/diy/ - Do It Yourself

>>1363100
sop > soic

this thread's digits brought to you in part by the HCPL-3100 power MOSFET/IGBT gate drive optocoupler

>>1363016
Half and half for me, been learning the formalities for the last three years doing a physics major, but picking up tips and tricks and more practical circuits from the google, /ohm/, and YT.

So I'm working on a project, and I have a question about ECL Logic.

I've used TTL mux chips before as a switch to select different analog signals, but can I do the same with ECL?
I know that ECL is differential, but can't I just use differential audio?
Or does ECL hard toggle to true/false?

I'm at a complete loss as to how I would do this.
How would I make the function 1 XOR X'Z XOR Y' using a 3 to 8 decoder made of two 2 to 4 decoders? I can additionally use 2 input NAND gates, 2 input NOR gates, 4 input NAND gates, and 2 input XOR gates, but need to use the decoder.

I promise I'm not trying to get 4chan to do all my homework for me, I'm just dumb.

>>1363210
1 XOR (something) is just an inverter, for starters

>>1363214
My issue is I can't figure out how to do it using the 3 to 8 decoder.
It could just be that I'm trying to do this late at night and I'm making a dumb mistake, but I'm not counting on it.

>>1363103
[dq]fn > sop

>>1363207
>TTL mux chips before as a switch to select different analog signals
are you sure? CMOS chips can do this because of transmission gates, but I have trouble believing BJTs make good general transmission gates.
ECL is current-based, not voltage-based. you should probably look at schematics of the chips you intend to use to see if you can feasibly get a linear signal through without distortion. I'm thinking no though

>>1363210
make a truth table of Q in terms of XYZ. if by decoder you mean a device that sets one output to a distinct value from the others as selected by the inputs, take all the outputs that would be selected when your logic expression is true and OR them together (assuming all positive logic, invert as needed). if you mean a data selector, connect the demux up so that XYZ is the selector and each input receives the appropriate value from the truth table

>>1363231
Sorry, yeah, they were CMOS. (74HCTxx).
I've never used ECL before so I'm not certain either.
I would continue to use the 74HCT chips, but I need that extra speed that ECL offers for this project.
I guess I could order a few chips and test, but they're kinda pricey.
I haven't had much luck looking at schematics or datasheets because they're all digital logic oriented.
Which makes sense. This is pretty non-standard use for logic chips.

>>1363234
what are your performance requirements? how fast does the mux need to switch? if it's only the controlling logic that needs to be fast, you could use a diff-to-single-end converter or just a fast comparator to control the switches, no?

>>1363237
It’s for a ham radio design.
Using 74HCT I can do the lower frequencies, but my goal is to also be VHF/UHF capable.
I’d be content with 2m, but I’d like to be able to reach 70cm.

The design is based on the Tayloe detector, which needs to be clocked at four times the radio frequency.
So for 500mhz, I need a 2ghz clock.

>>1363244
the fastest analog switches I'm seeing around have Ton+Toff in the ~5ns range, which is 6m at best. if MOSFETs with very small gates can't be made to do it, you're gonna need a downconverter (or a different architecture)

>>1363231
>ECL is current-based, not voltage-based
Maybe internally, but externally it is pretty much voltage-based: the inputs take very little current and the outputs source as much as your load wants.
If you bias ECL NOR gates carefully to their transition region, you should be able to pass low level analog signals with plenty of gain (and distortion). Basic amplifiers have been built this way, but using that trick to build an analog multiplexer sounds rather iffy.

>>1363244
Lol. So you finally get what the ham faggots were trying to tell you?

>>1363231
>make a truth table of Q in terms of XYZ
This, then simplify the algebra. Make sure to convert any XORs into their component gates, XORs don't factorise.

>>1363259
why, when there's a 1-of-8 decoder just sitting there waiting for us?

>>1363264
Well if that will solve his problems, but he may be able to do it with a lower overall propagation delay with discrete gates.

>>1363283
>I can additionally use 2 input NAND gates, 2 input NOR gates, 4 input NAND gates, and 2 input XOR gates, but need to use the decoder.
not if the point of the assignment is to give the student a more concrete feel for minterms/maxterms. this is why people need to read the original problem statements

>>>1363052
thanks anon. I think most of my classmates are retarded too, each in his own way. I am retard in the sense of getting very advanced and complicated stuff right intuitively but making stupid mistakes at very very fucking simple shit. feels bad man.

>>1363256
>Lol. So you finally get what the ham faggots were trying to tell you?
Get what? they just REEEEEEE'd at everything.
Zero useful information.
By your own post it sounds like it's possible.

Honestly, the only thing that needs to run at 4x RF is the mux.
The only other part that needs to run a high frequency is the op-amps in the receiver side, and it's not hard to find op-amps capable of 70cm frequencies.

>>1363459
Additionally:
I was reading a paper about how ECL works and all that, and it says this:
>ECL can be used single-endedly or differentially depending upon the application and noise margins needed. The drivers are low impedance open emitter outputs that generate a typical 700 to 800 mV output voltage.

Sounds like that would work, no? Low impedance is great, and they can supposedly be run in a non-differential mode.
I just can't find any sources for how to use them non-differentialy.

Also from the same paper:
>The output is typically terminated with 50 Ohms
Great! That matches the standard impedance for ham radio antennas.

>>1363459
>high frequency .. op-amps in the receiver side
>op-amps capable of 70cm frequencies
Are you sure you know what you're talking about? I'd suggest you re-read that article your image originally comes from.

>>1363459
Oh. Sounds like you REEEEE'd too hard to learn anything useful.
One of the points was that analog multiplexers are fundamentally very different from digital multiplexers.To begin with, they don't have dedicated inputs and outputs like digital ones. This is relevant to the circuit you posted.

What I said about ECL analog multiplexing was basically that there's some small chance of getting it kinda-sorta work with enough nigger-rigging. It is guaranteed to not work if you use it properly.

Well, whatever. Just out of curiosity, how are you going to produce the clock for your detector and what is going to be your tuning resolution?

>>1363475
>I just can't find any sources for how to use them non-differentialy.
Dude. That means just using one output instead of two opposite phase outputs (or inputs).
>That matches the standard impedance for ham radio antennas.
Why do you even care, considering it is relevant to the outputs? Weren't you building a receiver?

Is this all just trolling?

>>>1363720
>>>1363728
Thanks anon, for the explanation. My problem is less with the caps and more with the diodes and other non-linear elements. I can`t seem to fit in my mind and how I should analyse thigns with them.

I'd like to make a simple square wave generator with a single potentiometer to set variable frequency. Frequencies should be around the audio range, maybe 200Hz to 8kHz.

Here's the catch:
1.) single supply operation only.
2.) don't have any cmos logic on hand
3.) duty cycle should be close to 50%, I figure +/-10% is ok
4.) no uC based solutions

I tried some TTL inverter based oscillator on a breadboard but the edges were really jittery. Not sure why that is, maybe someone smart can help me understand.

I tried an opamp based oscillator but it didn't work on a single supply configuration probably because I'm a newb and don't know shit about biasing.

>>1363736
Well in the case of the voltage multiplier/doubler don't think about them too hard. During the charge pump cycle when the diode conducts the caps accumulate charge, when the diode isn't conducting the caps are acting as the reference for the other caps that are charging thus multiplying the voltage.

When you're talking about diodes being non-linear elements you don't need to think of them in some fancy domain like heterdynes, this application is literally a series of switches that are either open or closed depending on where you are in the charge pump cycle.

>>1363739
This is a light controlled application. Replace LDR with VR, remove piezo.

>>1363739
>don't have any cmos logic on hand
>>1363750
>TLC555

I mean...come on man. seriously.

>>1363750
>TLC

u did that on purpose didnt u anon

>>1363570
Yes, on the receiver side.
Those two op-amps are being switched between at RF frequencies.
And yes, it's not hard to get an op-amp capable of 70cm frequencies.
Here's one: http://www.ti.com/product/lmh6629
It's $4.25 on Mouser.

>>1363698
>how are you going to produce the clock for your detector
I found a couple of VCO ICs that will work for the different bands
>what is going to be your tuning resolution
Don't know that yet, I was planning on using a potentiometer as a simple voltage divider and then feeding that into the VCO, so It would depend on how precise the potentiometer is.

>Dude. That means just using one output instead of two opposite phase outputs (or inputs).
You sure about that?
Pic related.

>That matches the standard impedance for ham radio antennas.
>Why do you even care, considering it is relevant to the outputs?
If it's already supposed to be terminated at 50 ohms, then that means I don't need to do anything special to get the outputs to play nice with existing antennas or amplifiers.
>Weren't you building a receiver?
I'm working on a transceiver, since the Tayole detector can also be turned into a modulator.

>Is this all just trolling?
Can you see beyond your nose?

Look, I'm not married to the idea of using ECL, but it's one of the few logic families I'm aware of that clock fast enough for UHF.
I've also looked into using a FPGA, but I can't find any ways to get them working on 2m and up. They just aren't fast enough for that.
The FPGAs I was looking at all wind up topping out a little above the 6m band.

>>1363783
>I was planning on using a potentiometer as a simple voltage divider and then feeding that into the VCO
why yes, you are trolling

>>1363739
you need a low impedance V+/2 supply or the oscillator will pull V+/2 all over the place. generate it with a voltage divider and an opamp configured as voltage follower. then connect your "GND" points to the output, your "-V" points to 0V, and "+V" to +V. thus almost any op-amp circuit can be run from a single supply

>>1363790
You are the least useful faggot I have ever had the misfortune of coming across.

>>1363739
>but the edges were really jittery.
explain jittery. wasnt that pesky jerk fourier was it?

>>1363794
>>1363794
>>1363794
>Jitter
>Fourier

Fourier dealt with periodic signals, jitter is the description of how aperiodic a signal is. I mean how fucking retarded are first year electronics kids? I bet your 3 smelly virgin roomates that all have calc101 with you thought it was funny though.

>>1363739
use relays and timers, pot changes the RC constant

>>1363739
Either make another voltage rail for your opamp oscillator, which is pretty easy to buffer if you're using a 2-opamp IC, or use a 555. I've tried transistor-based astable multivibrators, but I could never get a favourable result. I actually had more luck using a reverse-biased NPN relaxation oscillator, though I don't think it will provide the duty cycle you're looking for.

>>1363793
<-- you
may I suggest you just settle for HF for now and, once you can appreciate what goes into a stable radio where the parts per million aren't quite as critical as they are at the next order of magnitude higher frequency, you can worry about extending to VHF or higher?

>>1363799
I think he's impugning the good character of whatever you were using to measure said jitter

>>1363720
>Capacitors store charge, don't think of it as storing voltage, but charge.
I think it's appropriate to think of capacitors as storing a voltage. A farad is a columb per volt. It stores charge as a voltage.

Holy hell, just made this custom drill to cut holes for a speaker, was quite the wild ride.

>>1363966
And here's the test piece. The actual diameter of the drill changes by 1 or 2mm because of the centripetal force being provided by the springiness of the metal. It has a terrible tendency to grab and it wobbles the entire workbench, it's a complete deathtrap, hence holding a sheet of plywood in between me and the press. Safety glasses needed. On this image I had to stop the one on the right because it grabbed, but the left one finished properly, before grabbing, hence the current state of the bit.

>>1363968
Right/left are reversed because the image is upside down.

>>1363210
The final result
It either works or I'm too brain damaged from trying to realize it doesn't.

>>1363966
They sell the adjustable hole cutters at that hardware store for like 10$

>>1363999
>buyfagging on /diy/
I bet you bought your drill press too.

>>1363983
you're on the right track... run through all the inputs and make sure the outputs match the truth table you built

>>1364001
>posting tools that don't even work and aren't even electronic on /ohm/
gj

>>1363983
I swear the XORs are unnecessary.

