/diy/ - Do It Yourself

>>2404248
>>2404426
i have used kicad to make pcbs in the past. i'm trying to learn SMPS and i want to do some of this shit on perfboard before i waste $15 and weeks of my time on waiting for a chink pcb that i did a bad job of designing. ive tried veecad before but the UI makes me want to kill myself. i suppose i could african american rig kicad to emulate a perfboard but it's a huge pain in the ass.

>>2404595
Ever tried eagle?

In my country this is called sneedlight.

>>2399471
anybody have anything on this?

>>2404642
If it stops working will you be the one doing repairs or will you buy new? If a hybrid inverter fails there is more extensive troubleshooting involved (unless you have the full schematics).

>>2404646
Im looking a getting good quality products, (think Victron), so Id claim warranty if anything croaked. My issue is the lack of solar inputs at different voltages for hybrid inverters, which is why Id like to do multiple solar-to-48v converters and add line inverters as I need them.

>>2404647
Do you have a MPPT charge controller?

>>2404651
Im building this from the ground up, I just want to be able to expand the system later, since were putting panels up now and will put up more later.

What are some 2U 20MHz + function generators like the Wavetek 175?

>>2404656
Philips Fluke PM5139 is 2U I believe

>>2404654
Discrete components will give you more flexibility at lower cost when you expand the system.

>>2404662
That was my thought process, especially since I have to split up the line into 3 phases. Do I lose much (efficiency wise) if I go solar-48v-line?

>>2404663
Several panels that aren't matched will not be as efficient as a balanced system, and there are losses in every buck/boost conversion and interconnection between the cables. MPPT charge controllers take care of a lot of that at decent efficiency (better than PWM).

>>2404647
I've been through what you're going through, and what I found out is that the real place that "hybrid inverters" (I usually see them called all-in-ones or AIO's) win out is that you can build the least expensive system with them by using cheap Growatt/MPP/EG4/etc all-in-one systems.
They aren't as flexible or configurable as using discrete inverter/chargecontroller pairs. And if you're planning to add more capacity later, it's easy to just add another charge controller or inverter in a modular way.
The other thing is that the all-in-one type systems tend to be low end chinese made units which have poor surge handling capabilities and other technical downsides like low quality loud fans and much higher idle current draw.

I've used Victron, Midnite, Magnum and Xantrex over the years and didn't have any serious issues with any of them, other than the kind of high price. Also victron has terrible technical support; their documentation is a mess and their dealers who are supposed to provide technical support are pretty useless.

If you don't have more than 2 real arrays, you can find an all-in-one that supports that pretty well.
Ultimately, you can run the numbers to see what the price difference will be for your proposed system using however many all-in-ones vs using discrete components, then try to weigh the tradeoffs of warranty support, build quality, whether the more complex configurability matters to you etc etc.

>>2404671
Ill have at least 3 different arrays, probably more since I have to split up a big one. 2 at ca. 1kw each, (1 east 1 west), and 1 big group facing more or less south. Since the big one will probably be about 5kw, Ill have to split it anyway, so I might as well go with 4+ arrays.
Can you recommend a brand thats good (ill look up the ones you already mentioned)? Especially one that will sell me rackable inverters, everything Victron has is wallmount.
>>2404668
Thats why I want discrete MPPTs for every "string".

>>2404673
>discrete MPPTs for every "string"
That's the way I'd do it.

>>2404674
Alright, Ill draw up a concept. Thanks.

>>2404676
NP, Anon. Good luck.

>>2404677
Thanks. I wish I dint have to make my own electricity, but these are strange (and expensive) times.

>>2404678
Necessity is the mother of all fuck. Maybe we'll see some new discoveries soon, made by /diy/ pioneers.
Hail the common man.

>>2404692
That might be, but I still prefer not paying 50ct per kwh. Call me old fashioned.

>>2404694
>Call me old fashioned.
old fashioned = sane

>>2404696
Its fading, buddy.

>>2404697
Everything fades. Welcome to hell.

>>2404703
I can only watch Apollo 11 this many times, and even OUATIH is losing its effect. I can barely study anymore because the future seems so boring and constricted.

>>2404705
Seems like a good time to go innawoods and imitate Primitive Technology guy to clear your mind. Don't forget to bring sausage.

>>2404709
I started surrounding myself with vintage test equipment, maybe that´ll have an effect.

>>2404711
It can't hurt. I long for the days when nobody batted an eye at photos of nude pinup models in workshops.

>>2404714
The amount of kvetching when someone uses 'Lenna' as a test image these days is mind boggling.

>>2404717
Political activists are sociopaths and men who believe they're women are psychopaths and/or sexually retarded. Yet they are only tools for people far more evil and cunning than they could ever be.
Wiemar 2.0

>>2404711
>conjure the aroma of old electronics equipment

>>2404731
Its in the attic, too, so its getting pretty toasty.

>>2404731
For me, it's the scent of Nintendo games on Christmas morning. Mmmmmm... plastic.

>>2404671
If you´re still here, do you have any communication between the devices in your system? I thought of connecting them up via serial/rs232 through a switch, but I dont know if its even really necessary.

>>2404630
Very simple use of a readily available module to solve a problem. Sometimes you don't need to design anything to solve a problem but just need to take a look at all the resources you already have

>>2404736
If you get equipment from one manufacturer (like all Midnite, or all Outback, all Victron etc) they normally communicate with each other. Everything I have communicates except the BMS for the battery-- it's a diy lithium iron phosphate battery.
If I'm reading this right, then your arrays are going to be two 1-kW arrays, east and west, and then a single 5-kW array pointing south?
If I was setting that up, I would use a single largish MPPT controller for the 5 kW array (probably a Midnite Classic 150, or a Victron Smartsolar 150/100) and then use some small cheap chinese MPPT units for the east and west arrays. But that's just me

>>2404922
Just so that I understand correctly, looking at the Victron 150/100. I cant just plug in 3 differently shaded "arrays" into it, right? It just has 3 different +- because you can only feed it 150V peak.
Also I thought about using those 24V RV replacement batteries, they´re around a grand each, but apparently have 200Ah and an integrated BMS.

>holes moving backwards through fused sand makes my computer work
you guys are retarded

>>2404929
Wrong board.

>>2404930
no its the motherboard

>>2404934
The motherboard is nothing without the fatherboard.

>>2404926
The Victron 150/100 only has a single actual MPPT unit (even though some of them have several terminals, this is just to make it easier to connect several conductors in parallel to make wiring the device easier compared to using a single giant set of conductors)
The MPPT algorithm/control will apparently not operate correctly when there are multiple different arrays with different operating characteristics attached, although I would think that the device would still be putting out power.

>>2404940
Alright, so its as I thought, multiple inputs but only one voltage. Should be enough to run my system. Thanks for the help. Sadly Victron doesnt offer a unit at sub 2kw with 48V.

>>2404945
Scratch that, 100/20 has 1160w at 48v, seems like a fit.

>>2404709
That's not how you spell "Ted Kaczynski".

I was thinking of making my own portable gps unit. What's the best chip to use for this that has good accuracy? I don't mind an external antenna. Do I just hook it up to a raspberry pi or some other small computer? I assume making my own board from scratch around the chip would be a lot more complex.

>>2405052
This may not help much, but I know mitxela on youtube used a GPS module to make a GPS clock. Since he goes into a lot about accuracy of timing, there's probably some takeaways from it that you won't get from an adafruit or sparkfun video, which should translate at least somewhat towards making a navigation system.

You may be able to do all the calculations necessary in an upper-end 8- or 16-bit micro, but I suspect a 32-bit micro is going to make them a lot easier, especially for also driving a graphical display. A raspi is probably overkill, I'd go for a blackpill or an ESP32 board of some sort. Getting street data may need either a cellular module or a large SD card packed with the relevant data.
If you want something as graphically demanding as a 3D map overlay, then an SBC like a raspi may be necessary.
If you hunt about a bit, maybe someone already has a dev-board with a GPS module on it.

See also: the microcontroller general. Also the remote control general guys may have a thing or two to say about GPS modules, I bet a few are plug-and-play with ardupilot boards if they don't have them built-in already.

https://www.youtube.com/watch?v=0nLT_bEaLlY

>someone made the half-life crossbow
>they literally just threw lipos at it until it worked
is this really all you have to do? just throw more battery until circuit works?

>>2405095
Anything's a resistor if you try hard enough.

>>2405033
America's Uncle™

>>2405095
>immediately cauterizes the wound and burns everything in sight except for the target

Anyone know if there's some acceptable limit to the path length between a motor driving MOSFET and its filter capacitors? I'll have 30-40mm on this design, which is similar to what a chinky motor drive I bought has.

