/diy/ - Do It Yourself

>>1483995
this thread's digits brought to you in part by the 1N3995 10 watt Zener diode

>>1484002
How are you supposed to mount those packages?

>>1484009
jesus, tone it down with the sexual innuendos

in a buck converter like this, there are in +/- and out +/- terminals.
The high voltage goes into IN and low voltage goes from OUT.

The converter has 12V going in and 5V going out.

So let's say i have arduino that can only handle 5V and i connect it's + pin to the OUT+ pin on the converter.

Can i connect the arduino's ground pin to the IN- terminal on the converter or do i have to use the OUT- terminal?

>>1484025
Can you already operate an ohm meter? If so, find the resistance between IN- and OUT-. If not, use OUT+ and OUT-.

>>1484032
unfortunately i only have a multimeter

>>1484033
>a multimeter
a multi meter measures resistance anon

>>1484033
Ok, then better use OUT+ and OUT- .

>>1484033
take care son

>>1484038
then why don't they just call it resistance meter? this makes it confusing as fuck

>>1484049
because the 'multi' has no unit

>>1484049
continuity-hfe-current-voltage-remote tester-capacitance-inductance-resistance-o-meter

>>1484009
consensually

>>1484074
*features may vary

Can anyone help me out in identifying what this component is? 2 years ago i ordered a 6502 clone from a local new-old-stock seller and a bunch of these were bundled with it in the package for some reason.

There is a window in the middle so i'm guessing it's an optical component. From what i've been able to find online it could be an IR sensor or perhaps a laser diode (though it seems too big for that).

(second pic)

>>1484105
it seems to be a emitter/receiver pair. post pic of window

>>1484049
The actual mystery is, why the fuck do they call it a meter when it can't measure any distance at all

>>1484109

Here you go. Im not too sure if it's part of a pair, the window seems pretty deep and narrow for that. The closest thing to it i've been able to find online is a thermopile IR sensor:

https://www.alibaba.com/product-detail/IR-Thermopile-Sensor-For-Contactless-Temperature_60519416357.html

>>1484106
Remove the cap and see for yourself.

Hey ohm, woodworking hobbyist here with only a passing knowledge of electronics.

A little while ago someone brought me a free drill press that had some working issues. I think the starting capacitor was shot but not sure.. Not exactly sure how to test it, but at first drill press stopped turning up itself, but could be push started essentially. However it lacked sustain and any kind of moderate speed or sized hole would slow the motor down. Now it just doesn't start at all. Maybe it's not the capacitor, maybe it's the motor. Does anyone have any advice on what I could check / try?

>>1484124
Uhm, weird. No markings of any kind? I said pair because it`s a 2 port thing.

>>1484143
try replacing the cap they are cheap

>>1484145

I think i narrowed it down - it's most likely a Melexis MLX90614, 30dg FOV variant:

https://www.digikey.com/product-detail/en/melexis-technologies-nv/MLX90614ESF-ACC-000-TU/MLX90614ESF-ACC-000-TU-ND/1948826

What's odd is that on all the pics online they have white markings on the side, but mine doesn't...

>>1484025
Depends entirely on whether or not there's a transformer electrically isolating the input from the output. If your pic is the SPS you're referring to then the ground is very likely common to the input and output, Use an ohmmeter to verify this.

>>1484170
He only has a multimeter.

>>1484169

Btw the Digikey price appears to be heavily inflated - you can find them on ebay for as little as 3$. They operate via I2C - two of the pins are PWR and GND, while the other two are SDA and SCL. There's an arduino demo sketch on adafruit's website that's worth a look.

I am chasing up an ignition problem on a bike, getting at best one or two sparks out of the plug for 10s of turning it over with the plug held touching the block.

tested the diode, 1 side infinite ohms, other side 600 ohms. did i find the broblems?

>>1484179
sounds perfectly normal for a diode

>>1484170
ever seen a buck converter that uses a transformer son?
thats not how they work

If i want to connect a device to my car battery what best place in the car should i pick for this? Seems like running wires directly from the battery would be too dangerous since when the car starts then lots of current is flowing

>>1484181
>The ideal diode has zero resistance for the forward bias polarity, and infinite resistance (conducts zero current) for the reverse voltage polarity

pls explain

>>1484189
Ideal diodes do not exist. The measured forward resistance depends on how your meter is built. A meter using 1mA test current, interpreting the measured 600mV voltage drop (typical value) as resistance would display 600 ohms.

the human body is supposed to have about 750k resistance so why the fuck is my meter showing 2M ohms when i place the probes on my finger 1 cm from each other?

>>1484186
current flow to other systems in the car doesn't interfere with other devices that way. it's absolutely the best place because it's electrically close to a big energy storage component that will absorb spikes and electrically distant from everything else

>>1484208
Are your fingers callused?

>>1484212
>callused
no, they are oily and smooth like the inside of a baby's foreskin

>>1484203
ohwell, thanks anon, the search continues.

i really got the urge to get a fuckton of transistors and a couple of these 7 bar clock displays(on pic) and make a simple calculator or anything really
i've done some low level programming so i could probably get the logic right and i'm definitely autistic enough to keep going at it

but i havent done much with electricity and have no clue how to even start
breadboard? what kind of transistors? where do i buy the shit? how much will it cost etc
i wanna make something like pic but way smaller

>>1484251
>the 60 w clock

>>1484252
it doesnt even need a timer, just make it count the ac cycles and go from there

>>1484257
ac cycles are very imprecise actually

>>1484267
is there a limit to how off they can be?
i measured it once and it was spot on 60hz

>>1484257
it can double as a heater

>>1484269
a lot, if you measure it say in europe, it can vary by up to 10hz

>>1484273
>>1484267
troll

>>1484269
ac cycles are very accurate and are commonly used for clocks.

>>1484251
194 Transistors
566 Diodes
400 Resistors
87 Capacitors
10"x11.3" Printed Circuit Board
Displays, transformer, etc
1250+ parts
2700+ solder joints
http://www.kabtronics.com/
~$200

>>1483995

LOL I see that OP finally almost admits that there's no such thing a s site for "problem ideas", whatever that could possibly mean, but he/she is not daring enough to completely correct the blunder.

>>1484284
For a clock it's not only the frequency (which varies with the load) but the phase that counts. You can watch the regulation process live at http://www.mainsfrequency.com/ and compare the grid time to UTC.

>>1484269
>is there a limit to how off they can be?
>i measured it once and it was spot on 60hz
In the US 60hz is held very close at all times.
Fine adjustments are made during a 24hr period to have the total cycles to be as near 5,184,000 as practical.

>>1484269
it is usually kept very close to the mark, or the synchronous generators will jambo their mambos. pic related is a medium pressure turbine, it`s speed must be also controlled.

>>1484347
It`s really amazing to think that all these fuck huge machines are spinning at the precise speed to generate the same electrical frequency. I mean the transformers of this alternator use 3x2MW engines just for pumping coolant.

>>1484344
In 2011, The North American Electric Reliability Corporation (NERC) discussed a proposed experiment that would relax frequency regulation requirements[27] for electrical grids which would reduce the long-term accuracy of clocks and other devices that use the 60 Hz grid frequency as a time base.[28]

>>1484185
Flyback converters do, and if you search for an isolated buck converter that's exactly what you'll get. Incorrect terminology or not. There's even a small chance that there's 2-3 isolated windings inside that core, though I doubt it.

I will more likely than not be called a humongous retard trying to do incredibly dangerous shit, but here it goes. I have basic (and I mean basic) electronics understanding.

I want to make a spiked taster shield, because LARP (I know, I know). I doesn't have to be actually protecrive. What I really want to do is make a curved wooden shield-shaped piece of shit, countersink some holes on the inner side, drive nails through so they stick out on the outward facing side, and wire every other nail up to some high voltage source so (ideally) the entire outward face goes up in a bunch of arcs between nails when I push a button. Slap generous insulation (more wood and probably a rubbery material) on the inside so I don't touch shit and then also a handle, call it a day.

My questions are the following; I could use a chink 1000KV high voltage generator or something, or maybe scrounge up a flyback transformer and get a transistor to switch DC from a battery pack into it REAL fast to get something sorta kinda like AC on the other side. I want the arcing to be on the nail points and not the nail heads (which would be closer together), is countersinking the spots where I drive nails through then sticking some isolating gunk on top of the heads good enough to prevent arcs there? Also, is the arcing distance dependant on how big the current is? I'd have alternating nails wired up "in parallel" to positive and all the other ones the same but with ground, if you consider air paths conductors. Because LARP shit, I'd like every nail to arc with it's neighbors and for the arc to be visible (doesn't have to be massive, just there), and while I'm pretty sure voltage would determine the maximum arc distance I wonder what the minimum current of each would be to get one. I couldn't find any info online about that.

hey, i have a marantz pm151 amplifier and the right channel is pretty dead.
was wondering if anyone knows what can cause this and if i could attempt a repair myself? or just take it to a repair guy?

>>1484420
>wondering if anyone knows what can cause this
yes
>>1484420
>or just take it to a repair guy?
yes

>>1484422
yes

>>1484318
>It is an art/science of applying principles to problems. Find problem, learn principles, design and verify solution, build, test, post results, repeat.
now do you understand why the word "problem" is used here?

>>1484414
Use a chink module, but it won't work very well anyway. Air breaks down due to it's dielectric strength, 3MV/m. Surpass this electric field strength and air will start conducting. Once it's started to conduct the voltage can be decreased to an extent, but this isn't exactly something you have control over anyhow. The main problem is that you've got a bunch of what are basically zener diodes in parallel all with slightly different voltage ratings, so once you surpass one of them then all the current will flow though that spark gap and not the others. Only way around this would be putting massive current-limiting resistors in series with each spark gap to stop one spark from shorting the voltage of the rest. Have fun finding resistors that work flawlessly at 150kV. Also the voltage ratings of those chink modules are often overestimations.

how exactly does static electricity destroy electronics?
My understanding was that if you push too much current through a too thin wire it will heat up and then something will melt and thus be damaged.
But static shock is too short to heat up anything.

>>1484523
The "wires" in chips are microscopic and the gaps between them are too.
>https://en.wikipedia.org/wiki/Electrostatic_discharge#Sparks

>>1484523
one failure mode is oxide breakdown. MOSFET gates are layers of silicon oxide on the order of 100nm thick. a large voltage potential between gate and channel will cause the oxide to break down, thus increasing gate leakage or shorting the gate to the channel
another failure mode is joule heating, where a p-n junction or other silicon feature conducts too much energy and melts, thus destroying the junction. it doesn't take much, a couple of hundred °C will do it
finally, thin metal on interconnect layers can be melted/burnt by an ESD pulse, resulting in an open and/or short circuit
all three failure modes can be complete, resulting in catastrophic failure of a device, or partial, resulting in degraded operation and shortened life
https://www.electronicdesign.com/power/identifying-eos-and-esd-failures-semiconductor-devices
>too short
joules = watts * seconds. a typical static charge on a human body is on the order of a millijoule or two. divide by the "too short" time of a static discharge and consider the micron-scale features called upon to dissipate all those watts, and you're looking at real energy density

>>1484523
>But static shock is too short to heat up anything.
The innards of ICs are microscopic.

>>1484531
>>1484533
>>1484534
welp i guess i'll add some of them zener diodes in my shit, so they can send any static directly to the ground

Hey guise, I just completed my first electronic circuirt and I have a few questions. It's an Atari Punk Console (a very simple oscillator).
The result I have sound a lot dirtier than what I got with the breadboard, why is that? Shitty soldering? Does the chip get hurt if I spend too much time heating it while soldering?
Also, there is one potentiometer in the circuit that doesn't work for its whole range (I get something for 25% of its range and then nothing), how can I fix this?