>>1363929
All you can store is energy. A capacitor stores energy in form of an electric field generated by charge separation and proportional to V^2.

https://www.aliexpress.com/item/MINI-T12-OLED-soldering-station-electronic-welding-iron-2017-New-design-DC-Version-Portable-T12-Digital/32839523934.html
this is on sale for $25 + shipping. any opinions on this vs the TS100? I'm about to buy it...

>>1364029
Well they "store" a charge separation, there will be negative charges on one plate and positive on the other, though if your entire system carries a net charge things get a little less clear. If anyone has an explanation for what happens when one electrode of a capacitor is bombarded with electrons that would be nice to know. Capacitors can certainly store energy, though noting that they do so in the electrostatic field of the separated charges helps for visualising how they work.

>>1364036
>bombarded with electrons
A vacuum tube (e.g. a picture tube) could be such a case where cathode and anode form the capacitor.

>>1364035
well, I ended up buying that soldering iron, though I wonder how it compares to this one:
https://www.aliexpress.com/item/Meterk-80W-Professional-Soldering-Iron-LCD-Digital-Temperature-Adjustable-Electric-Soldering-Iron-Mini-Soldering-Station-AC220V/32840002706.html

also, in your opinion, are these tips cheap or expensive? https://www.aliexpress.com/item/10Pcs-SET-T12-B-T12-D24-T12-C4-ILS-JL02-KU-K-BC2-ILS-BC1-Solder/32799109111.html

>>1364094
it doesn't look unreasonable for what you're getting. shipping options look disappointing from that seller tho

Does anybody know what the F.Mask layer does in kicad, is just this layer enough for some exposed copper stuff or does there need to be a F.Cu layer below it and is F.Mask just the place where the solder mask is removed

>>1364167
If you have a shape on mask layer with nothing on the copper layer beneath it you'll just get bare board, if you want copper, have it in the copper layer too.

>>1364188
tnx

I'm designing a nixie clock with K155ID1 chips driving IN-12A tubes.

According to the datasheet for the K155ID1, If I set the binary inputs to a value greater than 9, all the outputs should go to high impedance and none of the digits should light on the tube.

I wanted to be able to have the option to turn off the two tubes counting seconds just in case it got distracting.
I have a a switch connected to one of the inputs of my PIC and basically when it's pushed, the PIC shifts out all 1s to the K155ID1 driving the seconds tubes.
When I do this however, the tubes don't turn off, they just randomly light a few of the digits.

Did I goof up? It's as if the pins are floating but surely the cathodes shouldn't light unless something is sinking current for them?

The alternative is to just cut a track and wire a switch onto where the high tension rail connects to the anodes on those two tubes.

>>1364238
Anode section for reference

>>1364238
Just a guess, but have you tried lowering the anode voltage?

>>1364261
I stupidly put sealant on the pot so can't adjust it

>>1364261
>>1364268
My nigga.
I broke the sealant and decreased the voltage a bit.
It works now.
I owe you a pint

This 3.55mm panel mount for audio jacks has 4 poles.

What does this mean?

My cable only has 3 wires. What's the 4th pole for?

>>1364320
often, the sleeve has a switch of some sort, usually wired to connect the power source to the circuit ground. ohm it out to be sure

>>1364321
I found this Asian diagram. Are you able to deduce which pole I won't be using if my audio cable has 3 wires?

Also which one is the ground?

P.S. I genuinely appreciate your help. I know I'm electronically ignorant.

>>1364328
ohhh, it's a TRRS jack
I presume red, white and a common lead either bare or black? looks like you can connect the common to either R- or L- (R- is preferred for compatibility with mic-equipped headsets). red to R+, white to L+

what type of "electronics" fo you mean? Like engineering or repair or what

>>1364331
To be perfectly clear. I watched this video:

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

And he uses the exact same model jack. I can't see how he did the connections, but the headphone he fixed doesn't have a mic. So I'm guessing it's safe/possible to use that jack with those headphones. Right?

Here's how the wires on my headphones go:
Green - left
Red - Right
Black - Ground(?)

So I would connect the black cable to R-?

Are both L- and R- just grounds?

>>1364339
>Are both L- and R- just grounds?
yes
>So I would connect the black cable to R-?
right
>So I'm guessing it's safe/possible to use that jack with those headphones. Right?
almost certainly. the worst that might happen is your amp sees a short on its output, but a brief short on a headphone jack is extremely unlikely to cause problems

>>1364348
Thanks anon! I'm grateful for your help.

>>1364339
I really don't know why the hell that diagram is using separate grounds.
TRRS = Tip Ring Ring Sleeve.
Standard headphone connectors are TRS, with only one ring. The tip is left, the ring is right, and the sleeve is ground, shared by both. The other ring is for a microphone or other accessory.
The first ring, the one labeled in that diagram as R+, is your R. The second ring you can leave disconnected or connect it to ground.
The sleeve, labeled in that diagram as R-, is the common ground for both left and right.

>>1364353
And then L+ for the left wire, and leave L- completely alone?

>>1364356
In his diagram he has two ground wires, one for left and one for right. Most headphones do not have that, just have one ground wire that's shared between the two of them.
So in stead of having L+, R+, L-, and R-, your headphones just have L+, R+, and -.
You have one -, one ground wire, which goes to the sleeve (where R- is connected on the diagram)
Your L+ goes on the tip.
The contact labeled L- you can connect to ground or just leave alone. It doesn't really matter.

>>1364380
Thanks for the clarification. I'm doing it rn as we speak. I had a hard time stripping the cable despite having wire strippers haha

>>1364353
in theory there may be some high-end phones with separate grounds. it would be weird though

>>1364407
try burning it

>>1364408
>high-end phones with separate grounds
It would be monumentally stupid to do it that way. Sure, if you plug in a standard TRS, both grounds will be connected together to the headphones' common, but if you plug in a headset with a mic or with buttons that has a TRRS, you'd have no sound in the left channel.

>>1364094
>https://www.aliexpress.com/item/Meterk-80W-Professional-Soldering-Iron-LCD-Digital-Temperature-Adjustable-Electric-Soldering-Iron-Mini-Soldering-Station-AC220V/32840002706.html
samefag here. I wonder what could be the pros/cons of it. would you'd say using this thing is risky?

>>1364328
>Deduce
https://m.youtube.com/watch?v=HXJkLBj6G1k

>>1364333
Welcome new friend
ebin digits senpai

>>1364348
>>1364353
>>1364380
>>1364408
Alright, I finished it. One huge problem though. It's only playing out of the right ear.

I'm running it through a mono cable (1/4 inch with 3.55mm adapters), could this be the reason one ear isn't working? Or did I fuck something up 100%?

>>1364438
I tried resoldering to ensure good connections and now it only plays out of the left ear. Wtf.

>>1364438
>>1364457
IT'S WORKINGGGG

>>1364423
looks chintzy af. otoh, it claims to be temperature-controlled, is dirt cheap, looks lightweight, and looks like it takes the cheap and common Hakko 900-M series tips
if you're gonna get another chink iron, I would recommend getting a more established chink brand, but I think you'd be happier with this one for longer than that cutesy station-like thing. but now that you've already ordered the one, may as well save up for a better station and other equipment

Total beginner here. I'm looking for a solder iron to buy, but I don't know where to start given the overwhelming number of options available on the market.
My budget is around 40€.

I'm looking for a tool powerful enough to allow me to solder thick (few mm) electrical wires of some appliance's and brass tubes, but that at the same time won't fry an arduino circuit or smd LEDs when I will work with them.
I started looking at some 40W temperature controlled irons, but they seemed too cheap to me. The wattage seemed to low to handle the "heavier" work. My main concern with most of what I have seen are the plugs (for the Italian market): they only have two terminals and that means they won't be grounded. Since I'm going to use the tool for a hobby and I don't want to die while working on a arduino, I started looking at more powerful alternatives from known brands and the best I could find is this 80W iron:
www.weller-toolsus.com/soldering/consumer/irons/weller-sp80nus-soldering-iron-led-80-watt-120v.html

Now I think this will be able to handle the thicker wires or brass stuff and since it is a very known brand I assume it is safe to use. But now my main concern about this tool is that, since it has nothing to control temperature, it will fry the small components I am going to use, correct?
Any suggestions on what to look for?

>>1364552
>few mm wires
diameter? you should really be thinking about crimp terminals or welding at that level
>grounded
considered important for most semiconductor circuits
>it will fry the small components
it sure will
that budget seems a bit tight for an iron with the range you're expecting. consider getting an iron appropriate for small electronic work, and using something else entirely on the bigger stuff, like brazing or a spot welder (which can be made out of thicc wire and a couple of microwave oven transformers)

>>1364556
>consider getting an iron appropriate for small electronic work
What would be an appropriate wattage for a iron with no temperature control? Around 15W to 25W?

>>1364559
>no temperature control
I would strongly suggest you get a chink iron with temperature control, even if you gotta pay 100% import duty to get it. https://www.aliexpress.com/item/altalena-del-sesso-per-i-bambini/32840002706.html is cheap and may or may not be grounded but will probably handle anything you care to throw at it. you have money left over to get some extra tips, including some suitable for larger work

test

>>1364587
Wow looks like I can only post with an image now. I guess have fun with whatever is lying about on my desktop.

>>1364047
I meant having one plate bombarded with charges without the negative plate being bombarded with opposite charges. Like a spherical capacitor in space being bombarded with ionising radiation. I'm sure if you use Maxwell's equations and figured it out from there you'd find a relationship, but that would mean studying how Maxwell's equations work again. You're welcome to attempt this I guess, it should have real applications on existing spacecraft.

I wonder if you can surround a spacecraft with a layer of foil and attach it to the central chassis of your craft via a supercapacitor or two such that the central chassis remains at a neutral charge and the voltage increase between the inside and the outside remains tolerably low. Since charge over time is about constant, the voltage would increase at a rate of V/t = I/C, and the repulsion of ionising radiation will increase as a function of charge, the (limit) voltage at which no ionising radiation will encounter the craft will be lower, meaning less chance of the internal electronics being shocked from shorting to the outside. But what voltage means in this context is a little undefined, I'm not sure whether the central chassis remains at a net 0 charge or if it falls to an opposite potential of the shell.

Note that a spherical capacitor has less energy storage per volume dielectric than a parallel plate capacitor, though the two approach one another when the internal radius approaches the external radius. Maximum energy storage per volume dielectric of a parallel plate capacitor is a function of the dielectric's strength and permittivity, and is a constant for that particular dielectric measured in J/m^3 or equivalent dimensions. This does not cover capacitors with plates with altered geometries to give higher surface areas, like aluminium electrolytic and super capacitors.
E/V = 0.5 * permittivity * dielectric_strength ^ 2
The specific (per unit mass) energy is this divided by density.

is all this tariff talk making any burgerbros want to order up on components sooner than later?

>>1364597
no, i just bought 100 4148s for a dollar. i wouldn't mind paying a quarter extra if it brought us closer to the sweet release of death by nuclear holocaust.

Sup lads. I'm repairing this ancient east-german bandsaw. Got it down to a bad capacitor, of which I've attached a pic. Now I know a bit about electronics but certainly not really anything about capacitance. This is a huge capacitor with a pretty damn low capacitance, right? Will replacing it be as simple as finding a modern, much smaller one? There are 4 of these suckers and I'm pretty sure they're in parallel too, so I could just get a 48mF 250V cap instead right?

>>1364644
i'm not a motor aficionado and i can't tell the difference between a run and a start cap, but i'm pretty sure that's a polypropylene film cap (MKP) which are less energy dense than the electrolytic caps you want to replace them with. again i can't speak to motors but in electronics you don't use a film capacitor unless you need to. so i wouldn't try to replace that with anything but another film capacitor. you can replace all four with one larger cap. there are benefits to paralleling capacitors but they don't matter here.

>>1364541
thanks m8

>>1364653
Ok cool, thanks. I've found just a drop-in replacement for the faulty one here:

https://www.amazon.de/12%C2%B5F-Motorkondensator-Betriebskondensator-Anlaufkondensator-Anlasskondensator/dp/B00AFBVVEI

Which seems to be what I want. I think anyway. I don't speak much german but that's where I'm currently living (working on an old farm, helping renovate and make it eco-friendly and stuff).
As I understand it, it doesn't matter that it's rated for a higher voltage.