>>2404656
>2404656
HP3324A pictured. 21MHz sine, 11MHz square, much lower freq triangular, ramp, etc. (Not sure if it's working to spec.) 16 5/8" wide front panel. Rack mount ears optional. It is 20 inches deep.

i have a cheap dirt bike with no lights or indicators. all i want to do is hook up a neutral light to help find neutral when i stall (pain in the ass to find when hot and kickstart engine so required). im thinking about instead of trying to fit a small sla battery and making a bracket and everything, just using six 2.7v supercaps in series in a 3d printed case so that i dont have to try and fit a large battery just to power a 20ma led and i wont have to worry about a switch or unplugging it to keep from killing the battery (after a long long time).
>is this a dumb idea?
>do i need a resistor for the capacitor input or will the led and its resistor work since everything will be in series?
even though the bike has nothing electrical besides the ignition the engine has a power coil installed and wired as well as a neutral switch, so its all plug and play as far as that is concerned.

>>2405144
>a small sla battery
coin cells will last ages, but sufficiently well-matched supercaps will work i guess
don't understand your circuit though, is current meant to flow backwards through the voltage source?

My amp doesn't work. I poked many points along the audio signal path, and at least the passive components seem fine. I measured the op amp rails and found that nothing was going there. I measured the rails of the power amp though and a nice +/-15 V was present, and poked the inputs and it was indeed working, although maybe a little too soft? It's "only" 8 Watts or so but I remember it being really loud (last time using it was maybe two or three years ago). Pic related is my crude trace of the power section and other stuff. Any ideas?
As a side note, I applied some of the modifications you suggested and took come polar caps out. I bridged the traces, of course.

I'm going to try setting up an aquarium monitoring system for my betta fish. To do so, I need to measure temperature, pH, ammonia concentration, and possibly CO2/O2 concentration (I say possibly because betta fish breathe on the surface and don't really care).

Temperature and pH sensors are fairly easy to find, but does anyone know where I can get an ammonia sensor?

Also does anyone know what else might be useful for this sort of thing?

>>2405192
Do you mean you were probing the signal as it propagated through the amp with a scope? If not definitely do that. If you don't have a scope, you can probably bodge something to work using an external USB sound card. Chuck a voltage divider and some clipping diodes on it so you don't ruin anything, and a buffer amplifier is a good idea while you're there.

What exactly are the symptoms? No sound at all? When did this happen?

Personally I'd have holden off on the modifications until your troubleshooting was completed. The polar caps may well have been the issue, but I'd swap them for some non-polar caps until you've got it working. Arguably, the possible AC current going through C14 was causing the different sounds at different volume levels you initially complained about, that or it was IC3 itself. There's nothing else that could cause that difference from what I can see. The speaker doesn't have a cap in series, does it?

Also all the schematics were on the other thread, so either collate the three images into one and repost, or just link to those posts here so they're not hard to find (>>2400757, >>2402383). I'd lean towards the former, with some parts (especially that PSU) redrawn to be more symmetrical and easier to follow. Rectifier bridge becomes a block, grounds in the middle, +V at the top, -V at the bottom, etc.

>>2405244
A lot of those are nontrivial, even pH needs calibration. I've seen an anon here before use a greedy cup syphon in order to make repeated washings of the pH meter tip with pH 7 water between periodic measurements. I'm not sure if leaving a pH probe in the water would work. That said, I'm unfamiliar with the mechanism behind pH sensors.

Ammonia concentration can likely be detected using one or more metal oxide gas sensors. They heat a small piece of semiconducting metal oxide, and as gas flows over it decomposes the gas and produces an electrical signal. Different sensor element materials react to different gases, but none are particularly specific, so I'd use an array of different sensors (can get like 8 different sensors from ali for not much money). You can then waft some cloudy ammonia into the sensor chamber with these in order to characterise ratios between the sensors, and look for those in the noise with some sort of DSP algorithm. Or just find the one sensor that triggers on ammonia but nothing else you care about. I'd mount such a sensor chamber near the tank, with a fan blowing tank air through it to catch any evaporating ammonia. You'll need to correct for tank pH and tank and air temperature, but that shouldn't be too tough.
Chemically, ammonia complexes with copper, so you may be able to measure the electrochemical potential between a copper electrode and a non-copper electrode, though of course this potential will also change with pH and other aqueous factors, so maybe have an array of different electrodes to measure potentials across. At least for initial testing.
The other option would be an absorption-line sensor, which would be a pain to construct, but could function submerged instead of relying on evaporation. You can buy them as-is for CO2 though, and I recommend you do this instead of an ECO2 sensor even though they're more expensive.

As for oxygen, I imagine you'd have to measure it out of the water like my ammonia gas suggestion.

>>2405255
Oof. No wonder I couldn't find any basic ammonia sensors. I didn't realize how involved it would be to get it set up.

I think I'll put that idea on the backburner then. I'll just do my manual testing every few weeks like normal.

>>2405261
Don't let me scare you off. Make a little holder with a copper wire, iron wire, and some other wire, such that they're all submerged. Then measure the voltage between the three of them each time you measure ammonia and pH to write it down. Hopefully there's a correlation.

Also looks like the MIX2111 gas sensor on alibay will do ammonia, as well as a few other gases like formaldehyde, so it alone might be good enough for you, if you don't mind using a fan.
Not sure how those gas sensors handle humidity and if it impacts measurement accuracy, but I doubt it's much of an issue for sensor longevity.

I made a thing this weekend and filmed it

>>2405287
You aussie or kiwi? Either way you'd better put on a real drawl for the camera. Those STC-1000s aren't at Jaycar, are they? Wish they were, handy buggers.
>physical transformer brick
>for a digital project
>in 2022
why

>>2405303
Ausfag, got the controller on ebay for a different 12v powered project, it is surprisingly good actually.
Transformer because I needed 12v and had it lying around anyway. Simple and reliable.

>>2405311
>Simple and reliable
The bulk is the main thing I care about, means you often have to step up to a larger project box. Also for a given output current the required filter capacitance is way larger. Think I'm going to buy a bunch of AC-to-DC SMPS modules I can just drop into circuits so I don't have to worry about switching ICs. Maybe I should buy half a dozen 12V and 5V bricks from a thrift store to that end, they're probably fine.

I save transformers for audio and instrumentality projects, or anything where multi-taps are necessary. But if you have an excess it doesn't really matter.

>>2405317
Yes they are indeed heavy. Wasn't really a concern for me in this case since the finished item is just going to sit in one place (greenhouse)
SMPS is definitely a better option but the ones in Jaycar that I could see last time I was in there weren't quite right. A simple charging brick/wall plug type would also suffice but I've used all the 12v ones I had on various other things.

>>2404526
>super compact efficient
I decided to try surrounding coil with the magnets without using brushes.

This way the magnet field is converging instead of diverging.

>>2404526 >>2405347
>super compact efficient
Here's the powerhouse version.

>>2405180
according to 3 minutes of using online calculators at 5f the led will stay lit for 4 or 5 minutes with the engine off which is more than enough, and I just didn't want to have to deal with ever changing batteries Even though yeah a coin or 18650 would last 5ever basically.
also don't read to hard into the circuit, i just slapped it together to get the point across with no regard for correctness.

>>2405353
would this be better if both impellor [B] blades forced air in, and allow it to escape thru those other drilled holes?

.....to keep the spinning magnet rotor centrally balanced.

>>2405362
air bearings?

how do HVDC transmission systems invert DC to AC at massive power levels? i tried reading the wiki article but i'm kind of lost here. i know they typically use thyristors or IGBTs for switching but what kind of switching is used? modified sine wave or do they do "pure sine wave" directly? if it's the former, how is it smoothed out into pure sine wave?

>>2405372
>The MMC differs from other types of converter in that current flows continuously in all six valves of the converter throughout the mains-frequency cycle. As a result, concepts such as “on-state” and “off-state” have no meaning in the MMC. The direct current splits equally into the three phases and the alternating current splits equally into the upper and lower valve of each phase.

>>2405378
>A typical MMC for an HVDC application contains around 300 submodules connected in series in each valve and is therefore equivalent to a 301 level converter. Consequently, the harmonic performance is excellent and usually no filters are needed. A further advantage of the MMC is that PWM is not necessary, with the result that the power losses are much lower than those of the 2-level converter, at around 1% per end. Finally, because direct series-connection of IGBTs is not necessary, the IGBT gate drives do not need to be as sophisticated as those for a 2-level converter.

>>2405380
the MMC animation shows a modified sine wave. doesn't this have to be smoothed out before it can reach the customer's outlet (or be synchronized with the AC grid)?

>>2405385
The 300 submodules connected in series filter the line on each valve.
>and usually no filters are needed
Sometimes they are needed, apparently. Above my paygrade.

>>2405370
Designed as stand-alone hand held unit, without external compressor.
Using the reactors turn-force to get in-device impellors to provide 60psi just for the bearings to work seems excessive.
I do love their stiffness, thou.

I can see myself using this idea for larger models
cheers.

>>2405347
Uh, anon, do you wish to tell the NRC something?
Also please draw threads as two lines, not as a zig-zag.

>>2405144
>>is this a dumb idea?
A little bit, yes.
Your bike already has a battery so you can hook up the LED through the ignition switch. When the switch is off (aka bike is not in use) the LED will be off.
All you'd need for this is extra wires instead of some expensive supercaps.

