[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k] [s4s] [vip] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Board
Settings Home
/diy/ - Do It Yourself



Thread archived.
You cannot reply anymore.



File: 1506281496841.jpg (21 KB, 666x239)
21 KB
21 KB JPG
bump limit reached on old thread
>>1247660

https://www.wiki.printf.pl/index.php?title=Pasta (fresh)

>I'm new to electronics, where to get started?
There are several good books and YouTube channels that are commonly recommended for beginners and those wanting to learn more, many with advanced techniques. The best way to get involved in electronics is just to make stuff. Don't be afraid to get your hands dirty.

>What books are there?
Beginner:
Getting Started in Electronics Forrest Mims III
Make: Electronics Charles Platt
How to Diagnose Fix Everything Electronic Michael Jay Greier

Intermediate:
All New Electronics Self-Teaching Guide: Kybett, Boysen
Practical Electronics for Inventors: Paul Scherz and Simon Monk

Advanced:
The Art of Electronics by Paul Horowitz and Winfield Hill

>What YouTube channels are there?
mjlorton
paceworldwide
eevblog
EcProjects
greatscottlab
AfroTechMods
Photonvids
sdgelectronics
TheSignalPathBlog

>What websites feature electronics projects or ideas?
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/

>Where do I get components and lab equipment from?
ladyada.net/library/procure/hobbyist.html
Search the web for "hobbyist electronics sources" to find plenty.
In the US and elsewhere, mouser.com, digikey.com, arrow.com, newark.com are full-line distributors that entertain small orders.

>What circuit sim software do you use?
This mostly comes down to personal preference. These are the most common ones though:
NI Multisim
LTSpice
CircuitLab
iCircuit for Macs
CircuitJS (quick, dirty, interactive)

>What software should I use to layout boards?
Circuit Wizard
ExpressPCB
EAGLE
KiCad

>Q1. Lithium batteries
Will explode if you abuse them. Read and understand all relevant datasheets and be prepared for catastrophe. See
batteryuniversity.com/learn/
>Q2. My circuit doesn't work. Halp?
Check wiring, soldering, part pinouts, and board artwork if applicable, then post schematic.
>>
Is there any programs that allow you to simulate a breadboard?

I am new to electronics and want to mess around without bending up a bunch of shit and re do a lab I had today in school.
>>
File: 10938_13763_ACT.jpg (419 KB, 1809x1068)
419 KB
419 KB JPG
Chip of the thread: ADSP-BF608, because apparently we're doing that.
>>
>>1251611
In what sense? You can simulate the functionality with any of the OP's simulators. If you want something for planning the layout, you could try Fritzing.

>>1251612
Why?
>>
>>1251614
Layout is what im looking for.
>>
>>1251614
As with the last thread; in honor of the OP digits. 608 -> ADSP-BF608; 7660 -> 7660 charge pump.
>>
File: 4-way component switch.jpg (74 KB, 1000x377)
74 KB
74 KB JPG
>>1251189

4-way component switch, Dynex DX-CVS4, cost me $2.50 at the thrift store.

see if you can find something similar cheap.
>>
>>1251608
Oh, my image made it to the front page. Original title Fleur d'Ampère, carefully cropped to 666.
>>
File: sch.png (29 KB, 514x550)
29 KB
29 KB PNG
I'd like to build a controller for controlling some lamps via a single-wire data bus. The system would consist of a controller unit, which connects to a PC via serial and a couple of slave units (lamps) that connect to the controller unit via a single data bus.

It's really difficult to get started. I have no idea how to get started with the programming part. So far all I have managed to do with AVR have been blinking LEDs with various frequencies. Any advice/pointers?
>>
File: x-10 usb controller.jpg (9 KB, 266x260)
9 KB
9 KB JPG
>>1251624

X-10 is pretty old technology, which is why i use it, coz i can buy it at 1/30-th the cost of new units. so, i'd use a USB-to-X10 interface on the computer, and 1 receiver module for each lamp i wanna control. takes like 20 minutes to get it all running.
>>
>>1251624
>single-wire data bus
does not exist, a bus has at least two wires, mostly differential, sometimes signal and ground. Your diagram shows that all bus receivers are in parallel, which means they must be addressable. Look up DALI for a description of such a structure.
>>
>>1251608
Thanks OP. That pillow looks inviting and makes me sleepy.

>>1251611
Bending up shit is part of the fun.

>>1251454
Not the poster, but it tells me that something's not grounded that should be, either a lack of pulldown on the button, RES left floating, or CKEN/GND being poorly/not connected..
>>
>>1251636
https://en.m.wikipedia.org/wiki/1-Wire
>>
>>1251636
Dallas One wire.......

Uses earth for return and the one line for data/clock/pwr
>>
>>1251737
Both RES and CKEN are connected and grounded when the arduino is in the circuit, but when I clocked it with a button they were left floating.
>>
>>1251777
That wasn't a good test, then.
>>
File: 1481217544823.jpg (123 KB, 1001x1001)
123 KB
123 KB JPG
Need some help bros. I'm new electronics and I'm buying all kinds of shit for my mini-lab.

I ordered a bunch of wires like pic related for breadboard work. However, I need two things:

1) I'd like to buy some of these pins that go on the end of the wire so I can make my custom-length wires. What are these things called??? I searched for "header pins" on ebay but got shit results.

2) What gauge wire is the best for breadboard work?


Thanks a lot guys!
>>
>>1251798
I think they're called dupont crimp pins, or at least that's what a lot of them are called on amazon.
>>
>>1251799
Thanks! That search looks a lot more promising!
>>
>>1251798
1. Those are "Dupont male crimp pin connectors" and you need a specific crimper to apply them to the wire.
2. Just buy a spool of solid-core 24awg insulated wire. You won't need pins, just strip the ends and push them in. In casr you need to couple some cable in, grab some heat shrink and splice a small piece of the 24awg onto the end.
>>
>>1251801
Thanks! BTW, what;s a good wire gauge for flexible wire when you want to to use it for small PCB projects with perfboards and Arduinos etc?
I'd like to buy a spool of something that's gonna serve me well for all kinds of tasks (basically 5V and 3.3V projects)
>>
>>1251806
26awg is great for signals and modest power transmission, up to an amp or two. It isn't too hard to bend. My advice is to buy at least a couple, better three or four colors in case you need to wire up something more complex like a longish I2C bus.
>>
>>1251814
Thanks so much anon!!
>>
>>1251815
Cat5 wire is pretty nice. Blue orange green brown are pairs 1 2 3 4. Get it on a spool and it should be solid core. Patch cables are stranded. Solid is nice for wrapping terminals and poking into holes.
>>
This might be a stupid question but is there any kind of spray on phosphor coating or something similar? Like if the phosphor on a screen was wearing out could you just apply a new coating of phosphor?
>>
>>1251828
>phosphorus spray
that doesn't sound healthy
>>
File: phosphors.jpg (1.68 MB, 2006x3099)
1.68 MB
1.68 MB JPG
>>1251833
>implying phosphors need actually contain phosphorus

It would make for a pretty exciting spray though.
>>
>>1251828
>Like if the phosphor on a screen was wearing out could you just apply a new coating of phosphor
CRT? How the fuck would you apply it????
>>
>>1251828
>phosphor on a screen was wearing out could you just apply a new coating of phosphor?
it's on the INSIDE - of a VACUUM tube...
>>
>>1251848
he finally has an use for his masterful succ techniques
>>
>>1251616
>Layout is what im looking for.
KiCad, EasyEDA (webapp).
>>
Want to buy an arduino but im worried itll end up in my drawer collecting dust :/
>>
>>1251876
buy chink to lower your risk
>>
>>1251439
>>1251440
Still struggling with this, anyone? I'm thinking I'm just going to order a different receiver cause I believe the chinese sent me 5V instead (it works better on higher voltage)
Might rig a 5v powerbank and see if it works.
>>
File: 1485327848355.jpg (92 KB, 1518x1518)
92 KB
92 KB JPG
>>1251876
Every experimenter should have a handful of 2-for-a-burger step-up converters in their parts box..
>>
Me again >>1251876
Amazon is selling chink arduinos for $40 canadian. Does it look okay?

https://www.amazon.ca/gp/aw/d/B0188TO4VU/ref=mp_s_a_1_1?ie=UTF8&qid=1506702112&sr=8-1&pi=AC_SX236_SY340_QL65&keywords=Kuman&dpPl=1&dpID=61K5YR6UqZL
>>
>>1251940
Just got some, thanks for the reminder.
>>
>>1252002
i would advise against this. you can get chinese standard arduinos on ebay for $3 or nanos for $1. the nanos have the same capabilities but fewer pins so i'd suggest them unless you know you need more pins. buy a few of them because you'll kill them or want to permanently fix them somewhere.

example: http://www.ebay.com/itm/182601333242

the other hardware in that kit can be bought elsewhere in larger quantities for less. you can get those steppers for $1.50 and their drivers for $1. same applies for almost anything else. you should buy a large assortment of caps and resistors. the quantities there are useless.
>>
>>1251608
ohm senpai, I need some advice.

Let's say I have a simple 5V pulse output from a microcontroller, and I want to split this signal to multiple inputs.

I want to make sure that the signal is adequately delivered to each of the inputs (no drop in voltage, etc...).

What is the most ideal way to split a signal like this? Should I use an op amp buffer at the output? multiple buffers before each input? No buffers at all?
>>
>>1252090
what are the inputs, what is the nature of the connection, and what's the signal frequency? the answer is almost certainly that you won't need any buffer. an op amp on the output won't sink or source current any better than modern uc output drivers and may decrease your slew rate.
>>
>>1251954
Reverse log in US, log in Europe. Sauce: http://www.resistorguide.com/potentiometer-taper/

>>1252002
>$40 for an arduino
No.
>>
File: asdasd.png (1.33 MB, 1328x518)
1.33 MB
1.33 MB PNG
>>1252092
the inputs are control voltages for some synthesizers. What I'm trying to do is use one output to control the tempo of multiple step-sequencers. So, in other words, it's a sync pulse.

Pic is the pulse. Frequency is quite low.
>>
>>1252096
the rise time on that pulse is terrible, and i'm not a synthfag but that can cause oscillation on logic inputs for many ics. a transistor buffer on the uc output may solve your problem. it can either be complementary or a pnp (not npn)+resistor. is that a measurement of your intended circuit setup or do you have a wire looping across the room a few times?
>>
>>1252096

if you're using chips, you need to look at the datasheet to see the ''fan-out''. TTL, for example, typically has a fan-out of 10, meaning a TTL chip output can drive 10 TTL inputs.

so, it's likely you wont have to do anything, just wire them together. if the source has an unknown fan-out, use a buffer chip, or two series inverters in CMOS or TTL, then you'll know the fan-out from the datasheet. (since your output is kinda slow to rise, a Schmidt trigger would help clean it up.)
>>
>>1252099
>>1252102

agreed the pulse time isn't great, but that is actually the synthesizer's "sync out" line. And I wondered why my sequencers kept going out of sync lol.