>>1364662
voltage rating won't make any difference. capacitors last longer the further they're run below their voltage rating, but higher voltage caps at the same size will also have more self-heating (esr), so you can't say which is better.

digikey has cheaper than what you've linked but idk how much they charge to ship to germany. one example:
https://www.digikey.com/product-detail/en/kemet/C274AC35120AA0J/399-11427-ND/4833258

>>1363100
Using C, I want to write floats, ints etc over a UART. I can get it to work for single byte characters (have to load the register with the desired character, & it with the register).

Any idea on how I can do this for ints and floats? I tried to convert them to an array of chars but it didn't work, and the register will only take 1 byte at a time. So I somehow need to feed through ints or floats, 1 byte at a time, to be picked up by either PuTTY or a python script.

Here's what I currently got
>https://pastebin.com/3dqMVDE7

No idea why this doesn't work tbqh

>>1364707
> char arrayofchars[] = {intbyte1, intbyte2, intbyte3, intbyte4, '\n', '\r', '\0'};

What kind of results you're expecting? Human-readable text? Why not use sprintf or something similar to convert the numbers to text?
If you want raw binary output, then your transmit routine must be able to live with nulls in data. Also, it is generally easier to just take an (unsigned character) pointer directly to your int/float and then dump sizeof(data) bytes. This does not handle the possible endianness problems, though.

>>1364710
I'm transferring data from an STM32 microcontroller to my laptop, so if I can get the byte transfer to work then I can reassemble the integer on the pc.

Since it's on a microcontroller, there's no point in using sprintf since it doesn't have a console. Unless you can say
>char x = sprintf("My int")
or something?

>Also, it is generally easier to just take an (unsigned character) pointer directly to your int/float and then dump sizeof(data) bytes. This does not handle the possible endianness problems, though.
That's a good idea, I'll try that, cheers

begginer here, which ic or microcontrollers should i use to make a gamepad for usb,
should be 2, 1 for digital buttons and 1 for analogic gamepad
(although idk if joystick button must be attached to the digital controller or counts as analogic),any guides for identifying what controller you need?
also, i dont know how to code it other than fucking arduino, how should i do upload the code to the controller so i dont have to run the program everytime i plug the gamepad ?

thanks to whoever reads this

>>1364723
usb microcontrollers are not simple things. the first thing i would investigate is using a usb-uart controller (ch340g, ftdi, silabs) coupled with a regular uart capable microcontroller like the atmega328p in your arduino. the limiting factor here will be latency, but the math on that shouldn't be hard and i don't think it'll be very different from low speed usb directly.

>>1364724
one other thing, using a usb microcontroller lets you write it as an HID device which i think lets you do keyboard and mouse without writing any desktop code.

>>1364716
>since it doesn't have a console
If you specifically want to use console functions, then it's pretty common that their microcontroller implementations allow you to specify your own low level I/O functions. This way you can make, say, puts() work like "usual" with serial ports.

>Unless you can say
Why not just google sprintf, instead of making guesses?

>>1364726

Thank you very much, you are totally right

>>1364596
m8, there are easier ways of shielding things from waves. the field of electromagnetic coupling studies that. and I have no idea about your math because I have no time.

>>1364754
compatibility* not coupling

Last year I bought a akg k92 headphone
Today, the right side stopped working
Sadly, i couldn't find any sort of guide helping me with this model on google, lost of akg guides, but none applied to mine.

I'm a noob in terms of repair, but I heard that fixing this type of device was quite easy, so I gave it a shot
I spent 20minutes trying to open the damn thing and, after almost breaking some plastic, it opened

What i found out was: there are two wires connecting the speaker to the headband support(wich i assume is conducting the electricity, since it is made out of metal and sometimes it shocks the back of my neck) going through a plastic piece. Here are some pics

>>1364789
the metal here is losing its color where it meets the wires

>>1364789
>>1364790

>>1364791
>>1364790
>>1364789
my best guess is that the problem is here,
the red wire is scuffed but the white one is still in place so idk

how hard is it to fix this? do i need a soldering iron?

>>1364793
yes, you just solder that part. but it takes some practice. if you botch it, you could damage it for good

>>1364723
Teensy 2 would be a good choice. it has plenty of GPIOs and native USB. it is also supported by LUFA which has literally gallons of example code for each standard device class. it has a decent minimal USB bootloader. USB HID is a bit tangled so you might want to read through the spec several times, start from one of the example projects, have a test bench program on the host side that displays as close to raw values as you can get, and be prepared to iterate a lot

>>1364793
its not so hard, i have the same heaedphones and the same shit happened to me, tho the solder is a bit tricky

>>1364793
peel the black cable on tour pick a bit, try also to burn the cable with a match, but just the tip,and clean it afterwards, then try to solder it

>>1364843
is usb terrible? my intuition tells me that anything with so many artificial layers for transmitting bytes in a master-slave configuration can't possibly be good but i'm not educated enough to judge it.

>>1364931
it's not bad to use. the original standard was planned out well and has served nicely in practice for 20+ years. the layering is pretty natural imo. it deals in packets natively, unlike all the shit-tier serial port protocols that have to implement packetization for themselves. host/device/hub hardware handles a lot of the handshaking and routing so that the host or device doesn't have to do too much more than just buffer up a packet and hit send. most hardware isn't much harder to operate at the driver level than, say, an Ethernet interface
that said, I haven't dealt with USB 2.0 since the back-compatibility has been so good and coding to USB 1.1 still works great. in any case I would always use a library if one were available, just because there are many of them and why not enjoy the fruits of the effort of others who have gotten it right

Just got a new board in for a class project. How bad did I fuck up?
Reverse silk is the Prof's face

>>1364954
1. does it work? no -> fix it
2. is he offended? no -> fix it

>>1364954
Layout looks reasonable to me. Why so many chips though?

Would be nice to add some corner mounting holes, and a ground plane, and some test points, at least one for ground.

>>1364954
Maybe rotating the SMD ICs by 90 degrees could help with simpler connection so the pinheader. Id squeeze some SMDs between the pins of the top part too.

Looks okay tho.

>>1365078
>2. is he offended? no -> fix it
kek

Can i connect 2 pins of an arduino to double the current i can pull up to 80mA?

Id turn them on/off at the same time.

>>1365131
can work, but the GND/VCC pins of the MCU will also carry that same current in the other direction, and they have their own maximum current spec. see the datasheet

>>1365131
also be aware of per-port limitations and that you'll get a lower output voltage from the pins as you draw more current. consider a discrete transistor if you're making a permanent project.

>>1365135
Got it. 200mA max on supply pins so it should be okay.

>>1365139
I think im fine then. I might go with a transistor, im not sure yet.

Want to turn on a relay, that pulls 50-60mA initially and 40-45mA when its on.

>>1365141
Absolutely use a transistor. Microcontroller output pins are for signal, not power. You can get away with it for LEDs but absolutely use a transistor to drive that relay.
And don't forget a reverse-biased diode across the relay coil to protect against inductive kickback.

>>1363100
Does anyone have any recommendations for a book for the general troubleshooting for electronics?
Or a source to learn PLC preferably related to rslogix 5000?

Hi. Of the three years I've been at uni, this will be the first time I give in and ask for help on /diy/. I've wasted an embarrassing amount of hours trying to figure out what I'm doing wrong.
The idea of the assignment is that I need to do nodal analysis on the circuit pictured, then enter the resulting Y and I matrices into MATLAB using symbolic variables. I need to solve for the V matrix, and then use that to find Vo/Vi. The professor basically spelled out how to do this in MATLAB, and provided what Vo/Vi should look like. I'm following the MATLAB instructions to a T, and my Vo/Vi never comes out correct, so I'm assuming something must be wrong with my nodal analysis, and I have no clue what.
Here's the the Y and I matrices I got. If you're experience and would like to at least give the matrices a crack and report back, I'd really appreciate it.
Y:

[-G1 + GPL + C1s, -GPL, -C1s, 0
-GPL, GPL+GPR +C2s, -GPR, 0
-C1s, -GPR, C1s+GPR+G1, -G1
0, 0, -G1, G1+G3]
I:

[-G1*Vi,
0,
0,
0]

Sorry if those were unreadable or something. This is part of a bigger project and the professor basically gave us the answers, so I'm just going to move on at this point, and if I really can't figure this out I'll just have to say so on the report. Thanks.

>>1365337
what an ungodly waste of instruction time

>>1365392
would you prefer to work the resistor blob in that xkcd comic without them?

>>1365392
How so?

>>1365604
in my opinion academia's focus on teaching theory, often laser-focused on certain areas like his example, is depriving students of valuable knowledge both in the practical and also more broadly in the theoretical subjects that get glossed over while doing this tedious meme work. the system is geared towards future phds, not the future engineers that almost all students will become.

>>1365609
>muh job training
what? it's literally part of an EE's job to be able to solve and prove ANY circuit they might come across. an admittance matrix is one tool that is invaluable in circuits with many internal interconnects. how else are you going to solve this? do you take the term "engineer" lightly?

>>1365617
>do you take the term "engineer" lightly?
yes. who told you anyone solves these problems outside university? i'd much rather learn about snubbers or some other niche but interesting topic, not that they're related.

>>1365619
>learn about snubbers
learn about the oh so esoteric theory of damping and snubbers will be obvious, geez
>solves these problems outside of university
if you're not just assembling cookbook circuits, which is admittedly 90% of most practice today, you're probably using the harder stuff. if you're developing modules or ICs for sale, you're probably dealing in this stuff if only to write the cookbook
besides, this is literally second-year algebra, literally an extension of the Ra||Rb = RaRb / (Ra+Rb) formula we all know and use daily

>>1365628
>learn about the oh so esoteric theory of damping and snubbers will be obvious, geez
this doesn't seem to be the case at all for graduates, which is the point i'm getting at. matrices are absolutely useful but a university that takes some keys from trade schools will be putting out better students, though (maybe) worse phds.

>>1365630
>implying Ph.Ds aren't absolutely necessary to continue advancing the state of the art
pajeet, is that you?

>>1365631
>implying i implied that
anon

I'm moving on with the project but if anyone knows how to do this and try it instead of arguing, that would be nice.

>>1365078
I will take this to heart
>>1365125
It's a 24-input mux and 24-input OR gate. The middle column is 4 LM339's, and the sides are 7432's. Dealing with 24 input signals takes a lot of chips.
Mounting holes probably would have been a good idea. The header on the bottom is actually designed to drop onto another board, so if the connectors work as advertised it should hold up well enough. If that fails, I'm not too worried about drilling / gluing as necessary
>>1365128
Squeezing them under the photon would have been a good idea, I could've shaved off a bit more vertical space.

>>1365685
>dealing with 24 input signals
takes a modest CPLD
>LM339
16-pin? are you quite sure?

>>1365687
Whoops, you're right. I've got it breadboarded with LM339's right now, the SMD chips are 74151

>>1365685
CPLDs are magic, something like this should be able to do the same thing smaller and cheaper https://www.digikey.com/product-detail/en/xilinx-inc/XC9536XL-10VQG64C/122-1435-ND/966616

>>1365705
>smaller
>cheaper
https://www.digikey.com/product-detail/en/lattice-semiconductor-corporation/ICE40UL640-CM36AI/220-1959-ND/5129488

>>1365708
BGA, needs ~1.2V, maybe not enough pins, saves only 10 burgercents.

>>1365717
good points, and it does have nonvolatile storage onboard, cool
XC95xx is great then

>>1365628
>formula we all know and use daily
Except RPN users. They type Ra[i]Rb[i][+][i] instead.

How would I go about finding K? So far, I've tried making equations for loop 1 and loop 2, but I'm getting stuck on finding the mutual inductance.