>>2405675
i dont think a kickstart bike has a battery

Trying to fix a transformer with an open primary winding.
Found a blown thermal fuse in the first layer of windings.
Past it I get continuity.

Question, what kind of resistance should I see?

It's a 120v 60hz primary 23.2v-24.9v 3.2a secondary transformer. Only other comparable transform I have is 120v to 26.5v at 650ma. Its primary measures 36ohms. The transform I'm trying to fix measures 7ohms. The lower resistance is probably due to the higher amperage but want to run it past the thread before going on.

>>2405681
Then he'd have to add a small battery but my point still stands, there's no need to worry about forgetting to switch it off if you wire it through the ignition switch.
Also buying supercaps as stand in for a regular battery is tarded.

>>2405684
I dont think there is any electricity anywhere if the motor isnt running, and then he doesnt care about finding neutral (because he already started the bike). I think he wants something that indicates neutral effectively until the bike cools down, if he wants to start it again after shutting it down.

took apart an old microwave. Anything cool I can do with the microwave cannon component bullshit thing?

Tried to saw out the secondary on the transformer but I cant knock out the other piece since its glued or something in there :(

>>2405687
>dont think there is any electricity anywhere if the motor isnt running
Which is why I said there would be a need for a small battery. Nothing too fancy 2 AAs would do and last a fuckton for powering a single LED.
>and then he doesnt care about finding neutral
He literally said he wants to know where the neutral is so the bike would be easy to start.
>I think he wants something that indicates neutral effectively until the bike cools down
By wiring it to the ignition switch the indicator works at any time the switch is on, it doesn't matter if the engine ran recently or if it's stone cold.

>>2405698
mate Im just trying to apply what he wants, he doesnt seem to have trouble finding neutral when the bike is cold, only when its hot.
here >>2405361 he says he doesnt want to change batteries (ever), so the cap idea is ok (because you need no charge regulator)

>>2405696
If you crush and breathe the beryllium oxide disk around the magnetron you will get immediate cancer. If its a newer microwave it has anti spark features that would be neat to check out

>>2405683
You would need to measure
1. core area A
2. wire diameter
and then calculate N from
E_rms = 4.44*N*A*B*f
with B of 1.5 tesla if using standard steel core and then estimate winding length. It's a lot of work that is also based on approximations and assumptions. If you aren't seeing 0 Ohm then it's probably safe to plug in, as it's most likely that fuse did its job and melted before anything else. In the worst case if winding is already fucked it will start smoking and probably make your breaker pop. Just plug it in and hope for the best.

>>2405703
>so the cap idea is ok (because you need no charge regulator)
Nigger you still need a power source to charge the caps.
So far we've established that there is no usable power source on the bike. Do you expect him to charge the caps with his benchtop PSU every time he heads out? That sounds much more hassle than changing 2 AAs once a year.

>>2405711
If he has front facing lights, he should have an alternator/dynamo, right?

>>2405675
>Your bike already has a battery
no
>so you can hook up the LED through the ignition switch.
doesnt have one, kill switch button only
>When the switch is off (aka bike is not in use) the LED will be off.
you dont think i would have thought of this in the first place if it was applicable?
>All you'd need for this is extra wires instead of some expensive supercaps.
enough supercaps to do this costs as much as the smallest 5ah sla battery i can find

>>2405681
correct

>>2405684
>Then he'd have to add a small battery
see above why i dont want a battery to power a single led
>but my point still stands, there's no need to worry about forgetting to switch it off if you wire it through the ignition switch.
see above
>Also buying supercaps as stand in for a regular battery is tarded.
again, it is only to power
>one
>single
>led

>>2405687
I dont think there is any electricity anywhere if the motor isnt running
correct
>and then he doesnt care about finding neutral (because he already started the bike). I think he wants something that indicates neutral effectively until the bike cools down, if he wants to start it again after shutting it down.
exactly. when cool its piss easy, when hot it becomes a pain in the cunt because lolchina

>>2405698
>Which is why I said there would be a need for a small battery. Nothing too fancy 2 AAs would do and last a fuckton for powering a single LED.
i dont want to use non rechargeable batteries, if i already have to mount the electronics i want it to be as maintaince free as possible
>He literally said he wants to know where the neutral is so the bike would be easy to start.
REstart. HOT REstart.
>By wiring it to the ignition switch the indicator works at any time the switch is on, it doesn't matter if the engine ran recently or if it's stone cold.
*facepalm*

>>2405703
this nigga gets it

>>2405711
again, the bike has charging coils installed and wired. obviously i will install a regulator too because the coil outputs 10-60vac

>>2405711
scratch that, power coil is probably only for the spark, because no light. Hed need a bicycle dynamo.

>>2405715
as said i have 0 electrical implements besides the coil, but the engine is an off the shelf part and has a verified working charging coil

>>2405718
hah, I was right after all.

>>2405718
Attach a Peltier module to the exhaust and power it that way.

>>2405718
>enough supercaps to do this costs as much as the smallest 5ah sla battery i can find
*enough supercaps to do this costs as much as the smallest sla battery i can easily find (which happens to be 5ah)

>>2405727
this should be a thing for hybrids, im sure it would work greaaaaat

>>2405728
>*enough supercaps to do this costs as much as the smallest sla battery i can easily find (which happens to be 5ah)
I don't know why you're going full retard with this.
You're powering a
single
fucking
LED
Why would you need a lead acid battery in the first place? Why not an 18650?

>>2405728
>im sure it would work greaaaaat
For powering one LED? Yes.

>>2405705
I was hoping it'd be easier than that. I'll wire up the smallest fuse I can find and give it a go. Thanks.

>>2405718
>doesnt have one, kill switch button only
Wait, so the bike always has the ignition connected and you have to hold down a push button to disconnect it? That's retarded.

>>2405743
mate its a kick starter, it only has "wires to sparkplugs are connected" and "they´re not"

>>2405747
Mate my fucking weedwhacker has an on/off switch for the ignition. Most small engine equipment I've ever seen have the ignition on an on/off switch.
What the fuck kind of chinesium bike do you have that uses an NC push button for that?

>>2405734
battery=have to Install switch or light will always be one
capacitor=don't

>>2405738
I meant for hybrids

>>2405743
>>2405751
it's a no button, and it jumps the cdi kill pin to ground to kill it. idk what is so difficult to understand.

>>2405751
not my bike, but your small engine has a off(kill), on(run) and choke(start) setting. If not, it has that split into two switches. Theres no difference between these implementations, other than momentary vs constant switches. Dont know why youre so hung up on it.

>>2405753
>battery=have to Install switch or light will always be one
>capacitor=don't
Anon, caps don't switch off by themselves.

>>2405759
but they dont care when they run out

>>2405759
>>2405760 beat me to it

>>2405761
too quick on the draw

>>2405754
>Dont know why youre so hung up on it.
Push button implementation allows the engine to start at any time the button is not pressed. If this was something with a pull start like a lawn mower then that'd be fine but for a bike this is not very smart

>>2405764
but I enjoy videos of chinks running over baby chinks with forklifts.

>>2405764
ITS KICK START ONLY
HOW ARE YOU GOING TO *ACCIDENTALLY* KICK START SOMETHING
Jesus fucking Christ

>>2405764
why would the engine one a kick start just start at any time? That´s an engine id like to have, not just on a bike, also on a chainsaw it´d save a lot of work.

>>2405760
And why would you care if a battery runs out vs a cap runs out? I don't see the difference.

>>2405772
because a battery gets damaged if you run it dry, a cap doesnt

>>2405254
>Do you mean you were probing the signal as it propagated through the amp with a scope?
I was measuring resistance across components with a multimeter. I also poked at different spots with the tip of a plug (is that a retarded thing to do?) which is how I knew the power amp was at least picking up and amplifying stuff, and it does have power going to it.

>What exactly are the symptoms? No sound at all? When did this happen?
No sound between the input jack and the power amp based on poking around with a plug. No power going to IC1 or IC2. This happened yesterday right after I finished putting the pcb back together (I had taken some potentiometers and capacitors out). I did have to mock up VR2 with a few resistors because I literally broke it and don't have a replacement; also broke the pads just in case, kek. That's it, I think.
I took some parts from the power section for IC1 and IC2 and measured with a multimeter. R27 and R28 are fine. The zeners each put a 1.3Meg resistance in forward bias. I can't get a reading in reverse bias. The 470uF capacitors appear to be shorted.

>Arguably, the possible AC current going through C14 was causing the different sounds at different volume levels you initially complained about, that or it was IC3 itself. There's nothing else that could cause that difference from what I can see.
My complaint is the overall un-impressiveness of the sound coming out of the speaker, especially the distortion channel. The driver might be largely guilty, but the distortion sound could do with some refinement. I had installed some LEDs in the feedback loop of IC1b, but took them out out of concern that they were causing trouble (they're probably a decade or more old).

>The speaker doesn't have a cap in series, does it?
No, it does not.

I made mistakes with the pin numbers for IC3. It should be 2,1,3,5,4 instead of 1,2,3,4,5. I'll put the full schematic up later.

