it. I think they cost like $50 maybe less used.

[Anonymous No.1475345]

Finally. I have finished my joint 3d printer + fartduino project.

It uses a wifi module to check my web server and if it get's a specified response, i turns on a relay, which in turns on my gas heater at home.
Then the temperature probe measures, if the heater is getting hot and if not it means it failed to start
if fail is detected, then the 3d printed robot arm powered by a stepper motor descends down upon the heater control panel and pushes a reset button which resets the heater until it starts up (and the temp probe detects it getting hot)

oh yeah baby, this is some next level star trek tier shit

[Anonymous No.1475347]

>>1475345
Anon I... I...
Wow. Post another pic when it's mounted

[Anonymous No.1475348]

>>1475336
That's for later
Right now i'm still designing my electronic load and power supply
I'm not just going to buy shit like a retard

[Anonymous No.1475349]

>>1475347
It's literally flawless. I dare you name a single issue with it. You can't because there are none.

Worst case scenario is that the arm breaks, which is almost impossible, but then all that will happen is it not pressing a reset button so the heater won't start and the house will be cold but thats it.

[Anonymous No.1475351]

>>1475349
You misunderstand
I really like it.

[Anonymous No.1475352]

>>1475351
Fuck I just realized one major fucking flaw .. FUCK
This isn't stratrek, this is fucking starwars.


If a power goes out when the arm is in the bottom position, then after power comes back on the stepper won't wont know that because it is an open feedback loop, and therefore it will keep sitting on the reset button... FUCK... i just almost destroyed a $2000 heater
This means i have to add either a feedback sensor or a battery back up to make sure power never goes down

[Anonymous No.1475353]

>>1475352
You could add a battery and when you sense the power going off you move the arm into the resting position and turn off the arduino

[Anonymous No.1475364]

>>1475352
>This means i have to add either a feedback sensor

I mean...there's no reason t has to be anything other than a simple momentary switch. I don't even know why you'd consider a battery backup as an alternative.

[Anonymous No.1475369]

>>1475353
>>1475364
I'm going to make it super simple and reliable.
There is no good spot to put a button on anyway.
I am going to glue a reed switch like in the photo and a magnet on the arm, and when the magnet triggers the reed, i will know the arm is fully retracted. So every time arduino reboots i can run a simple homing loop.

[Anonymous No.1475370]

Can anyone explain to me why, on page 2 of this datasheet, https://omronfs.omron.com/en_US/ecb/products/pdf/en-g2r.pdf, under Coil Voltage there's 5 or 6 different values for rated voltage for each model?

[Anonymous No.1475376]

>>1475349
the only flaw is that it uses wifi. If you can run a wire to it easily, you might consider using an ethernet shield and hardwiring it.
maybe I've just always had shitty routers, but in my experience anything that uses wifi and stays on for long periods of time tends to lose its connection when I need it most.

[Anonymous No.1475378]

>>1475370

they have different coils. when you buy one, the part number also includes the voltage

[Anonymous No.1475414]

>>1475376
There is no cable connection in the heater room, and i don't want to run a cable when the wifi is working fine. Not to mention that the ethernet shield costs more than the arduino itself and i already own the wifi module anyway and it's just more convenient

[Anonymous No.1475528]

>>1475324
yep. SMT resistors and caps 5 for a burgerpenny in some chink assortments

>>1475336
PICs are slightly cheaper and more of a headache
fortunately for
>>1475321
AVR ≠ arduino and gcc is still free

>>1475369
micro switches were made for this shit

[Anonymous No.1475534]

>>1474893
Mr. Carlson's lab should be added to the bread as a youtuber

[Anonymous No.1475561]

>>1475534
there's limited room in the OP. whom do you propose he should replace and why?

[Anonymous No.1475594]

>>1475561
We need more racial diversity in the op.

[Anonymous No.1475595]

>>1475414
>the ethernet shield costs more than the arduino
I hate this, aren't there any MCUs that come with an RJ45 socket (and the correct ICs to modulate/demodulate)?

[Anonymous No.1475601]

>>1475595
if you don't bother about the fiberglass-wasting "shield" form factor, chinky W5500 ethernet modules can be gotten for around $3
>onboard ethernet
some of the more upscale STM32 devices have ethernet MACs in them, but you might have to provide an external PHY which adds some cost and only buys a little speed

[Anonymous No.1475616]

>>1475601
$3 is still twice the cost of an arduino nano, let alone any boardless MCU. I just want something that can act as a sensor array that can be plugged into a network so I can access the data from anywhere with internet connectivity. I find wi-fi is too unreliable and also a waste of power. Being able to power such an MCU with POE would be neat too, but I guess that would require a buck converter.

It's actually not a bad idea to have it all on one board, buck converter included, could probably design a PCB for it and market it to the IOT crowd.

[Anonymous No.1475624]

Why can't I find an AVR programmer that works? Does anyone have a cheap programmer that they know for a fact works with Windows 10?

[Anonymous No.1475625]

>>1475616
wired ethernet is an inherently expensive technology, sorry. transformers are expensive, MACs are expensive, IP is resource-heavy, Cat5 isn't exactly cheap, and wired Ethernet is almost as much of a power-hog as wi-fi
I'm sure you'll work out a tradeoff that's suitable for your particular application, with so many choices out there

[Anonymous No.1475628]

>>1475625
>transformers
I saw this on a datasheet, what are the transformers used for? Voltage boosting?

Sounds like a simple wifi ESP8266 is a much better fit for this purpose, thanks for the info.

[Anonymous No.1475630]

>>1475628
they're often called "magnetics." used for isolation and balancing.

[Anonymous No.1475638]

>>1475624
USBASP werks for me.

[Anonymous No.1475640]

>>1475628
no prob. two more notes: if you're concerned about range and reliability, there are ESP8266 modules with external antenna ports, which you might prefer on more distant nodes. if power consumption + range are really a big deal for you, consider nRF24L01+ which excels at both, but speak a proprietary air protocol which will require a gateway somewhere

[Anonymous No.1475644]

>>1475628
isolation, usually rated for 1kv or more, mostly so lightning or one computer fault can't fry an entire network.

[Anonymous No.1475756]

How can I go and make something like this? I want to create a Rom dumper, something that is able to read data from a GB/GBC/GBA and write it somewhere and vice versa. I think it would be a neat project for a beginner like to to work on.

[Anonymous No.1475781]

>>1475595
my favorites are the esp12 and esp32 modules.
They are basically small arduinos with built in wifi. with next to no power consumption
using ethernet cables today is like using leeches to treat wounds

[Anonymous No.1475783]

>>1475756
the first thing to learn about electronics design is how to look up information. about half of the hobby is reading shit
pro tip: if you don't know what you need to know, search the web to see if someone else has done it first
http://douevenknow.us/post/68126856498/arduino-based-gba-rom-dumper-part-1

>>1475781
nothing personnel, kid
>sends you a deauth storm

[Anonymous No.1475787]

>>1475783
>he doesn't have his apartment placed in a Faraday cage
kek
it's like you want to spread you ass for the government

[Anonymous No.1475789]

>>1475787
kek's on you, I don't even use wifi

[Anonymous No.1475790]

>>1475789
just because you don't doesn't mean the government doesn't. they literally have their satellite aimed at you this very moment and are reading your every thought and are corrupting you at the same time. do you really think it was your idea to fap to the dickgirls? no. degenerate population is complacent population. wake the fuck up son.

[Anonymous No.1475803]

Alrighty. I added a feedback sensor to created a closed feedback loop.
So now when the arduino reboots a homing cycle runs first, which works very simply. I starts rotating the motor until it recieves HIGH signal on an input pin.
The sensor is a simple reed switch, triggered by a magned kapton taped to the moving arm.
(I used kapton tape in case i need to push buttons in the vacuum of space)

[Anonymous No.1475804]

>>1475803
Fuck I can't even type properly with all this garbage on the table, i really need to clean up.

[Anonymous No.1475833]

>>1475790
>satellite aimed at you
No need, your personal satellite is your phone.
"Physiologically patterned weak magnetic fields [..] increase acceptance of false statements as true." www.ncbi.nlm.nih.gov/pubmed/19037785
That was 10 years ago...now duck /?q=synthetic+telepathy

[Anonymous No.1475841]

>>1475803
nice job mate.
If you ever want to do some more advanced automation of your heating system, you can get these little 1-wire temperature sensors called MAX31820. They're accurate and you can chain a bunch of them together on one bus. 1-wire communication libraries already exist for arduino.

[Anonymous No.1475846]

>>1474897
checkout microelectronics by sedra and smith. there's a good section on op-amps

http://gen.lib.rus.ec/

[Anonymous No.1475853]

>>1475841
The sensor i use already uses 1wire i will be adding more of them later so i can measure separate pumps

[Anonymous No.1475899]

>>1474893
i have a question, not sure if related

i want to make some /diy/ night lights that turn on/off automatically at a dusk/dawn

store bought plug in nite lights arent reliable


can I use a 555 chip to make it turn on a small light bulb at 8pm and turn it off again at 8am?

(basic timer controlled switch)

also, what a good source for small hobby project electronics enclosures?

[Anonymous No.1475902]

>>1475899
Use arduino mini and a 3d printer

[Anonymous No.1475903]

>>1475902
no, im being serious. i know i don't need an ardunio for a simple night light.
but this will be my first diy project like this.

[Anonymous No.1475906]

>>1475903
Then use photo resistor and a mossfet and a 3d printer for the enclosure

[Anonymous No.1475908]

How do you calculate how much phase shift a capacitor of some arbitrary capacitance value will introduce into a circuit at some frequency or over a frequency range?

Like if you were to do a magnitude/phase plot on a VNA. Just to clarify I'm NOT talking about the 90 degree voltage/current phase shift described by that ELI the ICE man maxim. I'm talking about pure voltage phase shift or phase delay/group delay whatever it's called. Like a phase frequency plot.

[Anonymous No.1475912]

>>1475906
i have a 3d printer
im going to put my device into a mason jar or something. basically a recycled-junk Fallout3 tier portable lamp

[Anonymous No.1475914]

>>1475908
Look into Bode plots

[Anonymous No.1475917]

>>1475906
i dont know alot about ic chips or electronics for that matter, but from what i read so far, i can use a 555 chip.

what is the diffeence or the advantage of a mossfet?

[Anonymous No.1475919]

>>1475914
I know what a bode plot is. I just wanna know how to actually calculate the phase shift of a component like a capacitor or some network of components like an RC filter at a given frequency or over a range.

Like what formula would I use if I wanted to generate a bode plot for a given network?

[Anonymous No.1475920]

>>1475919
You would find the transfer function of the circuit network and then generate a bode plot using that. Read into how you actually make bode plots.

[Anonymous No.1475922]

>>1475920
Isn't that just the ratio of your output voltage to your input? Where do the frequency dependent parameters come into play?

[Anonymous No.1475924]

>>1475922

http://class.ece.iastate.edu/ee230/Support/Transfer%20Functions%20and%20Transfer%20Characteristics.pdf

[Anonymous No.1475925]

>>1475917
In all seriousness a microcontroller (simplest option for you would probably be Arduino) would be easiest here. A 555 timer with a period of 24 hrs, (~85000 seconds) would require a huuuuuge cap and resistor - you're talking 1Meg resistor and couple of hundred milli-Farad capacitor (that's huge). At these values everything starts to affect your circuit - this assumes your cap can even charge up enough with such a small current.

Basically I'd ditch the 555 timer - best case scenario you'd end up with something that flashes, but it would be nowhere close to a 24 hour period and a 50% duty cycle. A microcontroller would save you headaches and money.

[Anonymous No.1475928]

>>1475924
Ah, this is more complex than I thought. I'll give this a read later, it's been like 5 years since I passed diff eq. though and I'm rusty. Sorry for the retardedness.

[Anonymous No.1475930]

>>1475925
okay i follow you. i'm watching some YT vids on how the 555 chip works and what it does now and i definitely see the problem with trying to use it for what i want to do. (time dalay switching events at 8 to 10 hours apart or whatever.)
555 is better for shorter time/faster switching it seems

thank you for taking the time to explain.
i'll look at micro controllers

[Anonymous No.1475931]

>>1475917
Mosfet is just a switch which gets turned on by the photo resistor when it detects darkness
Honestly personally I would use arduino mini with a clock module attachment

[Anonymous No.1475933]

>>1475930
No worries, if you google "arduino 12/12 led timer" you should come across a forum with the examples already done for you. Literally a few lines of code you can copy/paste.