Anybody here finding it increasingly harder ethically speaking to support chinese companies (alibaba in particular) and amazon by buying from them? What alternatives are there, if any (manufacturer wise, i know about western sellers like digikey etc)?

>>1365839
Convert your transformer to 3 K-dependent inductors first?

>>1365843
>increasingly
If anything, Chink ethics seems to be improving (very) slowly.

>>1365876
>If anything, Chink ethics seems to be improving (very) slowly.
Aside for the newly implemented country wide facial recognition network and social rating ledger, not to mention their questionable ties with N.Korea and Russia, borderline imperialism with Taiwan, HK, Tibet and India, as well as their recent decision to change their government from "democracy" to quasi-dictatorship and lastly their complete disregard for foreign intellectual property.
This is getting a bit off-topic though so I refer back to my first post >>1365843

>>1365843
This is /diy/. What is your personal course of action?

I know the basic idea of working with generating electrical currents via induction, but are there any good resources that go into detail about the formulas and specifics of how many loops in a coil makes how many volts and sizes and all that jazz? I'd like to do more than just looting copper coils from broken electronics and begin using my own designs in CAD and stuff. I'm not an electrician, but I had an idea to generate a high amp/low voltage pulse of electricity via a spring loaded magnet that would run through a coil when a button is pressed, which would then require manual resetting of the spring and magnet via a crank. The setup would be enough to power a series of LEDs that would flash on really bright for about 10-20 milliseconds and then disperse.

so I'm thinking about making my own hdmi switch

>HDMI is a fast signal which needs to be routed and switched carefully - you can't just use a few transistors on a protoboard!

why is this? can't a theoretical switch between two hdmi signals be super simple technically? all it has to do is pass on the preferred signals, right?

>>1366033
>HDMI is a fast signal
you answered your own question,
as frequency increases impedances that used to look like simple resistances start to look like inductance and capacitance.
things act like antennas when they are longer than 1/8th of a wavelength, at high frequency this becomes important you need to avoid antennas radiating shit everywhere.
things like bends are discontinuities in impedance and cause reflections which fucks you up.

Is this a good idea? I am thinking of running the lights on my backyard with my 12v solar cell + battery, but I want to keep the same switches inside my house, can I just use an optocoupler to sense when the switch is on and just use the power from the batteries?
Pic retarded, my current wiring vs the new wiring.

>>1365931
actually, I'm finding it much easier to buy from China for those very same reasons (minus the imperialism) because it gives the bourgeoisie heartburn. but thanks for asking!
2/8 b8

>>1366027
Faraday's Law is a good search term to get you started

>>1366058
sure, but I don't see that it'll buy you anything. perhaps consider powering the switching element from the battery, which would consume zero power when off.

>>1366058
solid state relay

>>1364596
You've just drawn a faraday cage there, m8.

>>1365839
How come? You can deduce the current on the first loop with the given info about the power dissipated by the resistor, so if you wrote your equations properly you should be able do find K together with the current on the second loop fairly easily.

>>1364793
I'm anon from this post

So, today I went out and bought a 10$ soldering iron and lead solder.

When I got home, I sat down and started trimming the wire, I first desoldered the loose part and put it away, since it would be easier to solder the rest of the wire to the metal than soldering two wires together.

It took me two tries to get it right, since the cable slipped on the first try.
After I was finished, I connected the whole thing to my phone and played some music and, surprisingly, IT WORKED.

Pic illustrates the final result, rate it.

>>1366218
That's really cool, anon, nice job!
Although quite simple, this kind of little repair feels really rewarding, doesn't it? Most people wouldn't even bother, they would just throw the whole thing in the trash.

>>1366227
>this kind of little repair feels really rewarding, doesn't it?
Truer words were never spoken

>>1366156
I don't see what that has to do with charges and capacitance.

Would it be easy to make a bode analyser? Guessing you'd need to use a microcontroller, for which the frequency measurement is relatively trivial provided you've got a sufficient sample-rate, but how to produce an arbitrary-frequency sinusoidal waveform up to high frequencies is a little beyond me. I want to use this for testing radio tank circuits.

>>1366296

>>1366315
Well 125MHz is a tad borderline for RF measurements, since I'd like to be able to see if my resonant frequency is some degree above the FM range. But all the direct digital synthesizer ICs I've searched through are somewhat pricy. $30+ on Mouser, and on ali I wouldn't trust these from them. They seems to specialise with arbitrary waveforms, something that I have no need for. A 1% tuning accuracy sinusoid is all I really need. I guess I could just use a digital variable oscillator and very high order variable filter to turn that square into a sinusoid, but that sounds like the long way around even to someone who doesn't know how PLLs and such work.

So I'm really just looking for the correct terminology for "digitally controlled sinusoid generator", and possibly some method of reading the amplitude of the signal if the microcontroller can't handle those high frequencies. If DDS is the right term, then how would I go about narrowing down the ones that can produce a sinusoidal signal by frequency and price? Mouser and Octo also include what are essentially digital square wave/clock generators.

>>1366343
network analyzers were never cheap. if you need a precisely shaped sinusoid, and I'm not 100% certain you do, I would think a 3rd order filter would be enough?
>and possibly some method of reading the amplitude of the signal if the microcontroller can't handle those high frequencies
there are all-in-one radio ICs that take RF in and give IQ out, see the HackRF One for a design example. with IQ you can read both phase and amplitude of the signal of interest

>>1366343
>above the FM range
Why don't you simply excite the LC circuit and use your FM radio to find out? Look for FM bugs to see how it is done.

>>1366349
Well if memory serves the next harmonic on a square wave is f*3, so at -60dB/decade (assuming the fundamental frequency itself is on this slope also) it would be diminished by -18dB, or a factor of 0.13 or so. 4th order (because active filters like multiples of 2) would give me -24dB, or a factor of 0.06.

You're right in that this is probably sufficient, I imagine any artefacts induced in the Bode plot will be fairly innocuous and ineffective near the circuit's resonant frequency. If I do use a filter with its corner set lower than the minimum frequency (100kHz or so I guess), I could simply not change its corner and use some sort of automatic-gain control to boost its amplitude, or otherwise simply measure the filter's output amplitude and that at the end of the circuit I measure and use that as the gain baseline, but that only works if the amplitude after a high frequency circuit is sufficient to be measured. Not using an AGC could lose me many bits on my output resolution, though using 2-3 switchable filters is an alternative.

>>1366373
at a guess, maybe he's looking for self-resonance on the L

>>1366377
might consider varactors, though tuning those up could be a problem all their own

>>1366373
Well I want to test on a variety of different circuits (many kHz to hundreds of MHz), and get myself a Bode plot, not just its resonant frequency. A Bode plot will tell me of any parasitics and such as well, and is functional for non-radio tank circuits.

>>1366378
>varactors
I didn't want to suggest those because I assumed nobody in their right mind would want to use them. But that sort of circuit certainly isn't off my list.

>>1366379
there was a fairly detailed diy rf network analyzer in circuit cellar like 10 years back

I'd really like to know of decent flux brands, especially chink-y that I could get from anywhere in the world from the likes of eBay or AliExpress.
Anyone could help me?
Thanks!

>>1363100
I want to make a little led blinking pcb business card type thing and I want to keep my part count as low as possible. is it a bad idea to use 1 resistor for 2 leds in parallel?

>>1366536
pros: smaller BOM
cons:
>possible Vf/intensity differences between the two
>if LEDs are kept on for a significant time, Vf/intensity differences increase and could lead to violating If specs

>>1366426
all the "Amtech" is either fake or blended

Is this the appropriate place to ask two stupid questions about electronics?
One about a resistor and another about voltage

>>1366642
probably

>>1366645
I have a blender which I took the motor and speed control board out of to look at
It still works as it did in the case
my analog multimeter shows ~120 volts from the wall socket as expected
When the motor is running and the meter's probes touch the input and output from the cord to the unit it shows ~20 volts, shouldn't 120 volts be showing? this unit as far as I can tell has no transformer and seems to only consist of a handful of resistors, a triac, and a capacitor

the second and much shorter question is relatively dumb; when you insert a resistor into a purely resistive circuit all that changes is the amperage going through that circuit, correct?

>>1366642
you must have over 200iq to post on this board

>>1366666

>>1366658
An analog multimeter will act akin to a true RMS DMM, at least approximately, so the voltage you're measuring (assuming it's after the triac circuit) will be the heavily chopped-up PWM waveform. Look up how triac dimmers work if you're unsure.

Depends on what you mean insert. But in general the voltage across the input terminals of your circuit (assuming an ideal power supply, which is relatively close to reality) will remain constant, and will not change by adding or removing a resistor. But the voltage across each component will change. If you have a resistor and add another in series the voltage across the original resistor will decrease. If you have two resistors in series and add another in parallel to one of them the voltage across the other resistor will increase. The current will change as a function of these changes. See: Ohm's law, Kirchhoff's laws, Thevenin and Norton equivalents, Millman's theorem.

>>1366751
Thanks, that answers my questions nicely

>>1366218
Hey, sameanon here

Today I went out again and bought a third hand, as soon as I got home I started working on a old lightning cable that was broken,

I stripped the top part and realized that the red cable was cut, so I trimmed it (which was difficult since it was so tiny) and soldered to the whole thing together... it didn't work.

So I watched a youtube video and decided to open the whole thing (took me 2 good hours, who knew that a knock-off cable would have such strong rubber). When I was about to start soldering, I realized that I had damaged the right side during the process of opening the thing, it broke.
Since I don't know how to repair such a tiny piece, I gave up for the day. Pic is end result.

I really want to start building some shit with boards and connections, reaparing stuff may become boring at some point, wich components do you guy recommend I buy first?

>>1366793

sounds like you're the right target audience for one of those 300-in-one kits (or 150-in-one if you're cheap). make sure to get one with a breadboard, not the retarded ''spring'' ones.

>>1366814
>300-in-one kit
looks like something you could 3d print the body of and make with aliexpress components

>>1366836
could, but they also come with a project/study booklet, which doesn't entirely suck

>>1366751
>An analog multimeter will act akin to a true RMS DMM
Typical analog multimeter measures (rectified) average and has its scale calibrated for sine input. It's quite different from RMS, albeit probably more useful in this particular case.

>>1366865
Hey I said approximately. The point is that it's not measuring the peak but some form of average.

hey guys, i need a new DC power supply, i was thinking of these two models
Rigol DP832 €500
Itech IT6322A €580 + 21% customs tax via aliexpress
they're both 3 channel 2x30V 3A 1x5V 1A 1mV 1mA resolution
the itech has the same hardware as the B&K 9130B so thats pretty neat, also i love the VFD display + remote sensing, but its still chink and more expensive, so what do?

>>1366981
Itech, no question. Rigol has a built in ground loop and the design is (as usual) an atrocity.

Hi people hopefully someone here can help I'm an idiot who is playing with electronics.

I was trying to install a Bluetooth adapter for my car radio and I ended up cutting the constant wire instead of the accessory wire. The radio didn't turn on with the BT adapter "installed" and after reconciling the wires back to their original configuration now my radio doesn't turn on at all. All the wires are how they were before when the radio was working. I've checked the fuses and neither the radio nor car's fuse is blown. What do to troubleshoot?

>>1367088
did you read the radio's manual to check for anti-theft unlock procedures?

What's the difference, or will it be a bad thing if I focus on a processor (STM32F446RET6) but instead use the dev board (B-F446E-96B01A) that uses the processor (STM32F446VET6) instead?

What are the differences between the STM32F446RET6 and STM32F446VET6? Am I losing anything here? Also, I thought Keil tools would have had libraries for this, maybe I'm being a retard and not looking at the right place.

>>1367165
packaging. consequently, some alternate pin functions may not be available and your choice of pinouts and feature combinations may be restricted as a result. see the datasheet for more details

>>1367177
Hmm, the thing is I have to pick a processor for a class something, but I have no idea where to look for. I've browsed mouser for a while and found some things, but are there terms to look for? For example, I didn't know that is called the "Ordering information scheme" anything else do you think I should look into other than the datasheet?