How do i record antenna signal with probes at 1+ GHz frequencies and save it directly to memory chip? I'd like DIY circuit way if such device isn't in production.
I'm only interested in few seconds of raw analog signal, for later processing on computer

>>2405836
>1+ GHz
>probes
>DIY
It's possible but it probably won't be cheap. What you need is a RF mixer that will shift signal down to lower frequencies and then digitize using ADC. This requires a lot of knowledge to do properly so if you're asking how to do it the you won't be able to do it. For $30 you can get a SDR dongle that will go up to 1.7 GHz, but if you need more you'll probably want to get something like Hack RF for $300 that can go up to 6 GHz.

any other way to measure radiation besides a geiger counter? I guess I can sink some money into one, but for a passing curiosity I'm wondering if there isn't some cheaper simpler method. Would a radio get super staticky in the smallest FM setting if the antenna picked up radiation?

>>2405855
You can visually detect neutrinos with a cryo-vapor chamber.

>>2405841
Well that's the problem, SDR aren't actually what i want because they're hardware FFT modulation devices and you're only getting certain bandwidth which is pre-processed for certain range of frequencies. SDR most of time offer "6 GHz" but the actual bandwidth is few MHz...
I know 1+ Ghz is beefy and not realistic that's why my goal is to directly write it cheaply to some kind of onboard memory and process at later time.
It's not about real time processing, but it also about raw signal.

>>2405860
yeah, tore apart a used mini fridge, repaired it, sliced the fuck out of my fingers tearing it apart, only to find out it doesn't get cold enough. You need serious liquid-nitrogen cold for that kind of thing

>>2405740
>I'll wire up the smallest fuse I can find and give it a go.

dont be a scaredy cat.
7 ohms is normal for 3A transformer.

>>2405696
>cant knock out the other piece since its glued or something in there

ask your boyfriend to do it for you.
or drill through with a 1/4-inch drill bit and pull off any stands that come loose.
like sodomy, it gets easier once you get going.

>>2405867
AOL keyword "cloud chamber"

>>2405869
>drill through
Hit it with a blowtorch for 7 seconds.

>>2405869
>>2405873

started drilling it, but a blowtorch is also not a bad idea. And hey i've beat the shit out of it where I am afraid to damage the whole thing, some of the plates have started to warp and bend a bit and I think I scratched/nicked the primary (not sure if that will affect it)

>>2405881
>scratched/nicked the primary (not sure if that will affect it)
It will if it's shorted. Enamel the copper and measure the primary resistance.

>>2405869
>dont be a scaredy cat.

You don't know how scaredy I am. I hooked the bare primary up to 24vac from another transformer and it was drawing over 2 amps so figured the coils melted. Thought about it some more and wondered if not having the iron core in it would cause it to draw more amps. Googled a bit and it seems it might because of however magnets work. So now I'm reassembling the whole thing to test again.

>>2405099
what

>>2405903
You just gotta shove enough current through it to matter, mate. Push harder with more volts. In short: try hard enough.

>>2405913
damn. i was hoping there was a solution to this problem that didn't involve buying $600 worth of lipo batteries. i guess everything cool is either illegal or expensive

>>2405914
There's a reason the hacksmith had to make a Youtube channel and get paid sponsorships to be able to make all the crap he does.

>>2405915
i will make an onlyfans and post racist memes on it, along with electrical and 3d printing projects

>>2405896
>>2405869
>>2405705
>>2405683
Update, ended up replacing the first 10 feet of winding and the thermal fuse. Got it put back together and welded up. I think it has a bit more vibration than it use to and draws 1 amp unloaded (don't know if that's normal) but is working so I'm happy.

>>2405948
how do you just "replace" winding? that shit is incredibly specific and you know full well that only advanced factories can make motors. you are shitposting now

>>2405967
Huh? I unwound the first layer of the primary which melted to the thermal fuse, cut it off, took some lacquer coated wire that was a little bigger, soldered it to the original wire, put a small piece of shrink tube over the joint and then wound the same number of turns I took off. It's not a big brained operation.

>>2405972
what are you even talking about

>>2405975
If you have no idea what I'm talking about then how would you even know if I'm shitposting?

>>2405975
look under there

>>2405863
Easy, just find a 2GHz ADC and hook it up directly to an EZ-USB FX3. Then watch as your $6000 ADC gets swamped by local AM radio signals, mains hum, and chinky SMPS appliances. If you spend a while making notch filters or can get your signal sufficiently isolated from noise then it might work, assuming such high bitrate ADCs are even available.

Do you think it's realistically possible to turn a salvaged laptop screen into an external monitor for an arcade cabinet?

>>2406116
You could if you can find a driver for it but it'd cost more than a used monitor off CL so there's little point.

>>2406123
Main issue is not really cost as much as size. I have a spare 22 inch monitor but I want to do a smaller bartop thing, but only if I can get away with this. What do you mean a driver would cost more?

>>2406136
Already have the laptop as well, it's an ancient piece of garbage besides the screen, so to me it's free.

>>2406116
So long as you can find a driver for it, definitely. The driver boards just have an HDMI port on them, probably a power socket of some kind too.

>>2406140
Oh okay that makes sense, wasn't sure if you meant driver board or software driver. I suppose it would be impossible to just directly wire up the wires from the laptop myself to an hdmi / dvi cable.

>>2406144
Yeah they do need proper data conversion. There's a surprising number of HDMI LCD drivers for sale on eBay and the like, though more for obscure LCD panels you might be out of luck. DIY Perks did a video about making monitors from laptop LCDs.

I'm half of the mind that laptop/tablet-made plug-in LCDs are more convenient than a tablet for watching movies, since they're cheap and lightweight, with no battery to worry about. Could even put one on one of those articulated lamp stands. Just have one of those remote control mice with a keyboard on the back, or a handheld trackball mouse.
Well I'm also the kinda guy to buy a cordless tool without the battery and use it with a battery adapter so I can plug it into the wall. Can't get brushless corded tools unless it's a 10lb Hikoki demolition hammer

>>2406148
Interesting, thank you. I'll check how much a driver would run me.

>>2405863
Some SDRs can capture raw IF data and if you're looking at signals from antennas why do you need to look at other part of the spectrum? You're basically asking for 2 GHz scope at minimum which start at 5 digit $ range. Even if you need only a couple of seconds you still need 2 GHz ADC, which alone costs more that $1k. If you want lots of bandwidth you're going to pay for it, so ask yourself again if you really need it. What are you even trying to see in analog signal? All you'll get out of such system would be a blob of noise.

>>2406028
>>2406202
My goal is software FFT processing, extracting various bands all at once on high-end GPU CUDA or GPU farms.

ADC is indeed almost non-existant(?) for such high frequencies, but if i am certain recording of the signal could be done purely analog form, without any conversion.
Thing is i am probably looking at some 100-60 years old tech via physical tapes or various high-speed signal imprinting via laser/light(?) nowadays such things are extremely rare to even research on but they have the benefits on extreme precision because they're heavily relying on pure analog/physical methods, i know i said "memory chip" but i just wasn't surely what is nowadays available when it comes high-speed (above GHz) analog recording, ADC seems like downgrade if i don't need real-time processing.

>>2406204
That kind of bandwidth for analog storage is basically nonexistent, though you may be able to attempt to store analog values on big digital backup tape systems. Make your own microwave frequency delta-sigma modulator or something, idk.

Oscilloscopes do go into the >1GS/s range, so there do exist ADCs that go up that far, but they are going to be expensive. You can also just use a DSO as a datalogger, though you'd need a pretty high-spec one to be able to continuously write high-bitrate data via USB3.0 or SD card. If the total amount of data is small enough you could write it to a DRAM or maybe flash IC, and offload it to something more convenient after the acquisition.

Can anyone who's familiar with PLC's redirect me to some resources for self learning? Interested in getting into the industrial electrician field and want to get a head start on this stuff if possible.

>>2406207
Ah well ADC really seem to introduce lots of noise because of their periodic sampling method, i was afraid that analog also only goes to certain freq because of actual tape/whatever being the limit.
I'll keep looking though, analog recording IMO has shitload of potential because you can run precise ADC on top of the recording later with way lower frequency just so you get most precise input for computing.

>>2406218
Ask the SQT too, PLCs are more industrial than /ohm/ usually caters to so you may have more luck elsewhere.

Someone, please answer >>2405362 about >>2405353's air in/out design.

I've decided to make a milliohm meter, using an MCU and some 7-seg displays because I have them lying around. I want to protect the MCU from seeing too high of a voltage across it (as will happen when open-circuit) so I need some sort of diode to prevent overvolting the MCU.
The MCU will be running at 5V but with an ADC reference voltage of 0.55V. There will be 537mA through the load at one range, and 0.537mA at the other range. So whatever diode I use has to be able to handle that high current, but with minimal forward leakage to not interfere with the low current (i.e. significantly below 52.4µA). My gut tells me that nothing will come close to that, but actually finding a forward leakage current value for a non-signal diode has been difficult. Picrel is from the 1N4148, which can't handle enough current to be used for this. I guess I'll probably have to measure it myself.