>>1484542
>doesn't work for its whole range

see Rule 3 in the top-most post

>3 Post schematic/picture/sketch/9001.5 hours in MS Paint with all part numbers/values/etc when asking for help.

yes you can destroy components with heat but since the thing works, that's not likely. if it sounds rough, it's because it is rough. square waves sound horrible and should be outlawed. you can smooth it out somewhat by putting a cap across the speaker. start with 10uF and work up.

>>1484535
To be fair, ESD damage isn't usually a big issue with hobby shit. If you check people's projects, most of them have absolutely no protection and they still work "just fine". Similarly, Tarduino and RPi have zero protection in their GPIO pins and yet they survive the handling, most of the time. When something finally gets zapped, there are usually plenty of other possible reasons to it, and so the ESD deaths have a good chance of going undetected.
The ESD damage becomes an important issue in mass production, where even relatively uncommon failures directly show as production rejects and you can attribute them properly to ESD, instead of "maybe I just shorted something with a screwdriver".

This isn't to say you shouldn't bother with ESD protection. It is still the right thing to do, even with hobby stuff.

>>1484542
>doesn't work for its whole range

see Rule 3 in the top-most post

>3 Post schematic/picture/sketch/9001.5 hours in MS Paint with all part numbers/values/etc when asking for help.

yes you can destroy components with heat but since the thing works, that's not likely. if it sounds rough, it's because it is rough. square waves sound horrible and should be outlawed. you can smooth it out somewhat by putting a cap across the speaker. start with 10uF and work up. (bonne chance, mon lécheur de bat.)

>>1484550
>if it sounds rough, it's because it is rough.
I meant compared to my breadboard version. When I first tried, I had a somewhat clean square wave, now it sound all weird (not that I don't like it).
Regarding the potentiometer, yesterday, it worked fine, did I destroy it while building the thing?

>>1484555

seems pretty obvious you didnt replicate the breadboard circuit correctly. i.e. you have one or more wiring errors, or the wrong components.

>>1484555
such appropriate digits for this project
obviously something changed between breadboard and dotboard. I suggest trimming those protruding leads, cleaning the bottom with alcohol, and testing for shorts between adjacent pins. also make sure you installed the electrolytic caps the right way

>that feel when you finish your cricuit, turn it on and it starts smoking

>>1484571
never had that, thanks for the chuckle though

>>1484555
I would include current limiting resistors (1K) in series with the pots to protect the wiper contacts in their end positions. The maximum current into pin 7 may be destructive.

>>1484571
Reminds me of one project where an IC popped its case open, produced lots of yellowish-gray smoke and when the smoking ceased, the IC chip inside was glowing orange.

I would like to use green energy in my house, but solar panels won't work, because it's constantly cloudy and rainy in this shithole.
So I was thinking, since there storms here all the time, what would be the best way to harvest lightning and use it to burst charge an array of batteries?
Now attracting the lightning is easy, but how do i handle the current and store it in the batteries?
I was thinking i could use parallel caps to store the burst and then trickle it down into the batteries.
What do you think?

>>1484574
>popped its case open
Was it a CDIP?

>>1484577
No, plastic. About one third of the top was blown off.

>>1484575
https://en.wikipedia.org/wiki/Harvesting_lightning_energy
It doesn't look practical. Look into microhydro if you're not in a city, that and wind power. But if you plan on selling back to the grid I'd advise against it.

>>1484575
don't forget to add a zener diode to divert any excessive voltage into the ground so you don't damage the caps

>>1484593
>a zener diode
I think a spark gap would be a better choice, unless your zener happens to have a large heat sink on it.

>>1484606
I'm sure a 10W zener like the 1N3995 ought to be enough :^)

Can i make a functional spot welder by just putting caps in parallel and charging them?

>>1484614
someone did, google

>>1484616
I'm pretty sure Google can afford just to buy one instead of building it

>>1484573
I think your solution could work. I noticed that as the battery goes down, the range of the potentiometer becomes larger.
I'll try to add a resistor later if I can find a place on the dotboard.

Can anyone recommend me a similar circuit to work on next? I'm planning to build the baby 8 step sequencer, but I'd like to practice a little more first.

60 watt clock guy, still have no idea what kind of transistors to buy and whether or not i should use a solder or solderless board
please help

Trying to repair a harbor freight drill, and I need help identifying a Dpak component.

These two DPaks failed (overheated, and shorted I think) on the main circuit board. I'm not sure if they are some kind of constant current driver, or just a power transistor. I don't think they are voltage regulators. When you press the button on the drill, 20V from BAT+ goes to the tab. Pin 1 goes to a few resistors, and pin 3 goes to BAT-. Anyone have any ideas?

There are some other pins that go to the MSP430 micro, the VDD label is the micro's power, which I think is lower than 20V. The green line is from battery to the micro.

The marking on the DPak was melted off, but I could make out a CS something. I found some On Semi adjustable VRegs, but I don't think those are the part (e.g. CS52015) as the pinout seems to indicate something else.

>>1484681
>what kind of transistors to buy
Second cheapest, in your favorite package.
>solder or solderless board
You want every part to be functional at the same time? Solder.

>>1484681

there's a sorta parts list at >>1484306

this is a dumb project. find something less insane. if you NEED a conversation piece in your boring life, get a dick piercing.

reposting because i still dont understand:

i have an AD7607 ADC and the datasheet (https://www.analog.com/media/en/technical-documentation/data-sheets/AD7607.pdf)) shows a plot of frequency response+phase response. pic related.

i feel like the phase response is fishy though. it says it is a second order butterworth filter, the sample frequency is given as 200khz and the cutoff frequency is 23khz. when i calculate the phase delay with matlab, it appears to be way more delay than stated. which would somewhat concur with my measurement.

could anyone possibly countercheck? or did i miss something obvious

>>1484705
i dont need anything, just figured i'd do something instead of playing vidya and fapping to shemales all day

>>1484708
You were told that the plot could be a group delay plot.
Pic related, 23kHz 2nd order Butterworth filter's group delay.

>>1484715
you mean this post?
>>1483780

did not know it was meant for me. got about the same result as you now, thanks.

now why wouldnt they state it as group delay tho?

What would be the easiest way to create a spark without using an ignition coil?
I need a simple circuit that will create a small spark between 2 wires so I can ignite an aerosol

I didn't realize electronics youtube community was such a cesspool of perverts.

>>1484700
They'll probably be switching FETs, and they look to be in parallel, which is odd for semiconductors. Since it's low-side switching they're NMOS, with the source being the pad, the gate being the left-most pin, and the drain being the right-most pin. At least from what I can see. I'd check what switching voltage is used, but it's probably 3.3V. So buy a couple of N-ch MOSFETs with a threshold voltage below 2.5V, V_DS above 24V, and as low an R_DS_on as you can reasonably get.

But she's a harbour freight drill, so don't expect any lasting performance even after fixing it.

t. never fixed anything

I bought some of these no-name combination charger and boost converter modules to see if they were any good. it seemed like a good idea, combined a small adjustable boost converter with a 4056 charger without cell protection (because I have cells that are internally protected).
unfortunately they draw 200-300uA from the battery which would drain it pretty quickly. I had to cut the trace connecting the boost converter so I could insert a switch in line.
If these things had a very low quiescent current and/or a pin header for a switch, they'd be great. too bad.
amazon B07DL6CMTK, a few other vendors sell these too under various different part numbers

>>1484724
A classical electromechanical ignition system is basically a flyback converter without rectification or filtering. A flyback converter is essentially a boost converter with a transformer. Without the transformer, the switch would need to take the full spark voltage.

The simple way to generate sparks is with a piezoelectric crystal. That's what cigarette lighters, hob lighters, etc use.

>>1484875
Well, calling it a flyback converter is a little disingenuous since it only fires once and isn't rectified at all.

>>1484724
You can always wind your own ignition coil/flyback transformer, since it will not be in need of the current-carrying-capability of a car's ignition coil. Buying a small ready-made transformer for some other purpose may also work, such as a strobe trigger transformer. But you're going to need fairly high voltage (3MV/m). If you have the luxury I'd go with the piezo ignition.

>>1484025
I have almost this same converter, same chip and looks like same components just no readout or buttons, and I'm trying to get my msp430 launchpad powered off of another board powered by 12 volts. The other board has an ic with pins that are meant to be connected to pushbuttons that connect them to ground for the button function. I want to use the msp430 to pull these pins to ground, or alternatively feed them high (I'm assuming whatever the Vcc is, I'll set the buck converter to like 3 volts) when I don't want them "pushed".

I checked on my buck converter and the ground is indeed common, so is there any issue just directly hooking up the microcontroller's output pins to the "button" pins?

The other board is a class D amp and they're volume buttons, up and down, and "actuate" when pulled to ground. I just want to make it controllable by IR. I think I have enough info from the IRremote documentation to do that part of it.

>>1484962
the buttons don't drive the inputs high so neither should you. let them float high
you would need open drain outputs on the msp430 that can handle the 12V. since you probably don't have them, you have to make them. you can use an npn bipolar transistor (2N3904, 2N4401, etc.) with a 10k resistor between the transistor base and the msp430 GPIO. connect emitter to ground and collector to the amp's button input. drive GPIO high to push button, low to release

>>1484857
Thanks. I didn't buy it, but I did break it, and I figured it would be fun to repair it if possible.
Funny how a couple power transistors can break a drill. I wonder if old drills even have pcbs at all... I doubt it.

>>1484989
old drills probably don't have tetchy but power-dense battery packs. the few that had adjustable speeds did it triac-style and certainly didn't have a high power factor. features generally require components, same as ever
(but a microcontroller? that one's new on me)

>>1484971
I was thinking of doing exactly this if the answer was no lol, so basically provide the button pin to ground connection via transistor?

>>1484992
you got it

>>1484989
Old drills have the trigger switch mechanically switching the power supply, while the modern ones use transistors such that they can switch far faster as a method of speed control; they PWM the output to change the power going to the motor. Though I doubt that drill has speed feedback.

>>1484991
>triac-style
Never heard of them doing that, though I guess that's since most teardowns I've seen have been of cordless drills.

>>1485000
yep just ye olde phase angle fired dimmer switch, with a trigger controlled pot
https://electronics.stackexchange.com/questions/167556/how-does-the-circuit-of-a-basic-variable-speed-electric-drill-work

>>1485003
I know how a triac dimmer works, just surprised that that's what they were using. My old Hitachi corded drill has dimming but it's a step-wise switch with 10 discrete speeds, unlike what I'd expect from a triac dimmer with a linear pot inside the trigger switch. So either it's a triac dimmer with a bunch of resistors switched sequentially by wipers or spring contacts within the switch, or it's some sort of less-analogue dimmer. I'd imagine it's sequential spring contacts since it's been going strong for a few decades without any sign of a linear pot wearing out, which is a good enough reason to use such a system. It was made before the common clamshell setup we've got today, the back 1/4 comes off with lengthways screws, with the front gearbox housing and the left quarter being moulded together, if I remember right. Big front-end bearing on it too, only one speed in each direction and certainly no clutch; nothing I'd want to drive screws with.

Not sure if this is the place for this, but I need to get my Galaxy S7 to report sensor data to an Arduino. Any idea how to do this?

>>1485045
Write an app for the droid. Easiest way from there would be to send it through wifi or cellular to a web database, and from there to the arduino. If you know what you're doing you could get it to send data directly to the arduino, not really sure which would be easier. But if you're writing an app for it then chances are you could write an app that does what the arduino is doing anyhow.

>>1485006
The most common way to produce the stepped effect is to add some mechanical stepping on the otherwise linear pot.
An old Bosch drill I opened a while ago had a thick film hybrid for its dimmer. The hybrid included a linear pot, or more accurately, just the resistive track. It would've been a simple matter to replace the track with a bunch of contact pads and some printed resistors. It wouldn't have increased the price either. IIRC it actually had one such contact pad at the low end of the pot for the "totally off" position.