[Anonymous No.1475935]

>>1475931
>photo resistor
ok. yeah the whole problem i'm having with store bought "night-lights" (which makes me want to build my own.) is that they are supposed to detect "darkness" and turn on the night light at night.
none of the night lights i have bought can properly sense light and will basically stay on 24 hours a day because they don't sense when its daytime because the room isn't fully lit with ambient daylight though the windows

so just as an excuse to diy something electronic, i want to make a few of my own small home made night lights. but i want them to reliably switch on and off at or around dawn and dusk

thinking of running it off a small 9v battery and mounting it inside a small enclosure of some kind, maybe an upside down mason jar

[Anonymous No.1475937]

>>1475933
great. this is just what i need to get started with small diy electrical gadgets. thanks

[Anonymous No.1475940]

>>1475928
No worries. Good luck with whatever you are working on. I wish I had a project to work on.

[Anonymous No.1475970]

>>1475928
For simple stuff like an RC filter, you can just look up the solution:
φ = atan(-ωRC)
https://en.wikipedia.org/wiki/RC_circuit#Gain_and_phase

For multiple of them in series, you would just add up the phase angles.

[Anonymous No.1475991]

>>1475970
>just add up the phase angles
not really, only if decoupled

[Anonymous No.1475996]

>>1475917
The prob with a 555 is the accuracy. You will spend forever screwing with RC values. Arduino is nth-degree overkill. Most clocks use the 60Hz mains freq as the time base, then you can just divide it down with a counter

[Anonymous No.1476025]

>>1475996
>buy $2 in logic chips
>instead of a $0.50 micro
74xx = hipster larp. just stop it

[Anonymous No.1476027]

>>1475996
>a $1.50 microcontroller is overkill
>so let's use a zero crossing detector
>and a bunch of hipster discrete logic
>and a whole bunch of other advanced design
>to a project for someone who doesn't know a lot about electronics
>instead of just getting the fucking job done and over with

[Anonymous No.1476034]

>>1475996
Using the 60/50Hz line would be an interesting side project, but the guy wasn't even aware of the limits of a 555. A $2 uC, 10 lines of code he can copy/paste and about 5 solder joints and his project works.

Even if you used some high count, programmable counter to convert your 60Hz to <1.17 microHertz, he'd spend more money on chips, protoboard, discretes, transformer etc... And would be playing with mains AC which isn't ideal for the beginner. I rarely use Arduino in my work, but it's perfect for guys that just wanna build something up safe and quick.

[Anonymous No.1476035]

>>1475930
The main problem with a 555 is accuracy. If you want something which drifts by no more than a few seconds per day, the timebase needs to be either a crystal oscillator or the mains frequency.

[Anonymous No.1476045]

>>1475928
It's not that bad. Find the overall impedance and get the phase of that.

[Anonymous No.1476148]

>>1476034
> A $2 uC, 10 lines of code he can copy/paste and about 5 solder joints and his project works.
You also need a crystal. Even calibrated, the RC oscillator built into a uC isn't accurate enough to use as a clock.

> And would be playing with mains AC which isn't ideal for the beginner.
You can get wall warts which output low-voltage AC.

[Anonymous No.1476150]

>>1476148
>crystal
$2 includes the crystal and the board
for 19 cents more one could get an ESP8266 capable of running NodeMCU, querying an NTP server, and in a small package. just add a MOSFET to turn the light on and off, and you're done

[Anonymous No.1476209]

>>1476150
>>1476148
It's not that hard to build a peak detect and a comparator that will detect that 60Hz that your power companies spends millions regulating.

[Anonymous No.1476212]

Let me state that only in some countries is the mains frequency accurate enough to be used for a clock: https://youtu.be/bij-JjzCa7o

Personally I'd make a clock that runs not off a crystal but off the time signal service. For the US there's all these:
https://en.wikipedia.org/wiki/Time_synchronization_in_North_America
Some of which I think are over copper, not the airwaves.

[Anonymous No.1476221]

>>1475935
>none of the night lights i have bought can properly sense light and will basically stay on 24 hours a day because they don't sense when its daytime because the room isn't fully lit with ambient daylight though the windows

I took one apart once and the PhotoResistor was in parallel with the light. It in effect shorts the lightbulb when there's light present.
Light present = low resistance

Seems to me that you could just put the right value resistor in parallel to make it turn off with less light. Add a potientometer to make it adjustable.

[Anonymous No.1476222]

>>1476221
>PhotoResistor was in parallel with the light
I don't believe it

>>1476209
it'd be easier yet to build a comparator that tests ambient light level and just turns the damn light on and off

[Anonymous No.1476223]

>>1476222
>it'd be easier yet to build a comparator that tests ambient light level and just turns the damn light on and off
That's not what he wants, though

[Anonymous No.1476229]

haha, that autist that was butthurt last thread that Dave Jones criticized that open source scope and its really shitty schematic is gonna be absolutely ass-blasted when he see's the new video

[Anonymous No.1476235]

>>1475899
sixpack for 10$
too cheap to diy

[Anonymous No.1476256]

>>1476222
>I don't believe it
Bigclive took apart one or two things that used this. Since the LED was being current limited by something with a large voltage drop across it anyway it didn't make much of a difference.

>>1475935
I'd actually just put a photoresistor (or photodiode) outside an equator-facing window and feed its signal to your lights. Even have a small calculator solar panel + capacitor and have the thing power itself and be the light sensor if you're going for wireless. Personally I'd have:
Light sensor > schmitt-trigger op-amp circuit (with potentiometer to adjust threshold) > rising and falling-edge detector > 555-timer oscillator > IR LED > > IR photodiode > band-pass filter op-amp circuit > diode detector > T-flip-flop > transistor > night light LED.
Here's an estimate at the timing diagram. The reason behind the edge detecting and flip-flop is such that you're not constantly wasting power by beaming IR all over the place. Using an oscillator instead of DC to power the IR LED because that way you can use a filter to seperate its high-frequency (perhaps >1kHz) from the slowly changing ambient light levels. You'd need direct line of sight from the transmitter to the receiver, though having relay stations with a receiver and transmitter would work fine too, could probably just have an IR photodiode > filter/amplifier > IR LED. Could replace all the filters and other logic with a tiny microcontroller, like a simple direct-USB programmed 8-pin ATtiny85, but it's a bit of a hassle and less fun.

[Anonymous No.1476275]

>>1474893
Does anybody know if it is possible in kicad to move multiple silk screen references to a differnt layer at once.

[Anonymous No.1476278]

Some op amps you can't get too close to one of the supply rails on the input. If you do, it goes to the opposite rail.. or something to that effect.
What's that behavior called again?
I'm looking for op amps that DON'T do that

[Anonymous No.1476280]

>>1476229
>that autist
Autists like dave jones though

[Anonymous No.1476281]

>>1476278
Phase inversion? Heard that term once but I've no idea if it's what you're looking for.

[Anonymous No.1476284]

>>1476281
Not sure
Don't think that term is ringing any bells
I just remember it's a thing on some op amps and want to pick one that doesn't have it

[Anonymous No.1476286]

>>1476235
inside

[Anonymous No.1476300]

>>1476278
>>1476281
Phase reversal.
By definition, rail-to-rail input opamps don't have it. Not that they're your only option, but still.

[Anonymous No.1476321]

>>1476300
Yeah that's the term. thanks anon
Yeah just trying to make sure op amps i want to use dn't have it. they dont' advertise them as having phase reversal on their datasheets if i'm not mistaken. Looks bad or whatever the fuck

[Anonymous No.1476336]

Hey bros, is it possible to connect multiple earbuds in parallel to one male 3.5 mm connector? While I'm aware that 5x the speakers doesn't equal 5x the sound I've got 5 half-broken/shitty/discarded earbuds and some free time so I'd like to see what it would sound like to run the same audio signal on 5x2 earbuds at the same time. Also, do regular pc or smartphone 3.5mm headphone jacks even output enough power to run more than 1 pair of earbuds?

[nite light anon No.1476347]

thanks to everyone for the info

here's a pic of my night light that doesnt sense ambient daytime light well enough to turn off during the day. I have two of these and neither one works properly

before i buy the $300 worth of electronics hobby kit components I have in my Amazon shopping cart, Im going to take this little fucker apart and see if I can fix it.

One simple idea i had is to physically move the sensor further towards the outside of the fixture, hoping more exposure to ambient light in the room will make it work.
Right now the sensor is kind of recessed inside the body and probably not getting enough light.

also I'll see the build layout inside and try to learn something about its circuitry

[Anonymous No.1476355]

>>1476336
try it and find out

[Anonymous No.1476385]

How linear is the relationship between the rated power in watts of a motor and the actual power?

[Anonymous No.1476414]

>>1476286
>not the airwaves
If the sensitivity didn't suit you, it could be adjusted via R3.

[Anonymous No.1476419]

>>1476355
Sure, I'll post the results when I do it

[Anonymous No.1476420]

>>1476286
>inside
If the sensitivity didn't suit you, it could be adjusted via R3.

[Anonymous No.1476434]

>>1476385
woops *torque

[Anonymous No.1476457]

>>1476420
>could be adjusted via R3
Obviously, but I don't fear darkness and need no night light. This is an even cheaper version, maybe >>1476347 will find something like that.

[Anonymous No.1476463]

>>1476457
This is not a good design. You're burning around 1W in each resistor which is wasteful and requires large power resistors. A 100-150nF capacitive dropper would be a better choice. A properly isolated supply would be best but that's probably overkill for a night light.

[Anonymous No.1476494]

>>1476463
>capacitive dropper
Not possible with a half wave circuit.

[Anonymous No.1476510]

>>1476457
>>didn't suit you, it could be adjusted via R3.
>I don't fear darkness and need no night light

the 'you' was directed at the snowflake who's trying to adapt his

[Anonymous No.1476512]

I have a 5 V regulator powering some ICs in my constant-current lamp driver circuit. Can I use the following way of switching the 5 V rail completely off with PNP transistor to minimize power consumption? I am planning to switch it on for a second or two every ten minutes.
The datasheet of the microcontroller says the max. voltage that can be applied to any of the GPIO pins is 6,5 V. I guess the output pin of the MC would get fried if the Base of the BD 136 would be connected directly to it?

[Anonymous No.1476521]

>>1476027

>a $1.50 microcontroller is overkill

Good idea Tony Stark; I'm sure he has lots of PCB material, layout/masking facilities, echant , plating, step-down regulated power supplies, support chips, programming experience, etc. etc. Usually, all that stuff is free isn't it?

Remeinds me of the adults that spent their lives living in their mom's basement, and have no idea that things like food and electricity cost money in the real world when it comes time to move out.

>so let's use a zero crossing detector
>and a bunch of hipster discrete logic
>and a whole bunch of other advanced design
>to a project for someone who doesn't know a lot about electronics

Well they certainly aren't going to learn about any of that stuff here. Nothing more hipster than the Arduino solution to flash a goddamn LED.

>instead of just getting the fucking job done and

Good idea: Just buy it. Welcome to DIY!

[Anonymous No.1476524]

>>1476034
Dave Jones has a whole video tutorial on this stuff, with a very simple way to get the (in his case, 50Hz) line freq that he built in the 80s. He built an entire clock, but you needn't go that far.

https://www.youtube.com/watch?v=5kGhoRuhlxM

[Anonymous No.1476526]

>>1476209
>that 60Hz that your power companies spends millions regulating

Embrace the coming green energy harmonic distortion apocalypse.

[Anonymous No.1476528]

>>1476521
you can get clone dev boards for less than $4 delivered. All he'd need to do it hook up a USB cable for power and make like half a dozen solder joints and he'd be finished. I agree the guy was a bit of a dud calling the logic hipster, but it's worth considering the time/effort/money to build a very simple circuit. I don't use Arduino in my work, and very rarely use embedded solutions at all, but component count wise it's hard to beat such a small solution with a ton of support out there for beginners.

Unless you're looking at this from a different pov than me, but I think the first guy was just wanting something to get him started quickly and easily. You gotta trade ease of design vs how much you want to design yourself - else why not go to the beach and gather sand to manufacture your own ICs with...

[Anonymous No.1476532]

>>1476524
Yeah I remember that video, and I had great fun buliding one similar when I was younger, but take even the components in that screenshot, plus a protoboard to build it on, plus a few spare for beginner's mistakes, and he'd be spending more than a couple of $.

I've got nothing against doing stuff without embedded solutions - it's my job to design RF front ends, but at the same time you have got to admit there are some things just quicker/easier/cheaper for the embedded guys to do imo...

[Anonymous No.1476534]

>>1476526
green energy is the biggest scam of the decade
they build about 100 windmills in my closer area
now they are shut down for 2/3s of the day
turns out the same people that want green energy, want them to be controlled and on a tight schedule to not disturb their residence and the local wildlife

[Anonymous No.1476536]

>>1476512

If you're willing to switch your regulator, some of them have a 4th pin that basically switches it on and off. I salvaged some from an old hard disk circuit board.

[Anonymous No.1476554]

>>1476526
>>1476534
If it wasn't for the damn hippies we could have nukes powering everything

[Anonymous No.1476559]

>>1476528
>>1476532

Getting back to the main problem, the clocks in real Arduinos are bad, and the chinese knock-offs are probably even worse.