I'd use three 1N4004 diode drops to give 2.1V or so before they short, which should be plenty of room.

>>2406256
A: it probably doesn't matter for balance purposes
B: those look like really tiny channels for cooling compared to the axial fans of even small motors
C: are you familiar with the biot-savart law?

>>2406260
>A: it probably doesn't matter for balance purposes
k
>B: those look like really tiny channels for cooling
drill lots of them around its circumference.
>compared to the axial fans of even small motors
impellor blades are on the rotating shaft.
can't waste axial-face on fans, nor get its rotation to the device's outside.
>C: are you familiar with the biot-savart law?
left to right field will look like pic
the shaft inside the bearings (magnets) rotates, not the coil.

Magnets & coil are as butt-end view on pic here >>2404526, except surrounding coil with the rotating arc-magnets without using brushes.
This way the magnet field is converging instead of diverging.

>>2406260
>>2406293
just to clarify
The many impellor (B] blades (around the rotating circumference) scoops air down their own hole. directing air-flow to the space between the coil and magnet.

I have a two coil latching relay with "nominal coil voltage" of 3v. This just means that I need to apply 3v to its coils to operate right? No need to add a resistor if i have exact that voltage

Here is its pin out. It has two coils between 1-12 and 6-7. Applying 3v on pin 1-12 will connect 3-4 and 10-9 and applying 3v on 6-7 will connect 4-5 and 8-9 right?

Datasheet if needed, i have txs2 lt 3v

Also what do you guys suggest? Distance between components's legs is smaller than breadboard. I dont want to solder it yet

>>2406338
Forgot pic

>>2406338
>>2406340
Breakout board with pin headers to match the breadboard.
>I dont want to solder it yet
Then just look at it.

>>2406338
>>2406340
Use wire wrap ?

>>2406338

easy peasy: dead bug wiring.

This frightens the European.

>>2406459
they're in every house here m8

my microcontroller (pic10f202) claims it can output 25ma on an IO pin but I am only getting ~5ma. Why would that be?

I wonder if it is because I am using 2.4V to power the microcontroller and 25ma is for maximum allowed input voltage (input voltage range is 2.0V to 5.5V)

https://ww1.microchip.com/downloads/aemDocuments/documents/OTH/ProductDocuments/DataSheets/40001239F.pdf

I am this anon >>2406334 and it can't drive the switch. I am using a transistor to drive my switch now (base is connected to my MC and transistor supplies switch using my battery)

>>2406519
>I wonder if it is because I am using 2.4V

Probably. And in general Absolute Maximums are shit you avoid. Look at figures 13-8 for output low and 13-10 for output high (pic related); that's what you work with, and 5 mA looks not too far off for VoH about 2.4v.

The mathematical designation of the forward voltage drop in a diode is V_FM, but what does the M stand for?

>>2406587

maximum?

https://file.elecfans.com/web1/M00/20/B0/oYYBAFmkyIGAJZVfAB5QmcdLswg641.pdf

>>2406587
Mastodon

>>2406595
Thanks that maxes sense

>>2406598
pic related is (you)

>>2406604
no u

>>2406340
didn't your breadboard come with jumpers?

what can you actually do once you learn electronics on breadboards and whatnot?

this isn't meant to be a put-down or anything, i'm just genuinely curious as to what you can do with basic circuit knowledge. can you actually fix appliances and shit, or is it all proprietary stuff anyway? how about modifying existing stuff? i guess for repairs it's a standardized part question. how about just "building" usable stuff though? like say if i wanted to make a weird homemade battery pack for a drill or some other mad max tier wacky idea. isn't this kind of thing just for big companies to do? what's really within the limits of a /diy/tard in his house with a soldering iron and a box of parts?

>>2406631
>what's really within the limits of a /diy/tard in his house with a soldering iron and a box of parts?
It's easier to say what he can't do.
1. Fabricate silicon chips. (I expect it costs a lot of money to get a foundry to fab a part of your design.)
2. Work with very small components.
3. Update the firmware of something with good security.

>can you actually fix appliances and shit, or is it all proprietary stuff anyway?
If there's a failed component, you should be able to diagnose and replace.

>how about modifying existing stuff?
Sure.
Get hold of the signals you need.
Produce the output you want to produce.

>how about just "building" usable stuff though?
Great Scott has a good playlist called DIY or Buy where he makes something and compares it against a mass market product.
https://www.youtube.com/playlist?list=PLAROrg3NQn7e3GQlBhuE_TIde0eJZHuzt

>if i wanted to make a weird homemade battery pack for a drill or some other mad max tier wacky idea. isn't this kind of thing just for big companies to do?
Doable.

>>2406644
>1. Fabricate silicon chips. (I expect it costs a lot of money to get a foundry to fab a part of your design.)
>2. Work with very small components.
>3. Update the firmware of something with good security.

those are remarkably wide limits. so it's true what they say, you can legit just *make* pretty much whatever you want within reason

>>2406622
Its legs were too thin for my female jumper cables.

I figured out something temp for now but i will probably solder it some fitting pins

>>2406657
Probably some of the widest limits out of any diy hobby, probably explains why it’s so popular.

Also these days it only costs $50 per IC or so to get custom silicon fabbed, though I think you have to do it in bulk. I recall someone collecting a bunch of silicon designs to put on a single wafer to bring the price down, but I can’t remember whether $50 each was their price or the bulk price.

You can solder some pretty fine parts manually, though it does require more expensive equipment.

And look at what hardware hackers do to glitch closed firmware systems into submission, oddly fascinating work.

>>2406766
nice. i'm sure i'll come up with something i need eventually, or at least something i would want to build just to learn how to build it. first i need to learn the basics though. i've been "modding" nerf guns (rewiring and now shoving RC car battery connectors where they don't belong) to get a hang of soldering, next up is the chinkshit amazon breadboarding kit. gonna be interesting at bare minimum.

>>2406260
>directly measuring the diff voltage with the MCU
it will be much easier to do with a current sense amp and it's i_bias will almost definitely be lower than an ADC input (unless the ADC already has an amp in it too). Power the amp off the same rails as your current source and that takes care of protection for that section
for the MCU you can use a resistor and zeners or diode clamps to the rails, see picrel.

>>2406334
check the typical coil resistance. if that divided by the nominal voltage is around the nominal current then yes you can just apply a voltage directly.

>>2406519
>ABSOLUTE maximum ratings
anon that is the max current that can be forced through that pin without the pad driver literally melting or latching up
yes you absolutely should use a FET or transistor. Don't forget the flyback diode or your MCU will probably reset every time it turns off the relay (I've seen it happen, it is very annoying...).
Also, since you are switching I presume high power signals, you have decoupling caps on your MCU supplies, right? ... r-right??

>>2404526
Double check my chinesium, anons.
I have a strip of 12V LEDs I want to power off 4 18650s in series, which will be charged with USB at 1A. I got a "4S1A USB C LiFePO battery charger", a 4S 8A BMS, and a 12V1A Buck Converter.

I figure I hook the USB to the BMS, the BMS to the battery, the battery to the buck, and the buck to the LEDs, as shown in my shitty drawing. Is this right? Am I missing anything? The buck only having a single ground concerns me. And desu I'm not even sure if the USB needs to be a dedicated board like I got or if I should just get a standalone port to hook into the BMS

>>2406800
Use a 19V laptop brick and matching barrel jack. Step it down to 16.8V or whatever with a ~5A CC/CV buck converter and feed that to the BMS and the rest. Then you can have at least 500mA charge current per cell.

>>2406800
uhhh what's the current on those LEDs anon
if it's more then let's say 30mA you may want to consider using a regulator that can do constant current rather than constant voltage, and ditch the resistor entirely. The buck having one GND isn't unreasonable but it probably means the entire control loop and stuff is internal (I assume it's a module or one of those drop-in 78xx replacement ones).

as to the BMS and charger, just do what the datasheets say and you'll probably be fine (assuming the BMS incorporates protection FETs to prevent overvoltage). But make sure to balance the batteries beforehand - charge them all to within the same voltage of each other with a CV/CC power supply (read battery DS for recommended charge voltage) then connect them in parallel with resistors overnight to balance the cell voltages before you integrate them into the pack. The overnight part might be overkill with newer BMS chips but might as well get them as balanced as you can beforehand.

>>2406807
1A. I guess technically it wouldn't be 12V, since that's for the whole 6m roll and I'd only be using 1m at most.

Also I didn't get datasheets, they're chinesium.

>>2406807
>>2406810
Also the resistor is built into the strip. All I'm wanting to do is take enough 18650s to power the strip at decent brightness and charge via USB, but idk if I want to wait on Aliexpress parts again. Too long a wait for being unsure of what I'm doing

>>2406812
You could get 20V USB-C PD trigger boards & PSU from Amazon. Then buck>BMS>LED driver.

>>2406815
Hmm, was hoping to be able to hand this off on its own to be usable with a wall wart.