Signal in pic is what the microcontroller will be replacing, though good call with the Schmitt trigger, that would help.
>>
>>1252002
>Amazon is selling chink arduinos for $40 canadian. Does it look okay?

it's not a 'duino, it's a kit, a pretty well stocked one at that, so it's totally worth the $40. especially with free shipping, and ''Fulfilled by Amazon'' so it's gonna be super fast.
>>
File: one-wire_bus.jpg (10 KB, 507x222)
10 KB
10 KB JPG
>>1251636
>>
>>1252163
They are in series, as far as the input is concerned. (Assuming a load of infinite impedance at the output, for simplicity.)
In short, the capacitor converts current into charge (and vice versa), which is reflected as a voltage difference between its plates. The resistor limits the current and thus the rate of charge accumulation. A higher-frequency signal input at the same level, therefore, leaves less charge stored on the capacitor and thus less voltage at the output.
I bet you would enjoy the CircuitJS webapp in the top post. It shows the current flows and voltage levels of most any simple circuit in a friendly, graphical manner that might aid your intuition of how this all works.

>>1252186
>t.Dallas Semiconductor
>>
I know this isn't strictly /ohm/ related, but /g/ is filled with mouth-breathing /v/tards

what are the best disassemblers for linux? Are there any FOSS ones? Or am I going to have to pay an arm and a leg for one
>>
File: IMG_20170929_211209.jpg (2.85 MB, 3120x4160)
2.85 MB
2.85 MB JPG
I bought this thing at goodwill for $8 a few days ago, I figured it was worth a lot more than that but I had no idea it costs like $700 new.
What use does it have for a hobbyist? It seems like it's more meant for testing finished products, ie nothing I do.
If I can't really use it, what are the chances of me actually being able to sell it as just a random, untrusted seller selling a used item?
>>
>>1252263
it's not useful unless you're into power supply design. it's only worth $700 new because it's ruggedized niche equipment, not because it's complex or tightly specced. you can sell it but you sure as shit won't get $700 for it. i'd bid on it.
>>
>>1252255
objdump
oh, you meant smart interactive ones? Not really. Apparently there's one called lida, but it's still in a very early development stage.
>>
>>1252255
Oh, here's another one that just popped up that seems to be coming along as well:
https://github.com/plasma-disassembler/plasma
>>
>>1251608
Best way to make a low quiescent current, low power output ~400V power supply that runs off LiFePO batteries? I've tried simulating a boost converter with zener diodes to regulate it and I get too much power dissipated in them. I've thought about a CW multiplier, but getting the supply waveform seems a little difficult. I don't want to wind my own transformer if I don't have to, but if a flyback converter is the way to go then I guess I'll go for it. Size is also a consideration to some extent, so no massive capacitors or microwave oven transformers.

Thanks in advance.
>>
>>1252294
if you're open to smd parts there's a neat 10:1 xenon flash transformer you can find on digikey. it was used in a micro geiger counter that eevblog took apart to generate the ~400v for the gm tube. obviously that'll only work for very small loads and you need a minimum competence to design that into your circuit. flybacks are definitely the only option for your situation.
>>
>>1252294
>>1252296
Also, consider EL inverter transformers such as used in laptop screens, or even the whole modules. How much power output did you need? Maybe a couple of these in series would do it.
http://www.goldmine-elec-products.com/prodinfo.asp?number=G13599
>>
just ordered 600 resistors for $5. hope it comes in one piece
>>
>>1252298
Some laptop screens use EL panels? I never knew.

So what kind of circuit do I need to use to drive the transformer? I imagine a normal square wave won't be too good, can I just put a (correct value) capacitor in series/parallel with the primary coil? I guess I'll use a variable boost converter to get the 40V needed to drive the 10:1 transformer.
>>
>>1252306
Could I use a tiny Xenon strobe trigger transformer? That would make it easier as they're cheap and made for high voltage.
>>
>>1252306
Brain fart, I meant CCFL. Still, transformers for that application could be of use.
A square wave is just fine for a switch-mode converter. You'll be varying the pulse width to regulate the output anyway.
>what kind of circuit do I need to use to drive the transformer
Look at switch-mode power supply datasheets for inspiration. I'd start with a low-voltage dual op amp with low input offset current like the TLV9062, with one unit as a sawtooth generator and the other unit as a PWM comparator, driving a low-voltage MOSFET in turn driving the primary. The sawtooth should probably be constrained to a maximum of 2V or so and bottom out as near to zero as is convenient. If you don't need isolation, feedback can come from a resistor divider in the 10-20 megohm range with a trimmer, just as you might with an LM317 or whatever.
>>1252315
You could, but flash triggers are low-duty pulse applications. If you need a sustained high voltage at more than a few microamps, you'll want something more sturdy.
>>
File: geiger_counter.gif (8 KB, 610x339)
8 KB
8 KB GIF
>>1252306
Look for disposable camera circuits.
>>
I'm having some trouble with an old car battery charger. The fuse keeps blowing on it when i turn it on, even when it's not connected to a battery. I assume it's a short circuit somewhere. I would appreciate any tips on troubleshooter it

The capacitors seem fine, and the transistors don't show signs of overheating. Checked the rectification diodes too they seem fine.

I'm pretty sure the transformer is borderline impossible to short out.
>>
>>1252320
>varying the pulse width to regulate the output
I was thinking more along the lines of a not-flyback converter, using a potentiometer to adjust the frequency until it resonated with the capacitor and just ran off a sine wave, but if there's some way of providing a well regulating feedback then I don't see a problem. The current I'm pulling will be minuscule since this is a Geiger counter circuit using a pretty small tube, so a tiny transformer like that should hopefully be fine. I assume you mean triangle wave not sawtooth, right?

>>1252371
That looks neat apart from the custom transformer. How much power do those Zener diodes waste?
>>
>>1252373
Do a continuity test to check for shorts on the board itself.
>>
>>1252377
>How much power
Multiply 400V by the base current the 2N2222 needs to shut down the generator.
>>
>>1252388
I guess that's enough current for the zeners to act properly? I'm surprised that the filter cap is only 10nF, though I guess the oscillation frequency of the thing is pretty high. Still can't get my head around that centre-tap transformer though.
>>
>>1252393
>surprised
Why? It's a geiger counter circuit, load is 10MΩ. Transformer is not center-tapped, not at all (pic). As I said, look into disposable camera flash circuits. There is no cheaper way to get the complete circuit including the 'custom' transformer. Modify to suit your needs.
>>
>>1252397
>load is 10MΩ
Real load is the 50µA base current that flows through the Z-diodes, the 40µA to the tube are short pulses whenever it gets hit.
>>
File: file.png (16 KB, 361x201)
16 KB
16 KB PNG
>>1252402
you can decrease the quiescent to basically nothing if you digitize the circuit. pic related senses the output voltage on switch shutoff. there's no continuous sensing though obviously which means you need a minimum load (or a known load) because sampling the voltage involves pumping a small amount of current into the output capacitor.
>>
>>1252285
>>1252277
I managed to ghetto rig something together with objdump and hexdump

Plasma looks real promising though too, I'm gonna check it out when I get home
>>
File: kodak_flashm.gif (12 KB, 707x393)
12 KB
12 KB GIF
>>1252427
Drive-side sensing does indeed work but it has its own problems. In this case a no-load output voltage could rise indefinitely because the D-C-R combo on the left side defines the 'probing' cycle and needs careful dimensioning. Monitoring the supply (battery) voltage may also be necessary. The problem with Geiger counters is that you have to keep the plateau voltage within ±10V or so for (almost) constant sensitivity while the load depends on activity. Pic: A very effective photo flash circuit from a disposable camera that can be easily modified for use with a GC.

The cheapest of cheap phone chargers also use drive-side sensing to omit the 'costly' opto-coupler but keep the isolation. They have a base load resistor but the 5V/500mA two-for-one-coin model I tested had 9V at no external load and 4.5V on a 10Ω load. The circuitry is of course much, much simpler than your approach.
>>
Long story short, I need a custom IC. Since they cost a shitton and are only manufactured in prototype batches a few times a year, I was wondering if I could make one myself/emulate one with a PCB of a reasonable size (meaning not room size, but biggest motherboard size would do, even if it needs to be a bit bigger). I know people have made their own ICs before, mainly on universities, but there are guides out there. I don't mind spending however much since it's a long term project. I know I would have to be very precise, get special equipment, be in a sterile environment and handle dangerous chemicals. I just wanted to know if you guys think it's possible at this point in time, or should I just drop it. If it's not possible, would emulating an IC with one big PCB or a lot of smaller ones, or a combination (I can even get flexible ones and shit) be a viable task, given the space constraints? I know I probably sound mad, but I do like to go full-diytard and do it myself all the way down, especially if it's something expensive I want to make myself, like in this case.
>>
>>1252502
you can't post something like this and not explain what you're trying to do with it
>>
I was thinking about getting a collie how would I make it a service dog?
>>
>>1252505
Some (not so) simple asic things. It's mostly just for shiggles. But it would make my year if it turns out to be possible.
>>
More advanced pseudo question would be "do you guys think I would be able to make it a service dog", embedded in some long-winded irrelevant story telling.
>>
>>1252502
Would an FPGA with external peripherals work?
>>
>>1252577
Hmm, I hadn't thought of that. Yeah, just not one, but many. Perfect solution.

I still do want to discuss the original question with someone though, as a backup plan of sorts.
>>
>>1252581
I've heard about a russian guy who tried making chips, but I don't think he succeeded yet, but he makes some cool die photographs https://zeptobars.com/en/
>>
>>1252377
>I assume you mean triangle wave not sawtooth, right?
Most commercial single-chip switcher controllers seem to use sawtooths. Either will do, really. I figured the sawtooth might take a few fewer components when the output waveform wouldn't be centered around the middle of the op amp's output range.

>>1252502
>a custom IC
Single MOSFETs have been created before with relatively low-tech equipment, which still cost them several hundred burgers. If you have money and can get the materials, anything is possible. Consider that the smaller the feature size, the more exotic the technology required. Ion diffusion stops working well below a couple of µm so you have to use an ion implantation system, the particle/wave duality starts distorting your patterns below about 0.35µm so you have to counter-distort your masks, you generally need to expose the work with a radiation wavelength a few times smaller than your max feature size, etc. It would make a lot of people's year if you pulled this off and put it on youtube.
>>
>>1252397
>no cheaper way
$12 for the camera or $1 for the xenon transformer. Unless the camera circuit contains another $11 worth of usable component, I don’t see how. I should be able to run a voltage divider and comparator (maybe a FET instead if I’m retarded enough) to give feedback to switch the transformer with some zener diodes like in that circuit.
>>
>>1252683
just be aware with any secondary side sensing you'll have a quiescent battery current in the milliamps.
>>
>>1252686
Not if I use a MΩ voltage divider. If I can get the whole thing below 15mA I'll be perfectly happy, but that depends on what FETs I use. If I can combine the feedback and resonant-capacitor ideas it should be pretty damn efficient, though I'll have to see how possible that is, especially with keeping it in resonance.

How would you even do primary side sensing? Or do you mean using that tiny feedback coil? I could always wind another few turns around the outside of the xenon transformer, but it might take some work to tune its dummy load to have the correct voltage waveform.