> http://jorisvr.nl/article/arduino-frequency
> "Neither oscillator is good enough to use in an alarm clock without calibration"

So, he'd be better off looking in a box of old stuff he's not using anymore for an alarm clock and tying into that. Probably learn more about electronics, too, just by doing that.

[Anonymous No.1476568]

>>1476559
Yeah, fair one, Arduino clocks aren't great, but that point in the conclusion is open imo... Maybe not good enough for an alarm clock, but even with a fairly bad couple of hundred ppm a crystal will only lose a tenth of a second a day. Mains on the other hand has a fairly reasonable accuracy long term, but can lose/gain 10s or more each day. Off the top of my head the mains has an Allan dev. of an order or two magnitude worse than those graphs.
In this application even that's probably acceptable, but it just seems to me to be more complex for little (if any) gain.

Thanks for that link btw, very interesting and I'm motivated to try some similar experiments out myself.

[Anonymous No.1476569]

>>1476347
it's literally just a regular photo resistor. just get a little variable 50k pot resistor and solder it in parallel with the LDR. adjust it until it turns off.

[Anonymous No.1476623]

>>1476536
Sounds nice but unfortunately at the moment I only have L7805 & BD136 at my disposal.
By the way, what's the type of your regulator?

[Anonymous No.1476701]

[Anonymous No.1476753]

>>1476568
86400 seconds in a day
11.5ppm = 1 sec deviation/day
otoh watch crystals are fairly available with 20ppm tolerance over temperature and can be easily added to any microcontroller with a compatible oscillator. for an application like this, no higher-frequency crystal is necessary
>the mains
where tho?

>>1476623
most buck converter ICs have an enable pin, you could mod most of the cheap chinese modules by lifting the enable pin and tying it to a digital output. boost converters have them too but most of them will just pass the input voltage through and not boost it
higher-current (> 1A) linear regulators with an on/off pin are a bit spendy (LT1963 for example), but lower-current linear regulators (≤500mA), often found in SOT23-5 or SOIC-8 packages, almost always have an on/off pin

>>1476554
nukes can't just be turned on and off on the fly, so are better suited for base load. e.g. natgas plants can be turned on and off quickly and are therefore better suited for peak load. you'd need some sort of power sink, whether that be storage or "smart" loads

[Anonymous No.1476767]

>>1476753
>you'd need some sort of power sink
just put two wires in the ocean

[Anonymous No.1476793]

>>1476767
>just put two wires in the ocean
choose between hydrogen generator or desalinization plant

[Anonymous No.1476795]

>>1476753
yeah my bad on the maths, thanks for correcting me. On the mains in the EU, there's a few examples where it has been horrendous, in Feb 2018 for example. Though I suspect in the US it would be somewhat better with a somewhat more coherent energy policy.

[Anonymous No.1476797]

>>1476793
i choose you

[Anonymous No.1476801]

Can anyone help me with this? I am a quality engineer trying to get back into the technical side of electrical engineering. I am having trouble with the problem below.

I have been assuming the capacitor is already charged to 10V before the first positive half cycle. The reason for this is VI-VC=0 where VI = -10 V. So on the upcoming the positive half cycle would jump to 20V and then slowly discharge through 2R resistor.

Then on the next half wave negative cycle the would start charging the capacitor again.

I guess I don't understand what is really going on here. When I build the circuit in spice it essentially mirrors the input voltage.

[Anonymous No.1476804]

>>1476753
You can pick up mains hum from anywhere, with just a radial inductor and a capacitor or two to tune it. It's easy to pick up without the band-pass filter too, did so with my elektrosluch with a couple of axial inductors feeding a high-gain op-amp, but if you can have the band-pass filter also be the receiver then you're saving on copper. Amplifier's probably still necessary though. Lighting may also get you to skip a beat.

If you're not in Denmark then Hydroelectric is a high-volume power source that can be turned on and off very quickly, in this way it is often used to supplement more periodic sources, such as wind and solar. Nukes as a baseline and hydro for the loads is a pretty solid way to go and doesn't make any emissions, but once you've got excess power emissions don't really matter anyhow. Imagine a future where it's more economical to make hydrocarbon fuels from CO2 and water than by pulling them out of the ground, not that we'd ever get to that point without petrol prices increasing from wells drying up.

[Anonymous No.1476806]

>>1476801
Having too large a capacitor leads to the R*C being too large, making the charge time far too long such that it appears as a straight line. In other words, the capacitor is so large that for all intents and purposes, the thing doesn't act like a filter at all.

[Anonymous No.1476808]

>>1476801
>>1476806
This is what it does (zoomed out a bunch) with the capacitor at 10µ instead of 100n: it passes everything except DC, the slow relaxation is the network slowly approaching the DC-isolated equilibrium as dictated by the 2 resistors. Since R2 is twice as large as R1, it gets twice the voltage across it since the current going in is current limited by the capacitor (in the long term).

I'm making this shit up.

[Anonymous No.1476820]

>>1476806
Okay, so in your example the time to fully discharge or fully charge the capacitor is about 20 seconds. Which is a super long time, especially when the half cycle is 1 millisecond. After hitting 20V the first time it is just never able to charge up again or discharge down fast enough. So it will eventually decay down until it hits a point where it will reach some standard pulse like in your example at 1 second and afterwards.

Thank you for help! I also realized I never switched my input voltage to pulse... I was using a sine wave for some reason haha.

[Anonymous No.1476823]

>>1476820
Sine wave wouldn't make too much difference to the end result.

[Anonymous No.1476844]

>>1476801
The graph assumes that R*C is so large that the change in charge is negligible (time constant much larger than the period). Every frequency above DC gets through without attenuation or phase shift, so the output is equal to the input except for the DC offset.

The DC component of the current must be zero, so the current during the positive half-cycle is equal and opposite to the current during the negative half-cycle. But the resistance during the positive half-cycle is twice that during the negative half-cycle, so the positive peak is double the negative peak, 2/3:1/3 ratio.

[Anonymous No.1476856]

>>1476844
>Every frequency above DC gets through without attenuation or phase shift
That is because the capacitor is a open circuit for DC voltages correct?

>so the output is equal to the input except for the DC offset
The DC offset is provided by the two resistors drawing current from the resistors, correct?

>The DC component of the current must be zero,
Why? Is it because there no DC input that can pass through the capacitor?

[Anonymous No.1476863]

>>1476856
No DC current can pass through a capacitor. Capacitors are defined as I = C*dV/dt, so to get a constant DC current through it you'd need to have an infinitely increasing voltage.

The resistors form a sort of voltage divider, but instead of two rails to one voltage, it's more like one rail to two voltages, where the voltages are connected but separated from one-another by the diodes. Kinda wonky, but if you split it up by where the current goes in each half-cycle (with the exponential charging of a capacitor equation) it should be fairly clear.

[Anonymous No.1476867]

>>1476801
This is showing how you have to take into consideration input frequency for a edge detector circuit.
Given enough time if the cap is discharged every change from +10 to -10 and vice versa will move the cap 20 volts which will then discharge through one of the diodes. If you make your time constant so high the circuit will never discharge and reach its steady DC state where it becomes some weird divider like >>1476863
said.

[Anonymous No.1476876]

>>1476863
>>1476867
Thank you two for helping me. I really do appreciate it. I can't help but feel like a brainlet. Maybe this comes from me misunderstanding capacitors in the first place. I think I need to go back and learn more about RC circuits.

I am having trouble visualizing on paper the charging and discharging as well as the weird voltage divider. In LTSpice, I look at it and it makes sense but I can't translate it to the page. As for the voltage divider, I just don't see it. I even got out of bed after seeing both of your comments to try again.

On the positive half cycle the current will discharge and on the negative it will charge up the capacitor. These discharge and charging periods are not long enough to ever finish.

>If you make your time constant so high the circuit will never discharge
If that were to happen would Vc always be stuck at +10 V? The polarity of the cap is throwing me off as well.

>but if you split it up by where the current goes in each half-cycle
This is where I am lost in the sauce.

>The DC component of the current must be zero,
Why wouldn't the current be different in a charging (negative half) vs discharging (positive half)?

[Anonymous No.1476880]

>>1476876
>Why wouldn't the current be different in a charging (negative half) vs discharging (positive half)?
The current is different, but it averages out the same. The capacitor is holding X charge, it doesn't matter what resistor you put across it, it will still discharge the same number of coulombs. Smaller resistor makes it discharge quickly, but with more current, larger resistor makes it discharge for longer, but with less current. Note that this is for a discharge that takes an infinite amount of time. Here it's finite, and this means the voltage left over after the slow discharge is twice that of the faster discharge. This amount is tiny, but it adds up over the cycles, causing the slight trend in the bottom image of >>1476867. Once the cap is charging and discharging to the point where twice the voltage is across the larger resistor, these charge times equalise and there is no net voltage.

The law that caps can never pass DC is only true at infinite time taken, anything else is technically just a very low frequency component. Look into high-pass filters to see how they work around DC levels.

[Anonymous No.1476887]

>>1476876
if it makes it any easier draw the circuit out in steps. if i have time at work tomorrow I'l explain with math but essentially the circuit has 3 ish states. Lets say the initial state is that the capacitor is uncharged. and we were at the -10V input part.
Remember V C T-1 = V C T+1

1. The signal switches to 10v. On the input side of the cap it sees 10V where it was just -10V. Obviously -10 =/= 10 so the voltage spikes on the cap's output side 20V (the delta of the difference) up to 20V.

Now because we have a positive voltage on the anode of D1 it turns on. This will start to drain the cap BUT because your time constant is so large the cap only loses a bit of voltage because the drain current is so low...lets say .1V before the input switches again.

2. Now the input switches to -10V. The cap is -10 on the input side and 19.9V on the output side . Take the -20 delta and now the output side of the cap is at -0.1V.

Now D2 diode turns on, this starts to discharge the cap BUT the time constant is so high it dosent discharge all that much. And hey, the resistor is twice as large so that's (lets say approx) only a .05 loss.

3. So after the first cycle we end up with 19.9 and -.05 volts. Then on cycle two it would be 19.85 and -.1 and so on. Eventually this will equalize out if you do the math.

These are just approximations for an example but essentially you have imbalanced resistors so one discharges faster than the other leading to the imbalance in the output signal.

[Anonymous No.1476915]

>>1474893
That SID chip has a 4th hidden channel, sending data to the volume control makes an audible click.
basically you can have a 4 bit sampler

[Anonymous No.1476920]

I wanna make a voltage controlled filter where I can control the cutoff frequency with some DC, ramp, or other periodic control voltage. Be nice to control Q too but not absolutely necessary, cutoff is more important.

I'm using a Sallen-key low pass filter. I figure varying the frequency will involve varying the resistors, capacitors, or both. Now I can do resistance with FETs in their Ohmic region or with digital pots but FETs in their Ohmic region are only truly linear over a vary narrow range as the curve tapers off as it's coming up to the linear region so that's a reject and digital pots would be great except I have none on hand and I want to breadboard this thing now. For varying capacitence there are varactor diodes but their capacitence is too small for low frequency filters necessitating very large resistors plus again I have none on hand and I want to build this now plus any I've ever seen for sale are all surface mount and I can't breadboard that.

I'm figuring my other options are current sources either pumping current into or out of some node in the circuit or the feedback path. Like if I have an RC filter block that when a signal is applied I develop a voltage across it then if I have a current source pulling some current away from that output it will change the voltage and the filter properties I think? I'm not really sure. Anyway a VCCS shouldn't be hard to build so it seems like the way to go but I tried a few current source/sink configurations with the filter and I can't really get it to do what I want.

[Anonymous No.1476921]

I know there's equations for buck and boost conversion.
But recently I remember watching some video, and the person who made the video was commenting on those cheap buck converters online and that they were bad and also pointing out that they "only have one inductor"
What does the number of inductors have to do with anything?

[Anonymous No.1476923]

>>1476921
Normal buck converter has one inductor.
Maybe he was comparing them to SEPIC, which allows both step-up and step-down regulation using 2 inductors.
Or maybe he was just a retard.

[Anonymous No.1476924]

>>1476920
transconductance amps are your very best friends in VCanything construction. if you're doing anything remotely analog-synth-like, you're really going to want to have some on hand
here's a non-OTA Sallen-Key design using plain old bipolars that might be worth playing with

[Anonymous No.1476925]

>>1476923
He's a pretty good electronics youtuber
Also, have to qualify myself saying I don't remember all the precise contextual shit he said. I just remember he mentioned that it regulated very poorly, and only had one inductor.
Beyond that, anybody here have experience with making switchers?
Blah blah better to buy one - humour me okay. I just wanna have the experience making one, or some, and learn that shit
I remember some application notes for a control chip and saying something about max current, and max duty cycle - I understand that.. I think. What I am not sure I understand is if max duty cycle and the inductor's inductance has any effect at lower loads
Also I"m kinda new at switcher shit so any common but not typically mentioned pitfalls to look for would be nice

[Anonymous No.1476928]

>>1476925
>He's a pretty good electronics youtuber

I'll be the judge of that, who is it?