>>1484875
>>1484895
I don't think I've a servo powerful enough to push the piezo spark from a lighter. I was thinking maybe to get one of the spark things out of a toy gun and run it off a small motor. Any idea if those Sparks are hot enough to ignite?

FINALLY
After weeks upon weeks of struggling with useless pieces of shit capacitive switches and their random fucking triggerings out of nowhere, after ripping my fucking hair out because no fucking troubleshooting worked, i fucking ripped it all out and tossed it into trash where it belongs.
And now i finally have a 100% reliable touch button that just fucking werks flawlessly and all it took was two strips of aluminum tape and a single 1 mega ohm pull down

>>1485219
Even with gloves on?

>>1485231
who the fuck wears gloves in the kitchen?

>>1485231
>Even with gloves on?

i dont wear gloves on my dick, silly boi.

> I was thinking maybe to get one of the spark things out of a toy gun

you mean the black powder caps for a cap gun? sounds hard to get that automated. one thing they use in gas powered RC planes is, essentially, a light bulb with the glass broken. you pass a current, it gets red hot and ignites the liquid or gas.

can a fuse be only half burned? i have a knife fuse that has a red dot indicator showing about 90%, the rest of it is white and when i measure the resistance it shows 0 ohms so why isn't it working properly?

>>1485250
I think he means those cylinders that spin and make sparking. I think they use something similar to those pyrophoric flints, though perhaps they're just a bastardised vann de graff generator.

>>1484681
You should probably get whatever complimentary NPN PNP pairs that are cheap. 2N3904 and 2N3906 are pretty much the standard jellybean BJTs and should work fine. Or you can do 2N5088 and 2N5087 or nearly any other small signal transistor pair.

Bear in mind there's more to this project than getting the logic right. Every transistor needs to be biased so that's a lot of resistors though honestly if you're getting a bunch of trannies anyway you're better off using current sources for biasing then. You'll also have problems with fan out so you need a lot of buffers. To mitigate issues with fan-out while minimizing the number of buffers you have you may want to consider BJTs with a higher typical beta, Darlington or Sziklai pair, or bootstrap circuits. You'll still need buffers though, just less of them.

I wouldn't recommend MOSFETs. Not only are discrete small signal MOSFETs like the 2N7000 or 2N6660 more expensive than similar BJTs but the higher gate capacitance could cause capacative loading issues and major issues with propagation delay and stuff.

As for breadboard vs solder. Solder. Definitely. Don't waste your time breadboarding something this large. You will give up out of frustration very quickly. Design a PCB, get it manufactured, hope it works, if it doesn't see if you can trace out a do some kinda bodge repair to fix the fault (ie jumper wire to pin, cut incorrectly routed trace, etc.), once you've nailed down the fault and confirmed that your bodge repair fixes it and makes the thing work the way you intended it then you have a new board made up with all the wrong shit fixed and populate that and you're done.

>>1485355
Oh and I'll add, you can breadboard smaller subsections of the circuit to make sure you got particular elements right or biasing right but doing the whole thing on a breadboard is madness.

If you got all the subsections working okay then the whole should come together if you put it all together right though that's not may not necessarily be the case, especially if you failed to take into account how one subsection loads another.

>>1485300
Try measuring the voltage across it while you put a current through it. Maybe it's not fully fused, the metal pieces that are left make contact sometimes but separate when they heat up?
anyway it should be replaced.

Is there any way for me to know how much heat an LED creates? Or do I just look at power draw, average heat creation and make myself an estimate how much heat energy that'll be? I want a decently bright led to glow for quite some time and not get too hot while doing so and have no idea on what data I need to look out to figure out what heatsink I need.

Are chink shit arduino starter packs sold on amazon any good for someone with previous experience? What's the best things to get to start out with arduino?

>>1485384
Without previous experience*

>>1485384
>
You don't buy chinkshit on Amazon. Look into eBay, AliExpress etc.
What do you mean are they good?
Good quality? Not really, but they are okay.
Good selection of parts? Rarely.
Good value? Almost certainly.

Regarding what to get:
I'd advise against kits unless it's for kids or similar. Maybe get some transistors, wire, resistors, sensors, potentiometers and LEDs. Then think about a project and get anything else you may need. Just starting woth a project is much better than starting with a kit.

>>1485384
>>1485386
To go into detail:

My first project was for example something for Christmas, a Arduino nano that notices movement via ultrasonic sensors and then curses at you ( pre-recorded audio files). It could also play songs etc. Strapped that to a small christmas tree as a gag gift.
Mainly built this for fun.

Second project read inputs ( potentiometers) and changes PC settings accordingly (and it worked digitally obviously)

Third was an alarm Glock that gives you mild electrical shocks to wake you. The possibilities are endless. I'd say don't start with some shitty "blink" app, as long as you know about basic electronics and how to divide projects and code into manageable chunks you'll learn quickly by yourself.

>>1485389
>alarm Glock

I imagine getting shot in the morning is a pretty shocking wake up call.

>>1485381
LED efficiency is never over ~50%, so just assume 100% is turned into heat as your safe estimate. Your LED datasheet should have something more concrete though, if you can find it.

how do i turn watts into square inches of heatsink aluminium? i have a bunch of old computer heatsinks and i want to calculate how many watts they can passively handle

>>1484185
>ever seen a buck converter that uses a transformer son?
>thats not how they work
this specific tripfag knows just enough to really fuck up whatever you're working on.

>>1484284
this anon is correct. AC is *intended* to be clock stable and very carefully managed to be so.
AC will be more stable over time than a quartz crystal.
>>1484329
>For a clock it's not only the frequency (which varies with the load) but the phase that counts.
talking out your ass and completely irrelevant to a clock.

>>1484551
>To be fair, ESD damage isn't usually a big issue with hobby shit.
Depends on the component. Most ICs have lots of protection, but still aren't bulletproof. A bunch of discrete components, eg 2N7000, are really fragile. They don't fail outright, they just start 'acting funny sometimes'....

>>1483995
I have a t420 thinkpad, it will not run on battery, but works on AC. I have bought it 2 new batteries, first one worked then it stopped, per the recommendation from here I bought a second one. Same issue arised, on linux and windows it says they are 100%, and I can get the stats, how many cycle it has gone through max power etc, how ever it will not run on the battery. How would i trouble shot to see where the short is?

>>1485413
>LED efficiency is never over ~50%, so just assume 100% is turned into heat as your safe estimate. Your LED datasheet should have something more concrete though, if you can find it.
Lighting LEDs these days are between 20% and 40% efficient. That said, this anon's rule of thumb can't go wrong.

>>1485565
a) if they're chinese noname retreads, don't fucking bother.
b) it's not a short, more likely an open. and god only knows unless it's OEM. If it's OEM, I might be able to help.

>>1485567
[more]
when you said 'it worked, then it stopped', do you mean one cycle succeeded then it wouldn't run on battery, or it worked for a few cycles then stopped?
If I had a nickel for every non-OEM battery that just 'stopped working' after two or three weeks, I'd have at least a case of beer.

>>1485554
Apply a known power load, measure temperature rise. Surface area isn't enough information for a sensible estimate of thermal performance.

>>1485569
>Apply a known power load, measure temperature rise. Surface area isn't enough information for a sensible estimate of thermal performance.
seconding. simple stuff like orientation and location of heatsink will change answers by an order of magnitude. Unless you're following a larger canned recipe, in which case you'd already have surface area spelled out.
Winging it is a harsh, empirical mistress.

>>1485567
I have bought a legit thinkpad battery,and a chink one they got the same issue, I wanna troubleshoot what is happening to the motherboard, where it can read all the autism stats of the battery but will not run on it.

>>1485571
OK, more info, good to know. And that's why I asked, because it sounded possible it could be either way...

So, if you'd blown a battery-related fuse it just owuldn't work ever. If you've damaged the battery power electronics (can happen) you're looking at some painful mosfet replacements.

>>1485575
you can actually debug this. If you're handy with a mutlimeter/scope and know how to read a schematic. The battery drive on the T-series is pretty simple.

That said, actually replacing mosfets is annoying because of the thermal nature of the motherboards. You need to be able to preheat the board and then use either an IR spot or hot air to get the damned things off (since they use the copper ground/power planes as heatsinks in operation).
It might not be a mosfet, but it's honestly most likely.

>>1485576
page 73 of https://people.xiph.org/~xiphmont/thinkpad/lenovo_thinkpad_t420s.pdf

Check the fuse F12 first, just in case it's that simple.

>>1485413
>>1485566
Thanks for the quick reply!

some stupid fuck was pressure washing a truck by an open workshop door and managed to cat this with the jet. No service manuals on the internet. None of the plasma cutter companies in my COUNTRTY want anything to do with it. The company who built the fucking thing folded years ago and well fuck.

It can be driven manually. But the the tracer isn't doing shit.
Any ideas or is this a great I dont want to do anyting today job?

I'm using a transistor to drive a relay that's delivering DC to a solenoid actuator. I have a diode across the relay coil to protect the transistor. Should I put one across the solenoid to protect the relay contacts from the solenoid?

>>1485646
Your relay will last longer and the solenoid will produce less interference if you do. It isn't strictly necessary, though, and usually people won't bother.

>>1485646
could also add a snubber capacitor

>>1485380
>anyway it should be replaced.
with a thick copper wire
fuses are for pussies

>>1484354
No longer an experiment, time error correction (TEC) has been quietly eliminated early last year by the Federal Energy Regulatory Commission.
Image: www.leapsecond.com/pages/mains/

>>1484504
>now do you understand why the word "problem" is used here?

No. I find it rather funny that your green text failed to use the term "problem idea" if it makes so much sense to you.

"hey man, I have a problem: my car will not start. do you have any problem ideas as to why it won't start?"

Do this: use the phrase "problem ideas" in a sentence.

>>1485678
jesus fuck are you guys insane?
is this what smart meters are about?

>join IEEE as a student member
>it's 90% indians
what the fuck

I'm a fucking idiot and reversed polarity to a 220uF 16V cap by using a reversed plug.
It started smoking slightly, so I quickly took the plug out. It still works.
On a scale from 1 to guaranteed housefire, how safe is it to keep using it?

>>1485806
What started smoking, the cap?

>>1485806
>On a scale from 1 to guaranteed housefire, how safe is it to keep using it?
You think it still works, but it's now out of spec and end of life. Replace it.

>>1485678
>(TEC) has been quietly eliminated early last year by the Federal Energy Regulatory Commission.
The REQUIREMENT FOR TEC AND FINE FOR NON-COMPLIANCE was eliminated.
Power companies still practice the corrections.

>>1485854
Yep, I was troubleshooting the board because the high voltage circuit wasn't working, I was out of ideas so I tried reversing the polarity.
Turns out the actual problem was that I didn't let the solder flow all the way under the big inductor, but now the 12V branch cap is shrekt.
>>1485871
Okay. I figured it would be fucked because the oxide layer already started burning.

>>1485928
>>1485871
>>1485806
Also, what are the chances I fucked something ELSE up while I was at it?

>>1485929
medium-low. Q2-L1-R34||R35 would have limited the reverse voltage on the 12V rail to about a volt or two, depending on current. D2-L2-C10(reversed) would have limited the reverse voltage on the 5V rail to under a volt. if you replace C2 you should be sort of all right for as long as it lasts

>>1484573
>>1484567
>>1484566
I didn't have time to work on it yet, but I will follow your advices and make a few modifications tomorrow.
I have a few more noob questions:
Does it make a difference if I use ceramic or electrolytic capacitators?
It's safer to use chip sockets because you can fuck up a chip with the heat, right?
I will build another oscillator soon, how can I blend the two oscillator signals like in a regular synth? (don't care about the tuning for now) Will it work with my small mixing console or do they use a different system in synthetizers?