>>1367158
I did, there doesnt seem to be any anti-theft procedures.
I just noticed that my car isn't making the usual beeping noise when the key is turned in the accessory position, or when i take the keys out of the ignition with the lights still on...

So it must be a fuse in the car? Wtf did i do ;_;

>>1367188
nah, the STM32F446?ET6 are equivalent aside from packaging. the V package is a strict superset of the R package. you can develop with the F446V board without issue, and can load the very same binary into an R package part as long as you don't use functions/resources that aren't pinned out in the R package. just compare the pinouts in the datasheet

>>1367195
Ah alright cool. Datasheets are amazing but there are a lot of stuff out there, it's hard to keep track since I'm kinda new to these. I went and got arm keil since we used it in one of our classes a semester ago, but I'm still trying to figure out if it even supports this devboard or the cpu itself. It has a arm cortex m4 section and an STM section, but it's weird because the devboard I chose isn't there.

I guess there is a different programming environment for this specific for testing? I'll look at the STM website and see if they have any tools specific to use, or any libraries that I can use gnu-gcc to compile with. Thanks for the feedback Anon.

>>1367051
Thank you for your input, I think I will go with the Itech.

>>1367196
peripheral locations may differ, but I don't see why Keil would care about that. it's hard to keep up with all those devboards. an STM32xy processor has the same core, for any given xy
>a lot of stuff out there
yep. ST's site has fair-to-good parametric selectors. if you narrow it down to the core you want (according to performance) you then choose peripherals and packaging to match your application's I/O needs
>specific for testing
not really, the dev environments just build differently for debug vs. production
>have any tools
Atollic TrueSTUDIO 9.0 is an Eclipse-based IDE that's free for STM32 targets
STM32CubeMX generates project files for several workflows, including gcc+make. it also helps massively with clock/pin planning and peripheral init code

>>1367201
Thanks, I'll have a look at these tools soon. I hope they have debuggers as well.

>>1367203
TrueSTUDIO does, indeed, have in-circuit debugging facilities. fwiw, it doesn't work very well with my cheap clone ST-Link interface, but ought to work fine with anything on an official ST eval board
otoh, the ST-Link tools for Linux do work nicely with my clone adapter, gdb, and the gcc+make workflow

>>1367051
>>1366981
I've never bought electrical testing equipment, but I can kind of justify to my mind that an oscilloscope would cost 300+ usd
Why is a unit that can output DC power at less than household ac voltage and power ~800 usd though?

a friend and I are making a fursuit. I want to embed speakers into the head because the things are so damn hard to hear in.
I want to make an amp out of an LM386. I don't need the speakers to be stereo, but everything I'm reading says I need a separate amp for each speaker. Can I get away with just daisy chaining the speakers together or will I catch myself on fire?

I have a motor hooked up to a pulley system that pulls rope similar to a ski tow rope. I want the motor to reverse directions after a set distance or time. What kind of control can I use wired in with my power source and motor that would send a reverse single to the motor without a switch having to be physically triggered?

>>1367218
accuracy, calibration, stability, low noise, remote control, all combined with low volume.

>>1367224
if you want any chance of locating the sources of a sound by ear rather than looking around bewildered, you'll want a separate mic/amp/speaker on each side

>>1367225
you'd need some sort of trundle wheel arrangement to sense distance. this is more than trivial.
what do you want it to do after it reverses direction?

>>1367224
>>1367227 is right about having a speaker for each ear if you can
I made an altyn helmet and I couldn't hear shit through the foam so I integrated a speaker setup into one ear which was okay but annoying
later I installed another after a friend showed me his asmr listening pillow
having simple mics discreetly placed on each side of your helmet/headpiece gives you great binaural direction-finding in my experience if you can rig them up

>>1367225
>time
maybe just an on-delay timer relay would be enough, but your system would be "dumb" in that it won't care how far you've pulled your rope, so heavier/lighter loads will give different results.
>distance
encoder and microcontroller?

can a good soul explain what the heck this does and how does it work? this is from a old test from my teacher, I`m trying to study for my electronics test but this and a couple of others got me stumped

>>1367240
Vi is a sine wave btw

>>1367240
looks like you would get a full wave rectified signal out of the "Vo" but I'm really really rusty on circuits

>>1367228
>>1367232
I want the motor to reverse forwards/backwards in repetition, the load size wouldnt change.

Okay so it seems distance is too technical, one based on time is much simpler. After looking around at on delay timer relays I think a dpdt with a timer on it is what I need. I cant quite tell if this product would work, but I may just buy to figure it out.
http://vi.vipr.ebaydesc.com/ws/eBayISAPI.dll?ViewItemDescV4&item=271968911625&t=1510085366000&tid=10&category=36328&seller=canton-electronic&excSoj=1&excTrk=1&lsite=0&ittenable=false&domain=ebay.com&descgauge=1&cspheader=1&oneClk=1&secureDesc=0

>>1367287
>>1367241
thanks anon, it seems to be it. And what about this one? He wants me to find I0/vin. The answer I found was -1/R. which is weird.

>>1363100
I am working on some solar panels calculations and I was wondering about the effect of the angle on the efficiency. Can I just multiply my solar panel efficiency with cos(angle relative to the sun)?

>>1367459
yes, both azimuth and elevation.

Why aren't you subscribed to the most underrated electronics youtube channel, /ohm/?

https://www.youtube.com/watch?v=CLnolhyT5SI
https://www.youtube.com/watch?v=hT0aG3SRpRY
https://www.youtube.com/channel/UCOepzhfgouq_rPxDx-kCn7Q/videos

Have a look, these guys have such good production value and interesting topics.

Hi I have some questions for my arduino project...not sure if this is the correct thread to ask these.

I am building myself a bee hive scale. It currently consists of 2 120kg load cells connected to an HX711 from where the arduino gets its values every 30 mins. These values (+ an optional DHT22 for Temperature/Humidity) are sent to a website via the GSM cell phone network (SIM 80x module). So far all the readings & sending the data is working. What I still have to do is the permanent cabling & figure out how to power it.

My questions:

a) How sensible is the arduino to humidity. Is it viable to mount everything on the bottom plate as shown in the picture? Should I coat everything in clear paint spray or sth like that? Put it in a box with silica gel? (connectors?!?)

2.) To conserve power I want to put the arduino into sleep between readings. Would you suggest some sort of external low power yC for simply turning on &off the power to the arduino board? When removing the LEDs on the arduino do I have to replace it with sth or is it fine to leave that spot empty?

3) In the pic you can see a "hole" in the upper lid for connecting the DHT22 (& maybe possibly a small solar panel+PWD should I not be able to get acceptable run times & possibly additional sensors e.g. "laser curtain" for counting bee flight activity etc.). How would you seal it? epoxy resin?

4.) Energy source: So far I have planned on a 10Ah power bank due to the advantage that it already has the 3.7=>5V converter built in and has 2 power outs (1 for the arduino and one directly to the GSM board since the arduino can't relay enough power). Any other ideas?

5) permanent cabling: I'm planning to soldering the cables directly to the arduino since I don't trust the cheap pins. How important is isolation? I would have simply gone completly without it since it is only 5V but it is my first time playing with electronics so what do I know...

Thanks in advance.

I'm sure you get asked this every other thread, but fluke 17b+ for $135: yea or nay?

>>1367635
a) not sensitive at all, it's condensation that will kill it. i would get an IP65 box (anything plastic with a non-foam gasket really). glue everything down inside. screw a flanged terminal block to the outside for connections (pic). seal holes with whatever you've got lying around. if you want to make it nice you can add some input protection. ideally diodes+resistors on data and a reverse protection diode+fuse on power. you can probably get an ebay board to do that but perfboard is fine.

2) you can leave the leds off. if you want to get creative you can desolder the 16MHz crystal + caps and solder on a 32.768kHz watch crystal + caps. you can run the atmega off its intosc at a few MHz for control, and then use the "power-save" mode to sleep while leaving timer2 running off the crystal as a wakeup. if you do that it's just one less gay little breakout board to have floating around.

your breakout boards will draw significant current as well even when you're not doing anything with them. set up a load switch for the arduino to cut them off entirely. in fact, everything draws annoying quiescent currents. the regulator on your arduino does. the boost converter in your battery does, right up until it switches off completely at light load. see 4.

3) no need to seal it as far as i can see if you put the rest of the project in its own sealed plastic box.

4) your arduino and hx711 can run off 4.2-3v. the GSM module probably can too but check. if so you can skip the power bank and just use a protected lipo cell with a charger board. the two problems with your power bank are that its boost converter will waste power at light loads and might even cut off if the load is low enough. if you get rid of it you can just wire the battery supply directly to the arduino's 5v pin. beware of oscillator frequency limits at lower voltages. run at 4MHz, not 20. i'm too lazy to check the datasheet to get the right numbers.

5) you're already isolated from ground so don't bother.

>>1367547
shill

Not sure if this is the right place to ask, but

Are there any (digital) multimeters that will do accurate true RMS measurements of AC voltage up to 20 kHz? (that also won't break my fucking bank)
I've been meaning to buy a decent multimeter to complement my el cheapo no-true-RMS one for a while and was considering getting something from BK Precision but they won't do anything over 1 kHz
I'd like to avoid buying an old school bench top AC millivoltmeter in addition to another DMM

Other features I'm looking for are frequency counter (don't need much over 10 kHz) and capacitance measurement

>>>>/g/65525413
Just looking at this thread on /g/, I can't figure out what happened. He hit the LED bulb, and damaged the socket such that it put too much current through the bulb and blew a trace on it after a short period of time. Screwing in another bulb caused it to blow in a few minutes also. Any ideas? Never seen a trace blow either, since usually there'd be passives or transistors that would open-circuit first. There are plenty of things that would cause this like damaging a dropper capacitor by the dent in the socket, but the second bulb had no damage before it blew.

>>1367920
Also I should mention that I do have two oscilloscopes but they're both analog (and uncalibrated) and thus not really useful for the task
Ideally I'd like to use the multimeter for that as well, patching it in line with one of my scopes
Also interested in suggestions for bench top DMMs if those are more likely to fit my requirements (particularly if they can be had for somewhat cheap on the used market)

>>1367920
cheapest way would be to build your own RMS-to-DC converter and hook it up to whatever meter meets your other requirements

>>1367920

>>1367920
Not sure how accurate you want it or what you call breaking the bank, but the Uni-T UT71A/B claims 100kHz RMS bandwidth, can get them for around 100 burgers.

>>1367923
Not looking to take on such a project, and I'd need a meter anyways to calibrate it

>>1367925
Manual says +/- 7% above 10 kHz, what's even the point then? I'd like to stay under ~200€/$
Also I've heard a lot of shit about Uni-T in general, what's /ohm/'s opinion on them?

I need it for calibrating and checking frequency response of audio equipment, so nothing NASA mission critical, but I really need more than +/- 7%

>>1367927
Doesn't require RMS. 7% is 0.6dB. Use a scope.

New to all of this, currently trying to implement a project into one power cord.

Since I have a lot of components in parallel with different current demands, what's the most reliable thing to connect them up? They are splitting between ac lights and a few power supplies i am just going to rip apart and connect to the main power line.

How would you guys better disperse the light so it hits the screen more? I was thinking a foil tent on top but I wouldn't know how to secure it without blocking some of the light from the top of the bulb

>>1367972
I don't know where the rest of the message went

I made a ghetto kindle light out of a straight-up LED. Being a bare LED the light likes to go a lot of places besides the kindle's screen. How would you guys better disperse the light so it hits the screen more? I was thinking a foil tent on top but I wouldn't know how to secure it without blocking some of the light from the top of the bulb

>>1367964
.combine voltages where possible by finding the actual voltage ratings of your devices instead of whatever you're happening to give them. run low power components off of linear regulators pulling from a higher voltage offline supply

>>1368022
This only works if you're putting a constant/equal current through them. Can't have 2 switched-mode power supplies in series across 240V if you only use one at a time. A multitap (auto)transformer may well be a better choice if you've got a lot of variable-current loads.