[Anonymous No.1476932]

>>1476925
Well, sometimes a post-filter is added after the switcher to reduce noise, but there's not much point in trying to guess what some youtube faggot meant.
Using even remotely modern switcher ICs isn't difficult. Datasheets usually tell what components to use and/or how to calculate them. Often even an example layout is provided.
Regulator's maximum duty cycle affects the maximum output voltage and its minimum duty cycle affects the point where it has to resort to pulse skipping. If possible, you should pick the inductor based on your expected input and output voltage and the expected output current. If your load is (always) light, larger inductor can be used to reduce ripple.

[Anonymous No.1476935]

>>1476928
Marco Reps
It's a video from 2 or 3 years ago I think.

>>1476932
>Regulator's maximum duty cycle affects the maximum output voltage and its minimum duty cycle affects the point where it has to resort to pulse skipping
Naively, what's bad about pulse skipping?
I perused a bunch of control ICs in the past and many of them have max duty cycles that hang near 50. some go as far as 90% and very few go as far as 100%.

>If possible, you should pick the inductor based on your expected input and output voltage and the expected output current.
What if the output voltage and current changes? (want it for bench power supply project, probably)

>If your load is (always) light, larger inductor can be used to reduce ripple.
How come I can use a larger inductor in this case, but not if the load is heavier?

[Anonymous No.1476938]

>>1476863
>No DC current..
If so you could never charge a capacitor. What about I*dt=C*dV --> V=I*t/C? The integral form is no less valid than the differential. The unit of charge is the ampere second (As) after all.. Pic: A less wonky circuit, AC in - DC out.

[Anonymous No.1476939]

>>1476938
>If so you could never charge a capacitor
>charge a capacitor
>DC current
wew

[Anonymous No.1476941]

>>1476925
>making switchers
yeah, they're pretty cookbook anymore. I once built a little breadboarding bench supply, including an Li+ charger/power path manager, a small trimmer-adjustable boost converter, and an I2C-controlled dual buck + dual LDO on one board, with a microUSB input. I pretty much just strung together the example application schematics *shrug*
a long time ago I made an inverting Vee supply for an LCD panel out of PWM and ADC on an AVR micro which also served as a USB-to-parallel bridge to the LCD's controller. that was kinda fun
too small an inductance limits the power your switcher can handle. for bucks, expect more ripple. for boosts, expect ripple and undervoltage. too large an inductance affects the feedback loop and may cause overshoots during line/load step changes. for most switchers on the order of 1MHz Fsw and a large fraction of an amp, a few to 10 µH is about right. there's an equation that I haven't yet memorized
pitfalls: layout, layout, layout. if the datasheet offers a recommended layout, try to conform to it. keep traces wide and short, keep components in close proximity. air wiring of minimum length may handle more current than solderless breadboards. don't try to use 60Hz rectifier diodes
if you want a learning experience, the MC34063 is a good, cheap starting point that isn't too application specific. it's slow (100 kHz max), but it's flexible as to topology, available in DIP (if you're in to that sort of thing), has current limiting (add a power resistor), and works over a wide input voltage range
oh look there's your second inductor :DDD

>>1476938
that's actually ac current of a continuously declining frequency

[Anonymous No.1476942]

>>1476938
Take a Fourier transform of your "DC".

[Anonymous No.1476944]

>>1476935
>what's bad about pulse skipping?
It produces low frequency, variable frequency noise. This is usually more difficult to filter away than the normal fixed frequency noise. Also, if the pulse frequency drops to the audio range, your coil (and the capacitors) might start making annoying noise.

>What if the output voltage and current changes?
Then you make some reasonable compromise.

>How come I can use a larger inductor in this case, but not if the load is heavier?
You can, but a pointlessly large (high current) inductor, in addition to being bigger and more expensive, also results in worse transient response. Again, it's a compromise.

[Anonymous No.1476946]

>>1476941
Thanks for info
>second inductor
kek i didn't think he meant inductors on the output stage but again i don't know that much about switchers, so maybe he meant on the output stage

>too large an inductance affects the feedback loop and may cause overshoots during line/load step changes. for most switchers on the order of 1MHz Fsw and a large fraction of an amp, a few to 10 µH is about right. there's an equation that I haven't yet memorized
For my project, potentially, I want it to output up to 10A
So knowing as little as I do, I'm not sure how that'd affect things if I want to output just 10mA for arguments sake

[Anonymous No.1476948]

>>1476944
Ah I see, thanks anon
How do commercial switcher bench power supplies have clean(ish) outputs? Or are they kinda noisy?
I was hoping to use a switcher as a pre-regulator stage before a linear stage

[Anonymous No.1476953]

>>1476946
>So knowing as little as I do, I'm not sure how that'd affect things if I want to output just 10mA for arguments sake
if you're following it with a linear post-regulator (which means what, up to 50W of heat) it will presumably soak up the excess, if any. looks like the 34063 will just skip pulses in case of no load

[Anonymous No.1476957]

>>1476953
My original project objective was to have it always stay 5V or so higher than the output of the linear stage, so that it could supply enough voltage overhead for the pass transistors that will be doing most of the heavy lifting
It also presents other challenges because the feedback loop for making sure the output of the switching stage is ~5V higher than the output of the linear stage might fuck shit up

[Anonymous No.1476959]

>>1476944
Not the anon you're replying too but you seem knowledgeable about switching power supplies. I'm working on building a power supply that can supply +/-350V @ 10mA and +180V @ 200mA from a 12V source. I'm just looking for general advice on topology and perhaps recommendations for a control chip. I was considering a TL494.

I do have one major issue with some earlier designs I played around with, the peak primary current is very high, nearly 20A. This is not acceptable. I'd ideally like a peak primary current of ~5A or less though I'm not sure how to accomplish this. I am willing to reduce the maximum output current I can pull on my 180V rail to no less than 100mA if I absolutely have to. Also it's unlikely the +/-350V rail will pull 10mA of current. The deflection plates it's driving pull a few hundred microamps and it'll probably total to 1-3mA. I want the headroom for now though in case I screwed up and I need the extra current somewhere.

[Anonymous No.1476961]

>tfw I use bang-bang controlled boost converters made with a 555 and comparator
>tfw never even tried to implement variable duty-cycle
Do you guys just buy a multi-purpose converter IC and use it for all your boost, buck, buck/boost, flyback, etc. converters? Ot do you have to get a different IC for each circuit topology and output voltage?

[Anonymous No.1476963]

>>1476961
I'm just trying to learn buck ones for now.
I think you need to get a different IC for each circuit topology, unless you get an IC that can do lots of them.
The basic ones are buck, boost, and flyback. Then there's a few other more exotic ones - cuk and sepic

[Anonymous No.1476964]

>>1476963
Buck-boost (swings both ways) may also have an incompatible topology, having an all-in-one IC would be pretty handy. Most of the ones I see commercially have the FET inside them, but I suspect for a DIP you'd want an external one since you don't have the vias for heat-sinking.

[Anonymous No.1476967]

>>1476964
try to look for a control IC and not a 'regulator' IC.. i think
Also don't buck-boost ones have odd behavior whenever it switches from being buck to boost or what have you?

[Anonymous No.1476970]

>>1476939
>>1476941
>>1476942
simple experiment

[Anonymous No.1476971]

>>1476970
DC current for a finite amount of time isn't DC, it's a bunch of low frequency AC harmonics. Put that simulation on for an infinite amount of time and you're not at a steady state, you've got infinite voltage.

[Anonymous No.1476973]

>>1476970
>simple experiment
if its charging its not DC now is it, genius
was that hard to grasp?

[Anonymous No.1476980]

>>1476973
Odd how direct current and alternating current are actually referring to the voltage, isn't it?

[Anonymous No.1476982]

>>1476980
I don't see how this is relevant
a capacitor that is charging is not a static situation ie not DC. which is why you'd expect to see anything. I'm sorry but your initial presumption is wrong

[Anonymous No.1476987]

>>1476957
you really only need about a volt or two of headroom if you use a pnp pass transistor (check its datasheet's Vce(sat))

>>1476961
I have a small collection of buck and boost ICs around, about two or three of each type. plus the multipurpose 34063s for quick and dirty occasions. and a dozen each of those little modules for bench purposes. flyback controllers are generally divided according to *input* voltage, and personally I haven't needed one so far. also haven't tried a buck-boost topology yet
most ICs for a given topology are adjustable-voltage, similar to the LM317. add the appropriate resistor divider and/or a pot to set the output voltage

[Anonymous No.1476991]

>>1476987
Using NPN
I suppose I can use PNP. But with NPN, the base current can go into the load, whereas with PNP, the base current is sunk by the control circuitry and not used in the load. I wanted to drive the transistors directly instead of just having a resistor between base and emitter

[Anonymous No.1476995]

>>1476971
>infinite amount of time
>>1476973
>if its charging its not DC
>>1476982
>not a static situation ie not DC

[Anonymous No.1477002]

>>1476995
>using memes as argument
its cute but it doesn't get you any points here, my dude

[Anonymous No.1477067]

>>1476347
>Amazon shopping cart,
order it from china you dipstick
or raise your kids properly to not be fucking cowards

[Anonymous No.1477073]

>>1477067
fuck you
the only thing i'm ordering from China is a Mail-Order bride

[Anonymous No.1477082]

>>1477073
you are such a cliché onions boy dad it hurts

ok then, buy your 500% mark up products which the store ordered from china, from the same factory from which everyone else in this thread buys shit in

[Anonymous No.1477098]

>>1476941
>Darlington switch
Absolutely disgusting

[Anonymous No.1477111]

Yo, I know this is a real basic question, but I've come across part of a schematic that I've never seen before and can't get my head around.

Obviously U1A and U1B are different pins of the same op-amp, but what exactly's going on with VR1A and VR1B here? Does this just show that two inputs are going to the same pin, likewise two outputs? Really uncertain what this represents physically.

[Anonymous No.1477118]

>>1477111
Dotted line represents a mechanical connection

[Anonymous No.1477119]

>>1477118
Ah, i understand now. Thank you!

[Anonymous No.1477122]

>>1477082
You're a self absorbed retard that just made a butt load of assumptions and started spouting retard opinions

[Anonymous No.1477131]

>>1477122
Maybe, but at least he didn't just spend $300 on electronic components he could've got for <$50 from the source, then get defensive because he didn't want to feel stupid.
¯\_(ツ)_/¯

[Anonymous No.1477167]

>>1477111
can you link the esp page you got that from? i'm mildly interested

[Anonymous No.1477169]

>>1477167
Sure, it's this one: http://sound.whsites.net/project155.htm

[Anonymous No.1477176]

>>1477131
>didn't just spend $300 on electronic components he could've got for <$50
neither did i.
that was just a temporary shopping list i put together so i could keep track of things i though i might want.
also, it wasn't all just resistors and diodes. i had a bunch of spools of wire, some bread boards, pc boards, lights and sockets, a few arduinos and some arduino add-on boards like relay boards and precise timer/clock boards and quite a bit of other stuff
this night light thing is just an excuse to start doing a simple electronics project
i have other more interesting and complicated projects in mind that i would eventually use alot of this stuff for

[Anonymous No.1477200]

>>1477176
If you take it apart, could you please post the circuit diagram? I'm collecting super simple solutions.

[Anonymous No.1477229]

>>1477200
ok will do. taking it apart is the first thing i'm gonna do before buying anything
of course its one of those stupid security screws i have to go find a special screw driver for

[Anonymous No.1477233]

>>1477111
It's a multi-ganged pot, in this case 2-gang. That's the term you would use if you were search for one to buy. Unfortunately they're kinda hard to find and ate expensive.

[Anonymous No.1477237]

>>1477233
>It's a multi-ganged pot, in this case 2-gang.
Also called a stereo or dual pot. They're pretty common and cheap for obvious reasons.

[Anonymous No.1477244]

>>1477237
Dual pots are probably relatively cheap yeah, I was going off my experience trying to find 4-gang pots once which was a bitch. I think I paid like $10 for four on ebay since none of the major electronics retailers had any in stock.

[Anonymous No.1477250]

>>1477098
What's wrong with darlington?
On that topic, what's wrong with pnp transistors? they seem way less common

[Anonymous No.1477251]

>>1477098
Like it matters, nobody actually uses the internal switch anyway. If you have any sense about you you use the internal switch to drive an external power MOSFET.