And surely a voltage divider will work better than those stacks of Zeners, right?
>>
>>1252702
(400V)^2/(3.6V*40 megaohms)=1mA quiescent, for example. only you know what an acceptable drain is for your application.
>>
>>1252707
Where did that equation come from? I was just thinking 400V/1MΩ, but evidently that isn't the case.
>>
>>1252719
(400V)^2/1 megaohm gives a quiescent power of 160mW. since it's converted from a ~3.6V source it'll draw 0.16/3.6=44mA.
>>
>>1252720
Ah shit, because its using a different voltage. Well I was originally planning on using a ~MΩ dummy load anyways such that the rectification diode actually conducts and charges the capacitor, though now I'm not sure if that's necessary. Anyway, I'll see about simulating a non-flyback resonant transformer again and see if that looks promising.
>>
>>1252735
I think I can get something like this working, provided my opamp works in the frequency range specified by the tank circuit. The two inductors make a transformer, fyi. This simulation gives me a power consumption of 50mW, which is definitely an idealised value but I might be able to get close to it.
>>
>>1252751
Shit it won't let me rectify it
>>
>>1252756
you'd might as well just retard proof it and use a 555 and 393 to make a shitty bang bang controller
>>
>>1252757
Would that be sinusoidal? I don't see what the 555 would be for if the comparator is the oscillator, and if you use the 555 as the oscillator, why bother driving it with a comparator (or even just a FET) if the 555 can drive the few mA through the transformer with ease?
>>
>>1252765
the efficiency gains you'll see from your conceptualization of resonant switching are essentially nil. why do you want it? the 555 was a half joke with the point being that you're going down a pretty odd path for the goal you're seeking to accomplish. all you strictly need is a pwm source to sink current through the transformer and a hysteretic controller to flip it on and off.
>>
>>1252769
As far as I see it, using a flyback converter requires some sort of feedback that inevitably wastes a small amount of power and complicates the circuit, and there's no reason not to make a (resonant) transformer if it's simpler. But I don't know whether the final circuit will be simpler or not.
>>
>>1252771
you cannot have a stable output voltage with a variable load without a feedback mechanism. this is true for resonant and hard-switched converters. if you have a truly static load then there's no reason to have anything more than a trimmable pwm.

at any rate i think you don't understand what a resonant converter is, because driving a transformer with an op amp won't get you one. what do you think the benefits are?
>>
>>1252772
If the cap and inductor are in resonance, it means I get a free sinusoid out of the deal. Sinusoids mean maximum efficiency when driving a transformer. Just look at this majesty.
>>
>>1252772
And I can't see how a simple rectified transformer secondary would have its secondary voltage sag under the tiny loads I'm talking about, especially with a large enough filter capacitor. It's not a flyback transformer any more than a microwave oven transformer is.
>>
>>1252774
okay so here's the deal. your sine is minimizing steinmetz losses in the transformer. this will save you microwatts. you're creating that sine wave by driving transistors out of saturation. this will cost you milliwatts. the reason resonant switching is used in efficient power converters isn't because of transformer efficiency, which is only an incidental gain. it's because it allows you to cut turn-on/off losses in the (saturated) primary side switching transistors. i can guarantee you beyond a shadow of a doubt that your scheme will decrease your circuit's efficiency.

>>1252775
in some configurations this is true, but in your resonant system the capacitance and resistance on the secondary impact the Q and resonant frequency of your LC tank. this means your output voltage will vary if you don't vary the drive's amplitude and frequency. with your dummy load you may find that the output variance is acceptable. if that's the case though it means your dummy is swamping your actual load which makes any pretense of efficiency seem silly.

your only options for boost conversion (ie overcoming the 10x voltage gain limit offered by available 10:1 transformers) are flyback and LC tank circuits. i hope i've clarified why your idea of a resonant circuit rules out the latter option.
>>
>>1252778
Would the output voltage really change that much based off the frequency changing? I thought it would tend to a constant ideal output voltage if you keep increasing the frequency (until other losses kick in) and the fluctuations in the frequency wouldn't effect the output voltage if the frequency was high enough.

But assuming you're right, how is regulation typically handled in a flyback converter? Through duty cycle? If so, pic related looks fairly nice, assuming I can get a high input impedance comparator and it's simple enough to change duty cycle based off an analogue voltage. Or I could just feed the triangle wave directly into the other side of the comparator, with a few resistors to set the voltage at the right level. The requirement of a simple transformer is also appealing.

Another question I've been thinking of is what determines the turns ratio of a flyback transformer? Theoretically its output voltage is infinite under no load, I think, so surely a 1:1 transformer would work. But that's essentially an isolated version of a boost converter. So there must be a reason why they typically use a higher turns count for the secondary, and I assume it is related to the flyback transformer not being an ideal transformer.
>>
File: flyback.jpg (51 KB, 720x450)
51 KB
51 KB JPG
>>1252784
>>
>>1252785
Wait, that's an inverting amp, not a comparator. Is that just a bang bang controller not a duty cycle controller?
>>
File: IMG_20171001_005117.jpg (883 KB, 3840x2160)
883 KB
883 KB JPG
Hello, wise men. I'm modernizing an old boom box from the 80s. I'm adding a Bluetooth module instead of the tape decks and installing some sound reactive LEDs in the tape compartments. The box has a blue and I'd like to have blue LEDs in deck 1 and red in deck 2. Is it ok to run then all on the same circuit if I give them all appropriate resistors? I'm going to run them in parallel from the boom box's internal 12.5v supply. The circuit I built has 2 tip31c transistors. Any help would be greatly appreciated.
>>
File: file.png (28 KB, 719x252)
28 KB
28 KB PNG
>>1252784
it's hard to ballpark the output regulation of an LC tank without just simulating it because there's so many variables. an unloaded tank will trend towards destructively high voltages limited only by parasitic resistances and mismatch between the driving frequency and the resonant frequency.

flybacks regulate through duty cycle. flybacks can technically have any turn ratio, but the ideal ratio is one in which the switcher has a duty cycle just below 50% at maximum load. that rule of thumb is only relevant because of its implications for more sophisticated control circuitry than in your pic, but it's still fine for general use. your minimum load is your feedback circuit.

>>1252786
that's an op amp control loop. black magic. they're fun to design and read about but i'd warn against going down that rabbit hole unless terms like transient response mean anything to you. also that simple configuration has problems at light loads which will be painful for your application. you would be better off using a bang bang (aka hysteretic) converter. it can be done with a single dual comparator ic, won't give you overvoltage transients, and stays quiet for light loads. it has a larger output ripple than a loop closed by an op amp though.
>>
I need a project that uses Arduino.
Can't be too easy and it must be done in one month
Any suggestions?
>>
File: file.png (8 KB, 191x189)
8 KB
8 KB PNG
>4chanx update broke the comment box
i can't think of the last time i was actually happy to get a software update

>>1252791
a plant nursery with a relay-switched grow light, a light sensor, a solenoid or pump to dispense water from a bottle, and a soil moisture sensor.

>>1252788
if pic is what you mean then you're fine.
>>
>>1252794
Thank you. I just bypassed the tape deck preamp and it sounds pretty good. Now I just wait for the Bluetooth module to arrive and I'm all set. I appreciate it.
>>
>>1252502
Universities have fab classes, where they use old cast-off fab equipment to teach students the ropes. They also participate in project wafers like those offered by MOSIS.

If you have a decent amount of money, going to MOSIS is not out of the question, but it's a one-off kind of thing. My aspiration is to someday have the capability to fab ICs in my garage, but note that I have degrees in EE and Physics and a severe case of megalomania. My thinking is that commercial IC production is geared toward mass production, and fab lines are designed for maximum throughput. If you're only planning to make a handful, you can get away with slow processes like e-beam lithography, rather than making expensive masks, etc. But like you say, this is a long-term project, and requires a lot of skill and knowledge.

There was this Fab@Home or something project, but it was pretty dead when I looked at it.

Anyway, the only reason to do this at home is for the bragging rights. If it's digital, then a modern FPGA is going to be miles better than anything you could reasonably come up with in your garage.
>>
>>1252636
>>1252798

Thanks for the input, guys. I appreciate it. As I said, I'm gonna go with FPGAs, but I still wanna build my own shit one day.
>>
>>1252798
Not the other anon, but, just for the sake of argument, let's say it isn't. Let's say it's something like several dozen copies of a several hundred transistor ±15V linear design. I don't believe Ferranti is still doing semi-custom analog designs anymore. What do?
>>
>>1252789
So one square wave oscillator, a comparator, and an opto-isolator if I feel fancy? I’ll give it a go, but if I can figure out a variable duty cycle I’ll give that a shot.
>>
File: 427-01.jpg (107 KB, 970x728)
107 KB
107 KB JPG
I'm looking into some small modular storage, and I'm thinking these things may be perfect for my use, but I can't find any interior dimensions. Do any of you guys use these little SMD boxes? Would you mind measuring the interior?
>>
>>1252807
MOSIS has analog processes too. But depending on your performance requirements, you can generally redesign older stuff from the ground up, do it with discretes. The parts are better these days, with a few exceptions.
>>
>>1252810
>regulation
I think I have some parts to change.
>>
>>1252821
Should I be putting a resistor in series with the primary winding to limit the current?
>>
File: 1498356074378.png (166 KB, 564x468)
166 KB
166 KB PNG
>>1252813
Looks like they're approximately smol x smol x smol mm.

>>1252816
My crazy idea was to clone some of the vintage Curtis synth chips that were going for $80 NOS about a year or two ago. It looks like some Latvian outfit has already gotten around to it.

>>1252822
Just shorten the pulse width?
>>
>>1252824
>tfw I wanted to use a 50/50 square wave generator
I guess it's just one more comparator.
>>
File: 1478810758681.png (1 KB, 194x124)
1 KB
1 KB PNG
>>1252825
literally what
>>
>>1252807
FPGA's analog equivalents and their semi-custom versions aren't entirely dead. They aren't transistor-level stuff, though.
>>
>>1252825
>tfw the primary coil is sinking 600A
I think I'll have to rethink something here.
>>
Anyone here know where I can find in-depth information on volt/octave usage?
>>
File: accident-kit.jpg (18 KB, 225x437)
18 KB
18 KB JPG
>>1252683
>$12 for the camera
No, four such kits for 10€. No need to re-invent the wheel, use one that turns and mod it.
>>
>>1252831
What's to know, really, one volt = one octave. Sometimes 3V = middle C. Usually but not always you can depend on whole volts representing some C note, except Korgs where it might be F.
sauce: analog-heaven mailing list archives via web search