If i have arduino connected with a wall adapter and also have a battery, how do i make it so that if the adapter power cuts of the arduino gets switched over to the battery fast enough so it doesn't reboot?
Using relay for this is too slow, and using something like a mosfet will leak a bit of current and slowly drain the battery

>>1485577
How do i learn to read this.,..

>>1486011
>blend the two oscillator signals
A mixer might be what you're looking for (else just a summing amplifier topology), I've personally got a bunch of SA602 mixer ICs on my ali wish list for messing about with both audio and radio frequencies, they're cheap (along with the half-dozen similar ICs) and look to be pretty reliable.

>>1486012
A good enough FET circuit will not drain the battery significantly compared to the battery size. 10µA is easily possible, which should be below the q-current of the arduino. But you could simply have the wall power at X voltage and the battery charged to X-y (but still within the tolerance of the device) with appropriate diodes so only when the voltage of the AC adapter drops by y volts will the battery start conducting through the diode into the arduino.

To switch fast it's a better idea to sense voltage before a diode which feeds a large capacitor (significantly larger than the PSU's own output capacitor) so that the capacitor buffers the supply in the few µs it takes to power on the transistor. The switching FET should go between the diode and the big capacitor.

>>1486011
some ceramic caps (esp larger ones over about 1nF) can couple vibration and shock into the circuit as capacitance changes, much like a condenser mic. if you plan on using your device in front of a wall of speakers or slamming it around a lot, consider this
meh, if you're new to soldering you might consider sockets for that reason, but with a good iron and good technique, it's a non-issue. another benefit of socketing is the ability to test out the rest of the circuit, check voltages and continuity etc. before you install a possibly much more expensive chip. machined pins >> cantilever spins
most commonly, just sum them together (see your favorite op-amp cookbook, try http://www.nutsvolts.com/magazine/article/op-amp-cookbook and skip down toward Applications Roundup if in a hurry or know this stuff already), but it depends on the exact sonic effect you're trying to achieve. the small mixing console should be fine to get you started

>>1486035
can I use this to make a microphone

>>1486053
probably not but we're not gonna stop you from trying it

>>1486054
maybe a trimmer cap would work well

>>1485890
Why not sell the control power required for the TEC? Shareholders would stone you if you don't.

>>1484025
Tie all the negative pins together. All ground must be the smae

>>1484114
Sure it can. If you know the resistivity of a piece of wire, you can run it between two points and measure the resistance to find the distance.

>>1484186
It depends. If you connect directly to the battery, it will always be on, regardless of whether the car is running/ignition on. That's great if you like to wake up to a dead battery. I've had good luck running a control signal from the fuse box to switch a larger current from the battery, so that you can control when the device gets power. The infrastructure is there--use it.

>>1486028
can you provide a simple circuit picture?

>>1484414
>Countersinking a nail

>>1485609
Pack it in rice.

>>1486150
>he doesn't follow ASME 19.72E to prepare the work piece for nail installation

>>1486147
Actually I don't even know if this will work, I think it needs a resistor in parallel with the first cap to drain it below the voltage across the second cap.

Design two is shit, but if you throw a diode in series with the battery I guess it could work without fucking up your day, if the battery voltage happens to be near enough to the voltage wanted by the linear regs on the ardy.

But more importantly, both designs completely neglect the concepts of battery protection circuitry, which are necessary if you use lithium ion cells. Both Lithium ion and NiMH require somewhat specialised charge circuitry (though the latter can be trickle-charged without much care). And since the existing DC wall supply is giving you a voltage that isn't a rounded battery voltage you'll probably have to have 1 or 2 power supplies in there for charging and discharging.

>>1486268
oh shit I think the FET is facing the wrong way

Has anyone here ever used scikit-rf? I'm new to RF engineering and just looking at some tools I can use in linux and came across this. However, it doesn't seem many people use it. I'm a sucker for using the most minimal solution.

If anyone with some experience can look over this and see how this stands out. Looking at the docs it's able to analyize RF networks, can plot smitch charts and can calibrate from vector networks

>>1486303
Just to add, only asking because of the lack of talk around scikit-rf. Don't want to waste my time learning this library if it won't surfice on the field/

>>1486303
>python
>minimal
eh

>>1486307
Well in comparasion of using programs that probably utilize a GUI

>>1486308
You're probably right actually. Only one I've even seen is GNU-radio, which if I remember right was programmed by dragging and dropping boxes. What kind of RF hardware are you working with anyhow?

>>1486310
Right now, just a rtl-sdr
Although gnuradio is a radio tool, it serves a difference purpose. Don't think skrf can replace it.

Can I use my atx power suply's 5v output to charge my phone?

>>1486316
most phone chargers are rated at 5V and above 1A. in theory i suppose you could, but chargers are so cheap and the peace of mind not damaging a $500+ phone makes it ill advisable

>>1486322
the phones dont actually charge at 5v.
the internal battery charging circuit bucks it down.

get a molex connector and solder a usb wire to it. make sure to get the 5v, not the 12v lead.

I'm sure it would work fine, but why?

>>1486012
what's wrong with two diodes?

>>1486310
kek. you drag and drop boxes, wire them together, and grc generates a Python script from them

>>1486316
why not, your computer does
you might have to connect some resistors to the D+ and D- lines to get the phone to draw max charging current
it's kind of overkill tho

So how will the new postal rules affect us?
https://www.washingtonpost.com/amphtml/business/economy/trump-ditches-144-year-old-postal-pact-that-boosts-chinese-retailers/2018/10/17/88aefea6-d234-11e8-8c22-fa2ef74bd6d6_story.html?noredirect=on

I have a few very basic questions regarding pic related.
Besides the ground and the output what are the other connections to this chip? None?
In the guide I use, there are 3 resistors in the component list, but I only see one in the schematics. Where should I put the other two? In parralel before the output?

>>1486390

the schematic and PCB definitely dont match. there's an obvious disparity in resistors, diodes, and the placement of the switch. in such a case, the PCB becomes the definitive reference.

>what are the other connections to this chip? None?

all the pins are used. what do you think those numbers on the gates mean? pins 1,2,3 for example are used as the leftmost gate.

>>1486390
There should (maybe) be a resistor between the switch and ground (where JC1P is written), which you'd think would be that second small one in the pic, but the trace doesn't really follow. Unless JC1P is leading to the IC's +ve and -ve pins, which aren't otherwise shown on the diagram. Either way the button needs a pulldown. The large resistor looks like 100Ω, which is uncharacteristically low for anything in the circuit. Anyhow the large amount of diodes is more of a mystery, and I don't think they're doing diode logic, so my guess is they're different circuits.

What is it supposed to do anyhow? The first gate looks like a sort of debouncer,

>>1486406
>>1486406
Thanks a lot for your guidance.

>what do you think those numbers on the gates mean?
It's the first time I have to read that ind of schematics. It will be a miracle if I can get it right...

>What is it supposed to do anyhow?
It's just some kind of oscillator.

>>1486406
>Unless JC1P is leading to the IC's +ve and -ve pins
they are. the button gates Vdd and the top input is tied high so that IC1A's bottom input is no longer a don't-care

Hey I just realised I can just hook up the two standard rotary encoder outputs (with appropriate pullups and pulldowns and debouncing) to the two inputs of a X9C104 digital pot. Since it has an up/down input and a hopefully edge triggered clock input, it should count up and down alongside the movement of the encoder. I could also send one pin to an oscillator before it goes to the CLK input to get multiple pulses each increment since it is a 100-step pot after all.

>>1486430
Well, sounds doable, but that way of decoding encoder pulses sucks shit.

>>1486406
>The first gate looks like a sort of debouncer,
It's a chain of gated oscillators.

how precise are hall effect sensors? i got one for my arduino and i wonder how precisely can it measure current going through a wire

>>1486443
>that way of decoding encoder pulses sucks shit
Well since I'm using the encoder because I can't get thin enough dual gang pots for this purpose (or ones with a button on them) I'd say it's a pretty elegant solution.

>>1486448
Hall sensors are used in DC amp clamps, not that those are terribly accurate. In a shielded enclosure with the wire a known distance from the sensor I'd imagine you could easily get to within 1% accuracy.

>>1486443
>>1486453
Oh and I plan on doing the same thing to a 4-bit up/down counter also, which will control an analog switch, which will switch between 4 audio channels. For edgy headphones covered in knobs and antennae.

>>1486453
1% is more than enough, i only want to use the hall sensor to make a simple customizable breaker where i can set a max current and if it is more than that it turns off

>>1486453
wait.. but clamp meter has to be wrapped around the wire to work
Does that mean that placing this thing in front of the wire won't work?

>>1486458
That's true with an AC clamp meter since they use a current transformer which is typically toroidal, and in general the clamp keeps the wire well inside the sensor range. I think DC clamp meters have 3-5 hall sensors around the inside of the clamp hoop, which sort of necessitates their shape. But there are DC clamp meters with just two plastic prongs instead of a hoop. Make sure you're aware of the direction of the magnetic field from a wire when you set up the sensor (I'd advise using silicone/epoxy/hot glue to keep them in the same relative position), since it's neither radial nor axial. I'm not sure which direction those sensors are good for.

Also take a look at how RCDs use what I think is a current transformer but with both the hot and neutral lines through, so it only picks up on the current difference, you could probably do something similar with a hall sensor. Putting your wire right next to the hall sensor may cause the ADC to hit its rails from too strong a magnetic field, in which case you'd either need to move the wire or change the passives' values. On the other hand, to get more sensitivity if your maximum current is a few factors of 2 below the ADC's maximum, you could change the passives' values to get more gain.

I'd look at existing hall-sensor current sensors (and their peripheral analogue circuitry) for inspiration. Also isn't that 393 a comparator not an op-amp, so isn't that just an on or off switch? Or is it just for the LEDs?

>>1486465
Well i need to measure DC so i will try if it will just werk as is. It has an analog output pin which should vary the output voltage based on the magnetic field and more current = stronger field

>>1486467
>It has an analog output pin
In that case there shouldn't be a problem. If you have a datasheet for that sensor you could throw a few numbers into some equation you can probably find on hyperphysics (something along the lines of B = I/(4πµ_0r^2) ) to see how the current will be measured by your ADC and if you need to attenuate or amplify it.

Say I've got some logic ICs and I want to preserve the memory on one of them (a series of latches) so I keep it powered constantly. Will it damage the logic after this if I cut power to them while leaving the output pins to them high? Or should I put a diode and resistor to shunt any power to the grounded Vcc rail? I see things like "max input voltage = Vcc+0.5V" but that doesn't usually involve Vcc = 0.

Looking for an IC with an output enable is a sensible alternative, so I'll see if I can do that. Putting 8 transistors or diodes to pull the lot to 0V would be a pain and this is meant to be a somewhat compact build.

Is DT9208A a good multimeter? Specs seem OK. I'm also looking for a soldering (soldering & hotair if combos are worth it) station that is a bit cheaper than Hakko. 100USD is way too much.

>>1486491
no idea about the meter, but
https://www.sra-solder.com/aoyue-888a-2-in-1-digital-hot-air-rework-and-soldering-station
has been good to me

>>1486491
>not automatic range selection
>no clamp for current measurement
it's garbage

>>1486498
Rec something within 50USD.

>>1486406
>>1486402
Eventually, I found an accurate PCB for this project.
I finished building it on the breadboard and it's working, even though I only did one connection on the jack. What are these numbers (1 to 5) supposed to mean?

>>1486570
terminal numbers. count and configuration vary widely by jack type and model
as long as you have signal to the tip and signal ground to the sleeve, you're probably in good shape

>>1486330
I want to put some banana sockets on it in order to have a 12v output, a 5v output and a USB port to charge my phone.

>>1486498
>autorange
>good

>>1486570
>accurate PCB
not kosher

Has anyone ever worked with Circuit Scribe? My boss wants me to do something cool with it for this kids open day thing we're doing but I have no idea what I can do with this stuff, I don't even know how to think about it.