Also what would that USB-C PD gain me over the setup I currently have? I already have my cells, and I'd still need the BMS and buck

>>2406810
>chinesium BMS with no DS
>chinesium charger with no DS
>1 motherfucking amp LED strip
OK, so first of all when you apply power to this circuit, you may want to do so outdoors away from other flammable materials. The 18650's aren't chinesium, r-right?

As to the strip, I would definitely consider constant current. With constant voltage being sourced to the strip small variations in output voltage due to self-heating, etc could cause wide swings in current, possibly enough to overload the strip, which if this is from China probably means it'll unironically light on fire. If your control loop is for current rather than voltage, you'll have way more control and can even change the brightness to some extent. A hack if you don't want to go full switched mode CC(P) supply is to keep your buck reg then have an LDO with a huge heat sink to drop those last few 100mV to keep the current stable. You can use a small current sense resistor for the LDO ref/adj pin instead of a voltage divider, many LDO's have in their DS (get one with a DS) example circuits for doing this.

>>2406819
>what would that USB-C PD gain me over the setup I currently have?
Up to 100W of usable power depending on the trigger board and power supply. Another option is to just use a 12V power bank and forego all the other stuff.

>>2406820
the cells are Burgerland, maybe Canukistani. Also as I'm a noob, how do I pick out an LDO? What specs will I need to be looking out for? I also would assume I'll need to get a different buck for this?
As you can tell, I'm noob enough to make a cuban look like a nasa engineer

>>2406793
Well if I can use the 2nd unit of my TL072 that could be nice, not that I think I’ll be able to.

Also the 200mV panel meter is like $10 so I may go for it instead. Even though I spent 2 hours yesterday designing a 7-segment display PCB to use up surplus THT shift registers. I’ll check it’s datasheet for input current.

So a differential amplifier will allow me to run it off a single supply, which would mean I could run it off a 12V SMPS instead of having to shove a multitap transformer in there. Yes, I’ll use chokes and capacitance multipliers before my linear regulators.

Apparently these things are good for hunting for short circuits.

>>2406826
>What specs
Max output current >1A
Max voltage > [the buck output]
Adjustable output voltage (many regulators are fixed V)
Dropout voltage < [buck voltage - LED voltage], or as low as possible

You'd probably want to get a TO220-3 footprint one and put it in a breadboard if you did it this way.

Yeah you'd probably need slightly higher than 12V for the buck to do that. Honestly if you don't already have the buck reg and there's a CV/CC one in module form (or even better a specialized LED driver module) that may be the easiest thing to do and switching only is the most efficient way to do it anyways. A straight-up all-linear current regulator doomping up to 4.8V of battery voltage to get down to 12V could work too and will be dead simple but it will get really hot, like you will need a heatsink & maybe even a fan hot.

>>2406841
Seeming as this will be in a wooden box, I'll try to avoid doomping. But anyway I'll just search Amazon for "Low Dropout Regulator 1A" or something, right? And then another search for a buck that'll take in 14-18V and can put out whatever the LDO needs?

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

is there any particular reason he used an LM317 in this? it seems like a costly and wasteful choice given that the IC dissipates excess voltage as heat, and the voltage input is a 9v battery. there's not much current to work with there, it's probably not wise to waste it. also why not add an LCD voltmeter in? i got a pack of them for like $7 on amazon. seems better than guessing which switches are on or off

>>2406848
>is there any particular reason
He saw it in the May 1982 issue of Popular Electronics after his wife's son's grandpappy kicked the bucket.

>>2406848
idk why that guy did it but I personally HATE switched mode power supplies and buck converters
I build everything linear power supply with good old fashioned voltage regulators

>>2406837
>if I can use the 2nd unit of my TL072
Is the first one used for the 537mA source? You might be able to exploit that to somehow get a pseudo-differential measurement, I can't immediately see a way but it might be possible. 4-wire measurement is probably ideal if you have any other amps lying around.

>these things are good for hunting for short circuits
they are also really good for hunting down inefficiency in power circuits
when you are dealing with big currents finding where the mOhms are burning away your efficiency can be useful

7-seg display would be pretty cool, hope you find a use for it

>>2406847
For a standalone LDO you'd probably want to search Digikey, Mouser, etc (component distributors)
If you are going the module route you may be able to find it on Amazon but Adafruit/Sparkfun/etc will likely be less shitty & have a datasheet. Actually, I wouldn't be surprised if one of those just has a dedicated constant current LED driver module so that might be a good first place to look.

>>2406863
>Adafruit/Sparkfun
Those cunts give money to terrorist organizations. Never got a dime from me, never will. Overpriced like crazy anyway.

Just bought the panel meter because it's cheap and I'll use it either way. It calls itself a CX102B, and apparently has 10MΩ input impedance. That's more than good enough for me. 86dB CMRR, ±0.1%, 100ppm/°C, and measures 200mV max with a display up to 2000 count. Not sure how much I trust it though. Also this B version requires a 10k resistor from in- to 0VDC, meaning it

>>2406857
>Is the first one used for the 537mA source
Yes. With a switch between two resistances for 2 ranges. But now that I'm redesigning it I may use a relay and have it be autoranging. I think if I use a BJT base-emitter junctions in addition to normal clamping diodes I'll be able to measure when I need to switch up a range. But I haven't selected a BJT that can handle that current, nor a good one for the current source.
Speaking of clamping diodes, at work I put 12µA through a 3A rectifier diode (with a 12V source and a 1M resistor) and got 0.26V across the rectifier itself, which I'll try to match or better for the clamping setup.
>You might be able to exploit that to somehow get a pseudo-differential measurement, I can't immediately see a way but it might be possible.
I very much doubt it, but I'll mess about with the concept a tad. Using two SOIC-8s isn't the end of the world.
>4-wire measurement is probably ideal if you have any other amps lying around
I thought I was already doing that. Or do you mean some sort of extra differential voltage measurement?

>7-seg display would be pretty cool, hope you find a use for it
Well it's just a bunch of those 5161AS 10-pin single-digit displays. Getting 4 of them to make a 1023-count display took three 74HC595s, and since those were surplus THTs they took up way too much room. FYI, that particular combination is a pain to route traces for, would be even worse if I multiplexed them.
I'm on the lookout for smaller SMT 7-segs with more digits, but it's hard to get small ones. I'd definitely consider just going for an OLED or LCD instead at this point.

>>2406865
Same sentiment here. I'd rather set money on fire than give it to them.

>>2406892
Ok here's a design for the mΩmeter, one that uses the panel meter. But there are a few things I'm not too keen on. The two-op-amp differential amplifier instead of a 3-op-amp instrumentation amplifier circuit isn't wonderful, also the damn TL072s are causing me to put a zener there to keep in their common-mode-range. This will result in a 12V power supply with only the inner 4V being usable. Having to add a virtual ground circuit to bump the panel meter somewhere the op-amps can output to is equally shitty feeling. That BJT circuit should work for detecting overflows, and maybe that can be used to energise a relay to change the range.

But now I'm thinking about that MCP3911 I happen to have lying about...
A 14-pin micro isn't enough to multiplex a 5-digit display, nor is anything I have high enough accuracy to warrant such a treatment, but it certainly is tempting.

>>2406793
>Also, since you are switching I presume high power signals, you have decoupling caps on your MCU supplies, right?
It is not really high power, i am using pic to turn on/off an arduino. I tried transistor/fet solutions but my 5v step up converter acts really weird if i connect it to a transistor

Yea i read about that voltage spike of coils when they are turned off and about adding a diode, i will do that

Can someone explain this circuit to me?

It's from a unit that measures power factor. The three phases are measured to ground through op amps and sent to an ADC.

Then there's three op amps set up as difference amplifiers which measure phase A to B, B to C, and C to A.

The outputs of these difference amplifiers go into three circuits like pic related (That's a comparator, not an op amp).
Is it for detecting zero crossing or something?

>>2407045
what an odd way of doing positive feedback

>>2407045
So, If I power it up and connect live to one of the Phase inputs, I get a 50HZ square wave coming out of the circuit (Euro fag)

got a bunch of these guys. are they any good?
What differantiates them from modern amplifier ICs?

>>2406855
is there a reason for it? space constraints, maybe?
>>2407068
tf is that

>>2407068
It's a low-power (5.8W) car radio amp IC. Do with that what you will.

anyone here built a joule thief yet? what did you use it for?

>>2406922
Wait I forgot about 7905s, which would save me an op-amp
but are they actually good for this? I'll have to test what they're like on an SMPS before and after a capacitance multiplier, in comparison to an op-amp regulator or TL431 regulator. Think I'll keep the current sink and zener as-is though, a TL431 wouldn't act too nicely with them.
That said, the base-emitter current will mess with me, so I may need to construct a current mirror or hope a MOSFET will just work.

>>2407045
>>2407047
Well the inverter acts as a buffer, so you can ignore the impedance of the pullup resistor when designing the RC part. Not sure what the RC part actually does, but I think the C will ensure a 50% duty-cycle while the voltage divider Rs give hysteresis.

>>2407221
I prefer to design my own switching power supplies, ones that don't require that I wind my own transformers.