>>1252828
Need a better blocking diode maybe, or possibly your transformer is out of phase.
>>
>>1252847
>analog-heaven mailing list archives via web search
Thanks
>>
>>1252824
>>>1252813
>Looks like they're approximately smol x smol x smol mm.
Pretty much. Mostly thinking about the bigger sizes. I've got a bunch of hobbies where little stuff might need storage, so if these are just big enough, sweet. Plus the obvious SMD use.
>>
>>1252852
Not for naked SMDs, but I use these for my Dupont terminals and larger items that won't go leaking under or over loose dividers
fieldandstreamshop com/p/field-stream-360-utility-box-4-pack-17fnsufs360tltybxtbx/17fnsufs360tltybxtbx
>>
>>1252821
i went to bed. your pulses should be in the low microseconds with a duty cycle of 50% to start.
>>
Thinking of making the jump to SMT parts in my hobby projects. What equipment do I need to do the soldering?
>>
>>1252949
a hot air gun. i have a hakko 851 which is great but i don't think there's any reason to spend that much since it's just an air pump and a ceramic heating element. others suggest a hot plate but i've lived without one. probably makes it easier not to melt leds. you'll want a usb microscope as well. i like my opti-tekscope. the software is pretty buggy but it's minimalist so no cancer bloatware.
>>
>>1252952
Well, shit, I already have a hot air gun. How do you apply the solder paste? Is there a particular kind I should use?
>>
>>1252954
the easiest way is to just buy a cheap polyimide stencil, tape it down over the board, and smear the paste on gently with a credit card. if you overdo it you'll get too much paste and fine pitches like qfns or usb connectors will bridge. oshstencils sells 1.6mm blocks to help fix the board and tape the stencil on, but i'm sure you can buy those elsewhere.

if you don't do that, then solder paste syringes have luer lock connectors that you can buy disposable needles for. they work fine. the ones i have are 1mm OD.
>>
>>1252959
Thanks. Do you have any recommendations for where I can pick up a cheap polyimide stencil? The only one I've seen is $50.
>>
>>1252962
oshstencils. depends on your board size of course.
>>
>tfw the guitar pedal you built works
I just took the schematic from a website but god damn it's satisfying. And thanks to the anon who gave me the correction a few weeks back.
>>
>>1253023
Congrats and happy stomping.
>>
File: s-l1000.jpg (40 KB, 808x573)
40 KB
40 KB JPG
My solder mask sucks, it takes ages to cure and it sticks to the transparency more than to the board. Is it me, or there isn't a good way to apply this shit?
Can anyone recommend me a better source or technique?

Pic related, my current solder mask.

>>1252949
Hot air gun, fine soldering iron, tweezers, solder paste, flux, copper braid and the components themselves. That's pretty much everything. I also recommend getting some dispensing nozzles because the original ones are too big.
I have a chines youyue hot air + soldering station, 45€, but it werks.
Some adapter boards might be useful to prototype smd stuff on a breadboard.
>>
>>1253077
Centrifuge spreading. Chances are, you aren't spreading it evenly or your curing technique is all wrong in terms of light intensity/distance/type or ambiance conditions.
>>
>>1253078
>Centrifuge spreading
This isn't PCB lab grade stuff, it is quite dense, I wouldn't have the equipment anyway. I usually put the transparency on top and spread it with a credit card. The transparency is needed not to expose the pads, so I can't get rid of it.

I'm going to buy a new kind of transparency sheet and try exposing it under the sun, maybe I'll have more luck with a stronger light source. But if there's a better solder mask available I would like to know.
>>
>>1253086
>wouldn't have the equipment anyway
Dude, just get an old hard drive, open it up, put a piece of double-sided foam tape on the platter center, put the board on the tape, spread solder mask with credit card, put bottle over board so as to not smother you in mask, and turn the platter motor on. 30 seconds and you have an even spread.
>>
>>1253025
thanks famalam. Now to figure out how the circuit actually does what it does.
>>
>>1253115
Battery or adaptor?
>>
>>1253120
just battery for now. Will include the adapter once i put it in the casing.
>>
>>1252263
For a hobbyist not much more useful than a cheap kill a watt or similar, I'd sell it, can probably get at least a hundred bucks.
>>
>>1253091
And modern art on your walls! Better yet, try an angle grinder!
>>
>>1253125
Battery should last about a week of continuous operation. The LED that comes up when you stomp it into action draws about 1.5mA while the rest of the circuit needs about 1mA. Remember to pull the input plug to switch it all off, even if the LED is dark (bypass mode, no fuzz).

Be careful with the 9V adaptor, it must have (-) at the center and (+) at the sleeve of the DC plug (modern convention is (+) at the center and (-) at the sleeve). It also should be well regulated else you may have a 120Hz hum on your audio output.
>>
>>1253086
That's pretty much "the" stuff. I hear you might consider diluting it with some solvent first, web search around for details. Success has been reported using PC case fans in spin coating, which would be less sensitive than a hard drive to imbalances throwing your board off of the jig.
Seconded on those adapter boards. Simple, small circuits can even be built upon those alone.
>>
https://youtu.be/a43LXqRwQC8
This came out the other day, and it looks fairly nice. But from a more experienced electronics/audio standpoint, what do you guys think of it? I'm wondering if leaving an closed back to the speakers was a good idea, not to mention the crappy little USB Bluetooth audio sender. The "power cells" are a cool concept, though made pretty shitty because of the hot glue. You'd probably have to make a custom charger for them too, and getting them aligned looks like a nightmare. Of course concrete and wireless seem to be a little bit antithetical; being not very portable but not wired.
>>
>>1253154
Even brazilians are smart enough to enclose the revolving pcb with a cardboard box.
>>
https://youtu.be/ePEAj7Q5bv0
>>
>>1253214
But that's no fun!
>>
what should I use to make holes on my PCBs? I read the carbide bits are best, but why? drilling at what speed? what size should I use?
>inb4 depends
i want to put the little legs of the components through the holes
>>
Servos BTFO

http://mashable.com/2017/09/20/artificial-muscle-lifelike-robots/#PslCf_pHJSqn
>>
>>1253266
> I read the carbide bits are best, but why?

they last. also, you can buy used ones in packs of 10 or more. they get them from pro manufacturers once they're dull. but they're still useable by hobbyists.

> drilling at what speed?

whatever works. 1 of 2 on the dremnel speed knob feels right

> what size should I use?

typically between 1/64 and 1/32 (#78 to #68). the packs will usually be the metric equivalent of that. occasionally you'll need up to 1/16 for relays and connectors.
>>
>>1253266
>I read the carbide bits are best, but why?
because they last longer
fiberglass boards are very abrasive
if you only have a couple of holes hhs bits are fine
>>
>>1253213
tl;dw
>custom charger
that's literally nothing more than an off-the-shelf chip and a few resistors.
>>
>>1253346
Pretty much, yes. Even more so since they're protected cells. But he didn't release a video on how to make an aesthetic "power cell" charger. It's like running your brand new drill off two pieces of copper stuck into the battery receptacles with a bundle of 18650s taped to them.
>>
>>1253351
exercise for the viewer ig. I am okay with that.
>>
>>1253356
Still, the project is simple enough that any old retard can follow it if they have $40 or so lying around, and a majority of the channel's audience can't tell electron current from conventional. If it was a Great Scott video then I wouldn't bat an eye.
>>
>>1253289
>also, you can buy used ones in packs of 10 or more
If there's a PCB factory near to you, you could just walk there and ask for used drill bits. Or you could ask your dentist, but dental drills aren't as nice.

>1 of 2 on the dremnel speed knob feels right
Yeah, you can run them that slow just fine, but they're actually designed for massive RPMs. If you run them at proper speed, they wander less.

>>1253266
>what size should I use?
Ones which fits for your components and pads. For example 0.8mm for random components, 0.6mm for stuff like IC legs, 1.0mm for thicker stuff like pin headers and then whatever your special components need. Optionally some small bit for vias, like 0.5mm.
0.5mm bits snap rather easily and 0.6mm bit can be usually substituted with 0.8mm, though.
>>
>>1253154
Did you actually read the whole post?

>put bottle over board so as to not smother you in mask
>>
>>1253372
>bottle
Over a PCB? What size bottle or PCB are you talking about?
>>
>>1253373
Does it matter? Get a sufficiently sized bottle. I'm talking about bottles cut in half, if you're thinking in that direction.
>>
File: 900 amps.jpg (176 KB, 1272x760)
176 KB
176 KB JPG
>>1251608
Ok anon, my electricity knowledge isn't so good. This video by the LockPickingLawyer has sparked some debate:
https://www.youtube.com/watch?v=6aNm3qWkW8A

Is what he is doing practically suicidal?
Is it totally safe?
He said that since he replaced the microwave transformer's coil with a large copper cable with much less resistance, the voltage of the current was reduced to only 2V.

I mean it seems really, really fucking dangerous.
>900 Amp current
>using bare hands to hold circuit in place
>circuit only isolated by thin rubber insulation
>>
>>1253174
got it. Thanks for the advice
>>
>>1253387
The biggest danger is probably the heat.
>>
>>1253387
Is this a bait?

Anyway, there's obvious fire and burn hazards and his unshielded mains transformer is a (mains voltage) shock hazard too.
The 900A output won't shock you due to the very low voltage. You need voltage to push current and even if you connected the output wires to open wounds, the current would be rather minimal.
>>
I'm looking for DC motors for my autonomous tracked rover and found these brushless https://www.amazon.com/Gartt4x-Brushless-Outrunner-Quadcopter-SKYWALKER/dp/B00JFR4YEC.
Says they weigh <60g but can deliver up to 230W which seems way, way more than any other motor of this size i've seen. Is this legit? Are brushless motors just OP or what? Could I get a better deal by buying brushed? I only need a minimum of 10W from each motor, it's just that according to the table on the amazon page I could run these at ~12V and 1.8A and be in the clear while also prolonging their lifetime.
>>
>>1253394
>Is this a bait?
Of course, /diy/ is full of it.
>>
>>1253394
Ok, that's what I wasn't sure about.
>>
>>1253395
What speeds do you need? BLDCs are generally used where you need really high speeds (like in quadcopters). The ones you linked are 920kV which means for every volt on the input, it will give you 920 RPM.

Generally the ESCs that are available on the market will not spin these any slower than 5000 RPM.

And yes, brushless motors are just OP, these are legit, I have a pair of them. That thing will draw 20 amps per motor easy at 12 volts.
>>
>>1253440
I was thinking something like 0.5 m/s max speed, 0.5 m/s^2 max acceleration, estimated mass <3 kg and should be able to climb 45deg slopes.
It's my first robot though so I'm not completely sure if these numbers are realistic.

I was planning on using some gear reduction to drive the tracks, possibly using worm gears
>>
>>1253450
Yeah you can use gear reduction with brushless motors to get some amazing amount of power and torque. Your robot's speed would depend upon the wheels / tracks you would use on your robot along with your gear reduction train.
>>
>>1253450
0.5m/s in a second is nothing compared to slope climbing
>>
Anyone running an isolation transformer on the bench?
What do you use for this task?
>>
>>1253483
>>1253567
I've studied the theory in class (dynamics and such), but it's like I don't really have a feel for the numbers, you know? Sure I can work out the acceleration resulting from a given torque but I wouldn't know if it's too much or too little.
>>
>>1253581
> what do you use as an isolation transformer

i use an isolation transformer, oddly enough. pulled it out of some old gadget with tubes in it. there was one at the thrift store yesterday for $8 with 4x150W outlets. just like the one in the pic. didnt buy it coz too heavy for my backpack.
>>
>>1253605
Having driven fast r/c cars before, I can tell you that acceleration will almost certainly not be an issue unless you set up your gearing to go like 80mph. Just set up your gearing to go whatever top speed you want.
>>
File: Nokia_E52_2.jpg (491 KB, 2104x2868)
491 KB
491 KB JPG
I happen to have acquired a used Nokia E52, which does not turn on, no lights, nothing. All it does is vibrates when I press the power button. No signs of mechanical or water damage. Battery seems fine, 3.4V tells my multimeter.