>>1486477

if it's TTL, you could easily kill a chip doing that. CMOS will easily survive, but it's not a good idea coz there will be internal diodes diverting the input power to where-ever. but putting 10K resistors in series with each input will limit the current these internal diodes have to handle to a very safe value.

Anyone here uses esp32? Is it 5v tolerant?
For example esp01 says in the spec sheet it is 3.3V only and 5V will fry it, yet i've had zero problems running it directly on 5V
But esp32 is too expensive for me to just try it and see

What's /diy/'s favorite book for exercises on transistors?

>>1486736
That's exactly the kind of answer I was hoping for, thanks! Though I'm still not set on whether I'll turn off all the ICs at once or leave them all on. Since they're all CMOS there really won't be that much power drain if they are all left on.

>>1486756
My favorite book is getting actual transistors and a breadboard and building the circuits then measuring and optimizing them. And when I'm too lazy to physically build I just do it in a simulator instead. Doing exercises is a waste of time.

>>1486756
It's a good idea to simply browse through as many transistor circuit topologies as you can: if you don't understand it at first, build it IRL or in a sim and take measurements until it makes sense. Eventually you'll reach a point where there are only so many common uses for transistors and any larger circuit will just appear as a combination of these mini circuits. I'm nowhere near that point myself, though I am closer when it comes to op-amp and digital logic circuits. That website that comes up when you google how to make (some variety of) a 555 oscillator looks like it has a bunch of circuits in it that you could probably learn well from.

>>1486942
how is that a book?

>>1486658
What's wrong with the PCB?

Yesterday, I tested it with different component, I'm satisfied about it, so I'll solder it to a dotboard today.
On the first PCB I posted, they used extra diodes, what are their purposes?

>>1486950
A trace looks to be disappearing into the ground plane and then back out of it again, following it to the right shows many places it could hit ground without going so far if it was intentional.

>>1486745
5V tolerance refers to signals, not power. your ESP8266 module may have a regulator on it to feed the SoC the 3.3V it wants for power, but still be damaged by 5V on logic inputs
neither Espressif SoC is rated as 5V tolerant and neither will accept more than 3.6V on its +ve pwoer inputs
anyway why are you still using 5V logic in 2018

>>1486950
can't tell you without seeing how they're connected, i.e. the back side of the board
maybe the designer put them there to protect against ESD, who knows

>>1486756
if you can't read the OP and see the book recs there, what makes anyone think you'll read a whole book

>>1483995
which t12 from ali is the best quality

>>1486954
>why are you still using 5V logic in 2018
Not him but 7400 series is ubiquitous on ali.

>>1486958
why are you still using 5V 7400 logic in 2018 instead of 74HC, 74VHC, 74AUC, etc.

>>1486961
Oh hey HCs can go below 5V I didn't know that. But my digital pots are still limiting me to 5V.

>>1486950
>On the first PCB I posted, they used extra diodes, what are their purposes?

since they're next to the caps, they're probably there to discharge the caps when the power is turned off, so they dont discharge into the chip instead, and maybe blow a gate. rare but possible.

>>1486967
Aren't resistors usually used for this? It's not like the power rail is pulled to ground when it's turned off allowing you to have a reverse polarity diode to Vcc or anything.

>>1486968
no, diodes are usually used for that. there are ESD diodes inside the IC but they aren't meant to handle a whole lot of current, such as when the first pot is set very low and the first gate is charging the big cap. the discrete diodes pass most of the current onto the declining Vcc rail and ensure the inputs aren't very far outside the rails, without overstressing the chip's protection diodes

>>1486967
You are probably right. Each time I turn the circuit, I hear a funny sweep. How should I install the diode? In series right before the caps?

>>1486979

How do I know how long a UPS battery will last?
It's 12V 7Ah Lead-acid battery, that invert to 220V AC, and the draw is about 0.6A.

>>1486990
>7Ah 12V
That's 12*7 = 84Wh
>0.6A 220V
That's 132W
84/132 = 0.64h = 38min
Not that long to be honest, and I'd cut ~20% off that to be safe.

>>1486991
I calculated it as such, but when I tested it, it lasted over an hour.

>>1486954
>why are you still using 5V logic in 2018
because arduinos use it and i use them heavily

>>1486993
Those batteries are probably a fair bit over 12V, 14 at best, but that still doesn't account for everything. Is the load a constant 0.6A? Might the battery be underrated for "expected lifetime decay"?

>>1486996
>Is the load a constant 0.6A
Yes, The UPS is powering a router.
>Might the battery be underrated for "expected lifetime decay"?
It's new battery.

what is the easiest way to limit current for an unknown load? I basically have a generic DC powersource and want to make sure that it won't supply more than say 500ma so that if i for example short the + and - wires together only 500ma will flow through them.
I do NOT wan't a constant current power source, i simply want it to be an ordinary power source on which i want to limit the max current available

>>1486998
I mean it might have more than the advertised capacity in order to account for expected loss of capacity over time.

But 132W sounds way too high for a router lest it has a heat sink on it, that's probably the current out of its PSU, not into it.

>>1487002
>the current out of its PSU, not into it
Nope, the input state 0.6A 220V.
I guess it's something to do between AC and DC conversion.
Any idea?

a computer psu needs a small load to turn it self on right? but mine works even without any load attached. i do have the 120mm fan connected, so that is enough then i assume?

>>1487001
>generic dc powersource
what voltage?
the very easiest way would be to find a USB power switch IC with current limit built in, if you're current-limiting a 5V supply
the next easiest way would be to place a current source such as Pic related on the output of the supply. if there is less than the current limit being drawn, the current source will simply turn Q1 on hard and pass as much current as the load desires. when there is more than the current limit being called for, the op amp will start adding resistance via the MOSFET to limit the current draw to the set current limit

>>1487013
older ones, yes, but newer ones seem to be stable with less of a load
a fan is usually enough load for even older ones to run steadily

>>1487004
it's hard to measure SMPS input power. the 0.6A measurement must be way high
does the UPS not have diagnostics to show load power consumption and estimated run time?

>>1487016
>does the UPS not have diagnostics to show load power consumption and estimated run time?
No.
But I doubt a router uses more than a light bulb, So let's assume 20 watt?
But if I use that nath from >>1486991
I get over 4 hours, and in my testing it lasted just short of two hours.

My ghetto-lab bench power supply is almost finished!
All i need to do now is repair the spray paint to fix the scratches and add labels.
It outputs direct and clean 3.3V, 5V and 12V and then fully regulated 1V - 34V that can be easily set with thew knob
The blue number shows the total current drawn
I am really happy with it. It saved me a ton of money since i got the PSU for free and it's even better than regular lab bench supplies, because they usually have only one output, where this thing outputs 4 different voltages at the same time and can easily handle like 5 times more current that a typical lab bench PS

>>1487018
consider that the actual power capacity of a lead-acid battery (and most batteries, really) depends on how quickly it is drawn down, and that the nameplate rating usually assumes a 20h discharge time to 1.8V/cell
since they're not really designed for efficiency, the UPS probably has a sizable fixed component to its power consumption, relative to the tiny load from the router
the empirical method is the most reliable way of determining the run time under any given load

>>1487020
berry gud
BUT... you don't necessarily want maximum current on a bench supply, especially when experimenting with expensive components that you don't want to let the magic smoke out of too easily. see current limiter in
>>1487016

>>1487025
>. you don't necessarily want maximum current on a bench supply,
already planned for that. i am going to make a simple customizable fuse with a hall effect monitor where it will be isolated from the powered device and if it detects over current it shuts down a relay, turning off the device, the hall sensor will probably be plugged into an arduino mini so i will have lots of options to set the current limiting.
maybe later i will buy ssr instead of a relay because it is much faster

>>1487020
Should have made it a dual supply so you could do negative voltages.

>>1487030
i don't need those anyway

>>1487031
You probably will at some point unless this PSU is the last electronics project you ever plan on building.

>>1486950
This faulty layout was published at https://andreassiagian.wordpress.com/2013/02/22/quad-oscillator-using-4093-tutorial/ and as depicted the circuit has never worked because pins 5 and 10 are grounded. Maybe the Eagle files don't have this error. C5, R5 and LED1 do not make sense either.
If you want to build the circuit better look at the original at http://www.fluxmonkey.com/electronoize/4093_Oscillator.htm
More stuff at http://www.electro-music.com/forum/index.php?f=160

>>1487033
if i ever need it all i have to do is drill one extra hole under the voltage knob and add one more output terminal, because there is a blue wire inside of the psu which supplies -12V and i didn't cut it in case i needed it later

>>1487037
That's handy. Negative variable would be better but a fixed -12V rail is probably good enough for most applications.

>>1486984
The diodes to Vcc in the GetLoFi kit are four 1N5230 4.7V Z-diodes. No idea why that is.

>>1487041
If your "most applications" are build around the stuff from the seventies, then maybe.
For anything made on this millennium -12V is quite a lot.

>>1487051
I meant that it's more than enough. Like you're designing an amp that will run off +/-5 V when testing it you can can use your PSU's +5V rail and you can use the -12V to at least give you a negative rail, even though it's more than you need. That way you don't need an inverting SMPS or a virtual ground divider.

>>1487028
Ah so you're autofuse anon. I don't see the issue with using a normal current shunt, but go for it. Make sure you latch the thing so it doesn't just linearly regulate current via negative feedback and make a fuckton of heat.

>>1487004
It could have a really shitty power factor.

>>1487028
I wouldn't digitize overcurrent protection. you don't need isolation. high-side MOSFET switches are easy enough to use

>>1487051
other than analog synth projects, which generally like ±12-15V, anything digital made in this millennium is probably fine with a single 3v3 supply

>>1487049
hmm, still doesn't look like exactly the same board in each photo. the 47µF cap and the 1k/100 ohm resistor leads should be exactly in line but they're not
anyway, assuming the - leg of the 47µF is lifted and connected to a pot for output level control, and there's nothing going on underneath the socket, it's a different circuit entirely, with four apparently independent oscillators feeding a common output thru the zeners. probably some circuit bender jamming parts into places until he finds something fun

>>1487028
> if it detects over current it shuts down a relay, turning off the device
That's going to be slow. It won't be fast enough to protect whatever's connected to the PSU, it may not even be fast enough to protect the PSU itself.

If you need strong over-current protection for the load, you're stuck with a linear regulator. Anything with a filter capacitor on the output will deliver a current equal to the output voltage divided by the capacitor's ESR.

If you just need to protect the PSU, the simple solution is to build the current limit into the regulator's control loop.

>>1487063
plus a [current shunt > comparator > SR latch > FET] setup would do the job pretty well and probably faster

How to draw power from two usb outlets for the same eg. lamp?
My 5v->12v step up males can handle 0,65A and I want to run a lamp drawing 1A. I'm thinking current limiters will go half way but it's small comfort if I can't draw power from several outlets in parallel. Pic is using a weak lamp but even using the stronger lamp all power comes from a random of the outlets.

>>1487066
you generally can't parallel voltage regulators. each of them is going to have its own slightly different output setpoint (due to process variations etc.) and the ones with the higher setpoints will carry most of the load
if your 5→12V converter will also convert 10→12V, try connecting the power banks in series

>>1487064
adding, make sure the comparator's common mode range allows the voltages in question
>SR
y when you can get DFFs in SOT23-6 packages

Most stuff I build is analog so I use +/- 12 and 15V rails frequently. Most digital stuff I do is with LS TTL or that 15V CMOS and I usually use 5V if I'm mixing CMOS and TTL or I just use my analog supply voltage if all the digital stuff is CMOS. I pretty much never use 3.3V. It's inconvenient. And I avoid microcontrollers because I'm shit at programming and I find it tedious. I have one PIC micro sitting in a drawer somewhere, it's literally the only chip I have that requires 3.3V.