[Anonymous No.1477254]

>>1477250
bjts are far less efficient than fets in switchers, especially at low current. pnp and pmos are less common than their N complements because physics (electron vs hole mobility, don't ask me what that means practically) gives them improved performance on the same die area. and because of that, more N type devices get made so they're cheaper too. you'll see a lot of buck switchers use nfets even on-die for the former reason because they're still better even after considering the extra bootstrap circuitry to drive them.

[Anonymous No.1477277]

>>1477254
>pnp and pmos are less common than their N complements because physics (electron vs hole mobility, don't ask me what that means practically) gives them improved performance on the same die area. and because of that, more N type devices get made so they're cheaper too. you'll see a lot of buck switchers use nfets even on-die for the former reason because they're still better even after considering the extra bootstrap circuitry to drive them.
Gotcha
built in bootstrap is fucking nice. doing it externally sounds like a pain in the ass - would the circuit even be fast enough?

[Anonymous No.1477283]

>>1477277
>>1477254
Also, for most hobbyist use, I probably won't be affected by whatever small limitations are imposed by PNP transistors right?

[Anonymous No.1477285]

On this data sheet for a 2-digit seven segment display it's got this pinout, I'm assuming that DIG1:3,18 is to mean that for digit 1 pin 3 is the positive voltage and 18 is the ground, is that right?

[Anonymous No.1477292]

>>1477285

nope, it means 3 and 18 are the common anode, i.e. where you apply the positive voltage.

[Anonymous No.1477295]

>>1477285

nope, it means 3 and 18 are the common anode, i.e. where you apply the positive voltage.

[Anonymous No.1477297]

>>1477295
O

So since there's two of them does that mean I choose one or do both

[Anonymous No.1477300]

>>1477295
If these displays have little PCBs in them, why don't they include the resistors with it? It wouldn't be as multi-purpose, but surely enough people want to run their displays off 5V at 20mA to warrant such a product.

[Anonymous No.1477302]

>>1477297

they're connected together internally, but you use both coz that's best socialist practice, so one wire doesnt get all the current all for itself.

[Anonymous No.1477305]

>>1477300

coz it's silly to limit yourself to one voltage, the same way people who make tampons limit themselves only to the female market.

[Anonymous No.1477331]

why is Physics II - Electricity and Electromagnetism considered to be so difficult?
I've yet to take it, but I've heard nothing good about it.
The only thing I can guess is that it's hard because it uses integration

[Anonymous No.1477341]

>>1477283
not at all. the only situation that comes to mind is if you're making a buck converter then you'll still be better off using an nfet rather than a pfet at a given package size or generation/price.

[Anonymous No.1477344]

>>1477341
Gotcha
all this old school talkage that i've read before has made me really hesitant to use p channel bjts and fets kek
Basically i want to make a power supply that can output up to 30V. Only way I can think of to do that without using an op amp that can swing past 30V is to use a P channel bjt. The liner section anyway. switching prereg section will just be regular n channel shit,whenever i figure out the sort of shit I need for that anyway

[Anonymous No.1477345]

>>1476991
that's fair as long as there's enough voltage headroom on the input to drive the base and you're using rail-to-rail I/O op amps. but you'll need to reflect a control current from the low side (where you do your voltage measurements) to the high side anyway. consider the Sziklai pair, Pic related

>>1477283
ceteris paribus, a pnp will have less gain and be a little slower than an npn. not usually a big deal, especially in small signal work

>>1477300
>but surely enough people want to run their displays off 5V at 20mA to warrant such a product.
what century are you living in

[Anonymous No.1477348]

>>1476991
>>1477345
>pic related
>doesn't pic related

[Anonymous No.1477349]

>>1477345
Or 3.3V, or any single voltage. 7400 series are still used somewhat often if you aim at the hobbyist target market.

[Anonymous No.1477357]

>>1477349
>at 20mA
multiplexed displays are usually driven at a higher current
>somewhat often
so are CMOS. oh, neat, different Vol and Voh specs invalidate the assumptions
resistors are ten for a penny ffs. if you need your hand held that tightly, buy a module, which will almost certainly be cheaper than including power resistance inside the display

[Anonymous No.1477364]

>>1477357
It's for space saving, not anything else. And ~3.3V power rails are pretty common for microcontrollers, just have a few (~3 or 4) common resistances that you can mix and match for wanted currents and voltages. 150Ω, 75Ω, etc.

[Anonymous No.1477368]

>>1477331
Anyone who really knows and understands anything about electromagnetism is far too autistic to relate that information coherently

[Anonymous No.1477381]

i experimented with using a ULN2004's clamp diodes as a bridge rectifier because i had some spare pins. its case got up to 85C while passing 114mArms. i calculated that it was dissipating 290mW so its die temp got to around 100C per the datasheet. seems pretty doable. the biggest error there would be my trash thermocouple.

[Anonymous No.1477407]

>>1477345
>>1477348
kek that's roughly the schematic I drew too. It's the only way I can think of to efficiently power NPN emitter output without also sinking few tens of milliamps for no reason when i want the output almost zero volts
What's the point of the 220 ohm resistor between the base and emitter of the 2n3055?
Is the 4k7 resistor on the base of the MJ2955 to take off some voltage from BC237?

[Anonymous No.1477410]

>>1477407
>also sinking few tens of milliamps for no reason when i want the output almost zero volts
this is an interesting concern when you're already using a linear postregulator that'll have a 1% efficiency with a near zero output

[Anonymous No.1477411]

>>1477381
Maybe you should just get an actual bridge rectifier package, faggot

[Anonymous No.1477412]

>>1477410
>1% efficiency with a near zero output
How do you figure?
I want to put extra efficiency onstraints on it to see if i can do it anyway

[Anonymous No.1477425]

>>1477364
>It's for space saving,
so are Pic related. if you're not using these, I assert you aren't really worried about size reduction

>>1477410
he's using a tracking buck pre-regulator with a small offset

>>1477407
220 ohms helps the 3055 turn off quicker in case of a sudden negative change in load
yes, 4k7 absorbs a bit of voltage. the resistor values were first estimates and probably wouldn't survive a second round of analysis. you could probably lower them if Miller capacitance were to become a problem

>>1477412
I think it was a ballpark estimate but not too far off really. assume a system output voltage and an output current of your choice. figure your preregulator out voltage at that system output voltage. math it out, with the understanding that P = IE in any power element
say 0.1V @ 10A, pre-reg offset is 1.5V
so pre-reg output is 1.6V
neglecting pre-reg efficiency, Rbepass, and Ibdriver, Pdpass = 1.5V * 10A = 15W, while Pload = 0.1V * 10A = 1W so Ptotal = 16W, therefore 1W/16W = 6.25% just for the post-reg
also noting that power transistors tend to have poor gain, so the driver could be passing several hundred mA at 0.6V which you might or might not care about
one more note on power supply design: if you really want to test the fuck out of it, take a rat-tail metal file and run it across the output terminals every which way at various points of operation. if it survives the show of sparks will be very satisfying, probably get a bunch of likes on youtube

[Anonymous No.1477428]

>>1477425
Hopefully I'll rarely be using 0.1V 10A output. But that is a nice thing to reach for. Maybe in mk3 version
Nevertheless though that's necessary losses. me buring 50mA when its totally off is unnecessary if i use pnp so pnp is what i'm going with
I hear you with the file though. that's to test a bunch of transient conditions, yeah?

[Anonymous No.1477443]

>>1477428
the loss across the pass transistor should be the same at any voltage, if the pre-reg tracks exactly. maybe in the operating version you can get that down a bit closer to Vbedriver + Vcepass(sat)
>transient response
yep, also tests for tendency to oscillate at various frequencies as you scrub the teeth of the file across it while watching the scope

[Anonymous No.1477450]

>>1477411
every IC on this board is going to be a DIP14 or 16 and you can't stop me

>>1477412
your linear regulator has a dropout voltage. for npn regulators that's at minimum 1.2v. so if you want to supply 0.1v you'll need to buck to 1.3v (more realistically 2v+, and your postreg PSRR will get better the higher that is). so you're burning 1.9v of your 2v buck output or thereabouts. 1% was hyperbolic, with those numbers as an example your efficiency would be 5%.

[Anonymous No.1477454]

>>1477450
what about discrete diodes

[Anonymous No.1477457]

>>1477454
i like the idea of fitting them into an existing component.

[Anonymous No.1477469]

>>1477450
>efficiency
as stated above, necessary and therefore acceptable losses

>>1477443
Oh don't worry i'm already looking at Vce sat on datasheets. and then i'd probably, ideally, test each transistor I was going to use to see what my actual saturation figures are

[Anonymous No.1477484]

How do I maximize my odds of getting companies like TI, Analog Devices, Maxim Integrated, Linear Technologies, etc. giving me free samples?

I'm a nobody with no career in EE and am only a student in a shitty community college. I'm figuring using the college email will probably slightly increase my odds but I'm thinking I'll still get denied. If they do somehow send me free samples how many chips can I get?

[Anonymous No.1477487]

>>1477484
Aren't the prices on ali good enough? Or do you want samples of some pricy ICs, like FPGAs or RF electronics?

[Anonymous No.1477489]

>>1477484
>I have no plans for any projects but help me accumulate free shit
no

[Anonymous No.1477493]

>>1477487
I do want samples of some rather expensive analog ICs and the last time I used ali I had some rather frustrating issues with my payment being frozen by my card company because I was paying someone in China and they thought it was fraudulent. It wasn't but the guy I spoke to on the phone from mastercard was very reluctant to let me allow the purchase to go through. It wasn't fraudulent and I received the thing I purchased a few weeks later and my account was never drained of more than I was supposed to pay but the whole thing experience sucked. On top of that I don't wanna have to wait a month or two to build the shit I want.

>>1477489
I do have plans for projects I want to build but I need some components to prototype stuff and I don't have much in the way of cash. The free samples would allow me to test and see if the circuit I've designed actually worked and if it did I would go and buy more and then design a proper PCB for the final design. Unfortunately these are just personal projects not stuff I intend to sell or anything but it's not like I just want free stuff to fill my parts drawer.

[Anonymous No.1477495]

>>1476887
Thank you very much for providing the steps. I greatly appreciate it! I am going to go back and review RC circuits. After that I will revisit this problem and attempt it once I have a better understanding.

[Anonymous No.1477509]

so im trying to learn about radio electroincs, but im having a hard time finding beginner resources. anyone have anything?

[Anonymous No.1477516]

>>1475345
NICE!

Challenge, someone find OPs webserver , hack it and shut off his heat in the middle of winter.

[Anonymous No.1477523]

>electronic load
>hmm, being able to multiply and/or divide to get values for power or some such would be nice
>google
>apparently its tedious and requires matched transistors and shit
>there are dedicated ICs
>ohcool.jpg
>digkey
>10 dollars per IC
well fuck

[Anonymous No.1477524]

>>1474986
Thanks, looking into a syllabus for Electromagnets and Microwaves they used ADS, CST, and EagleCAD. What are some FOSS alternatives to this? I already use KiCAD so I can get around Eagle

[Anonymous No.1477528]

>>1474986
RF is disagreeable because your'e starting to fuck with wave equations and shit

[Anonymous No.1477555]

>>1477524

found a torrent for ADS - dunno if legit, downloading now. Eagle has (or had until recently) a free version.

[Anonymous No.1477559]

>>1477523
>can't divide with a resistive divider
>can't multiply with an op amp
did you consider a couple of $1 OTAs?

[Anonymous No.1477566]

>>1477559
>sense the current
>sense the voltage
>divide voltage by current to get resistance for constant resistance load
>divide with a resistive divider
okay there

[Anonymous No.1477579]

>>1477509
I'm assuming you mean simple hands-on stuff without much math. Check hobbyist books from the sixties or seventies.

[Anonymous No.1477580]

>>1477509
I got mine from qsl.net. Booklist and links at: http://radio.radiotrician.org/2017/01/rf-design-books.html

[Anonymous No.1477599]

electronicloadfag here
if an electronic load is set to constant current load, and you attach a constant current source to it, and the settings on the load and the supply don't match, wouldn't the control circuitry of the load and the supply go completely in one or the other direction (depending on which one is set higher than the other) causing the power supply and the load's protection circuitry to trip?
And same for constant voltage load + power supply outputting constant voltage

[Anonymous No.1477603]

>>1477599
I was wondering the same thing, but it all comes down to the unideal behaviour of a real current source/sink. A current source will have a maximum voltage, and if the effective resistance across it is too high it will just act in constant-voltage mode, which will occur when the load wants a lower current than the source. I assume the opposite happens when the load is lower than the source, though it isn't exactly intuitive.