What is likely to have died here? The screen or something else?
I assume if it were the screen, the phone would at least play the welcome tune and the keyboard LEDs would go on, right? Kinda most feature-phones do that when you turn them on.
>>
>>1253395
For land vehicles you generally want lower kv rating, often inrunner motors instead of outrunner, and you want ESCs that can handle lower speed and reverse, ones intended for aircraft use usually don't.
>>
>>1253695
>Telenor PK
>probably only ever called one number
>>
File: small_dc_motor.png (308 KB, 584x700)
308 KB
308 KB PNG
>>1253605
>I don't really have a feel for the numbers
Same here. I tried to recover from memory long forgotten stuff and see where it leads me.

I started with energy and time. When you accelerate a mass, the power*time you invest appears as kinetic energy of the mass: E=P*t=(1/2)mv^2. Since v=a*t and t=v/a you get P=(1/2)m*a*v. In your case it would take one second from zero to top speed and the power required is P=(1/2)*3kg*0.5m/s^2*0.5m/s which is 0.375W. As the other anon said, at such low speeds acceleration doesn't seem to be an issue.

Now slope climbing. Your mass gains potential energy E=P*t=m*g*h or P=m*g*h/t where h/t is the vertical component of the speed. If you want to drive uphill at ground speed v you get P=sin(slope)*m*g*v which in your case would be 0.707*3kg*9.81m/s^2*0.5m/s or 10.4W. You will arrive at the same result using P=F*v which may serve as a connection to torque.

Torque is the force that turns the wheel, the angular variant, M=F*r, unit newton*meter. Like P=F*v so P=M*w where w is the angular velocity (w=omega=2π/t, rads per second) and likewise v=w*r.
So M=P/w=P/(v/r)=P*r/v. If a single wheel with a radius of 1cm drives the vehicle up the slope the (geared) motor would need to produce a torque of 10.4W*1e-2m/0.5m/s=0.208Nm or 20.8Ncm. Units: (Nm/s)*m/(m/s)=Nm. So at least the dimensions add up..

The wheel would rotate at about 8 turns per second which is about 480rpm. The power is the net mechanical value and the real electrical power will probably be twice to three times as much. I'm afraid you will have to carefully verify all this, too long ago.. I'm adding the specs of a small DC motor that shows the typical performance curves. Good night for now.
>>
Do lipos on aliexpress really have the capacity they claim? Or would I just be wasting my money
>>
>>1253695
>>1253751

Do you even read anon?
>does not turn on, no lights, nothing
It's just a random wikipedia pic.

Looking forward for quality advice.
>>
>>1253859
>Looking forward for quality advice.

fat chance!

did you measure battery voltage under load? voltage readings are meaningless otherwise. and while the battery is out, hit the battery contacts and phone contacts with a tiny eraser to clean oxidation.
>>
>>1253695
>What is likely to have died here? The screen or something else?
Yes,
>>
>>1251932
Can you link the model of the receiver? And how much current does your fog machine draw that you have to use a MOSFET for it?
>>
File: yes.jpg (63 KB, 625x564)
63 KB
63 KB JPG
>>1253952
>Yes

>>1253910
>did you measure battery voltage under load
I didn't but from my experience phone batteries that are dead don't even charge up to a decent voltage, do they

or if they do, the voltage drops fast without load, but this phone been lying around an day at least since it was last charged (maybe way longer), and it was 3.3 or 3.4 like that, and I tried charging it, and the battery voltage raised very little or not at all
>>
File: LiPo-charge.png (100 KB, 546x789)
100 KB
100 KB PNG
>>1254013
>3.3 or 3.4
playing dumb or is it genetic
>yes yes
>>
>>1254022
why you so bitter anon, I honestly don't remember what it was precisely, but it was smth like that, either way it's in ok health according to your graphs isn't it. from what I know batteries that are below 3V are pretty much dead, this seems to not be the case

after some 15h of charging, the battery was 3.58 and the phone's behavior had not changed, and now, after 2 or 3 hours the battery lying on the table, it's still 3.58 and the phone still only vibrates

so fuck me if this is a battery issue
>>
>>1254137
If you don't have basic troubleshooting knowledge, you almost certainly do not have case opening tools, flashing software, JTAG pods, or SMT rework tools.
Consumer electronics repair is beyond this general's ambit. Try posting a thread in /g/
>>
>>1253695
>What is likely to have died here? The screen or something else?
This guy >>1253952
>Yes,
knows.
It's the screen or something else.
>>
>>1254146
>don't have basic troubleshooting knowledge
Yes I don't. Hence am asking this here.
See I'm not about to attempt to repair it myself. Well, if it was just about replacing the screen perhaps I would, since in these kind of phones it's all about screws and connectors afaik, stuff isn't badly glued together as in modern smartphones.
If it's (apparently) not, I just want a ballpark approximation of what to expect if I bring the poor thing to some workshop.

>>1254146
>Try posting a thread in /g/
A thread? Maybe I'm too modest, but this doesn't seem to me as a matter worth having a thread. I asked at /g/sqt/ already, got zero response.
>>
I only have access to 200V MOSFETs, to switch more than that can I just put a couple in series? Because that doesn't seem like it would work reliably, even if I got the gates to trigger them fully.
>>
>>1254386
Ok I found an SCR that can handle enough voltage, what's the difference when it comes to practical use? Is there a reason PWM circuits use MOSFETs instead of SCRs?
>>
>>1254388
Regular SCR can only be switched on, but the current through it must drop to zero for it to turn off.
>>
>>1254438
Well shit. After a little sifting I found I can get a BUK854-800A IGBT, which should do the job, though I'm wondering what to expect when using an IGBT over a FET, apart from no body diode. But for the purpose of a switched-mode power supply I don't think the difference matters.
>>
>>1254448
you need to care if you're switching faster than 10khz. igbts have a current tail on turnoff which makes their switching losses significantly higher than most mosfets.
>>
I'm interested in making myself a relatively cheap amplifier to play my bass guitar with.
The LM386 based Ruby amp seems like a decent starting point
http://www.runoffgroove.com/ruby.html
However it only outputs 1W whereas I suspect i need at least 10W. Power amps however seem to be quite cheap so would it be possible to create a 'pre' amp with the LM386 since it seems fairly flexible and easy to work with.
Would this be possible or would the amplified output from the 386 cause problems when input into a power amplifier (lets say 100W)
>>
File: pimped_bike.jpg (51 KB, 468x348)
51 KB
51 KB JPG
>>1251608
i got an amplifier kit so i can listen to music on my bicycle. its powered off +12V and -12V and looking at the schematic theres a bridge rectifier. I want to run it off a 12v rechargable battery whats the simplest way to do this? i was going to just plug it in to the +12v and grnd but maybe it needs the full 24v difference? thanks
>>
>>1252002
With all this stuff included? Good enough.
>>
>>1254692

It will need both + and -. Easiest way is to just connect two 12V batteries in series and then just attach ground between them.
>>
>>1254692
You want to run it off 12v?
So how are you going to power it?
Unless you mean you want to run it ON 12v?
>>
>>1254721
nigger
>>
File: super transistor.png (7 KB, 506x804)
7 KB
7 KB PNG
Very high-power bipolar transistors tend to have very low current gains at high collector currents, around 5..15. I wonder if it would be possible to use a medium-power P- or N-channel MOSFET as a "driver" stage to make it easier to drive. It would be roughly similar to an ordinary Darlington transistor pair.
>>
>>1254728
This is basically what a IGBT is anon. "Insulated Gate" as in the gate is a MOSFET and "Bipolar Transistor" as in the output stage is a BJT. Its just packed in the same package.
>>
>>1254680
The input impedance of your power amp is probably higher than your pre expects to see at its output, so characteristics might vary from those with a proper speaker load. Have you looked at the Albert Kreuzer bass pre designs? I've built the onboard pre (with some changes) and it sounds bretty gud.
http://www.albertkreuzer.com/electronics.htm
>>
can someone redpill me on programming stm32's?

I got a board with a stm32f103 and i got a chineseium stlink v2,got a hex file.
could not connect with jtag using stlinkutility.
I'll look to see if i made a mistake on the board, but i dont think i did.

some others tell me i should ftdi.
i suppose that works natively on these fuckers. im used to needing a bootloader for that to work.
>>
File: pn.png (4 KB, 288x266)
4 KB
4 KB PNG
>>1254728
many ways to boost current gain
>>
>>1254386
What are you trying to switch?
>>
>>1254680
Ruby amp as preamp is perfectly possible, the LM386 needs no load to be stable.
>>
i accidentally shorted the positive and negative lead of a laptop battery (was trying to jumpstart it) and something on the controller board burned, so now laptop just flashes the lights and goes off. what could it be and can i somehow repair it?
>>
if these stungun modules are so cheap why do real stunguns cost so much compared to them?
and is there real benefit of building a stun gun out of CCFL inverters, Pulse transformers and basically from real components instead of just buying one of these modules, sticking a battery and a switch and putting it in a box.
>>
>>1254868
To begin with, a stun gun is supposed to stun instead of being somewhat unpleasant. There are probably some legal requirements for them, too.

Those advertised voltages are comical.
>>
Go
>>
>>1254877
yeah i just ordered bunch of components for an "arc lighter" project just for fun and in retrospect looked at those super cheap things and wondered if i took the wrong path
>>
>>1254865
Not that you could do safely. Looks like you're going to need a new battery pack.
>>
>>1254881
>those super cheap things
have been there for ages and the fine print says that they self-destruct when operated without load
>>
Is a Tektronix 2236 100MHz Oscilloscope worth buying for cheap?
$75?
>>
>>1254610
I planned to switch at 70kHz, but since my total current is pretty low, the switching losses shouldn't be too high. Provided the switching action is enough for the boost converter to work properly, that is. The datasheet says that the maximum turn-off fall time is 800ns, which is a little more than 1/20 of a full cycle, which should be ok.
>>
>>1254868
>why do real stunguns cost so much compared to them?
Liability. It's expensive. One lawsuit can make you bankrupt.
>>
File: file.png (72 KB, 599x481)
72 KB
72 KB PNG
>>1254985
just be sure to do the math. you're looking at very high losses. 10+ watts on the turnoff alone.

2A->300V turnoff:
0.1mJ*70kHz=7W
>>
>>1254386
>200V mosfets
Still trying to build something useful from jaycar's non-selection?
>>
>>1254988
>2A
Not happening, this thing isn't going to be pulling more than a few hundred mA.

>>1254989
I'll have you know it's Surplustronics' non-selection.
>>
>>1254994
can you explain your circuit? you may well be right but when you say smps, 70khz, and milliamp primary currents you make me think you have no fucking idea what you're saying.
>>
>>1254386
>can I just put a couple in series?
Does your video game let you?
>>
>>1254994
Kek.
Wait, you need more than a 200V MOSFET for a boost converter? How does your switch even see those voltages?
>>
I need a soldering iron. I have one of those cheap $20 ones from a hardware store but the lack of temp control is pissing me off. It also takes something like 5min to come to temperature.

What should I get? I'd like to do some throughole and SMT repair work on some of the broken pieces of electronics that I own.