>>1367964
If everything's regular mains power, just use a regular power strip.

>>1367964
>what's the most reliable thing to connect them up
ultrasonic splice weld or hydraulic electrical crimp
>but that's not really what I meant / it's out of my reach
or use hand crimp connectors, twist connectors, or a busbar (common) screw terminal block

I have a current shunt that has a voltage drop around 90mV max. I designed this thing to amplify the 0-90mV range to around 0-5V. I built it and it only amplifies 0-70mV to 0-3.6V and then it stays at 3.6V no matter how high the input goes.
The amp is an LM358, with a 5V supply. The bottom resistor to 5V is to offset the lowest output to some 100-200mV instead of 0V

Why doesnt it go full scale?

>>1368168
>LM358
They only go up to Vcc- 1.5V. You need a rail-to-rail amplifier, or better yet, use a higher supply voltage. A 9V battery or 12V PSU are good choices. Typically audio circuits don't run off such low voltages, hence the terrible range when you do want to put them on 5V. Changing your gain to 35 should do the trick. Looks like the 358 can handle 0V, but otherwise I wouldn't want to feed 0V to its input instead of a ground rail, assuming that the opamp negative rail is the same as the AC 0V rail.

>>1368168
If you want to hit 0-5V you'll either need an opamp that can get closer to the supply, or a higher supply voltage.

>>1368057
not what i'm getting at. pic related is an example of four wall warts compressed into one.

>>1368222
that would work

>>1368222
So long as your 5V/3V devices aren't that current hungry it's fine. If they are it'd be a better idea to use step down converter modules rather than linear regulators, or get a power supply that does multiple output voltages such as a PC PSU which does +12, +5, +3.3, -12.

Going to try printing my own arduino because I want the signal cords and sensor cords to be fastened to the board. What type of connection should I use? Haven't ever tried to make something permanent before.

>>1368364
absent any special requirements, 2.54mm "Dupont" pin headers and the matching crimp-on terminals are as good as any. lolduino did it backwards (of course) so put the male pins on the board and the female connectors on the wires, except for the USB, which is better left USB

>>1368364
"JST" connectors are a respectable choice to "Dupont" that's lower-profile and as easy to find. there are dozens of whole catalogs of connectors if you decide you need something a bit more special

>>1368370
>>1368371
Cool, thanks for the suggestions. Depending on what I find locally, might do both.

I also have some screw terminals floating around, opinions on using those as well? It's not like I'm going to be removing these all that often, I just don't want to solder it to the board.

>>1368389
crimp-style connectors and tooling are best sourced chinkshit. screw terminals are more reliable, easier to field-service, and easier to source locally, at the cost of density

>>1368173
>>1368176
Right. Thanks.

Was tinkering around with a Samsung wireless fast charger and somehow it's shorted on the input. I originally had plans to add a fan and 3D print a new case so it holds my phone better. Now I'm trying to figure out how its shorted out. Any tips on finding the culprit? I have a lab bench power supply, a multimeter and a very strong magnifying glass at my disposal.

>>1368417
Perhaps you blew its reverse protection diode into closed-circuit? Show the PCB otherwise all we can do is guess. A 1080x1920+ camera better be in that arsenal of tools. I'm also a little interested in seeing if they differ much from an induction cooktop (and whether the two are interchangeable).

>>1368573
All I have is my phone and shitty lighting. I've never done SMD work but I'm not opposed to learning! I don't have any SMD tools right now, but I might be able to spend up to $100 if you have any recommendations for a beginner.

>>1368573
>>1368584
Oops

>>1368584
Well it will either be a quick fix or a throw-away and buy another job.

>>1368586
The bottom also? Or does that 2x4 array of vias lead nowhere? I'd do a diode check on the two components (diodes I'm assuming) near the USB input, and a resistance check on those 3 resistors near them. From there continuity on any +to- capacitors (probably just the larger ones, but there are a dozen), all to ensure nothing's gone short. Make sure you're not making a circuit through a reverse protection diode when you measure continuity of a cap though. How were you powering the thing?

>>1368594
Wouldn't I have to remove each component to test them? And I originally powered it with the wall plug that came with it, but when it wouldn't power on, I plugged it into my bench power supply to see if it was drawing any current. The bench power supply hit the current limiter instantly.

>>1368594
>Or does that 2x4 array of vias lead nowhere?
Why are you trying to help this person when you have absolutely no idea what the fuck you are talking about?

>>1368601
It should be fine to use the tester on it, a volt won't surpass any breakdown voltages. Besides, it's already broken.

Now is that dark mark on the back just an artefact of the lighting? If not it's evidence of overheating, and it's coming from the transistors. It looks like the array in the left-middle is a half-bridge since those 4 "B1H" transistors appear to be two lots of parallel pairs, which I guess works since R_DS ON increases with temperature, but it's cheaping out on a larger package. Since all 4 are connected to the positive(?) rail, two must be for one coil and 2 for the other, meaning the transistors to the right of that must be the positive transistors of each half-bridge. One of these looks like it's got a pockmark on it, though at first glance it looked like a smattering of conformal coating. Now this could have happened from putting the whole thing in reverse, as the FETs' body diodes would conduct and cause a mess, or perhaps it overheated in normal use, either way the 10W maximum that a USP PSU will output is more than enough to blow one of those tiny SMD packages. If in reverse, chances are one of the 8 transistors was going to hit thermal runaway worse than the others, and once this had happened the PSU cut out, but not before frying a transistor. Since the two lots of transistors are labelled differently (and it's only running of 5V) the transistors on the right are bound to be P-type, and so require 0V at their gate to turn on, so the thing was shorting before the MCU/driver even got time to boot.

So I'd check those diodes because (if they are reverse protection diodes) they were probably fried also when you blew the transistor, along with any possible fusable resistor, assuming they're across + and -.

>>1368618
Well they're current carrying for the +ve line, not heat-sinking like I originally thought. I assumed that the main current-carrying traces were on the top.

https://www.aliexpress.com/item/High-quality-220V-Portable-BGA-Rework-Solder-Station-Hot-Air-Blower-Heat-Gun-8858-Better-Hand/32808359589.html
+
https://www.banggood.com/ANENG-AN8008-True-RMS-Digital-Multimeter-AC-DC-Current-Voltage-Resistance-Frequency-Capacitance-Test-p-1157985.html?cur_warehouse=CN

Y/N? if not, why?

>>1368636
cont.
So if you can find out what model of transistor that is, which isn't always straightforward, you can buy another one and replace them. You can brute-force the old one off with wire-cutters if you can't desolder 3 traces at once, but one at a time should be sufficient with a little patience and wick. But I'd replace it with a TO-220 or equivalent package suspended off the board with some 1mm solid-core copper wire, since it would avoid this problem in the future, if there's room of course. But matching the values of the components is a bit of a pain. There's also 2 more big FETs on the left side I didn't take account of, but it probably doesn't matter. Some of the other components, especially the neighbour FETs will possibly have a decreased lifespan thanks to their exposure to high-heat, so be prepared to replace more in the future.

Do you know if you could have put a reverse voltage across it at any time? If not and it's just a result of overheating through normal usage/shorting something when it's hooked up correctly then you're probably fine without checking any of the diodes or other components, the status LEDs wouldn't light up regardless because it's shorting before boot VIA the FET, causing the USB PSU to shut down. A test would be to desolder that FET and try turning it on, it's (probably, check the traces) got another one in parallel and will work at low/no-load, or if it doesn't have one in parallel (like 50% likely) it should boot fine but without working. But in that case I'd put a freewheel diode across where the drain-source diode used to be in case you get a shot of current flowing through the coil as a result of the corresponding N-channel FET.

I asked this in /hard/ but i'll ask it here too

Can anyone recommend textbooks for digital electronics? Particulary focusing on things like cmos logic, transistor sizing, synchronus and asynchronus signals, setup time, hold time, and propagation delay, metastability, and verilog, if possible

>>1368640
I have the AN8002 and I'm very pleased with it. I think it's reasonably accurate

>>1368594
>diodes I'm assuming
TVS, probably. not connected to the power rails, therefore irrelevant

>>1368601
check to see if the USB jack itself isn't shorted

>>1368640
>rework station
looks bretty gud, can't speak to quality or lack thereof. parts should be cheap, in any case

>>1368644
>TO-220 or equivalent package suspended off the board with
just go away and play with your arduinos please

>>1368743
>I have the AN8002 and I'm very pleased with it. I think it's reasonably accurate
>>1368744
>looks bretty gud, can't speak to quality or lack thereof. parts should be cheap, in any case
thanks m8s

has anyone found a ready-made cables kit in aliexpress? I want to buy many parts at once, and I'm too lazy to look for them atm...

BTW, I don't know much abaut electronics, RF and stuff, barely the basics, but I found these cool things I wanted to share just because:
network analyzers:
https://www.rtl-sdr.com/rtl-sdr-tutorial-measuring-filter-characteristics-and-antenna-vswr-with-an-rtl-sdr-and-noise-source/ https://www.youtube.com/watch?v=SAAF2FvV9g4
https://www.reddit.com/r/RTLSDR/comments/55nzdf/antenna_analyzer/
https://hackaday.com/2018/03/29/putting-a-poor-mans-vector-analyzer-through-its-paces/


spektrum
https://www.rtl-sdr.com/poor-mans-spectrum-analyzer-with-an-rtl-sdr-and-noise-source/
https://www.rtl-sdr.com/spektrum-new-rtl-sdr-spectrum-analyzer-software/
https://github.com/pavels/spektrum/releases

#188 Antenna Tutorial incl. cheap DIY Antenna Tester (LoRa, ESP32)
https://www.youtube.com/watch?v=J3PBL9oLPX8
#191 Optimizing Antennas using a cheap N1201SA VNA (Vector Impedance Analyzer), LoRa
https://www.youtube.com/watch?v=ZpKoLvqOWyc

>>1368744
>just go away and play with your arduinos please
Oh boy is it hot garbage, but it's not like you can SMD solder on a package with higher heat tolerance, assuming normal use is what caused the fault. An overkill package is about all I can think of to stop the problem from occurring again, though I guess one of those tiny heat sinks slapped atop the plastic package could do something. What would you do instead? Try to get a transistor with a lower R_DS? I was assuming they picked a near-optimal transistor for the job. Also it looks like the parallel FETs are not activated at the same time, suggesting they're really pushing the bounds of the frequency handling ability of some of their components, else some sleazy switching logic.

>not connected to the power rails
Ah you're right, the +5V enters the ICs though a via halfway down the main power trace.

>>1368748
>cables kit
like what kind of cables?

>>1368771
>assuming normal use is what caused the fault
the anon had been fucking with it. abuse can't be ruled out
>tiny heat sinks
could also use thermal glue to attach a thin aluminum plate or a piece of copper roof flashing across all the transistors, and hope you don't ever have to service it again
>Try to get a transistor with a lower R_DS
maybe, if in fact it wasn't just infant mortality from one of the FETs popping and the others having to share a greater load, or an issue with the device being charged being too far away from the coils, or something metal laid on top of it
what you see as conformal coating looks like fried FET to me

quick and dirty layout review, /ohm/fags?
Tayloe detector. upper is layout: left is bottom, right is top with bottom in background. SMA inputs on left, from top: RF, clkQ, clkI. box to be assembled from copper sheet.
this is my first RF board. in particular, I'm interested in whether I did the ground separation even close to right, with separate ground planes for: bias & RF, mux detector & caps, diff amps, all meeting at one area near the regulator. should I put a copper shield sheet on the bottom side too or can I get away with using the ground planes as shields and just box the top?
other thoughts?

>>1368794
Well all I know about RF boards is that you should keep your traces bending with round corners and not changing diameter, of which you've done 1 of 2. Is it common practice to make traces that change diameter just before entering a solder pad?