>>1487063
>It won't be fast enough to protect whatever's connected to the PSU,
dunno about that, every time i connected something wrong it went
>smelly
>smoky
>glowing
>broken
in that order and it took at minimum 5 seconds to do that. the fuse will engage in like .1s

>>1487060
Most stuff I build is analog so I use +/- 12 and 15V rails frequently. Most digital stuff I do is with LS TTL or that 15V CMOS and I usually use 5V if I'm mixing CMOS and TTL or I just use my analog supply voltage if all the digital stuff is CMOS. I pretty much never use 3.3V. It's inconvenient. And I avoid microcontrollers because I'm shit at programming and I find it tedious. I have one PIC micro sitting in a drawer somewhere, it's literally the only chip I have that requires 3.3V.

>>1487057
>>1487063
Well the problem is that i want the fuse to be easy to build, but also i want to have the ability to simply set the current limit with a simple knob.

I tried to design a fuse which doesn't use a micro-controller but even the simplest solution was way too complicated.
Pic related is a shunt based fuse. It uses small resistor to prevent heating and thus has to use one op amp to amplify the voltage and another one to trigger a mosfet if limit current is reached, and the mosfest (bottom right) will then either turn off a relay or just directly shut down the circuit (and also the fuse it self to prevent infinite looping).
It's just so overcomplicated for such a simple function. I like to make things as simple as possible. And if i use arduino i can make the fuse with literally just 3 plug and play components: hall, arduino, and a relay/fet/ssr. Don't even have to solder shit.

>>1484025
What sort of arduino? The uno can take a 12v supply

>>1487093
never trust what the sheet says.
i once tried to use arduino mini with 12v on the raw pin, which according to the sheet it was supposed to handle without any problem and it burned within seconds

>>1487073
fair enough, you analog guys get a pass

>>1487086
the amp stage is unnecessary. you could use a forward-biased 1N4148 diode as a shunt regulator. supply it with ~1mA to get ~650mV, then place the 100k adjust pot across the diode, for a 0-6.5A range (approx)

I tried my hand at building a high gain discrete transistor amp with class G output stage but I can't seem to get it to work right and I just have no idea where to start with fixing this thing. Anything obviously wrong here that I'm missing?

hello, do phase shift oscilators and other kinds of oscilators generate the half negative part of a sine wave even on a single supply rail? I`m trying to get started in electronics and I have no waveform generator or scope (due to gomunism they are expensive and second hand ones are hard to come by because people don`t mess arround with electronics here so much)

hello, do phase shift oscilators and other kinds of oscilators generate the half negative part of a sine wave even on a single supply rail? Or it clips/offsets it? I`m trying to get started in electronics and I have no waveform generator or scope (due to gomunism they are expensive and second hand ones are hard to come by because people don`t mess arround with electronics here so much)

>tfw you finish soldering the cable and it is fucking flawless
BUT YOU FORGOT TO PUT THE FUCKING SHRINKING TUBE ON

>>1487139
Hot glue. Even better than shrinktube, because it forms a solid joint to the board without tension concentrated.

>>1487138
It generates the negative half cycle by DC biasing the amplifier halfway between ground and the supply voltage. You see in this example the transistor's collector is has a DC bias of 7.5V which is 1/2 of 15V. The output is capacitvely coupled using a capacitor chosen to have a low impedance at the oscillator frequency. As you know or will learn capacitors do not pass DC and so this allows the AC component of the signal while blocking the DC and you get a positive and negative swing.

In the image yellow trace is output after cap and blue is output before cap. You can see blue has a DC offset and yellow does not.

>>1487144
Thanks my man. Oscilators, criteria for oscilation and similar things are the areas I`m weak at. The guy that invented truly was a wizard.

>>1487134
your input stage looks really funky, not sure what the Q10/Q11/Q5/Q6 arrangement is supposed to do
anyway try opening the loop and inputting a very small signal, and seeing where shit looks way too fucked up. also plot the feedback node voltage on the scope, which should of course equal 6V ± an extremely tiny signal
Bob Widlar discovered the principle that it was unwise to try to replicate macro circuits on the micro scale. I'm sure the reciprocal is also true

>>1487151
Thing is disconnecting the class G and taking the feedback from the output of the buffer right before it gives me something more normal though it's a bit lower than the gain of 9.1 that I should have. At least it's a mostly distortion free sine wave. So the voltage amplifier stage seems to work okay. I don't know if the power amp is loading it down or what.

>>1487134
are D1/D7/D2/D8 backwards?

I want to become an electrician, what books should I buy to help me on this path, or online resources? I would prefer a book.

Also, if I study my shit on my own and am smart enough to retain it all and expand upon it past a normal guy, do you think a company would hire me without school accrediting they thought I really knew my shit?

>>1487166
No they are supposed to be that way. I'm not sure why but it's different than a class AB amplifier. I'm not using Zeners cause my supply voltage is a lot lower than +/-70V. I've already tried it with one diode instead of two as well.

>>1487168
Also, is there a specific branch of electrician I should be aiming for if I don't care how hard the work is or complicated if I want good job security and a lot of potential pay?

>>1487166
okay, but the voltage drop of those 1N4148s in reverse is going to be what, a couple hundred volts?
turn 'em around. they're the reason your top driver isn't getting dick

>>1487171
industrial/commercial electricians rarely run out of work
this isn't really a sparky thread doe, maybe post in /qtddtot/ or make a new thread in /diy/

>>1487180
>okay, but the voltage drop of those 1N4148s in reverse is going to be what, a couple hundred volts?

Okay, we'll it's definitely a different results but I'm not sure I'd say it's better...

>>1487183
hmm, well, those zeners are supposed to be greater than the b-e drops of the driver and pass transistors. maybe add one or two or three more diodes in series
I would also suggest taking R18 up to maybe 10k, since R16 sort of does the same thing. if that doesn't work you might need to replace Q24/Q25 with a class (A)B configuration instead of a class A

>>1487192
Taking R18 up to 10k basically makes my output zero. There's a tiny nub that's a fraction of a mV that you can barely see on the scope on the 2mV range but that's it. Adding more diodes and even replacing the diodes with 3.3V Zener's made no significant difference. Maybe I'm just gonna have to do more research on how to design these amps. There isn't a ton of reference material since most power stages use Class AB because it still gets good efficiency while being less complex.

I wanted to do a class H amplifier which is even more complicated and involves varying the power supply such that it is always just slightly above your signal voltage which greatly minimizes transistor dissipation. I guess they use a bootstrap capacitor of some sort to hold the supply voltage just above the signal voltage for a short time and that's sort of how they get the rail tracking. It's too complex though. I can't even get class G to work yet.

>>1487198
another thing you should be checking is current through your pass transistors, in case your bias is just way the hell off
what happens with R17 and R16 10k and R18 back to 1k? or when you replace R17 and R16 with actual constant current sources?
I also notice your inner-rails are at 1/3 and 2/3 instead of the original's 1/4 and 3/4, in case that matters
not sure this class G business is going to be more efficient at 12V anyway, especially if you have to generate your inner rails

If I want an LED that is bright enough for a reading light (one on the book, not in the room, pic related) (and doesn't take too much power) anyone know what power I should look at? I struggle with a feeling for how bright any LED is going to be/ how much and how well it will illuminate somewhere. Any help is appreciated.

>>1487298
Anecdotally, I've used red LEDs at about 10mA and dead on, from around three feet away, they're too bright to really look at. I don't think you need to look for particularly high power for this.

>>1487298
on the order of 1/4W
considering the average by-the-bucket white LED is about 1/13W, get 2 or 3 or 4, maybe a constant-current step-up converter if you feel like saving an extra bit of battery life
you will probably want flat-top round LEDs for flood coverage, the dome-top ones are better for spot coverage

>>1487304
Was a bit surprised by that, but I just hooked some random RGB LED I had lying around up and it's plenty bright. Thanks, I'd have gone overkill for sure.
>>1487311
Thanks a lot for the information, I considered a diffuser for a dome LED, didn't think about flat LEDs. Thanks for the quick and competent answer!

>>1487311
I'm wondering though, why does a constant-current step-up converter save battery life? Isn't the power draw all that matters and all a step-up converter does is changing the voltage while reducing the current so power stays constant?

>>1487322
because you're not burning energy in dropper resistors. now if you're using e.g. button cells with a significant internal impedance, you might not need dropper resistors at all and can save yourself the effort
what a step-up converter does depends a lot on how you hook it up. in the usual configuration, the feedback voltage comes from a voltage divider connected to the output. however, one of the tricks you can play on four-terminal regulators in general is to connect the - leg of the LED string to the feedback input, and connect a resistor from there to ground, whose value is chosen to drop the regulator's specified feedback voltage at the desired string current (R = E/I). the regulator will work harder to increase the output current until it sees its specified feedback voltage on the feedback pin

>>1487086
My circuit used a single op-amp and two NPN transistors as the latch, maybe with another NPN or two in there.

>>1487330
Thanks for the clarification

>>1487086
as for the latch, consider Pic related

Might anyone recognise the name of this tip type? I'd like to get a chisel replacement for finer work. The company I got this iron from (Jaycar TS-1540) doesn't sell different tips for it, let alone specify its fitting type. Best I can get from it is "collar" and/or "locking screw". I found a similar-looking tip on amazon that says it's for a Weller, though I can't find anything in a proper reseller's catalogue. "Wall Lenk soldering tip" shows me something more promising, but there's little to no presence of this online. If it isn't a common tip then I'll just bite the bullet and get a T-12 meme station I guess.

>>1487595
fug

Hello /ohm/
I've been doing some designing of a circuit and been having difficulty over something I feel should be simple.

I'm designing a circuit from nothing in logisim to control multiple 7 segments displays. I'd rather not use 7 wires for each display and would like to find a way to control multiple. Google hasnt been helpful since I'm not buying an arduino or external parts to program when I'd rather simulate a circuit I've designed.

>>1485219
I just bought some inline dimmers for LED strips from China for ~$1 each and soldered bigass strips of copper tape to their capacitive switch pads

>>1487615
not sure but i think you want the magic of multiplexed displays. google it. see top of pic.

>>1487596
see bottom of pic. there's also tutorials on making your own cheap tips using #6 copper wire. google that too.

>>1487018
consider losses in the inverter too. A typical consumer UPS design is optimised for being as cheap as possible with the most possible power output for ~2-5 minutes, efficiency for running hours long with low load isn't a main concern.

>>1487648
Multiplexers have multiple inputs and a single output where the inputs used are decided by an OP code.

From what I got from your pic I see a counter connected to RAM. The ram is then deciding the input for the multiple displays out of one Segment control display.
I've already made a Segment Display Control.
It only counts up to 9 and gets a little screwy after that. I've always planned for multiple segment displays.
However, I still fail to see how a multiplexer could help in this case. Wouldn't a demultiplexer or a decoder work instead?

Also, as I've been looking online most of the segment displays shown have a K input. I'm staying really simple (relatively) and using one with only 7 pins 1 for a decimal point

>>1487648
It's threaded on the end, did you see my image? Unless you want me to tap some copper wire, which I don't think would work very well.

>>1487664
>It's threaded on the end, did you see my image

yeah, i saw the thread, and i also saw that there's a dent from a retaining screw. which means the thread is not really needed, coz the screw can hold it tight.

and you can find pre-threaded tips. it was a very common thing on Radio Shack irons, and other super-low quality irons. that includes the red-handle Wellers they sell at Home Depot. (model SP23 for example).

as for tapping copper, it's very doable.

Redpill me on getting charge/discharge boards (seperate or 2-in-1) for non-standard li-ion sizes. Small lipos in particular.
I'm seeing TP4056 boards that only do charging but look to be made for 18650s without much room for current adjustment, I'm seeing 134N3P(?) modules with USB in and out (which is what I want) but it's impossible to find documentation for them and it looks like they're missing almost all their passives (for better or worse), and then there are some discharge boards with 5V out that are marketed as 18650 UPS modules, with both 8-pin and 16-pin ICs.