Thanks.
>>
>>1255001
active PFC is usually boost configuration and needs 5-600v switches. it's used in most(?) new computer power supplies. that's not what he's doing though, just a fun fact.
>>
>>1254959
Does it have a warranty?
It's old enough that it could be totally fuckin' fucked m8.

>just got a 7704a for free
>Broken PSU, cold joint on a resistor was causing heat, board obviously burned.
>POTs al needed to be cleaned and repaired
>Z-axis is still broken in someway
>I should probably just replace all the caps.
>Screen wobbles slightly.
>>
File: oki_ps900.jpg (18 KB, 321x400)
18 KB
18 KB JPG
>>1255002
>What should I get?
>>
>>1254997
>>1255001
400V geiger counter, still, figured since the load current will be almost zero that a boost converter would be as efficient and easier than a flyback converter. The whole thing pulls 0.75W quiescent, so I think there is some optimisation to be made, but that depends on the amount of ripple voltage I can withstand. Higher frequency and larger inductance means smaller voltage steps each pulse so less ripple, but also lower maximum voltage and more power. Ideally I tune it such that it reaches maximum voltage after a fairly long time, maybe a few hundred ms, but that means any sudden changes in the load (by a radiation count) could mess with the output voltage and perhaps take some time to get up to voltage again, and miss out some counts. To this end, if I use a 1mH inductor (largest I own) and 25kHz with a 1MΩ dummy load to represent a worst case scenario, I theoretically get up to voltage in 11ms, have a ripple voltage of 2V, and burn 0.6W quiescent when getting up to voltage (0.3W once stable at 400V and way less if I remove the dummy resistor). To get this power and frequency lower I would need to use a larger inductor, which I guess I could do by winding my own or putting a few in series. I also haven't tried changing the duty cycle or the capacitance, which might be able to make the changes I want without needing to increase the inductance. The simulated comparator is also setting a lower limit on the ripple voltage and I have no idea how that will turn out when I build the thing; maybe using Zeners is a better idea than a voltage divider.

>>1255000
I've got enough practical experience to know that its probably unreasonable to expect them to switch at the same time without some sort of matched pair, and since their resistance increases with heat (or so I've been told) they'd be astable in series. I wonder if using NPNs with an opposite thermal coefficient would work, since they would split the voltage through heat dissipation?
>>
>>1255017
Ideally I want 3 or 4 clock pulses to trigger the FET each time it gets under voltage at minimum load, to ensure I don't overshoot, and this depends on the sensitivity of the comparator. I'd also want an almost-full duty cycle when at maximum load.
>>
Anons, How can wind my own high amp transformer? Which is better ,winding a toroidal core or MOT?
>>
>>1255010
>oki_ps900
$250 ? come on man. that's beyond ridiculous.
>>
>>1255024
Get a chink hakko knockoff
>>
>>1255025
Now that's what I'm thinking... got some names/models or maybe a link to aliexpress/ebay, anon?
>>
File: circuit.png (17 KB, 1212x565)
17 KB
17 KB PNG
Can someone critique this circuit?

I want to make an automatic plant watering thing that will last a month a minimum, hopefully more.

Separate power sources are to maximize battery life and keep voltage to the micro clean.

Relay is supposed to provide isolation for the micro as well but the 6V circuitry is still connected so I dunno...
>>
>>1255026
Yihua is generally pretty trustworthy, I have their FX-888D knockoff and it works great. I wish I had gotten one of those two in one hot air/iron stations though now that I've started making my own PCBs.
>>
>>1255028
Thanks anon! Will take a look.
>>
>>1255027
Why use a relay over a FET?
>>
>>1255027
Use a MOSFET instead of the transistor and just wire the low end of the motor directly to it. The relay (800mW) is only wasting power. Look for an Rds(on) < 300 milliohm and a Vgs(on) < 2.2V.
Consider a low-quiescent-current 5V LDO regulator off of the 6V supply such as the LM2936-3.3. If there is no significant current draw (and with a MOSFET there won't be except for charging and discharging the gate), the power loss in the regulator will be small.
>>
File: tard.jpg (42 KB, 640x426)
42 KB
42 KB JPG
>>1255031
>Use a metal-oxide-semiconductor field-effect transistor instead of the transistor
>>
>>1255027
I agree with the other anons that there's not much point to using a relay.

Calculate your integrated power draw over a month. Even a mA adds up like crazy.
>>
File: 1485692778324.jpg (79 KB, 640x480)
79 KB
79 KB JPG
>>1255038
fine, then, instead of a blow job transistor
>>
>>1254902
i can disconnect the cells pretty easily, can that help? whats the reason for not being safe/possible to do?
>>
>>1255030
>>1255031
>>1255039
Thanks I'll look into a MOSFET.

From what I was looking at before it's hard to get one that'll turn on completely with only 2 AA's though. Still I guess I could still use a BJT to turn on the FET.

My other concern is that transistors can fail in short which would be bad.
>>
>>1255081
That's another reason to run the micro off 5V. 2.5V-rated MOSFETs are not super hard to find, check out the selector guide at vishay com. I grant they are hard to find in through-hole packages, but if you can work with TO-263 surface mount packages, either with a professionally made breakout board or just lifting up some copper by hand, your options broaden significantly.
If MOSFETs or BJTs are appropriately protected, which you have done with the diodes in your circuit, the odds of failure are maybe one in ten million over the next ten years.
>>
>>1255081
You should mention the duty cycle.
>>
>>1254775
I was just wondering if there would be any advantage to have a FET instead of BJT sourcing current to the big power bipolar transistor.
Maybe I could make an experiment with this sometimes. I've got one BD318 and a couple of MJ15004 that I could test with this arrangement.
>>
I was given and Beaglebone black and I want to make a robotic arm. But I have zero experience with hobbyist electronics or electronics in general. How should I start?
>>
>>1255144
Zero drive current (at DC) is one obvious advantage.
Smaller total voltage drop in >>1254728's cases is another, at least if you omit the resistors.
But like the other anon said, this is basically the same as IGBT. IGBT's selling points are easier drive than a corresponding BJT, better speed than a corresponding darlington and lower conduction losses (at high voltages) than with a mosfet with similar die size (=price).
>>
>>1255146
>How should I start?
Start by reading >>1251608
>>
File: circuit.png (11 KB, 1218x565)
11 KB
11 KB PNG
>>1255107
Yeah I'll just run everything off 6V and use a MOSFET then.

Now I just need to keep the power down. Looking into just using the internal watchdog or attaching a RTC. Probably need to add a crystal too if I want it somewhat accurate after a month.

>>1255143
It only needs to turn the motor (pump) on for a minute or so once a day.
>>
File: 2017-10-05 15.42.33.jpg (1.39 MB, 2560x1920)
1.39 MB
1.39 MB JPG
>>1254902
here's the picture of the burnt component, can someone identify it? im no sure if its some kind of resistor or what
also its thinkpad s30 battery pack which cant just be replaced like that (at least not in a cost-effective way)
>>
File: 1479296298028.png (43 KB, 800x444)
43 KB
43 KB PNG
>>1255196
I'm pretty sure you wouldn't need to run at more than 32kHz for this application. Consider the ATtinyx14 which has an RTC onboard, can clock the CPU from the same crystal as the RTC, and consumes negligible power when idle. For the LDO, consider an ultra-low-quiescent-current regulator such as the MC78LC50.
>>
Hey there, building an underwater propeller system for a project. Can anyone tell me if my battery/motor/controller set up wont fuck my shit up?

Motor
>6 Pole Brushless, 1000Kv
>Max Amp- 161 A
>Max Volt- 50V
>Max Power- 1000W
>Max RPM: 50k

Controller:
>SeaKing 180A Brushless ESC
>UBEC: 8A *

Battery:
>5000mAH, 40C

*Controller will be powered by bluetooth, do I need a UBEC as well? 8A?
>>
>>1255357
Better ask in the rcg general.
>>1249313
>>
>>1255107
>run the micro off 5V. 2.5V-rated MOSFETs
>>1255196
>run everything off 6V and

you mean 'ON'
run the micro on 5v. not off 5v. on 5v.
run everything on 6v. not off 6v. on 6v.
>>
>>1255390
I mean what I say. Go troll somewhere else.
>>
hi /ohm/

i bought a 2.5w d class amp from adafruit and a 10ohm speaker (i have another amp intended for this speaker coming)
and was wondering why im getting a knocking sound from it when plugged in.
for the record if i remove one ground wire from the amp to the FX board the knocking stops. pics will follow.
>>
>>1255341
I'm guessing it's a diode, resistors don't usually blow out like that without smelling bad for a while and possibly catching fire. Either way, desolder it and see if it conducts in either direction. Chances are its either non-conductive in both directions, in which case go ahead and replace it with a similar-sized high power component that has the right-sounding values, or it conducts in both directions, in which case it might have blown short and blown something else in the board, hopefully a fuse and not an IC. If you post well-lit pictures of all (both sides) the board I could probably determine the rating that the diode should have, assuming I'm not looking at 1/10th of the whole board.
>>
>>1255393
ok so how are you going to power it then? if you are running it off 5v or 6v then what are you using to power it instead?
>>
>>1255407
>plugged in
To what? Probably a ground potential difference.

>>1255411
A power tap OFF OF the power source in question. Stop being an autie and lern2vernacular.
>>
>>1251608
lol
>>
>>1255409
thanks for the reply, ill post the pictures tomorrow
also, when i accidentally shorted it for a milisecond, it actually caught fire and smelled bad, it even burned a hole the size of that popped out part in the plastic, so it may be a resistor, shouldnt a diode have a marked anode? on other battery boards i saw resistors of similar shape
continuity test sometimes gives no connection and sometimes a short, so im not sure if im even testing properly
what about the 50-something marking on it? the end of it is burned so it cant be read all the way but perhaps it can help in identifying it?
>>
>>1255341
I think it's a current sense resistor, possibly less than an ohm, which would pass a lot of current under a short circuit load. I think I see Kelvin connections on the top edge but it's hidden in shadow.
>>
>>1255422
better pics coming tomorrow, thanks for the help
>>
File: 1495071262317.png (73 KB, 625x626)
73 KB
73 KB PNG
>>1255428
>>
>>1255414
When plugged into ada FX board
>>
File: IMG_20171006_123118.jpg (1.16 MB, 2688x1512)
1.16 MB
1.16 MB JPG
>>1255465
>>
File: IMG_20171006_123053.jpg (1.22 MB, 2688x1512)
1.22 MB
1.22 MB JPG
>>1255466
I do get sound from the speaker if I connect it directly to the FX board via 3.5mm audio jack but it's very low.
>>
>>1255468
did you check for ground loops?
>>
Im posting this so you guys would tell me whats wrong with my calculation.

I have a 19m long heater wire, with a 1.6diameter which is almost 2mm^2 of area.
The resistivity of the wire is 1.45 Ohm*mm^2/m
Mains is 230V.

The above gives me a total resistance of: 1.45/2*19 = 13.70Ohm
Current drawn from mains is: 230/13.7 = 16.70A
Total power is: 230*16.07 = 3840W
>>
>>1255465
are any of the power supplies joined in any way? the signal ground out of any particular device may not equal V- or V+ especially if there are any "popless" switching games going on.