>>1368799
afaik, sure. in some cases I had to do it on the 0603s so that the board would pass DRC, but I see the RF input trace looks messy and could use cleanup. I could shorten the neckdowns a bit on the clock traces too

>>1368776
>like what kind of cables?
all kinds of cables (jump wires, probe cables, crocodile cables).
I'm buying electronics equipment for learning, and I don't even have decent cables, plus I don't know what kinda cables I might need in the future, so if you can suggest some, please do so

>>1368904
>all the cables
those kits are frequently stocked as separate items, so you could just buy them separately. there are loads of sellers on aliexpress who sell these things. ElectronicFans/ModuleFans ships quickly and accurately ime, and usually they're my first stop when looking for something within their purview
besides, you'd almost certainly be annoyed at the choices someone else has made for you were they to package everything all into one kit. I frequently am

>>1368904
you can spend several hundred burgers on wires and cables and still not have one of everything. if just getting started, some male-male, male-female, and female-female "Dupont" cables would be good to start, so that you can connect to and from the numerous breakout boards available. I survive well enough with a set of 10 alligator-clip cables, never used them all at once. it helps to have a set of test leads for your meter that have screw-on alligator clips. that's a good generic start. from there it really depends on what you're doing, some people need SMA cables and other RF goodies, others need a ton of USB power cables for power, ...

>>1368636
>>1368644
Holy shit that's a lot of info, but thanks for your help. The charger has been getting really hot, which is why I was going to add a fan to it, which is one of the main reason I took it apart.
>>1368776
>the anon had been fucking with it. abuse can't be ruled out
All I did was take it apart. I found out it didn't work after plugging it in to test a layout design. I never messed with the circuitry.

I think I'll just end up buying a new one. They are only $15 on ebay, so not a big hit. Still disappointed that I only got three uses out of it.

>>1368925
>>1368926
thanks a lot, guess I'll have to be patient, because having to look for every single item (their quality, length, etc.) will take me a lot of time...

crossposting from /gg/ so not strictly electronics but i hope you guys can help

I have a crate 220h amp head which has 2 6.3mm outputs to the cab which also has 2 inputs. I've only had one cable connecting them and it works fine so far (haven't really cranked it up)
I also want to use this as a speaker but my 4x12 cab can't do the high end.
Can I connect a different small speaker/cab to the second output from the head to try and get some treble without blowing shit up? The second small speaker is 8Ohm, it's only 10W but I barely leave 1/10 so shouldn't be putting out much power. Furthermore could I solder a simple high-pass filter to the small speaker so only 5/6k up get transmitted and cut down the power going to it that way?

>>1369055
10W 8 ohm will go pop if you crank up the volume, but if you're leaving it low you should be fine. You're probably going to want a high pass but it's not absolutely necessary. May want to add an attenuator/big resistor too for level matching.

Does anyone use graph paper with 2.54mm grid spacing for PCB drafting, or does it probably not exist?

Whats a good dcdc converter that allows 12-50V inputs and gets me 5V? The ones i find on chinkbay dont go that high.

>>1369409
It exists. Just search for 1/10 inch spacing.
Or you can print your own.

>>1369427
How much current do you need?

Do opamps have a minimum gain or minimum amplitude?

>>1369568
>minimum gain
depends on the circuit and the opamp. iirc some can become unstable at very low gains
>minimum amplitude
not that I've heard of, but mind your noise budget

>>1369569
Ah, noise. It has an input noise level given by a nV/√Hz value, but this doesn't really mean much in this situation since the input voltage can't have been lower than a few tens of µV. Even still, it bottomed out at -55dB.

>>1369467
No more than 200mA.

>>1369611
Can probably make your own at that point, but modifying one made for 120VAC is another method, they rectify the AC anyway. They're good for ~80V to ~260V, which corresponds to 110 to 370 after the rectifier and filter cap, so just decrease the primary windings ratio by a factor of 10, same with the dropper resistor that powers the controller. Theoretically at least, I've never tried it. All you need to make your own is a dual comparator IC, a FET, and a handful of passives. It's a little more if you want a variable duty cycle though, but if it's a constant load that doesn't really matter and you can just tune it to be as smooth as possible at that particular current.

I'm trying to figure this out. Would the "design with D flip flops" map just be a copy of the first set since the input has to be a 1 for the output to be 1?

>>1369611
you could roll your own switcher using something like an MC34063

>>1369611
>>1369838
actually, this might be exactly the chip you need http://www.ti.com/lit/ds/symlink/lm5008.pdf

Does anyone have any advice on buying a second hand variac? What is there to fail & how hard are they to fix?

>>1369611
You could use one of the linear regulator replacement style modules https://www.digikey.com/product-detail/en/xp-power/SRH05S05/1470-4531-ND/8021459

anyone look through basic electronics by radioshack?

I've a Nikon D3100 DSLR, and I'm wondering about home brewing an external/IR trigger. Pic related is an existing wireless one, that square plug plugs into the side of the camera. Firstly, does anyone know what sort of port that is, and secondly is it possible to just try touching one conductor to another for that port and expect to trigger the camera? That thing costs $200, but this thing could be promising:
https://www.aliexpress.com/item/Im-not-really-into-rape-culture-I-think-its-a-little-forced/32799512765.html

>>1370375
http://pinoutguide.com/DigitalCameras/nikon_d90_pinout.shtml
you will probably want to take some account of the half-pressed state too, if you want shutters actually synchronized
the connector appears to be Nikon proprietary, the easiest source may be one of those pendants which you would then reverse-engineer
dat link doe :D

>>1370386
Hey I've never seen that site before, that's pretty neat. I've bought one of those wired pendant things for $10 locally, and I'll throw in some timing circuitry to focus for 2s before taking the shot. This way I can trigger it with a single LED. Thanks for the help!

>>1370398
if shutter timing isn't critical, as it would be if you were trying to sync with a flash or another camera, you can just close both switches at once and the camera should take the shot as soon as it acquires focus.

>>1370399
I prefer a longer focus time, my camera is a little finicky over focus. I think I'll need a latch of some sort to store the data IR pulse. I might be able to use a 555 for this, though I'd still need another timing circuit. A 556 would be good, but I would rather throw 5 transistors and a handful of passives at it instead.

>>1370411
you might want something more robust to noise. those little nRF24L01+ boards are good for this but you'll need a uC to configure them
if you don't want to use a uC, how about an RC circuit and a resistor ladder of comparators? you could generate a square wave output to your IR transmitter, and use a standard op-amp bandpass filter to tune it in, followed by a diode and integrator to detect the presence of the square wave. the nice (or maybe not) thing about this is you have to hold down the IR transmit button until the shutter is released, so you can abort a shot just by letting go. Pic related is a very rough, tired-eyed sketch of the sequencing half

Im repairing a machine that is hermetically sealed, stuff is pretty moisture sensitive.
I have to bodge in a psu like pic related, problem is this needs airflow. The machine has water cooling running around inside i can tap, the water temp varies from 5-15C. What would be the best way to cool the psu with this water?

One idea is to run water through a heat sink and place it in front of the fan to blow hot air on it. Maybe even put a fan on the heat sink to suck in the warm air.
The other idea is to flatten copper pipes with water running through them and attach it to the sides of the psu where the power devices are.
I dont really want to mess around inside the psu.

>>1369838
>>1369840
>>1369941
>>1369825
Thanks for the replies, ill think about the options.

>>1370443
get as close to the power devices as possible. any residual heat distributed by the fan will probably be minimal

>>1370429
Optical noise you mean? I just did a 7 page lab report on designing such a filter. I'd use a band-pass filter centred at around 10Hz or so, getting rid of both 100Hz lighting noise and any DC or slowly varying parts. I'd then run that filtered signal to a comparator (or spare opamp) to turn that voltage peak into a square pulse, with a potentiometer on it to adjust its sensitivity.

I'd prefer a single (>100ms) click to activate it, and perhaps an adjustable delay via potentiometer. If using a 556, I'd have the first monostable doing the focus, and have it trigger the second timer once it finishes, which would just be there to pulse the capture line for 100ms or so. Putting the pot to control the focus time would be the way to go as there's no downside to holding that on for even a minute.

>>1370463
oh, I thought you were considering RF. nm that part
the action of the two-step shutter button is that the focus switch is closed at the half-press, and both switches are closed at the full-press. experiment to see if the camera doesn't mind having the focus switch open when the shutter switch is closed

please teach me how to get op-amps to work?

>>1370470
Will do. RF might be a better idea but for now I'll keep it simple. Not like I have any RF modules lying about anyways, but they would be nice for longer range shots. I can always use a couple of RTL/RDL logic gates to keep the focus line down, though I may need a little delay buffer also in case the focus line flickers between the two timer cycles.

I'd also like to be able to be able to use this for night-time long-exposure shots, so adding some sort of timer to do that would also be neat. A switch on the second timer should do the trick, but I've heard that long-period RC networks are bad. I wonder how possible it is to try to demux an RTC? I just want to throw this together on some perfboard and hardly have a micro to have onboard. The alternative being throwing a few clock-dividing shift registers in series and powering them with a 0.1Hz clock or so, which I might be able to do with one of the 556's timers. Well since I have more than a hundred shift registers lying about that's probably what I'll do. I don't have any idea how to make/get a digital comparator that works on whatever demuxed data format the RTC uses anyways.

>>1370518
Then again, the only reason I really want one for long exposure times is because it takes forever and it's hard to not bump the camera while holding down the button, so just putting a toggle switch on it instead is what I'll do.

>>1370511
http://www.instructables.com/id/Op-amp-Basics/

Retard here.
I'm trying to figure out how to power 8 servos (Tower Pro SG90). I don't care if its by power supply or batteries. I just need all of them to work at the same time. I know they require 5-6V and it looks like up to 1A each (not sure) under load.
Right now I'm leaning towards buying a fuckload of AA batteries and powering each individually with 4 AAs. I'm pretty sure this is not the way to go. Please help.

>>1370559
You can do it with a power supply. Datasheet says 4.8V but 5V will work almost certainly.
If a servo draws 1A when loaded, then use a minimum 8A supply because you have 8 servos.
Wire them in parallel.

>>1370564
Thanks anon.
What are the chances I burn my house down if I get one of these?
Also what if I used a power supply with more than 8 Amps? Would I have to put a resistor (5 ohm I think) in front of each servo or something?

holy shit you can just press G on your keyboard and LTspice will give you a ground node ready to go

>>1370572
That sort of supply is fine, cheapo chinese ones can be a bit shitty though, if you don't mind the price getting a proper brand one like this would be nicer https://www.digikey.com/product-detail/en/mean-well-usa-inc/LRS-50-5/1866-3358-ND/7705050 Note you've got live exposed 120/240V terminals, so no touchy while it's plugged in.

Higher current rating is fine, don't need to change anything.

>>1370612
Thanks for the help anon.

>>1370518
why don't you just trigger the second multivibrator from the leading edge of the first?
>demux an RTC with TTL
pointless
>a few clock-dividing shift registers
reasonably possible but if you're trying to do this with junk on hand, only you know what's in your junk box and your balance between buy vs. make do

>>1370534
the other advantage of using the level of the received signal instead of the edge is that bulb shutter settings are as easy as holding down the transmit button
>not using manual focus
wew

>>1370572
quite low

>>1370572
>Also what if I used a power supply with more than 8 Amps? Would I have to put a resistor (5 ohm I think) in front of each servo or something?

5 ohm you think? Where did that number come from?
Short answer no. I said minimum 8A supply, a bigger supply will also work fine.
Think about it this way: Your wall outlet can supply 15 amps (more for brief periods) before the breaker trips, but you don't have to "put a resistor in front" or do anything special if you plug multiple devices into one power strip.
When you see a power supply rated for say 5V @ 20A it means that can supply *up to* that many amps while keeping its output at 5V.

>>1370701
>the level of the received signal instead of the edge
Come again? You mean the received IR signal? I was going to put the toggle switch for long exposure times on the camera side, no point in having that wireless. In fact I'll build the entire thing wired with the wireless section as an add-on since it will require quite some tuning.