Do I need to store active components or ICs in anti static bags or is it bag jews scaring people to buy them?

>>2407247
Only for shipping. Store them in ESD safe storage containers if you're worried about it.

>>2407247
ESD damage isn’t common, so you can likely get away without the bags and never notice a problem. So long as you don’t live in a really dry climate with wool socks on. Though I’d at least keep expensive things like MCUs in antistatic bags or foam.

>>2407234
>I prefer to design my own switching power supplies
damn. that's awful technical isn't it

>>2407293
He's a 2000 man. His kids don't understand him at all.

why are there capacitors in literally every circuit? there's always one or two capacitors in both negative and positive leads. is there a reason for it or is it just so i have to buy more electrical components

>>2407307
Big Capacitor is run by Yakuza.

>>2407307
its because power supplies aren't perfectly stable and often introduce fluctuations or spikes that could be harmful to the steady operation of your circuit. These "filter capacitors" are seen in almost every circuit to hopefully absorb and neutralize this noise
Even if you're using a pretty stable voltage source like a battery, the nearly instantaneous loading characteristics of a lot of devices can cause temporary voltage drop at their supply connection

More filter caps = more stable

>>2407307
>is there a reason for it or is it just so i have to buy more electrical components

I think you are on to something, but you will probably get higher quality answers if you start a thread in /pol/

>>2407307
>is there a reason for it

you'll find that if you build 100 identical circuits, 99 will work just fine, but one will do stupid shit unless you add a specific cap at some specific point.
since you dont know which one is gonna go stupid, you put caps on all of them.
thus you keep your job, and you arent mocked by your fellow engineers.

>>2407307
>fluctuations or spikes that could be harmful to the steady operation of your circuit
Capacitors keep the riffraff out of the nice neighborhoods.

How did they obtain sinc at the last step?

>>2407370
e^ix = cos(x)+i.sin(x)
e^-ix = cos(x)+i.sin(-x) = cos(x)-i.sin(x)
e^ix-e^-ix = 2i.sin(x)

>>2407311
>>2407313
i'm a /pol/ native lmao
>>2407312
makes sense. so it's just storing up a tiny amount of power to level out any kind of power fluctuation.

>>2407378
awesome, thanks.

>>2407293
I copy reference designs. I pick 2 or 3 different ICs and inductor sizes and a handful of power FETs which I can mix and match for certain situations. The built-in gate drivers are kinda ass though so for high-power shit I end up adding my own, which also happens to be an easy way of making a converter use synchronous rectification since my gate drivers are all half-bridges. Also any old boost controller IC can do SEPIC, so no need for fancy buck-boost circuits.

But anything isolated? Anything that wants a transformer at all? Fuck that. I’ve dabbled in resonant LC transformer drivers but they’re a total pain to get working properly. Flybacks need specialty transformers that have an air gap to reduce their coupling efficiency so they can store energy like an inductor. Until I have a way of measuring and simulating that I’ll be assuming they’re black magic and will avoid them to the best of my ability. Forward converters are a different kind of arcane, which I should probably read into.

ZVS drivers are interesting, but they get complicated if you want actual regulation or feedback. Might still be good for driving EL wire.

>>2407419
Oh and synchronous rectification on a SEPIC is some strange shit. There’s like one forum post linking to a TI reference design on the topic, which probably works, but I don’t know WHY it works. Also his SEPIC (like many others) uses a transformer and I haven’t yet decided whether transformer SEPICs are worth attempting. I briefly looked about for info on the advantages on doing so (compared to having to source an arbitrary transformer ratio) but came up blank on the specifics.

Also I go for switching controllers with a wide frequency range, makes it easier to match with my inductors. I think current-mode converters are the way to go, though I should probably have both.

Bros, look at this fft of a 1khz 3v probe calibration square wave in the SD1202X-E.

>>2407307
the better question is why capacitors are much more widespread than inductors even though they are the dual of each other and either of them should do the job.

>>2407452
Because inductors need to be bigger/are more expensive? I think that's what I heard being mentioned in passing in a video about filters.

>>2407452
Capacitors buffer voltage, inductors buffer current. Modern electronics run constant voltage rails, not constant current rails. If you used an inductor as a filter before an H-bridge, you’d fry the H-bridge with inductive spikes. Ensuring that a constant current flows into a microcontroller isn’t good for it, it’s a surefire way to kill it.

Where do I learn to understand and work with low voltage relays like in pic related?
I have a solid grasp on electronics, but do not know much of anything about this.

>>2407565
Do you understand how electromagnets work? Because if not I'd start with electromagnets. The wikipedia page should be good, but there's also Applied Science's video on magnetic circuits. A relay just opens and closes a mechanical switch using an electromagnet.

>>2407565
Youtube, Mehdi has a video on them

>>2404526
Hello can I use a BC548 instead of the BC337 for this single transistor class A amp?
https://www.instructables.com/Building-the-Simplest-Best-Sounding-Yet-Most-Ineff/

>>2407638
So long as the collector current is less than 100mA it should be fine. The 337 is rated at 800mA maximum while the 548 can only handle 100mA. hFE may also be different, but they're more dependant of what batch you've got (548A, B, or C) than on whether you've got a 548 or a 337.

Once that project is finished you should learn how to use op-amps. They're more linear than plain old BJTs and very useful for audio circuits. BJTs still have their place, but they can be pretty arcane to fully understand their implementations, op-amps are a lot more beginner friendly.

>>2407656
I already build it on breadboard but I just used the specified values without thinking and doing my own calculations, so I'll redo it. Though it is a bit louder connecting my headphones to it.
I found this website which has a good explanation: https://learnabout-electronics.org/Amplifiers/amplifiers20.php
I would have thought bjts to be more simple than op amps. I guess it's just the appearance of the packaging that puts me off.

>>2407656
But thanks for the suggestion though

What price should I expect to pay for photo resistors, in bulk?
For my senior design/bachelor's thesis project, I want to make a rudimentary 128*128 pixel camera out of a 128*128 grid of photo resistors

However I can't find any for less than 15¢ each which is still $2457. I really like the idea, but if I can't find them cheaper the idea might be DoA.

>>2407763
You will have to deal with chinaman directly on alibaba, considering you need to buy 16k pieces you might haggle your way below 1¢. But why would you want to torture yourself with this? Do you realize that you will have to solder 32k connections? That's like several kilos worth of components.

>>2407763
>For my senior design/bachelor's thesis project
I would reject your thesis and flunk you so hard in the a.

>>2407763
>2407763
Not what you're asking about, but the least expensive photo diode at DigiKey is $0.09 in bulk (3k pc).

>>2407763
how do you intend to solder that? you'd probably be talking another thousand in solder. you'd kill yourself with lead poisoning before you finished the project

it's a really cool idea, i just don't think it would be a really feasible thing. maybe go for a 64x64, that would still be a massively impressive project but it would be a lot more practical and less expensive.

>>2407763
That is retardedly infeasible

>>2407763
Here is an idea: use 128 diodes but slide it using a motor take 128 of those slices to make a 128x128 image.

>>2407774
>>2407779
>>2407781
The labor is going to be divided between 3 people, and the plan is to get PCBs made that will vastly streamline the process. (Though 256
128x128 is because I wanted to have one of those little OLED displays on the back of the thing, and you would be able to see your picture in real time.
I should mention this is me doing a feasibility study. Nothing is set in stone, especially if this turns out to be wildly impossible.

>>2407777
I did manage to find some photoresistors for about 1.5¢ each

>>2407776
excuse me Mr. Helpdesk I think my mouse unplugged itself again chop chop I have work to do

>>2407782
A backup idea I had was to make a really shitty "CRT" display using a laser pointer and glow-in-the-dark paint which is in a similar vein. It's a backup idea because I have no idea how to move/aim the laser

>>2407783
>(Though 256
Though 256 PCBs isn't going to be cheap either

>>2407783
>Mr. Helpdesk
Sir, please do the needful and unplug the 128x128 photodiode for 1 hour and 15 minutes and plug it back in again.

>>2407783
So you're just going to copy
https://www.youtube.com/watch?v=PaXweP73NT4
4 times and call it a day?

I need a physical person-operated switch that I can reset with an electrical signal, what am I looking for

>>2407980
Gonna need more info there, Hoss.

>>2407980
Transistor?

>>2407986
>>2407985
Like a lightswitch type switch. But with maybe some kind of electromechanical relay built in that will actuate it. So say I come along, I flick the switch on with my hands. I can just flick it back off again with my hands, or I could signal the relay and it will flick it off for me.

I'm just wondering if there's some COTS product I can buy or do I need to make one myself, it seems like the kind of thing that would exist.

>>2407992
https://www.digikey.com/catalog/en/partgroup/et-series/26208

https://www.farnell.com/datasheets/2145064.pdf

>>2407997
Ah yes, this is exactly what I'm talking about
Thank you my friend

>>2408003
np

>>2407452
>>2407533
It should be noted that inductors (chokes) are commonly used to stop conducted EMI between devices, sometimes even on the same PCB if you have digital and analog chips on the same power rail.