>>1487624
i did that too but it misfired because the ic wasn't calibrated for such large pad

>>1487680
I'd need to file down the area to go into the tapped bit by quite a bit, since the internal diameter of the threads is ~2.5mm. And I just tightened both screws in and could unscrew the tip just fine, so it seems it's just being held in by the threads alone. Or at least it is now; it may be the product of a year and a half of thermal expansion. I think the screws are mostly for keeping the two parts of the heater together, since they're not doing anything else any more. One screw is also mysteriously shorter than the other.

I'm overdue for a digital soldering station in any case.

>>1487690
The classic: fridge switches off - some shit triggers.

>>1487660
the multiplexer is at the bottom. RAM is just for demonstration purposes. if you switch the common lead of only one display on while presenting the desired pattern to the segment leads, only that display will show anything. the idea is, if you rotate between them really quickly (several tens of times per second), persistence of vision makes it look as if they were all on at once

>>1487698
>overdue for a digital soldering station
yeah it's time. good tips on shitty irons are kind of an insult and won't work out as well as you'd hope anyway

anyone remembers the faglord making a water conductivity sensor using two inductors? So I got it working, I`m still having problems with the salt water shorting things out now and then because the prototype is just a 20mm pvc pipe with a buttload of hot glue but using a 1,5kHz oscilator I manage to get a signal through. It`s kinda hard because I have no function generator or scope so to see stuff I either use the multimeter AC measurements (which aren`t very reliable) or amplify and listen the signals using a speaker. In water with a 1.2-ish Volts AC signal I get a 1.4-ish V output on the secondary coil. As I add salt that increases. I`ve been testing with 29 AWG wire and 40:80 relation. I`ll make some other sensors with 22 AWG wire and other sizes when my package arrives. Thanks for the guy helping me with oscilators.>>1487144

>>1487722
Also, fucking sadistic EE teachers sucking your time and soul and not leaving time available for cool stuff. And I`m really regretting dumping a shorted 60W fan motor away.. All that sweet wire...

>>1487724
>>1487722
The theory is that the changing field int he primary makes ions move in the water (an electric current) that in turn induces a field in the secondary, the more ions in the water, the more salty it is, the more salty it is, stronger is the current and field induced. This is in contrast to a conductive meter, which would have two conductors in contact for a long time in the water, shit would grow on it and it would rust and stop working.

If I massively increase the current going to a laser diode and decrease the duty cycle.

Do the same for like 10 other laser diodes.

aim them all at the same place.

does that give me a powerful laser that wont burn out instantly?

Pic related but in reverse.

>>1487732
I ask this because I want to do an experiment, but can't afford the CO2 laser that I'd like, and I know that getting a coherent beam from several lasers that doesn't interfere with itself is a fucking pain in the ass.

>>1487726
Ah, interesting stuff. Though I'd like to see about using two insulated capacitor plates in water as a flow or conductivity meter. I was thinking that as the plates polarise the water between them some current flows, but as the water flows back out again this polarisation will relax but not between the plates. Since new water will keep flowing in and getting polarised there would be a small DC current flowing through the capacitor (not via direct conductivity since the plates would be insulated from the water) that would be proportional to water speed and conductivity. You could also sense the conductivity of the water by putting AC through a single inductor, the eddy current losses will be proportional to conductivity. Though I imagine people use the 2-inductor method for a reason.

>>1487732
The actual silicon in a laser diode is minuscule, such that the heat buildup over a single on-cycle can't dissipate quickly enough and is so concentrated that it can blow out the diode in a single cycle. Take a look at the thing's junction to body thermal resistance and calculate from there.

What experiment do you want to do anyhow? Because I'm well versed in linear and nonlinear laser physics (and laser diodes fucking suck). If you want a strong pulsed laser you could try to obtain a slightly damaged ruby rod reject, plus TEA lasers are somewhat of a diy project. But if you have the money, go for a Nd:YAG, they're the shit. CO2s have a bunch of inhomogeneous broadening due to the gas phase of the active medium, which will interfere with long-distance coherence, moreso for diode lasers. Gaussian beam optics is something to look into if you're wondering about focusing laser beams.

>>1483995
hello, guys
I've found some old soviet russian electronic components
thought to share my finds with you

also, would like to find details on components in english, I know only very basics of russian

pic related what's in the box I've found

>>1487760
closer look at some components

>>1487760
>>1487762
so, are these useless or can I make something?
as you might see, I'm new to electronics, but this find makes me wanna learn

>>1487764
>I have junk, what do?
http://lmgtfy.com/?q=мycopнaя+кopзинa

>>1487762
>rogget benis chips :DD
transliterate to Latin and google it, my lad. see if it's junk or not
>K224GG1
universal multivibrator
https://kiloom.ru/spravochnik-radiodetalej/microsxema/k224gg1.html
>K224UP2
amplifier/color limiter, probably useless in the digital TV age
https://kiloom.ru/spravochnik-radiodetalej/microsxema/mikrosxema-k224up2.html
>K155ID1
Nixie decoder/driver, neat
https://kiloom.ru/spravochnik-radiodetalej/microsxema/k133id1.html
>155ID3
TTL 1-of-16 decoder
http://www.chipinfo.ru/dsheets/ic/155/id3.html
you get the idea. not entirely useless, most of them are equivalents (if not pin-for-pin replacements) to Western chips. could be fun to see if you can build something with pure Soviet technology, if only to shut up the muh indispensable nation shills on /pol/

>>1487739
polarizing the water fucks up the very things you are measuring. That is why even using two electrodes you have to use AC current.

>>1487739
>What experiment do you want to do anyhow?
https://www.youtube.com/watch?v=AoWi10YVmfE

>>1487792
AC*

what is a "capacitative sensing" gpio pin?
I know i can rig up simple sensing with regular gpio pins (tinfoil + wire), so what is the diference between them and pins marked as "capacitative sensing pins"?

>>1487789
>making something with pure soviet technology to shut up /pol/
sounds very interesting
will see what can I do with this, thank you

>>1487797
hardware, of course
rather than driving current into the touch pad capacitance (to what?) and measuring the time to a logic transition, e.g. Microchip's mTouch system connects the chip's ADC hold capacitance in a capacitive voltage divider arrangement with the touch pad, the voltage which is then measured by the ADC

>>1487732
> If I massively increase the current going to a laser diode and decrease the duty cycle.
You'll destroy the laser diode. They're limited by instantaneous power, not mean power. The main failure mode for laser diodes isn't thermal; exceeding their rated power can destroy them in less than a microsecond.

>>1487835
If that's true then how does the death star do it?

>>1487807
>ADC hold capacitance in a capacitive voltage divider arrangement with the touch pad
so it's the same way as just adding a cap to the regular two-pin (output input) capacitive solution, where the human skin just steals some of the charge? Is that resistant to ghosting? I want to use capacitive switches hidden in furniture so badly but they always ALWAYS start to trigger themselves randomly on and off within hours or days after i install them and not even dedicated chinkboards touchpads i tried aren't immune to this.

>>1487842
no, the S/H cap and the touch pad are charged to opposite rails, then disconnected from rails and connected together for a time to equalize. then the sampled/held voltage is read by the ADC. there is no logic drive applied to either cap during sampling and no comparator involved
for more info see Pic related. Microchip also has a lot of other good app notes on touch which are worth reading

on esp32 the VIN pin goes into the voltage regulator right?
Just checking so i don't accidentally fry it like the arduino last time

>>1487764
>so, are these useless or can I make something?

make some art project, or glue them to you face for halloween. go as Ruskie election hacker.

>>1487926
more

>>1487760
>>1487762
Write 'HДCКIИG CHIP' on them and stick them on election booths.

>>1487793
To make femtosecond laser pulses you need a mode-locked laser. A CO2 laser will probably work for this purpose, but a laser diode definitely won't because of its broad energy bands. Helium-neon will also work and may be cheaper.

>>1487921
that depends on your module. if you're that scared, go in through the micro USB port
more than half of this hobby is reading documentation anyway

>>1487936
Кeк

What would be the best approach if I want to build a circuit that detects when water starts boiling in a pot? I was thinking either a humidity and temperature detector, or a sound detector, I'm still fairly new to electronics so I don't know too much about components.

>>1484726
pls elaborate

>>1487926
>>1487927
https://www.youtube.com/watch?v=cxlsaMlI_ng

>>1488070
any temperature probe will do if waterproof. place it near a vent in the lid so that steam will blow on it (and heat it rapidly) when the water reaches a good boil but won't just rise around it when the water gets hot
or, just put the temp probe into the water and pretend that when the water hits 95°C (adjust according to altitude and mineral content) it's "boiling"

Anyone got CAD files for a single-sided board? Something simple like a LED blinker with big SMT discretes would be ideal. I'm working on my home fab process.

>>1487726
>>1487722
I'm surprised it works.

https://www.mdpi.com/1424-8220/15/9/20990/pdf

>>1488243
So it works off the changing permeability? That sounds a bit far-fetched since in E&M we always assume materials to have unity relative permeability, but by the looks of things they've got a measurement system accurate enough to work. Though I'm surprised that the permeability difference is more significant and easier to measure than the eddy-current losses.

>>1486357
>what's wrong with two diodes?
this was the battery and power source will work in parallel and the battery will be drained

>>1488299
Only if the battery voltage is higher than the PSU voltage.

>>1488300
are you sure? Seems to me that logically if the source voltage is higher then the battery would supply proportionally less curent. So say if source is 15V and battery is 10V then out of 10A the battery would supply about 3A just like water pipes

>>1488301
Diodes only conduct if the voltage on the cathode is 0.6V lower than that of the anode. Otherwise they simply don't conduct. If your PSU is at 10V, the voltage after it's diode will be 9.4V or so (depending on charge rate) so if the battery's voltage is less than 9.4V then no current at all can flow through the battery's diode (aside from minority carriers). Test it in spice if you're worried. A small current will probably flow the battery and PSU voltages are within a diode drop of each other, so avoid this situation.

You could also ditch the battery diode all together if you have a fairly resilient battery chemistry, like Pb-acid or to a lesser extent Ni-MH, such that it keeps getting topped up by the PSU. But you'd need some sort of current limiting for the charge stage to prevent nasties when power comes on again after a power cut. Ni-MHs are fine with low-current trickle charging without control, and Pb-acids too are pretty tough.

>>1488301
It does not work that way. If the DC power source is conducting then the diode connected to the battery is reverse biased and no current flows. The only way any current flows is if the battery is ~0.4V higher than the source for a Schottky diode. If for whatever reason the battery IS higher than the source by at least that much the battery powers your load until such time as the battery voltage is about 0.4V below the source and then the battery diode becomes reverse biased and the source diode will conduct and will stay that until the DC supply is disconnected or cut off.

All that said while Schottky diodes have a lower forward voltage they do have high reverse leakage usually in the 100s of micro-amps to low milli-amp range. Regular silicon diodes are better if you want to minimize leakage current as much as possible with typical leakage currents ranging anywhere from a few nano-amps to a micro-amp or so depending on whether it's a power diode or a small signal diode. If you want to minimize leakage and you can afford to burn some extra power in the diode then don't use a Schottky. A FET solution might still be better than either of these though. The leakage is extremely small and you'll burn less power, especially if you get a FET with a very low Rds(on).

>>1488309
On second thought leakage with the diodes probably wouldn't be an issue. It is very small and it'd be leaking into the battery when the battery diode is reverse biased effectively charging it very slightly. I'm not sure how much different battery chemistries will like this but it's a pretty small current so it's probably fine for most. I'd look into this a bit further though if you're using any lithium based battery technology though.

>>1488304
it's a regular old car battery

didn't know about the cathode thing that will be extremelly useful. Also does it have to be a shottky diode? I own some regular diodes.