>bulb shutter settings
>holding down the transmit button
Are you implying that I can change the F-stop and aperture size remotely via this port?

>>1370731
sorry, the level of the demodulated, decoded signal (asserted) rather than the edge (asserting)
>bulb shutter settings
nope. I'm suggesting that, if your camera has a "bulb" shutter speed setting and is set to it, the shutter stays open for as long as the shutter button (or equivalent input) is depressed, and that a simple delayed-on action based on the demodulated/decoded signal level (R/C/diode/comparator) would allow you to utilize that feature and control shutter open time remotely, assuming the IR wouldn't interfere with your shot for whatever reason.

>>1370701
> just trigger the second multivibrator from the leading edge of the first
I'm triggering it with the lagging edge of the first so I only need to adjust the period of the first one via potentiometer.

>>1370732
>the shutter stays open for as long as the shutter button (or equivalent input) is depressed
That's the feature of the camera I was planning on using, yes. But I just don't see the point in making it remote activated. I'll be leaving the thing for hours on end, likely pointed at the sky, so simply having a toggle switch taped to my tripod just sounds like a better idea. And when taking photos remotely I don't really have to worry about an IR flash from my surroundings, a wasted shot or two is hardly a big deal. But when I leave the camera for hours I don't want my shot to be ruined by a similar flash of IR.

No clue how else I'd take such long exposure without a bulb mode, my camera only goes up to 1s exposure times without it. The problem with using the entire IR pulse as my trigger period is that the background IR will change over the day so I'll need to constantly calibrate it. And a quick thought leads me to believe that adding a calibration button is not an easy task. Sample and hold on the base IR level, use that plus a little as the level on which to discriminate a high or low. It would work, but it would just make my circuit that much more finicky.

Pic related is the basic idea for the wired model. The left is the single-press photo trigger, the right is the long-exposure toggle. Since the Photo and Focus lines are active low, I'll throw the matching pairs of outputs into NOR gates. And by that I mean 2 diodes, 2 resistors, and a transistor.

>>1370740
>background IR will change over the day
>calibration button
add a blocking capacitor. background IR is basically dc, compared to your signal
>And by that I mean 2 diodes, 2 resistors, and a transistor.
>not just two resistor-transistor pairs in parallel
hmm, also, don't pull up the output, let the camera do it
>focus
if you're taking sky shots it sounds like you would prefer to manual-focus to infinity, no? so you could skip the focus step entirely and just trigger the shutter, no?

What exactly is the use of a d flip flop with a clock? It can store a bit just for one time origin period, right?
What good is that?

I can't quite understand voltages. It's the electric pressure between two points, right? In a logic gate, where 5 means yes and 0 means no, 5v is between which two points?

>>1370832
control the clock, and you control when the FF latches its new state

>>1370836
referenced to the "ground" point

>>1370840
Double thanks!

>look up synthesizer schematics
>most were made/drawn in the late 90's/early 00's
>half the components are obsolete
>don't know enough about components to know what I can replace them with

what do

>>1370863
look up newer schematics
check aliexpress for old stock
check coolaudio to see if there's a modern clone
lern2substitute
approximately in that order

>>1370863
>most were made/drawn in the late 90's/early 00's
>most
find the ones that weren't?

>>1370873
they're paywalled

How do you know which way up a resistor is?

Is this 220ohms or 10kohms?

>>1370900
the space between the tolerance band and the rest of the code is usually wider, though it can be difficult to see on some resistors
also, silver isn't a proper first-band color
so, probably 220 ohm

>>1370903
Thanks. It is 220ohms. I was confused for a while.

>>1370900
some oldfags on here are going to blast me for this but I got a component tester (pic related) from my uncle and now I just stick the resistor, cap, whatever in it and it tells me exactly what it is in a couple of seconds.
Fuck markings honestly, most counter-intuitive shit ever. Only electrolytics managed to do it right.

>>1370907
I was thinking of getting one of them.

>>1370907
how's it do on inductance? I bought an LC100-A and it doesn't even work, and I don't have a scope to debug it and find the fake component(s)

>>1370914
honestly I haven't tried that yet. I never use inductances

>>1370920
wew
maybe wind 12 turns of insulated wire closely spaced on a pencil and see if it recognizes it?

>>1370907
2.83V at which current?

>>1370925
I'm sorry I can't right now. I'll try it when I can and post the results.

>>1370926
That's not my photo, and I don't actually know what current these things use. I'm pretty sure mine's on a 9V 1A(?) supply if that helps.
Anyway I think you can normally just assume the diode characteristic to be a vertical line, unless you're doing something that requires more precision in which case I'd look at the datasheet.

>>1370928
what color LED is it and we can work it out from there
I'm guessing white or UV so 1-2mA

>>1370746
>add a blocking capacitor
That will also block all but the rising and falling edges of my IR signal. Unless I modulate it at a few hertz that is.

>hmm, also, don't pull up the output, let the camera do it
I was thinking about that, but I'll mess about with that sort of thing before I build the circuit.

>if you're taking sky shots it sounds like you would prefer to manual-focus to infinity, no?
You always have to push the focus button since it's just the first step on the trigger button so I'm doing that here to ensure the camera behaves the way I want it to, but for looking at the sky I'll just flick the auto/manual focus switch on the lens.

>>1370907
I got one, but while shoving it in my bag I broke a small corner off the top-right of the screen, which contained all the traces for the bottom dozen pixels. I can still read most of what the components are though so it doesn't really matter.

>>1370914
I run mine on inductances and it's pretty handy for that, but it does show my 1mH axial chokes as 500µH. Not sure whose fault that is, but since my 470µH radials are fine I'm blaming the chokes. Highest inductance I've measured was 30mH or so which was measured fine (though I hardly have another way of testing it). But on its listing it should tell you the range for each component.

>>1371055
consider the relevant frequencies. natural IR variations maybe 0.5Hz at most. signal is what, 100Hz? surely you can find a happy medium
>LC meter
hrm, what's the smallest? I'm hoping to measure nH, but if i have to build my own oscillator and a frequency counter to read that low, it is what it is. just gotta figure out these STM32F1 timers and ST's inconvenently documented HAL

>>1371062
Well transmitting IR at 100Hz would be pretty foolish, since that's where all the lighting noise is. If below 100Hz is too close to DC, then I'd run 10kHz.

I'm putting some electronics outdoors, should I pot all of my electronics with hot glue? Would it be a retarded idea in the first place to pot an ESP8266 with hot glue?

>>1371278
Well if you need to worry about heat sinking then make sure the heat sink sticks out of the glue. You may need to put heat sinks on things that wouldn't require them if they were in open air. But personally I'd use one of those ubiquitous IP65 ABS project boxes. They even come with brass threaded inserts for fixing the lid and a few slots and such to mount PCBs within. Not to mention the existence of waterproof cable glands.

What does it mean when your 0Ω "resistors" have a 5% accuracy rating on them?

>>1371278
>pot all of my electronics with hot glue?
Good idea. If something doesn't work you can throw it all away in one piece.

>>1371307
post pic

>>1371309

>>1371307
0 +/- 5% = a range of 0 to 0. EZPZ. Realistically it's gonna be within a few tens of miliohms, technically that makes infinite percentage error, you should demand your money back for this outrageous error.

It's just an easy way for them to tell you it'll be in the same size/price/mechanical spec as their other 0.25W carbon 5% range.

Cameraboi here, just made the right-most circuit on >>1370740, and it works a treat. Though I can't use BJTs for the inverters that pull the camera lines to ground, my circuit's "Focus 2" line has a stupidly high output impedance that prevents the RC circuit from charging unless it's pulled on with more than 1MΩ between it and the BJT base. And since doing this prevents it from pulling the camera focus line down far enough, I've decided to replace both inverters with MOSFET based ones. And I'll keep using the IRF840s because I don't have anything else, though if someone has a nice small-signal FET to suggest that would be nice.

I spent a while trying to figure out while it wouldn't take a photo on bulb mode lest I had the lights on, it kept telling me "too dark", but a bit of google suggested that it was just too dark for the autofocus to work. Switching it to manual focus made it work like a breeze, but I tested it and it still requires the focus button to be pushed beforehand. But it also works when I just tie the focus line to the photo line and pull them to GND together, so it looks like I'll just replace all of right circuit with a single switch. The more you know.

>>1371311
that's qaulity

>>1371403
BSS170 is a cheap, available, not terrible small-signal N-channel MOSFET to keep in stock
>focus line to the photo line
yep, it's just as if you slammed your finger onto the shutter button

>>1371407
>it's just as if you slammed your finger onto the shutter button
Well it's hardly impossible for the camera's microprocessor to spy the 50ms gap between the motions of a swift button push, but I guess there's not really a point in ensuring the focus button is depressed first.

I just pulled some motors out of an RC car
they both have a 100nf ceramic capacitor bridging the positive and negative terminals

what does this serve to do?

>>1371425
filters noise from getting back into the rest of the circuit
it also seems to stabilize running speed

Is there a rule of thumb for how big the cap should be? Also, if I put an electrolytic instead of ceramic will it blow the moment I turn off the motor, due to flyback current?

>>1371438
If it's a DC motor then it won't blow up from reverse voltage, but I'd put a freewheel diode in there if there isn't already lest you overcome the cap's max forward voltage. Though if C*V_max^2 > L*I_max^2 then this isn't actually an issue since it will only charge up the capacitor some of the way. However on the return stroke (since you've made a tank circuit) you'll probably hit the reverse voltage of the cap.

Anon with the wireless charger here. After doing a little research, I figured that the guy sold me a fake one irregardless of it not working, so I just put it back together and opened a return request on eBay. Not 5 minutes after doing so, I got a full refund. Welp, at least I got a free wallwart and usb micro cord from this.

>>1371438
you mostly just want to absorb commutation spikes. too lazy to math it out as a function of motor inductance

>>1371444
noice
and checked

public service announcement: polyimide tape is the shit for soldering pin headers and other wobbly THT components

>>1371444
Neat.

What kind of molex connector should I use for a 50W heating element? Also, what kind of crimping tool do I need to make a cable for it?

>>1371527
connectors and contacts are typically rated in amps. what voltage we talking here?
>random molex
depends on your budget and neatness preferences. you could use anything from automotive "bullet" connectors and the cheap crimp tools from Home Depot, if you're looking for a quick, cheap solution. or you could get Pic related which will crimp just about everything in the small electronics world and use e.g. AMP Commercial Mate-N-Lok (incoorectly known in the PC world as "Molex") or compatible connectors, with the bonus that you can plug them right into any PC power supply

>>1371533
24v, so it's about 2 amps, but lets assume it's actually around 4 or 5 because transients

>>1371539
CMnL connectors are rated for up to 250VAC/VDC and 15A per circuit, should be fine

>>1371551
>Mate-N-Lok
Do they come in a single size and you just use more of them for more current? Also, how about 10 wire IDC connector / ribbon cable?

>>1371558
that particular series is rated for 15A. you could parallel circuits to get more current handling capacity according to the table, e.g. assuming 14AWG wire, a 4-pin CMnL with two parallel circuit pairs could handle up to 30A, a 6-pin with 3 parallel circuit pairs could handle up to 39A, 8-pin in 4 pairs up to 50A. if you need much more than that, you might consider thiccer connectors.
>10 wire IDC
nope. that's like 28AWG wire which isn't enough to handle any serious power, plus you'll lose a ton of power due to resistance being inversely proportional to cross-section.

How sensitive are analogue oscilloscopes to magnets? Do the CRTs degauss when switched on like tvs?

I want to mount a somewhat strong permanent magnet (for tool holding) near it.

>>1371624
not really, degaussing is important mostly because of convergence. CRT oscopes just have an angle control that just rotates the trace, so you can align the trace with the X axis. you can use this to tune out most influence of a permanent magnet too, I imagine

NEW BREAD
>>1371574
>>1371574
>>1371574
NEW BREAD