[Anonymous No.1477608]

>>1477603
>A current source will have a maximum voltage
Well true enough
>CC load: set 1A
>CC supply: set 1.5A
the supply attempts increasing voltage until such a time that it passes the current that it is set for, while the electronic load reduces load (increases the effective resistance) to try and reduce the current that it senses passing through the load?
The load would eventually be completely off, while the supply would keep upping and upping the voltage until it hits its max limit. If the voltage source could supply 1000V, couldn't the case where the voltage exceeds the rating of the mosfets in the electronic load and you let out the magic smoke?
Obviously, I don't have nor am i making a supply that can supply 1000V but just playing with hypotheticals here

[Anonymous No.1477609]

What's the max current you can safely put through perfboard?

[Anonymous No.1477612]

>>1477609
Depends on your idea of safety, but you can expect 10C temperature rise at 3A or so. I would go with that.
Doubling the current to 6A would result in around 40C temperature rise and doubling the current again to 12A should be enough to cause immediate problems.

[Anonymous No.1477613]

>>1477612
Not him but wouldn't you just need to have super wide 'traces' on perfboard to make it work?
on production PCBs, don't they handle double digit currents by having like 10cm wide traces or some shit?

[Anonymous No.1477614]

I don't have the a or a schematic beside me right now, but recently I bought a wireless doorbell which I modified for some of my purposes. The transmitter was powered by single 12V primary battery, but when I looked inside at the pcb it was immediately regulated down to ~3.3V.
Why would you do that instead of using more standard (and possibly rechargeable) batteries?
Does maybe regulating down the voltage allows you to draw more current? Because I thought those regulators (or at least some of the older ones) basically wasted the power from excessive input voltage.

[Anonymous No.1477615]

>>1477614
Because switched mode power supplies are incredibly efficient, and yes switched mode supplies allow you to draw more current at 3.3V. If it was a linear supply it wouldn't

[Anonymous No.1477616]

>>1477509
I found a good one with some fundamentals for beginners

Essential & Practical Circuit Analysis: Part 1

https://www.youtube.com/watch?v=sqxzQkAdJm0

[Anonymous No.1477618]

>>1477615
>switched mode power supplies
>If it was a linear supply it wouldn't
thanks

[Anonymous No.1477620]

>>1477609
Be the first to find out.

[Anonymous No.1477663]

>>1477613
One common way to determine the maximum current is to decide the allowed trace temperature rise (usually some convenient value like 10C) and see what current is enough for that. Since the temperature rise is caused by I^2R losses, doubling the current quadruples the losses. On the other hand, doubling the track width only halves the resistance and thus losses. Power dissipation from a track isn't directly proportional to area and you end up needing more than twice the width to handle twice the current.
Another thing is that even that 10C temperature rise means quite high voltage drop and power dissipation. This isn't always acceptable, forcing you to use wider traces. Also, if the track is carrying high frequency current, you might need to take the skin effect and inductance into account.

One track width calculator: https://www.4pcb.com/trace-width-calculator.html

[Anonymous No.1477679]

I have a meter (a bunch of them) in the picture which measures heat and stuff and it has an infra port on it where you can somehow use IR to read the data it measured.

Do you think it would be possible to read such data with arduino? What sensor i would need for it?

[Anonymous No.1477687]

>>1477679

[Anonymous No.1477690]

Looking for good EM simulation and Antenna deign software that is FOSS. Came across this https://en.wikipedia.org/wiki/Comparison_of_EM_simulation_software
http://emlab.utep.edu/opensource.htm
But wondering if there is anyone out there with a bit more knowledge can give a recommendation.

NEC, openEMS fail to have GUIs which isn't that big of a deal breaker, but will like to fall onto something else. Elmer looks nice, but not really just for EM as it's an entire physics suite lacking tutorials on the EM part.

>>1477555
To be FOSS is to be free to distribute, edit, and open source. Sadly ADS and Eagle aren't that.

[Anonymous No.1477692]

>>1477679
At least:
- Infra-red sensor diode
- operational amplifier circuit

[Anonymous No.1477701]

>>1477692
What sucks is that is can find almost no information on this on the google, it seems that the irda uses some retarded code and many meters use their own blinkie code to cummunicate, it will be probably easier to just replace the meter with a flow sensor

[Anonymous No.1477737]

What does the 1x/10x thing mean on oscilloscopes, and why does the 10x setting generally have higher bandwidth?

[Anonymous No.1477740]

>>1477737
it switches between resistorts
10x is 10 megaohm i think, it removes interference

[Anonymous No.1477741]

>>1477740
What are the benefits and drawbacks of using a higher resistance probe? I mean, it obviously has to have drawback in some situations, otherwise people would only ever use the 10M setting because of its higher bandwidth.

[Anonymous No.1477742]

>>1477737
it's gain.

[Anonymous No.1477749]

>>1477741
Not loading down your circuit under test

[Anonymous No.1477752]

>>1477737
10x probes have 10 times higher input resistance and 10 times lower capacitance, but they attenuate signal 10 times.
If you use a straight probe, then the device you're measuring sees all the oscilloscope and probe cable capacitance (around 100pF).

[Anonymous No.1477757]

>>1477348
>massive gain
>no compensation
that turd is going to oscillate, I guarantee it

[Anonymous No.1477838]

I got a bunch of these rodent repellers.
But since the chink made a mistake and sent one of them with a US plug which is useless to me, i decided to open it to see how it works and....

..was i just fucking scammed?

[Anonymous No.1477841]

>>1477838
Yes, yes you were.

[Anonymous No.1477845]

>>1477838
you don't have any clue how these things work do you? the idea is that it burns your house down which will effectively repel the rodents. stupid gwailou.

[Anonymous No.1477852]

>>1477741
Using x1 mode gets you more sensitivity for measuring tiny voltage. x10 gives you better bandwidth and less loading.

[Anonymous No.1477853]

>>1477852
what is tiny?

[Anonymous No.1477854]

>>1477757
>that turd is going to oscillate
very probably. loop compensation is left as an exercise for the designer

>>1477838
perhaps rodents are extra sensitive to 60Hz flashing lights
if not, then holy kek

[Anonymous No.1477855]

>>1477853
depends on your scope
in x1 mode, multiply the scope reading x1
in x10 mode, multiply the scope reading x10

[Anonymous No.1477857]

>>1477838

looks like a perfectly good night lite. it'll help the rats find the cheese.

[Anonymous No.1477858]

>>1477841
>>1477845
>>1477854
The package claimed it uses the electrics wires from the outlet as transmitters and turns them into an impenetrable force field against mice.

I thought it maybe uses the cap with the led to oscillate in some sort of RF waves way reacting with the water inside of the mice and bugs making then uncomfortable, the way other modern repellers do it

[Anonymous No.1477860]

>>1477858
And this drivel didn't sound like a scam to you?

[Anonymous No.1477861]

>>1477860
not really since this kind of technology actually exists and is used

[Anonymous No.1477863]

>>1477690
>To be FOSS is to be free to distribute, edit, and open source.

we got a purist here. i found pirated copies of both CST Studio and ADS. it's not likely you'll find open-source equivalents that are anywhere near as good.

[Anonymous No.1477864]

>>1477860
i dunno, I can get pretty nauseous under cheap LED buib flicker

[Anonymous No.1477865]

>>1477858
In order for a circuit to oscillate it needs to fulfill both conditions of the Barkhausen stability criterion. The circuit needs to have a total loop gain greater than or equal to 1, AND the loop must provide 360 degrees of phase shift. This capacitor LED nonsense does neither. The wires do not act as an antenna any more than your home wiring acts as a large low power 50/60Hz antenna.

[Anonymous No.1477867]

>>1477858
what was the actual retail price of that fine item?

[Anonymous No.1477876]

Perhaps it's not a hoax?

[Anonymous No.1477881]

>>1477876
Oh but there's a filtration capacitor across the LEDs anyhow so they won't flash. Still, this RF method shows a little promise, though it's doubtful that these models do anything in particular.

[Anonymous No.1477945]

>>1477679
I really want something like that, connects to a garden hose, can turn the valve on and off with a microcontroller.

[Anonymous No.1477950]

>>1477945
Latching solenoid valve + half-bridge + MCU? Flow meter and thermometer not included.

[Anonymous No.1478006]

I'd like to fabricate my own pcb for a custom keyboard, and the microcontroller I'm going to be using is an atmega32U4. The only issue is that its a 44tqfp package and I'm not sure how to program it. I have a USBtinyISP, is my only option to buy a weird ZIF socket like in pic related, or could i put headers on the board to program it once its soldered on there? There are also pseudo-arduino pro micro boards with the chip on there, would it be possible to program it mounted on the arduino board, then desolder it and attach it to my final pcb? These boards are also cheaper than actually buying just the chip, so thats another benefit. I've never designed a pcb with a proper bare microcontroller in such a tiny package so I'm a little lost.

[Anonymous No.1478050]

>>1477858
anon, force fields dont exist....

[Anonymous No.1478053]

>>1478006
add an ISP header

[Anonymous No.1478061]

>>1478050
Gravity is a force field.

[Anonymous No.1478073]

>>1478061
but he said "impenetrable" force field.

[Anonymous No.1478076]

>>1478073
Either way it's a descriptive term anyhow. A real, effective electronic pest control system that is unnoticeable to humans could be said to act as an impenetrable forcefield.
https://en.wikipedia.org/wiki/Electronic_pest_control

[Anonymous No.1478078]

>>1478053
So you mean to say I should just add header pins to my board, so then i can program it once its assembled? That sounds like the most elegant method at least. The idea of not being able to test it before its all together scares me, but perhaps its still the best way.

[Anonymous No.1478080]

>>1478078
I'd do more primitive testing with an arduino-style board (wipe the bootloader) or a DIP MCU of a similar enough family. The board sounds like the more likely option since I've never heard of a DIP-40. Once you've got proof of concept down, you should be able to make the PCB and tune it via ISP without too much worry.

[Anonymous No.1478083]

>>1478078
sure. there are fancier ways to do it, like bare pads and a fixture of pogo pins, but a pin header is more than sufficient for a one- or two-off
you could write some test firmwares to help yourself out. since it's a 32U4 you can pretend it's an arduino leonardo (withuot bootloader) and easily create tests for whatever ails you, then upload them to the board with avrdude and the tinyISP

[Anonymous No.1478088]

>>1475325
MEASURE BEFORE DISCONNECTING
DISCONNECT
MEASURE AFTER DISCONNECTING

[Anonymous No.1478093]

>>1477867
$4

[Anonymous No.1478094]

>>1476920
Try https://waset.org/publications/12213/artificial-voltage-controlled-capacitance-and-inductance-using-voltage-controlled-transconductance

Also yeah, transconductance as someone mentioned already, but also gyrator circuits.

[Anonymous No.1478099]

>>1476959
Why can't you use a normal switch-mode design? The MOSFET driver current would be pretty high, but if you use a cap in parallel with an inductor, and then use a comparator to initiate a charge cycle, sounds like it would work as long as everything is voltage rated (and the inductor is amp rated).

[Anonymous No.1478101]

>>1478088
>>MEASURE AFTER DISCONNECTING
for what purpose

[Anonymous No.1478103]

>>1478101
so you don't do what the guy he's replying to did

[Anonymous No.1478104]

>>1478103
if it's not under power before disconnecting it won't magically energize after

[Anonymous No.1478105]

>>1477254
"holes" are vacancies of electrons, so the only way they can "propagate" is by an electron filling the hole randomly. Being positive charged helps, but it's not actual charge like an electron, it's push response by the electrons around the hole. Costs more time and more energy (a hole isn't excess in a sea of charge, so there's always slightly more energy needed than negatively doped regions)

In practical terms positive stuff is always less used and more expensive for a given application, and worse, once you run compound devices other than a diode, you at minimum double your, uh, "charge resistance". It will take at least twice as long for a PNP barrier to transfer an electron as an NPN device.

[Anonymous No.1478109]

>>1478104
If you only measure afterwards you could have a failure of your voltmeter and you wouldn't know it.

If you're working with 115-230 house current behind a breaker you can be okay if your tool absorbs the shock and the breaker trips properly.

When you're working with 230-460-480 volt industrial sources with no breakers, you are likely to die and it will be quick and painful. You'd be lucky if your limb exploded before your entire body was cooked.

[Anonymous No.1478111]

>>1478109
just hire a mexican to walk around with you and touch any wires first before you touch them

[Anonymous No.1478113]

This might be a stupid question but what advantage does a DAC chip have over just putting an R2R ladder on my microcontroller pins? I feel like I'm missing something here.

For example here's a $12 12 bit DAC with a parallel interface. Couldn't I do the same thing for a lot less with a set of 24 high accuracy resistors and maybe an op-amp for good measure? I know more individual parts means more manufacturing costs but that's not the only reason is it?

[Anonymous No.1478119]

>>1478113
better accuracy and (usually) guaranteed no missing codes. where you gonna get 0.02% resistors for a 12-bit DAC? 0.05% are about 64 cents each in 10-lots
also driving them with a low-resistance source is another source of missing codes and non-monotony
do you really need that fast and that deep of a DAC?