>>1488309
nice, thanks

>>1488252
no, by inducing ion currents

>>1488312
>regular old car battery
you should have said so. it's much simpler now
use any power supply of adequate current to supply your device plus a few amps, adjust it for 13.5V at its output terminals, insert a diode rated for that current plus a bit (probably a large one, or use the diode in a power MOSFET) between power supply and battery, then just run your shit right from the battery, which stays charged automagically and no other electronics required
the only caveat is that your load needs to be able to handle that 13.5V. if not, insert a sufficient buck converter

>>1488349
>which stays charged automagically a
every time i mentioned here connecting a car battery directly to a voltage source without any acid charger controller in between them, everyone here lost their collective shit, painting a grim picture where everything around the battery ends up on fire covered in literal burning acid because lead acid batteries need very specific charging cycles

>>1488352
>acid charger controller

A what? There is no such thing. Lead acid batteries are one of the most robust battery chemistries. They don't need special controllers or anything. Just think about it, you don't need any complex electronics when you jump a car right? There IS an explosion risk if you seriously overcharge them but I'm assuming your voltage source is about the same voltage as your battery which means it isn't gonna get overcharged. If your source is higher than the battery voltage than check your battery's datasheet for what kind of overvoltage it can safely tolerate.

>>1488361
For example https://www.amazon.com/Sealed-Lead-Battery-Charger-D1724/dp/B001G8AIMU
Just to be safe i will add a diode to prevent battery charging from the source. I only need 5v anyway so i can use step up to raise the brick to say 14v to make sure the cathode thing blocks the battery

>>1488380
I don't think there is particularly sophisticated charge control circuitry in there, at least not like you'd get with a Lithium-ion or LiPo battery.

First of all that particular charger is designed for charging lead acid batteries off mains which is something you definitely can't do directly. It'll have a transformer in there to step the voltage down, that particular one can do 6 or 12V so it's probably a center tapped transformer and the switch selects the tap. From there that's rectified and filtered (though the reviews for the model you link suggest there is no filter cap as it seems to output AC which is bad) to give a DC output voltage. It's doubtful there's any active electronics at all. Perhaps a regulator but there's no real need. The bulk of that charger brick simply the transformer which I'm guessing is a 60Hz mains transformer because if that thing used a offline flyback converter or something it'd be a LOT smaller.

>>1488352
that depends on the voltage source. by keeping the voltage at a reasonable level, you don't damage the plates or electrolyte. Li+, on the other hand, do pee fire
see https://www.electronicspoint.com/threads/lead-acid-batteries-and-float-charging-question.132004/
>If you're using an adjustable supply, if it has current limiting, set the
voltage 13.5v with no load, limit the current to whatever the battery will take
safely, and forget about it.

>>1488361
they don't *need* them, but a CC-CV algorithm will keep the battery healthier than just jamming current into it

>>1488380
>i will add a diode to prevent battery charging from the source.
you could do this, if you're willing to put up with the self-discharge of the battery and give it a proper charge every few months

>>1488385
>they don't *need* them, but a CC-CV algorithm will keep the battery healthier than just jamming current into it

Yeah well, the lead acid battery "charger" he linked to on Amazon sure as shit ain't doing any of that.

>>1488386
>single-stage
my vulcan sides

Is it possible to wire up a 3 phase 220V variable frequency drive to a 220V single phase induction motor?

Why do induction heaters, tesla coils and some other "big" circuits use some relay fuckery, ZVS and tank circuits for oscilations? why they do not use phase shift oscilators or another simple oscilator and amplify the signal?

>>1488661
The oscillator has to be resonant within the active coil, otherwise you're running at suboptimal efficiency. Just amplifying any old sinusoidal resonator not only means you're wasting power in the amplifier, but also that you're attenuating the signal in the load because it isn't at the resonant frequency.

This is not actually a hard question, just wanted to ask if someone could look over the pic and tell me if that would work. (the light grey LED is not gonna happen most likely)

U battery: 4.5V
U of LED: 2.8V-3.4V
I of LED: 20mA = 0.02A

Therefore: Rmin = 55ohms, Rmax = 85 ohms
Parts: R0 = 50 ohms
R1 = 100 ohms potentiometer

Regarding the capacity:
3*140 mAh = 420 mAh
420 mAh/20mA=21h lighting time until empty

Would all that work? Am I retarded? should I go with a bigger potentiometer? Thanks!

how fast can mosfet flip on and off?

>>1488992
LEDs, Diodes, Transistors and other junction stuff follow Shockley`s law. (They are exponential) So when the tension in the LED gets under their junction drop they`ll stop working.
So if your LED is ON at 3.4 V, the battery must have 3.4V+Rdrop (20mA*resistance). When it gets under that value it`ll shut down.

>>1488995
2* Sum of (Switching on time [ton], raise time [tr], falling time [tf], switch off time [toff]) I'd guess. Look into the specific mosfets datasheet to get a better idea.

>>1488998
i just need a rough idea

>>1488997
So I should just skip R0? Without it It will be over the 3.4V and work?

>>1489000
IRLB8721PBF has ton=9.1 tr=93 toff=9.0 tf=17 =>128.1 ns , so roughly 256.2ns for this specific mosfet.

>>1489001
Look for the model of the batteries you are using, usually the datasheet has a curve showing V vs Usage. If you know your kirchoff you can think of the diode as a constant current source

>>1489004
does gate pulldown affect this?

>>1488997
>shut down
show me the discontinuity in the Shockley diode equation and I'll believe you
they will always emit light when forward current is going through them. maybe it's not enough light to see but it's there.
>being this new
people under-amp LEDs to dim them all the damn time

>>1488992
yeah should be fine as drawn. you can put a resistor in parallel with R1 to make R1 seem smaller
don't omit R0 unless you're absolutely sure the battery won't supply too much current when the pot is at 0 ohms. some button and coin cells have a high enough internal resistance that you can get away with it, but I wouldn't

>>1489009
the gate is basically a capacitor, so the usual RC equations apply

>>1489011
Thanks, If I went with a 100 ohms potentiometer I'd not need to lower the resistance of R1 would I? Regarding the R0, I'll just test it when the parts are there most likely, thanks for the tip.

>>1489011
Are you retarded? He said 2.4-3.4V I just took the worst case.
> maybe it's not enough light to see but it's there.
But as he is using a LED, I imagine he wants to see the light, so if he dips under a bit of the diode drop it`ll die out very quickly, I imagine that will be arround 2.4V

>>1489025
>>1489013
What I`m saying is that your resistors are ok, but the time you calculated is off. It`ll be shorter. But as you probably are making a reading lamp or a prostate iluminator, a few hours don`t matter much. But as we are on the subject, usually the current changes 1 decade for every 60mV at room temperature. From 2.4V to 2.2V the current will shrink in a factor of almost 40 thousand times

>>1489013
come to think of it, 100 ohms should be fine, but 200 or 250 ohms might be better, giving you a 4:1 ratio. still assuming a 56 ohm R0 for a maximum current of 20mA

>>1489027
>usually the current changes 1 decade for every 60mV at room temperature
for silicon. other materials vary. Pic related
stop shitposting forever

>>1489032
>slower rate of change
>as if it matters
You said under-amp LEDs for dimming, as in controling current (decades to swing without apreciable change in Vd), I`m talking about supply voltage(tiny changes arround a certain point = big changes on current and on light) .You are clearly confused, study more anon.

>>1489040
the battery curve should taper off to the left instead of climbing, I forgot to correct that but still you get the idea.

>>1489032
Thanks again, I'll propably go with 200 ohms, I can always not use the full range of the potentiometer if I have to after all.

So esp32 has this nice deep sleep mode to save power.

I want to use it to turn off and off light in my kitchen. Basically i need it to check capacitance on a pin every 1 second.
Do you think think there us any benefit in putting the esp32 to sleep for 0.8s? since it takes only about .2s to read the capacitance and turn shit on or off.

>>1489151
you should consider using an external touch sensor like the TTP223, where you can leave the processor in deep sleep almost all the time. the TTP223 can be jumpered for toggle mode so you can connect it directly to the switch
>esp32
>for an on-off switch
I sincerely hope you're also using the radios on it

>>1489152
>radios
why in the world would i use wifi or bluetooth for an on/off switch?

Also don't even remind me of that piece of trash TTP223. I ordered whole lot of TTP223B and those cunts only work when actually directly touched, making them useless. And attaching a large sensing pad makes them go off randomly. I wish i could strangle that anon who recomended them to me the first time around.

I am done with such gimmicks and that is why i will be using the capacitive sensing pins on the esp32 from now on without any IC gimmicks like TTP223.
I am sure it looks nice on paper, but after you get woken up by your lights turning on for the 5th time in a night, because casper jacked off nearby, you would smash that ic into pieces as well.

>>1489155
>why in the world
dunno, ask the millions of people who want to be able to turn shit off from their phones
ESP32 sounds like a heavyweight solution for this, especially when there are PICs with touch sensing that should work just as well. but if it all works, I'm not gonna argue too hard against it
>strangle that anon
yeah, that was me. fite me
anyway there are other sensing ICs using different sensing principles. don't dismiss the whole genre just because you hit some bad chinkshit that didn't work for your application

>>1489171
name one scenario when you need to turn on light at home when you aren't at home except to ward off tyrones

>>1489178
eh, think of a use case more like a universal remote for when you're all comfy in bed and don't want to have to walk to the wall switch, or you want to start the coffee maker going in the morning while you catch a few more winks

>>1488992
>Regarding the capacity:
>3*140 mAh = 420 mAh
>420 mAh/20mA=21h lighting time until empty
when batteries are connected in series the ampere rating stays the same - only the voltage is increased
to increase the ampere rating batteries are connected in parallel

>order some boards on the 21st
>still no shipping notification
ffs make my boards already chinks

>>1489246
sounds like you're getting great feedback from them. which board house is this?

>>1489215
mAh are not current but a representation of charge.

Is gyroscope and an accelerometer enough to measure movement in space? Like i want to know that from the point the device was at a second ago it moved up 10cm

>>1489302
Yes if you have the processing power. Gyroscope tells you the angle, accelerometer gives you force/acceleration, I'm assuming you use a microcrontroller, so you should be able to get time, if not you'll need to get that somehow. If you have a graph or dataset for those three things you can get movement in space from those.

>>1489264
elecrow

>>1489302
you may need a high sampling rate to integrate the accelerometer data for best accuracy. it should be simple, some phones use FPGAs to do it

>>1489321
thanks for the warning. I'll preemptively strike them from my personal approved vendor list

>>1489302
It's called inertial navigation, and it's a bit shit.

>>1489332
>some phones use FPGAs
kHz sampling rate + trapezoidal numerical integration is probably all the cheap 6/9-axis sensors are good to anyhow.

Dude use the accelerometer to detect the position of the moon and sun to determine your time of day, latitude, and longitude.

I need to into LoRa communication between arduino and pc.
What am I even paying for if I get this $90 shield
>https://www.cooking-hacks.com/lora-radio-shield-for-arduino-868-mhz
or this $20 module
>https://learn.adafruit.com/adafruit-rfm69hcw-and-rfm96-rfm95-rfm98-lora-packet-padio-breakouts
instead of either of these $3 and $4 chips?
>https://www.ebay.com/itm/192272728108
>https://www.ebay.com/itm/173340521360

>>1489402
What a joke.
Buy a gsm shield for $16 and you will get planetary signal coverage

PAGE 9
BUMP LIMIT EXCEEDED
PANIC

>>1489508
except where GSM has been shut down

>>1489402
>€90
you're paying for the weird specialty site markup, VAT, a modestly better antenna, and that goofy Digi form-factor baseboard
>€20
you're paying for a multi-band module, 2.54mm spaced holes, plus Adafruit markup
>€4
some assembly required

>>1489526
this isn't /g/, stop panicking
I got dis