>>1255472
16.07 ≠ 16.70
looks right otherwise, what makes you think it's wrong?
>>
>>1255470
>>1255470
>>1255477
Ok thank you after looking up what ground loops were I realised the vin on the FX board was not outputting power it was to power it. So the amp was never even powered. After some trial and error I have a working reverse car alarm.
>>
>>1255419
I don't really have a sense of scale, but it definitely could be a (fairy big?) resistor, as the smell and lack of anode band seems to indicate. A current shunt (>>1255422) is definitely plausible, and if so it's almost certainly safely blown open-circuit. If you can read the PSU's model number you might be able to find a google image of the PCB and simply read off the resistance to replace it, otherwise some reverse engineering will be necessary. Hopefully there's just a comparator to read the shunt voltage and regulate based off that, but if it goes into an ADC then it will be pretty impossible to replace unless "50..." is enough to go by.
>>
>>1251876
>>1252002
Got my arduino a few days ago and I finally got around to opening it. Cant wait to dig in
>>
File: 20171006_013440.jpg (3.24 MB, 4032x2268)
3.24 MB
3.24 MB JPG
>>1255533
Picture to go along
>>
>>1255477
>what makes you think it's wrong?
Idk, better make sure. Im tired too. Thanks.
>>
>>1255501
>I don't have a sense of scale
Sure you do. The LM35 SOIC-8 body is 5mmx4mm. The putative resistor is about 6mm long and about 3.2mm wide, so likely 2512 case code (inch), so up to a 3W resistor. I don't think it's a fuse because of that catastrophic failure mode. Best guess from the one pic related only is 50mohm 1%.
>Hopefully there's just a comparator to read the shunt voltage and regulate based off that
That device is probably for overcurrent protection. They don't regulate, they just monitor and interrupt, like a circuit breaker. You may need to software-reinitialize the protection controller.
>>
>>1255560
>They don't regulate, they just monitor and interrupt
A bang-bang controller is still a regulator of a sort, if it turns on again with hysteresis. If it really does require a reflash of the MCU that would be a pain.
>>
>>1255565
From what I've briefly searched on the net, you might not need to reflash per se, just set some parameters via I2C. It depends on what's on the other side.
>>
File: Untitled.png (2.73 MB, 2287x571)
2.73 MB
2.73 MB PNG
>>1255501
>>1255560
>>1255565
here are the pictures of the entire board, i really hope i wont need no flashing for this one though
it still gives ~12v on the output but it shuts down when i power up the laptop (perhaps when it draws more current?)
>>
>>1255002
yihua 936
just make sure the ground is screwed on.
>>
File: asdfasdf.png (29 KB, 786x536)
29 KB
29 KB PNG
I've got this fairly simple circuit, that works fine on my computer, but when assembling it on my breadboard and powering it with a 1.5V battery, the LDRs (simulated by a R and potentiometer) have no resistance. When removing the red Rs the LDRs have resistance, just like when I disconnect the battery, what am I missing here?
>>
>>1255664
Is your opamp suitable for 1.5V operation?
>>
>>1255665
First I had the OP-AMP connected with 8 batteries (so +- 6V), but I disconnected it and still had the same problem.
What's also probably relevant: Somehow one of my OP-AMP battery packs got pretty hot (some of the plastic melted). Probably because everything is lying on my desktop, maybe a short circuit?
>>
>>1255628
Looks good, no digital distraction and incredibly cheap. No ground to screw on though, always connected to PE.
>>
>>1255667
>maybe a short circuit?
most likely, diagram is unreadable, better draw it by hand.
>>
File: DSC_0001.jpg (643 KB, 2113x1000)
643 KB
643 KB JPG
>>1255678
Here's a handdrawn version.
The battery to the left supplies 1.5V, while the batteries on the right supply the power for the OP-AMP.
>>
>>1255683
Much better. The input section seems to be floating. Do you want to evaluate the difference between the two LDRs?
>>
>>1255689
Yeah, I want to "show" the difference with +-1V at Vout.
What does floating mean? No ground?
And how can I fix it?
>>
File: 1507305324660-.jpg (65 KB, 1057x500)
65 KB
65 KB JPG
>>1255691
>No ground?
Yes, maybe like this?
>>
File: dCYJg[1].png (1 KB, 200x200)
1 KB
1 KB PNG
>>1255692
I assume it's a fairly stupid question, but how do I exactly "ground something"? Will connectiong the 0V parts of the batteries sufficient? What is the actual equivalent of this symbol when building circuits?
>>
>>1254721
you kno what i mean dont play dumb language is fluid things change meaning
>>
File: gnd.png (1 KB, 200x200)
1 KB
1 KB PNG
>>1255696
This symbol is normally used for earth, like the PE on a mains connection. The simple ground symbol (pic) denotes the common (zero) reference of the circuit. All voltages refer to this unless otherwise described, like |<--V-->|, a voltage between two points.
>>
File: Untitled.png (67 KB, 1024x768)
67 KB
67 KB PNG
Well, fuck, tiger mosquitos just became endemic in my area, their bites hurt like a bitch for days or even for a full week before healing.

Such hate for those little bastards drove me to make this. A modern version of the electric insect killer.

The way it works is simple, a laser shoots a low power beam, the beam bounces back and forth on some mirrors creating an invisible mesh, until they reach the sensor. If an insect interrupts the path of light the sensor causes a variation in the voltage, this alerts a microcontroller, which fires the laser at full power for a fraction of a second, enough to fry those fuckers.

I have some questions about this invention.
1) Is it feasible? Or will the laser loose too much power after bouncing off a couple of times?

2) What power would be needed to make it feasible?
>>
>>1255757
Mosquitoes aren't big enough to disrupt your rays.
>>
>>1255757
Why do you want to reinvent the wheel anon? Get one of those high voltage double mesh type mosquito zapper. The real issue is getting the mosquitoes to your termination chamber in the first place.
>>
>>1255760
octenol or co2 generator with a heater
>>
>>1255759
It just need to detect a small variation, even if I need to use an amplifier I think it's very feasible.
>>1255760
Usually mosquitos seem to avoid the meshes. I had one of those UV insect killers which didn't kill a single one. I think the plastic mesh to avoid getting your fingers also prevented mosquitos from getting inside.
>>
>>1255757
Hmm, interesting. I'd never considered making a laser fence like this.

Here's some numbers :
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4758184/

It would seem you'd need a couple of watts and hit it with a beam only a couple mm wide (if you actually hit it you won't need to track it with a 25 msec pulse, it will only fly a couple mm at most in that time). Trying to bounce a mm range beam down without would require extreme precision and rigidity for the mirrors though, seems almost impossible.

That said, it seems they get scared away at much lower intensities ... but where is the fun in that?

https://web.archive.org/web/20110626095044/http://techventures.columbia.edu/blog/profile/mosquito-repelling-light-barriers-show-great-promise-containment-malaria

I wonder if you couldn't use an airblast to shoot down a line of salt when a mosquite flies through. Have a bit of water down below to catch the salt.
>>
Guys, I just got into Arduinos and electronics and I'm looking at stocking my drawer with various components. I went on chinese sales sites and noticed there's a bunch of transistor grab bags for not much.

One seller has this kit:

#1
>BC337, BC327, 2N2222, 2N2907, 2N3904, 2N3906, S8050, S8550, A1015, C1815
>Quantity: 20pcs each type

and the other has

#2
>S9012, S9013, S9014, S9015, S9018, A1015, C1815, A42, A92, 2N5401, 2N5551, A733, C945, S8050, S8550, 2N3906, 2N3904
>Quantity: 10pcs each type

Which of these is more useful for 3.3/5V Arduino/RasPi breadboarding? I know fuck all about all these types.

Thanks!
>>
>>1252813
did you not even see the tac switches its holding, have you not by now realized its size or are you that new.
>>
>>1255487
why the fuck did you bend perfectly good jumper leads?
>>
>>1252502
This is worth a read: https://electronics.stackexchange.com/questions/67598/how-are-integrated-circuits-fabricated

The first comment explains what you'd be up against.
>>
>>1255487
Holy shit, dude!! Get some header pins on that thing!
>>
>>1253822
"""""Maybe"""""
>>
>>1256017
Flat profile best profile.
>>
File: 1477501907645.jpg (99 KB, 1000x1000)
99 KB
99 KB JPG
>>1255980
The second has a few higher-voltage devices which you might not need. The first has some higher-current devices which you might have a use for. I'd probably pick #1 for a beginner.

>>1256031
Fine, have some Pic related.
>>
>>1256044
too thicc

I'd go with some flat flex connectors for ultimate space saving.
>>
>>1255593
bump, help me anons
>>
>>1251611
Tinkercad circuits
>>
>>1256078
It's a fairly complicated circuit. Googling the name silkscreened onto the board did't help. Half the components not having readable names doesn't help. The thermal fuse probably hasn't gone, so I'd go with replacing the resistor for a 50mΩ one, though I don't know the standards for naming sub-ohm SMD resistors. Pic related gives me some hope. Googling the legible IC names doesn't get me anything like a microcontroller, but since there's what I think is EEPROM labelled bq2040, there probably is one.

If there's an easily available reset line on that microcontroller then you could simply power that for a second if replacing the resistor doesn't work (due to current sense failing I guess), but I doubt it would be that simple. Though since it outputs 12V with no load, chances are it just checks each time whether there's any current through the sense resistor when it detects that it's been plugged in through the other battery-pack terminals, so replacing the resistor should do the job.
>>
What is the ground terminal on a comparator even for? It might be why my circuit doesn't work, since I've got 5 V on the (+) and 120 mV on the (-) and it's outputting a logical low when running off a ~12V source. LM319, by the way. The datasheet says:
>Do not tie the ground pin to V– when the power supply voltage exceeds ±9V
So am I just to leave it floating or should I tie it to 5 V or some other rail?
>>
File: 1502998091497.png (76 KB, 856x677)
76 KB
76 KB PNG
>>1256046
>not just rawdoggin it

>>1256100
>bq2040
Multi-chemistry smart battery system (TI's site has a datasheet)

>>1256078
Can you unsolder the broken part to see what sort of stuff is underneath it? I agree with the above anon that you should just replace that blown package with a 50 milliohm 3W resistor in 2512 (inch) form factor and see if it works.

>>1256101
That's an unusual comparator. It looks like the ground terminal is just used for the emitter of the open-collector output. Pic related
>>
I've been thinking of making a ghetto oscilloscope substitute by playing signals through a speaker so you can hear the frequency, like an electrical stethoscope. Has anyone already done this sort of thing? I know about people making things to hear EM radiation, but nothing bout directly hooking up and measuring voltage.