[Anonymous No.1478120]

>>1478113
Enjoy 5-10 bit accuracy unless you're laser trimming:

https://en.wikipedia.org/wiki/Resistor_ladder#R%E2%80%932R_resistor_ladder_network_(digital_to_analog_conversion)

[Anonymous No.1478122]

>>1478113
That's a 20MHz *current* DAC with a bunch of weird features. Did you even check the sheet?

[Anonymous No.1478123]

>>1478113
Can't you just use high-frequency PWM and a low-pass filter as a DAC? A bit like a class-D.

[Anonymous No.1478125]

>>1478123
"Analog" includes digital signals. Class-D amplifiers also require feedback, either to the originator or the driving circuit. You could pre-build it with a current-sink, in a sort of sample-and-hold configuration, but it'd get messy pretty quickly in a non-integrated design.

[Anonymous No.1478128]

>>1478125
and then there's the small matter of the application frequency range which anon hasn't actually specified

[Anonymous No.1478129]

>>1478125
But that's a class-D that's meant to drive a load, they need feedback because of the less-than linear output impedance. If you're just using it for small-signal stuff then buffer it after the filter with an op-amp, using the op-amp to make a nice -40dB filter without inductors.

[Anonymous No.1478143]

>>1477861
No, dude. It's just another magical mouse repeller. They don't work.

[Anonymous No.1478153]

>>1478123
You'd normally use delta-sigma conversion rather than PWM. Most digital audio DACs work this way.

[Anonymous No.1478162]

What chink product should i look for if i want to power arduino (5V DC) directly from the 230V AC mains power? I want ideally some small wart which can be easily hidden inside walls.

[Anonymous No.1478163]

>>1478123
The DAC needs to be capable of outputting waveforms from a few hertz to a few megahertz so a low pass filter isn't really an option.

[Anonymous No.1478174]

>>1478163
A class-D amplifier outputs at a constant frequency which is significantly larger than the highest wanted frequency, and filters this PWM square wave into a soft audio wave. In this case having this low-pass filter is not effected by the low output frequencies you want, though it is effected by the MHz range you're seeking; either you run it at a frequency in the range of hundreds of MHz (not going to happen without spending a bunch) or run it at maybe 3 times your macimum frequency and filter+buffer the fuck out of it. This not infeasible, but unlikely to be an effective solution. I think the same frequency limitations apply to >>1478153's ∆∑ conversion, though I'm unsure what (if any) filtering they use.

[Anonymous No.1478192]

>>1478162
Found this.
https://www.aliexpress.com/item/New-Power-Module-AC-DC-Converter-110V-220V-230V-to-5V-Isolated-Switching-Power-Supply-Board/32901800794.html?spm=2114.search0104.3.9.438078cdGw5oBq&ws_ab_test=searchweb0_0,searchweb201602_5_10065_10068_204_318_319_5727220_10059_10884_10887_10696_100031_320_10084_10083_10103_452_10618_10304_10307_10820_532_10821_10302_5727320,searchweb201603_60,ppcSwitch_0&algo_expid=50fa4fed-fec4-449a-953a-ebe5d7305ab8-1&algo_pvid=50fa4fed-fec4-449a-953a-ebe5d7305ab8&priceBeautifyAB=0
Is there any better solution?
This is only like 2cm big so i think i can't get any smaller

[Anonymous No.1478204]

>>1478192
You posted your private session data. Do you get chinkcoins for product placement?

[Anonymous No.1478229]

>>1478119
You can get R-2R ladders in SIP or DIP resistor arrays like pic related but I don't know how high accuracy they go. Pic is 2% is which is two orders of magnitude worse than the 0.02% you're talking about.

[Anonymous No.1478337]

Hmmm.. what is going on here? I have arduino mini and i connected it to a power source adapter, putting 12V on the raw pin.
The arduino worked for couple of seconds and then it turned off and didn't turn on again.
I thought it might be bad wire connection, so i went ahead and soldered the wires to the pin holes.
Then i connected the arduino to the adapter again and same story. Worked for couple of seconds and then turne off and won't turn on again.

I thought it maybe can't handle 12.5V which would be stupid since it is rated for 12V officical and up to 16V in unofficial specs. And arduino uno can take the same adapter just fine.

So i tried to connect it to a USB FTDI programmer which supplies 5V over the vcc pin from the USB port and my computer displayed a warning that the device needs more juice than it can provide (which is bullshit i can power arduino uno + a motor connected to it from an usb cable just fine)
also when i push the ftdi programmed into the arduino mini it immediately turns off.

It almost appears as if there was a short or some shit, but that can't be because nothing smoked, or smelled and like i said the arduino is rated up to 16V and i only connected 12.5V

So what the fuck is going on?

[Anonymous No.1478371]

>>1478337
Sounds like a short somewhere to me. Check with a DMM if when the MCU shuts off what the voltage after the regulator is.

[Anonymous No.1478373]

>>1478371
i checked the wires i soldered and there is no short there
if the short is somewhere on the chip i can't repair it anway since it's so fucking tiny nor would i bother to, the things costs $1.5 so i will just throw it out like the piece of trash it is and replace it

Buying minis was a fucking mistake. I will stick to unos so i don't have to solder shit

[Anonymous No.1478381]

>>1478373
Did you measure the voltage of the regulator?

[Anonymous No.1478383]

>>1478381
it's not the regulator, when i connect FTDI it powers the VCC directly bypassing the regulator anyway and that didn't worked also

[Anonymous No.1478387]

>>1478383
Then you've probably got a faulty IC. Get half a dozen, chances are you won't have another bad one. I've got a couple micros and they're perfectly functional, I don't see any need to use unos. Once you solder the pin headers on (with plenty of flux) there's no extra soldering to do, and they're much smaller.

[Anonymous No.1478392]

>>1478387
it pisses me off chinks don't offer minis with pins soldered on. doing it manually is tedious as fuck

[Anonymous No.1478398]

>>1478392
It's not that awful, provided you have a temperature controlled iron and some extra flux.

[Anonymous No.1478402]

>>1478398
nah i just have an iron with a power cord coming out of it and my solder has flux built in

[Anonymous No.1478403]

>>1478402
You can get power-cord irons with a small temperature knob on them, that's what I've got and it's working fine.

[Anonymous No.1478451]

>>1478403
For what purpose? You just need it to melt solder

[Anonymous No.1478464]

>>1478451
So you don't fry the shit out of everything. Irons work best at a particular temperature. I've got a non-temp-controlled iron and if left on for a minute without touching to solder it gets up past 400°C and drops below 200°C with heavy use. The temperature is far too dependant on the thermal load. A temp-controlled iron will have a high maximum power rating that it turns on and off in accord with the thermocouple in the tip.

[Anonymous No.1478475]

where do you guys go online for surplus electronic components?

[Anonymous No.1478484]

>>1478475
>surplus
interested in this also

[Anonymous No.1478564]

>>1478464
>tfw want a jbc station but they cost a fuckton

[Anonymous No.1478569]

Want to make a circuit that uses momentary on switch to do on-to-off and off-to-on is such a pain in the ass. especially one that doesn't oscillate or go back to a previous state if you hold it down. no wonder people do switch debounce and enable the functionality in software fuck

[Anonymous No.1478570]

>>1478569
just use a microcontroller :)

[Anonymous No.1478590]

>>1478570
Goddamn it kek

[Anonymous No.1478592]

>>1478475
surplus what exactly? try
halted.com
allelectronics.com
sciplus.com
google.com

>>1478569
try putting a cap between the switch/pullup and the comparator, and make sure you have enough hysteresis
or TTP223 touch ICs have a toggle/jog mode select input
or may as well just use an ATtiny4 or something like that

[Anonymous No.1478595]

>>1478592
I suppose I can use logic gates. I have very meager practical experience with logic gates though.
Also I have no idea how to program an at tiny. don't those ICs also need a bunch of supporting components? Every time i look at an arduino there's a few ICs, a crystal oscillator, bunch of resistors and capacitors

[Anonymous No.1478599]

>>1478595
>>1478592
Also what's this about a comparator? can you show me the circuit you have in question?
I've already looked at the eevblog one and that one definitely oscillates or goes back to the previous state for one of the states. even some of the improved ones in the comments. just can't wiin

[Anonymous No.1478615]

>>1478595
not really, the important aspect of the ATtiny is the onboard oscillator and counters which help you wait out the switch's bounce time
arduinos need the supporting components because of the serial communications and stuff requiring a specified precision that the onboard oscillators don't provide

>>1478599
nm look what I just found: two Rs, a C, and a 555
http://www.555-timer-circuits.com/switch-debounce.html
I suggest you consider the debouncing separately from the toggling. a D FF circuit with ~Q to D and the debounced button to CK should do it

[Anonymous No.1478648]

Dave Jones' latest video on that 3c MCU is pretty interesting. Not only because of the MCU itself since it looks like getting a working development hardware+software set for it would be a bit of a task, but because he shows a site for buying obscure import electronic components that aren't even on Octopart.
https://lcsc.com/

[Anonymous No.1478669]

>>1478648
>lcsc
pretty based, have found this to be the only source of English datasheets for a few of the Chinese jellybean components

[Anonymous No.1478671]

>>1478669
From what I saw on the video they had Chinese datasheets for that micro, so you had to go to the manufacturer's site to get the English datasheet.

[Anonymous No.1478672]

If /diy/ is so slow, why don't we increase the bump limit? Maybe for registered generals only (give the janitors something else to do)?

[Anonymous No.1478689]

>>1478671
>for a few of the Chinese jellybean components
... just not this one
that little micro does look intredasting. I guess who needs flash if they're only 3 cents each and frankly smarter than the average 12-bit PIC

[Anonymous No.1478713]

>>1478689
>smarter than the average 12-bit PIC
Those things are pretty old by today's standards, not that I've used one more than twice. For the price there are certainly cheaper MCUs.

Open question: apart from bloatduino, what microcontroller (family) is the easiest to program, both the software and required hardware? Could include the ATmegas/tinys through other methods than the ardy bootloader. I don't think PIC is in the running since their ZIF-socket PICkit 3s go for $10 minimum on ali.

The Arduino environment lets me program a digispark board with the USB data lines running directly into the IC which is really convenient, but I'm assuming the bootloader wastes quite a bit of extra program memory.

[Anonymous No.1478718]

>>1478713
how fast are arduinos anyway?
Like lets say you program it so that one pin goes high when it detects this other pin goes low from external sources
How soon after sensing a low condition will it set the appropriate pin high?

[Anonymous No.1478719]

>>1478718
Sounds like a job for an oscilloscope that isn't my $15 DSO138.

[Anonymous No.1478725]

>>1478713
>cheaper
than a 10F200? other than the Padauk devices Dave was just hustling?
sometimes the SOT23-6 package is nice. I've seen them implementing a timer function in coffee grinders, with an analog input for the time control knob, a digital in for the start/stop button, and the triac trigger output
>easiest to program
that's kind of a hard question. ESP8266 requires only a USB-to-UART chip with 3.3V levels, and includes a gcc cross-compiler, but the dev kit could be a little weird and it's kind of a power hog
AVR also is a gcc cross target and avrdude supports a bunch of programmers (you could flash a USBasp programmer firmware into a digispark, apparently)
STM32 has a pretty decent configurator that lets you set up the chip's internal and external resources and clocks via a GUI and generates C code for a few different IDEs
STM32 ROM bootloader facilities depend on the exact chip but the common F103 that you can get boards for about $2 in singles has a UART bootloader and can also be programmed via ARM SWD which is almost commodity

>>1478718
>arduinos
if you're in a hurry you probably would want to use bare metal AVR assembly and interrupts, rather than putting code in your main loop like
>if (!digitalRead(x)) digitalWrite(y, 1);
without looking at a datasheet, I guesstimate that it would take 8-10 clock cycles from edge to edge, maybe more if you need to save/restore regs etc, or if you disable interrupts for any other reason
if you're really in a hurry, maybe you wanted discrete logic or programmable logic

[Anonymous No.1478726]

Does anybody know how I can get my hands on tunnel style heatsinks? looked on google, ebay, reddit. got nothing i wanted
There were one or two hits on google that sold them as one of two halves with a weird groove or some such down the middle. but I'm looking to attach TO-247 transistors on them so that's no bueno

[Anonymous No.1478728]

>>1478725
>without looking at a datasheet, I guesstimate that it would take 8-10 clock cycles from edge to edge, maybe more if you need to save/restore regs etc, or if you disable interrupts for any other reason
if you're really in a hurry, maybe you wanted discrete logic or programmable logic
Just wondering how much I should rely on arduino or something for overvoltage/overcurrent protection
>programmable logic
like what?
I'm definitely going to try to implement all that shit in hardware if I can

[Anonymous No.1478729]

>>1478725
The Padauks look like a pain to program, plus they're one-time programmable. A little too inflexible for just messing about, I could easily go through 100 of them just refining code.