If not, is there any simple way of dealing with widely-varying voltage and frequency ranges? It's not hard to come up with voltage dividers, and even frequency dividers, if you know what you're going to get, but I want something more universal. Currently I'm imagining I might need something like a multimeter where you switch through a bunch of different ranges.
Besides that, would it be better to try to stay analog and just drive a speaker with an opamp, or would going ADC/DAC be better/easier/cheaper?
>>
>>1256102
>what sort of stuff is underneath it
Oh yeah, there might be silkscreen on it that says "50mΩ" or something, though that's being optimistic. Chances are there's just a shitstain.
>>
File: o silly scope.png (7 KB, 800x600)
7 KB
7 KB PNG
>>1256103
>ghetto oscilloscope
I got ya homie

Yes that will work for frequencies between 20Hz and 20kHz, give or take a few %, but remember that human hearing is logarithmic, so a difference of 10kHz at the high end will sound the same as a difference of 10Hz at the low end. Good for listening for changes in frequencies, not good for knowing how much unless you're really musical. Different waveforms could be determined with a little practice. But playing it and listening back to it with a microphone on a computer will let you FFT it and pretty much use it like an oscilloscope with the right software. I think you can only do frequency divisions digitally, so if you want to get out of audio range you'll need to do that, though I've no idea how those RTL-SDR frequency dividers work since they detect really high frequencies. ADC it and put it through a few optocouplers and DAC and you've got yourself a scope that you just plug into your computer's microphone port.

If you want to make a display for it then a good feature would be to trigger the scope on falling and rising edges, which would be pretty trivial for digital only stuff with just a latch, a (big) shift register and some sort of variable oscillator to determine the frequency at which you clock the registers. Add an LED to each ADC output and you've got yourself a switch bounce o-meter. I should make one of those...
>>
>>1256106
*Add an LED to each shift register output
>>
>>1256104
Terminal configuration and footprint, mainly.

>>1256106
>how those RTL-SDR
They don't divide, they subtract. Look up frequency mixing. It's a little bit magical.
>>
>>1256101
>kW
Am I supposed to pull the output up through a resistor? Because it doesn't say, but that's what I get when I search images for "LM319 circuit". Damn datasheet doesn't say shit.
>>
File: help.png (10 KB, 1000x380)
10 KB
10 KB PNG
Can anyone tell me why the motor won't turn?
I don't know why it doesn't turn when there is a R.
And the circuit to the right is a voltage divider, right? So the motor should get half the voltage (aka same voltage as the circuit to the left).
When removing the R, the motor works fine.
>>
>>1256116
Use your multimeter and look at the voltages.
>>
>>1256116
This doesn't work, the motor asks as a small resistor in parallel with the left resistor which basically makes that left resistor like 5 ohms, making it worthless
>>
>>1256121
*acts
>>
>>1256110
>datasheet doesn't say
Read again, it says "The uncommitted collector of the output stage..." i.e. output is open collector.
>>
File: help2.png (5 KB, 654x380)
5 KB
5 KB PNG
>>1256119
I checked the left circuit, like I pictuired and got 3.21V with and without the R (I used two batteries).
>>1256121
Thank you!
>>
>>1256123
I'm not well versed in terminology such as common collector, what does that entail? Also now the 5V regulator that I'm comparing to is acting like a dead short and pulling my A23 batteries down to 4V. It never fucking ends.
>>
>>1256135
The regulator was fine, so I guess the 20 kΩ on the output of the comparator to 12V or the floating ground inputs caused the (+) input to be tied to 0V? Fucking weird. I'll mess about with the other 319 instead of burning myself on this dot-board.
>>
>>1256135
uncommitted collector, also known as open-collector (or open-drain in CMOS land). The general idea is that the chip only pulls down or not at all, and doesn't actively or otherwise pull the output up, so you can (have to) pull it up how you want it, which gives you more flexibility with your load e.g. LED, TTL logic, CMOS logic, or something more complex.
>dead short
You sure you got the pinout right? You're not using a big output cap, are you? Some 78xx regulators can be destroyed rather easily by reverse current such as by input shorts.
>using a linear regulator on batteries
Don't do this. Get a better power supply to save yourself some headache. 12Vdc or 5Vdc wall adapters should be easy to find.
>LM319
tbqh that's kind of an "advanced" comparator. Any reason you went with it instead of the single-supply, dead-simple 339 or 393?
>>
>>1256106
That top LED is upside down right?
And aren't resistive opto-couplers really low bandwidth? Like 250Hz low?
Still, wow I hadn't thought of optocouplers, that's insanely simple and way easier to use than I thought it would be!

Is there a way to avoid having to make a connection to ground in addition to the signal? I don't think you could use optocouplers for that though, so it would probably be a lot more complicated and/or add a lot of distortion for a really minor convenience. Or am I missing something?
>>
>>1256139
The regulator isn't dead it worked after desoldering and testing on a breadboard. I'm using 100nF caps, and I just tested that they haven't shorted.

The 5V regulator is just because it's more available than a voltage reference for the comparator; I'm not pulling any current through it. I'd just use a voltage divider if I didn't expect the voltage to decrease over time as this is a prototype for one running off 3 Li-ion cells. All I noticed was that it started heating up a bunch when I plugged the thing in, and the V-out was 0, which I'm blaming on the comparator because I don't understand it. Though the comparator wasn't warm, so who knows.

I picked the 319 because it said high-speed and I had no idea what switched-mode frequency I'd be running them at. Turns out it's probably audio frequency (I can hear a whine) so it wasn't really much of a concert after all. All those NCs though.
>>
>>1256143
It's a very bad circuit because of the resistive optocouplers and forward voltage of the LEDs, but one LED is the right way to catch the negative volts while the other catches the positive volts. It wouldn't be linear, but at voltages significantly higher than the forward voltage of the LED you would get some idea of what's going on.
>>
File: ferite bead.png (2.03 MB, 1222x719)
2.03 MB
2.03 MB PNG
What you see here is the primary side of the power supply module that belongs an old projector.
According to the markings the component that blew up here was a smd ferrite bead. The CXD9841M IC also has some scorch marks on the side facing the bead. I'll try to get a good pic of that with a microscope in the couse of the week.
Anyone got a clue what might have blown up that bead? I can't find any other obviously bad components on the board.
>>
I got this little 2 cell lithium protection module.
I figured I need to connect the first cell's negative terminal to B- and the positive to BM together with the negative terminal of the second cell. And then the second cell's positive terminal to B+.
But I don't get any voltage reading on P- and P+.
>>
File: protection-board.jpg (120 KB, 751x750)
120 KB
120 KB JPG
>>1256231
Connection is correct. Can you charge the batteries?
>>
>>1256231
>>1256241
Weird. I put a tiny bit of charge through P- and P+ to test if that works. And now I'm getting a voltage even though the cells were almost fully charged and were nowhere near the cutoff point before.
>>
>>1256247
That's why I asked. Some boards need to primed.
>>
>>1256251
*need to be primed
Same process as after undervoltage shut-off.
>>
File: burned.jpg (15 KB, 225x185)
15 KB
15 KB JPG
>>1256209
>pic of that with a microscope
No need, traces of the bead-eating current would tell more. Symptom known, cause wanted. Further diagnosis requires circuit diagram and multimeter.
>>
>>1256100
>>1256102
thanks a lot, i was googling around for a while and i found out 2 things: 1) markings are <value> mohmF and 2) there are current sense resistors with the same notch in one angle
so im now 99% sure its a 50mohm resistor, ill replace it in a few weeks probably, once i go back home
huge thanks again
>>
>>1256103
Buy a 2$ stm32f103 board from aliexpress, and make a USB oscilloscope from it. It has two 1 MSPS ADCs that can be overclocked up to 4 MSPS. There are plenty of examples, best one is
https://geektimes.ru/post/278106/ It is in russian, but it's the one where ADC is overclocked, but even without that it's still miles better than a sound card, which is only 48 ksps.
>>
File: 1487946796300.png (6 KB, 546x491)
6 KB
6 KB PNG
>>1256153
>more available than a voltage reference
TL431s should be pretty easy to find. You could also stack up a few diodes and use them as a shunt regulator, which, if the current through them is fairly stable, will be as precise and repeatable as the 78xx, albeit less accurate. You could also use a Vbe multiplier circuit, Pic related (top).
Vout on the comparator won't be anything but 0 unless you pull it up and connect Vgnd to the "low" voltage you want. Behaviorally, it's like Pic related (bottom), with some restrictions on Vgnd.
>high speed
393s should be good into the high hundreds of kilohertz. Also, dI/dt and diode recovery times tend to favor lower frequencies for higher output voltages.

>>1256377
Also, soundcards are often ac coupled, which is not a feature for dc measurement.
>>
How do I find out how much mA my Raspberry Pi 3 B uses? I'd like to get a battery pack so I can strap it onto it and put it my shed. Not sure how many hours it would last on a common cell battery charger pack...

I don't have a fancy scope or a bench power suppy... just a DMM.
>>
>>1256448
You can get cut open an USB cable and put your multimeter in between the red wire, or you could power it from a 7805 and measure current there, or just search online these numbers are pretty well known but it does of course depend on what kind of stuff the raspberry is doing but it is good for a ballpark figure
>>
>>1256448
https://www.google.nl/search?q=raspberry+pi+3+b+power+consumption&client=ms-unknown&tbm=isch&prmd=sivn&source=lnms&sa=X&ved=0ahUKEwj04b-y5t_WAhUMKMAKHdIuDusQ_AUI_QIoAg#imgrc=MpGKwvkl8wyzeM:
Here you go
>>
>>1256448
There are "USB charger testers" that provide a USB-A passthrough and will read out voltage and current.
It probably uses a lot more than you would like. Consider something smaller like an Orange Pi Zero that you can power-manage down below 100mA, or better yet a no-OS microcontroller like an STM32F10x board.
>>
>>1256455
>>1256448
I have also tought about a raspberry pi running on battery on a battery. But it is basically worth it and won't last long unless you get a really big battery. Arduino would be a nice idea but in my case I needed wifi so I used a wemos d1 mini, this is basically a $5 esp8266 with build in USB to serial. An esp8266 is a microcontroller with wifi that you can program with the arduino ide.
>>
>>1256451
>You can get cut open an USB cable and put your multimeter in between the red wire,
That sounds like a good way to go forward. I'll try that.
>>1256452
I plan on strapping a camera on it and do some processing. I doubt those BS marketing numbers will come close to real-world usage.
>>1256455
Thanks for the USB tester tip! I'll order one.
I have a RasPi that I used for other things but would like to see how good it is for surveillance and notifying me if anyone sneaks behind my shed. I don't have a fence there right now and wonder if there's any trespassing though the night. I installed a motion sensitive light as well.
>>
>>1256457
Yes, the esp8266 is patrician-tier, if still a bit power hungry for a microcontroller.
>arduino ide
drake-nah.jpg
>NodeMCU
drake-yes.jpg
>>
>>1256459
It is a bit power hungry but you can run it for ages if you program smartly around the different sleep modes.
>>
>>1256425
A regulator will still have lower q-current than any of those solutions, right? Either way that's what I'm stuck with, the only voltage reference I saw in store was a 2.5V diode and that was too low a voltage. Since I'm dividing down from 400V to compare the voltage reference to, using a higher reference voltage means more precision. The 7809 required too high an input voltage, but either way this is just a prototype.

After some testing, it looks like the ground rail determines what the low output voltage is and can't be left floating. I guess the "Do not tie the ground pin to V– when the power supply voltage exceeds ±9V" doesn't refer to Vcc being higher than ground by 9V, so there's no problem anymore if I just tie those ground pins back to 0V.
>>
>>1256470
It's still not working...
>>
O shit we bumped out
>>
>>1256512
i got this
>>
NEW BREAD
>>1256544
>>1256544
>>1256544
NEW BREAD




Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.