[Anonymous No.1478736]

>>1478728
that depends on how fast you need to react to the fault, which depends on your application. you'd need to quantify that. you'd also need to know whether any libraries you're using are turning off the interrupts, and for how long at most they would be out, which you would add to the ~8 cycle latency
if you don't know or there is any doubt, probably better to not try to do it in software

>>1478729
meh, not too much of a pain for what they're meant for. probably a pain for messing about, but they could be really nice for when you have an easily and well defined task you want to do (like the aforementioned timer function, or debouncing a switch, or power supply sequencing, or spinning a stepper motor, or...) and you don't necessarily want a special purpose or very large chip for it. great for pcb business cards

[Anonymous No.1478744]

>>1478736
>pcb business cards
What do people put on these? Flashing LEDs? Could make a nice VU meter business card with one I bet.

[Anonymous No.1478747]

>>1478204
It's not private anymore, is it now?
But it doesn't matter anyway, if you feel like logging in and ordering me some nice shit then go ahead.

[Anonymous No.1478750]

>>1478747
clean up your links before you post them next time pl0x. anything that looks like a referral link is strongly discouraged

>>1478744
sure, if you can find an SMT mic that takes well to calibration
flashing LEDs or small arrays of LEDs are one popular choice. some place a USB connector on one edge and serve a small filesystem via the mass storage class, or provide an HID device that types something. my favorite is Pic related found in a ddg image search

[Anonymous No.1478755]

Do you think a 2.5W laser could be used for drawing silk screens on PCBs without damaging the traces?

[Anonymous No.1478756]

>>1478755
Buddy, you're going to need something with a lot more power behind it than 2.5W if you are hoping to damage copper
As for drawing patterns, I dont' know

[Anonymous No.1478758]

>>1478750
This isn't csg, it might not be common knowledge how to trim an ali link. That might be something to add to the OP.

Anyhow, the massive coin cell on that card is a bit unappealing, I wonder how far a calculator-style amorphous solar cell would go to powering such a card? Could recess it into the board if it stick out too much. The crystal on there is also probably not very necessary on something that requires no precise timing, an RC oscillator should be possible instead, maybe even a tiny SMT inductor, or one printed on the PCB itself for an LC oscillator if the native circuitry of the MCU can't handle it. Alternatively, insetting the axial crystal into the PCB would be fairly flush.

These things look pretty cool, though impractical when sticking out the side of a computer. NFC cards are also pretty appealing (though impractical for me since my phone doesn't support it), like the N-O-D-E store's NFC key: https://n-o-d-e.shop/collections/frontpage/products/nfc-key which has two 8kB NFC storage ICs. Buttons are pretty big though.

[Anonymous No.1478762]

>>1478758
>csg
most people seem to pick it up along the way. adding anything into the OP is tantamount to cutting into muscle. what would you have me cut?
>massive coin cell
yeah, kinda, but meh, to be added by the user for demo purposes and otherwise not take up obnoxious room in a wallet
>The crystal on there
that's a tilt sensor
>practical
they're demos. they're the ideal sort of application for those 3¢ micros. they're not meant to be practical. like any business card, they're meant to amuse, be transferred to address book, and filed away

[Anonymous No.1478767]

>>1478762
What was cut for the first 5 rules? If there's room LCSC could be added to the part sourcing, though Octopart should probably have a higher priority since it links to Mouser and the rest already.

I guess I'm a little stuck on the "it goes in your wallet" angle, where wallet space is valuable. Those credit card multitools have always looked either too flimsy or too bulky, with little practicality to either, but an NFC key could fit pretty well.

A credit-card PCB AM (energy harvesting or crystal-style) radio might also be possible, which is an interesting thought. A coil for the antenna, handful of passives, and perhaps some sort of headphone jack that's made of plated thu-holes partially open to the side of the card (pic). Though I'm unsure how tuning would work for the crystal (no IC) variant. A rotatable piece of shimstock or FR4 pinned atop (or inside) the PCB? A long thumbscrew into the side of the card surrounded by vias (pic)?

[Anonymous No.1478768]

>>1478756
well it can easily cut wood and plastic, and pcb traces are really thin so i wouldn't be surprised if it cut them

[Anonymous No.1478770]

>>1478768
How are lasers used to silkscreen? Isn't silkscreening done with a silkscreen, or am I showing my age?

[Anonymous No.1478771]

>>1478768
The reason you probably wont' is because metal is both shiny and thermally conductive
Any damage you'll do is from carbonizing the plastic and having some soot to clean off
I mean hey, I could always be wrong, but having played with 1W lasers before, my intuition is that you'll be fine

[Anonymous No.1478772]

>>1478770
they aren't but i have a i don't have money or space to buy a silk screen machine, but i have a small laser engraver

[Anonymous No.1478773]

>>1478767
You could probably get a radio AM radio into a credit card style form factor yeah, at least the receiver and demodulator sure but... you're gonna have a bad time when you actually wanna listen to the damn thing cause where's the power for the power amplifier gonna come from? Stick any cheap four ohm earbuds on that and watch your output get loaded down to the ground, and forget about driving higher impedance cans.

[Anonymous No.1478775]

>>1478770
nope, still silkscreens

>>1478772
if you back the power WAY off maybe you could use light-curing ink of some sort

>>1478773
a USB port, natch

[Anonymous No.1478776]

Hello I need to make a sensor to measure conductivity of water. I`m thinking of making a inductive one as it`ll be out in the open measuring all the time and in contact with water. Anyone has any warnings, hints or tips for me? I never built coils before, anyone has experience with them? They`ll probably be wound up arround a plastic tube as to allow water to pass through.

[Anonymous No.1478778]

>>1478773
The size of the card is perhaps a quarter the area of some small crystal radio antennae I've seen, so the volume with high-z/crystal/piezo earbuds would likely be sufficient, if barely. Would need lots of turns to get a high enough voltage to work the diode detector, perhaps a rectenna to run a MOSFET ideal diode would be worth doing.

Whether using a rectenna and amplifier would work to get more output power is doubtful, unless you could harvest energy from other radio bands somehow. I don't think the form-factor plays nice with multiple antennae each resonant at a different frequency, even if the silicon to run it would be worth the power cost. Perhaps a tank circuit with multiple resonances?

Running normal ~30Ω earbuds off such a thing is probably possible with a matching network.

>>1478776
There are such sensors on AliExpress, though I think they just have two wires in the water and measure the electrical resistance. Does an inductive sensor put AC through a water-core inductor and measure the eddy-current losses? Stainless steel would probably last perfectly fine for a low-voltage direct contact sensor.

[Anonymous No.1478784]

>>1478778
Even measurements with wire need to be in AC as DC would polarize salts in the water and the measurement would be useless. Metal is hard because there is corrosion and build-up of stuff than can corrupt data. Inductive sensing of conductivity is based on the fact that the induced voltage on the secondary coild relates to the magnetic permeability of the water (which has a thin connection to connectivity due to dilluted salts); in using another calibrated sensor or a solution of known conductance you can calibrate the thing to give you a nice mS/cm result. I`ll also need to make a turbidity sensor, a temperature and humidity. The last ones can be bought for cheap but conductivity and turbidity will need to be DIY. Also I`ll have to code something to record the data on a SDcard and that will last a couple months on two AAs floating on a lake. I can do that, but building transducers is new to me. I`ll look in ali to see if there is anything.

[Anonymous No.1478785]

>>1478784
Also the bad thing about conductive.. conductivity measurement is that is related to current path geometry. That involves something called cell constant, which given proper calibration can be found. But if some piece of dirt gets between the sensors or there is algae buildup it gets fucked. With M a G n E t S, it doesn`t really matter, because the field sees most stuff equally. (With the exception of the dilluted salt, which is what I`m going for)

[Anonymous No.1478798]

>>1478784
>>1478785
I don't think you're going to have much progress trying to measure conductivity of water using magnetic fields

[Anonymous No.1478807]

>>1478798
I know it is possible and done in many industrial applications, my question is not about that. Is about making the damn coils, anyone has experience with that?

[Anonymous No.1478809]

>>1478807
Development of a Conductivity Sensor for Monitoring Groundwater Resources to Optimize Water Management in Smart City Environments and there other articles.

[Anonymous No.1478810]

>>1478809
Also ISO7888 mentions inductive and conductive methods

[Anonymous No.1478835]

>>1478807
That paper's existence and its recent publication suggests it's not "done in many industrial applications" sir

[Anonymous No.1478854]

>>1478835
Comercial conductivity/concentration/salinity meters in many plants are inductive. They are used in enviroments that damage conventional electrodes. Hand held ones can be inductive sometimes and those boys marine biologist use have inductive sensor. As he mentioned a lake I can imagine he is trying to make one for a buoy.

[Anonymous No.1478866]

Do i have to solder ALL of the pins?
Or can i just do the ones on the sides? Will the rest of them have signal if i don¨t solder them?

[Anonymous No.1478895]

>>1478866
>Will the rest of them have signal if i don¨t solder them?

i once built a test jig where 12 pins were just inserted without solder, and surprisingly, it worked like 80% of the time. the rest of the time i had to tilt the connector to make contact on all 12 pins.

but this is only good for temporary testing. for normal work, you have to solder everything, obviously.

[Anonymous No.1478900]

>>1478895
But i hate soldering ;_;
It's tedious, my hands shake and the vapors give me cancer

[Anonymous No.1478902]

>>1478835
My man, I literally gave you the ISO thing about that topic.

[Anonymous No.1478963]

>>1478900
Use something to brace your hand against to steady it, and keep a window open.

[Anonymous No.1478976]

>>1478807

[Anonymous No.1478978]

I am so done with the fucking shitty small arduinos
>turn it on
>it burns even though the voltage is within limits
>get another one and use buck converter to feed it
>takes hour until the programmer makes a proper contact to upload my shit
>works for 10 minutes then turns off forever
>always breaks only AFTER i attach all the wires and components too

fucking hell i am so done with this piece of shit arduino mini series
it's going straight into the garbage bin inside of my filthy room with the rest of my cum covered tissues, where it belongs, fucking piece of shit

from now on i am using unos and rpis for everything

[Anonymous No.1478983]

>>1474893
So I want to make a single pixel scanning camera. How do I figure out how much current my photodiode is going to produce based on scene brightness and stuff?

[Anonymous No.1478988]

>>1478983
experiment and measure

[Anonymous No.1478989]

>>1478983
How about using a waterclear RGB LED as your photo sensor so you get 3 colours? I think that will work.

[Anonymous No.1479037]

>>1478718
It's not exactly what you asked but this guy measured the timing of digital reads and writes:
https://jeelabs.org/2010/01/06/pin-io-performance/

[Anonymous No.1479042]

>Thinking about making a milliohm meter
>real ones use Kelvin clips
>look up kelvin clips on digikey and mouser
>like 50 dollars for a pair of clips
my fucking face

[Anonymous No.1479053]

>>1479042
If you're sure you really need them, you should try and make them yourself. Start with a regular (big) aligator clip, use a plastic pin through the center and some plastic washers to keep the halves apart.

[Anonymous No.1479077]

>>1478648
almost cheap enough to buy a bunch just for process characterization and soldering practice

[Anonymous No.1479093]

>>1479077
>soldering practice
I wonder if you can buy bulk empty packages (or packages with a thermal sensing diode or two inside) solely for soldering practice. It would be neat for tracking the lead and die temperatures for different soldering techniques (wave, different tips for iron and nozzles for reflow).

[Anonymous No.1479094]

>>1479053
And make sure you plate it with gold. I hear plating with cyanide is pretty easy.

[Anonymous No.1479135]

>>1479093
mechanical samples/dummy components exist but are expensive and hard to get

[Anonymous No.1479139]

>>1479053
I'll just buy cheapie ebay ones
jesus christ why do these things cost like 50 dollars, or like a hundred for something from like bk precision

[Anonymous No.1479166]

>>1479139
low volume vs. amortized NRE costs

[Anonymous No.1479188]

>>1479166
I only have a vague sense of what that means
low volume so they are priced higher, and on top of that, they aren't made much by them nor their vendor which further adds to their cost? is that what you mean?

[Anonymous No.1479214]

>>1479188
NRE = non-recurring engineering
they spread their tooling costs over fewer units, plus they pay more to setup shorter runs rather than have 20-year-old tarnished product littering up the warehouse

[Anonymous No.1479271]

>>1478403
I got one for around $20 in a kit iirc and I think the temp knob is already fucked after two uses. I have another one around here I need to test.

[Anonymous No.1479281]

>>1479271
Mine has been my longest lasting iron (it's a shitty jaycar model). Anything that's gone wrong with it can probably be repaired, assuming you have the right screwdriver to get into it.

[Anonymous No.1479294]

NEW BREAD

>>1479293
>>1479293
>>1479293

NEW BREAD