In an audio amplifier circuit board I have there's a large electrolytic cap filtering ripple out of a rectified power supply. Will I have improved filtering if I add a ceramic cap in parallel with the electrolytic? Should this be standard practice?

>>2407763
>>2407783
If you use SMT photodiodes (or even LEDs) you might be able to get LCSC to fab the boards for you.

>>2408021
Chokes are not simple inductors, they are designed to be extremely lossy (low Q). You don't get inductive spikes to anywhere near the same extent, such that a capacitor after it will usually buffer any resultant ripple without issue. I think the resulting LC filter is (ideally) critically damped, if not significantly overdamped.

>>2407565
The data sheet.

>>2408030
>Chokes are not simple inductors, they are designed to be extremely lossy (low Q)
due to thin wire with high resistance? why?

>>2408060
core permeability

>>2408060
Loss in the core, not in the wire itself. Like how an inductor with a solid copper core has very little DC resistance, but is very lossy when you put AC through it due to eddy currents.

>>2408075
hmm, but Q = wfL/R. wouldn't increasing the frequency also increase Q? R can't be a function of f, can it?

>>2408085
*wL not wfL

>>2408085
We're talking about coupling of an inductor's magnetic field to lossy elements (basically a secondary winding) so I don't think a simple inductor's Q factor equation fully applies.

Can someone confirm if I understand leds correct.

Say my supply voltage is 2.4v and I have a red led with 2.0v voltage drop and rated safe for 15mA. I need a resistor that gives 15mA for 0.4V right? So in this example it is ~27ohm

If I had 3.3v supply then it is (3.3 - 2)V / 15ma = ~86ohm

/diy/ claims another one

>>2407992
you can use a dual coil latching relay. a button goes in one coil to turn it on, and your electronic signal goes to other

>>2408103
YAY! Welcome to the club! Endless fun guaranteed.

>>2408102
The series resistor limits current, and excess is dissipated thermally.
It just looks wrong on the negative side. Like a man wearing a lady's dress.

>>2408109
>Endless fun guaranteed.
yes, at least once my extra parts get here. i don't think there's much i can do at the moment with a box of resistors and LEDs, but once my LM317s get here i'm gonna make a shitty and wasteful portable power supply

>>2405353
>>2405347
>super compact efficient generator
>>2406293 >>2406308
>Magnets & coil are as butt-end view on pic here >>2404526,
>except surrounding coil with the rotating arc-magnets without using brushes.
>This way the magnet field is converging instead of diverging.
Here's the axial view.

Any suggestions?

>>2408134
I wonder what pumping a ferrofluid through the core would do (or some exotic nanoparticle solution). Impeller, heat, or EM pump?

Got my oscilloscope shitting out properly formatted HPGL into some software CERN wrote like a decade ago and somehow it just works. Mostly. That bottom line is supposed to say:
>P_10x 5V
>0.5V
>A=0.5µs
Where the _10x means it's in subscript (all done as a pair of user characters).
Also there's meant to be a ∆ before the V1 at the top, which is also a custom character.

I initially wanted to write an HPGL reading python script from scratch to get around those issues, but that was a massive waste of time, so instead I'll just try to mess about with this software to figure out what kind of text I can actually feed it. Hopefully I can get the custom characters to work as a general case with some small modifications, else I'll need to just hunt through the filestring and replace all excerpts of user characters with something good enough from ASCII or possible alternative character sets. Hope there's a greek one.

>>2408137
>I wonder what pumping a ferrofluid through the core would do (or some exotic nanoparticle solution)
>core?
coil?
>Impeller, heat, or EM pump?
?
Not sure what you mean, but I guess its an EM Pump.
Alternator coil & rotating-magnets are re-designed to produce electrical output for the full 360deg rotation duty-cycle;
Instead of just the split-second every time coil + magnets line up.
This coil has a far greater fill-factor, and doesn't waste the volume between coil-windings.

I can easily get this config to be 3-phase, but I'm not sure if its worthwhile as these versions do not include my speed regulator inclusions.
>Impeller, heat
I will have both left+right rings of >>2405353 impellor [B] drawing air in.

I have some smt caps in these two little strips. No idea what the value is. The only marking available is shown in pic related. I want to solder a bunch of them together and use them somewhere.

>>2408027
>Will I have improved filtering if I add a ceramic cap in parallel with the electrolytic?
In theory yes. The reason is the ceramic cap will most likely have a higher self-resonant frequency than the electrolytic, and this would therefore result in a wider bandwidth of ripple that can be removed.

>Should this be standard practice?
I have heard at least two different EMC experts give different views on this topic.

One approach was, to use multiple sizes of caps with increasing SRF. Of course you will have to parallel more of the higher freq ones as they will be physically smaller and have lower values. This approach has the drawback of "resonances" where there are some sensitive frequencies due to different values of caps.

The second approach is to pick ONE size of cap that has sufficiently high SRF to cover your needs, and parallel as many of them as you need to achieve sufficient EMC performance (low enough ripple, sufficient transient response if at the load, etc). The drawback of this approach is you may need a shitload of caps depending on how high an SRF you're targeting.

I can't really comment on which is better since I haven't done any rigorous characterization of the effectiveness of either approach, but in practice I've seen the first used more. Rule of thumb I use is, big caps at the supplies, small ceramic caps at the ICs - honestly, if you need something more rigorous than that, you should probably be focusing on your measurement setup so you can accurately diagnose what part of your design is the cause of the noise or EMC issue.

>>2408181
Forgot pic

>>2408181
>>2408185
Measure with your DMM's capacitance setting. With any luck they're 100nF ceramics so you can use them for 99% of your circuits as bypass caps.
>inb4 shitty DMM

>>2408186
>inb4 shitty DMM
I have three multimeters. None of them has a capacitor tester. Maybe instead of buying another multimeter I'll build a capacitor tester based on an Arduino board. I saw one here that shows promise: https://www.circuitbasics.com/how-to-make-an-arduino-capacitance-meter/

>>2408196
>I have three multimeters. None of them has a capacitor tester
Must not be very good multimeters, that or they're specialised for particular purposes like some sort of automotive or HVAC DMM. Even the $30 AN8008/9 has one.

A capacitance tester is definitely a good project, and there are a lot of different ways to go. Resonance, time constant measurement, capacitive ramping or coulomb counting, etc. I'd look more at what the pros are doing and simplify those, instead of starting at arduino trash and trying to make it good. You'll be measuring possibly very small ceramics remember, so just just relying on the input impedance and capacitance of your microcontroller being negligible likely won't get you far. Though you can probably extrapolate a decent value from any measurement setup so long as you measure it with a few different resistances/currents, and have a sufficiently high SNR.

>>2408161
Well I hard coded the 5 symbols I could see during a cursory glance through past plots. I may be missing one for XY mode though, which I'd definitely want for getting plots of my curve tracer. I should also export all the datapoints as a CSV file of actual data while I'm at it.

FYI, only some text editors show ASCII character 0x03, and it is not the same as any other whitespace. Took a dozen frustrating minutes to figure out that one.

>>2408253

And now it's working in XY mode. Found a couple of extra user characters while I was at it, including "Hz" and "1/∆T". Also I managed to get the font size for the 10x to be smaller, without too much difficulty. Not sure if I like the gridlines though, be nice if I could change the pens for them to make them a different colour.

But the CSV data output is more important.

Someone qualified please feedback on this project design approach >>2408134.

I'd appreciate advice before continuing.

>>2408285
idk man, try simulating it with one of those magnetics cad packages
or in python or whatever

>>2408288
>Someone qualified
>idk man, try
sorry for that annoyance, anon.
I'm just looking for someone to say yay/nay my theory is sound.
Doing all that stuff you suggested is next, and I'd rather not waste my time learning new package if my theory is flawed, or missing possible improvements.
*my bad

>>2408293
>............jpg
Ok let me rephrase that less nicely. I've studied electromagnetism enough to be skeptical when you're wrapping a coil around the rotation axis. Without mincing words, your theory is complete bunk. You complain about the poor fill factor of Tom Stanton's amateur 3D printed alternator, but your design is nothing like a more professionally designed axial flux motor (pic related) but rather has a completely different geometry. FYI the coils in my pic should be filled with magnetically permeable material.

I tried to say this 2 or 3 times the first time you came up with this idea, but let me say it again. The magnetic flux density integrated across a coil's area needs to change as a function of time. Your coil is symmetric about the axis it's rotating in (from what I can tell from your simplistic 2D diagrams) meaning by definition the flux density won't change as it rotates. Adding the Halbach arrays like I suggested won't help that. If you still want to keep up the pretence of not understanding what I'm describing but won't 3D model it in something shitty like sketchup or tinkercad, or on the off chance that I'm actually wrong, I'll fucking 3D model some cross-sections of your design myself.

Then there's also the appeal to "engineers are smart" saying that if this kind of motor topology was smart we'd see it all the time. I don't usually rely on that kind of logic, but considering BLDCs are pretty ancient tech and the chinks have been refining the technology for drones and ebikes for the last decade or two, I'd say it holds at least some water.

>>2408299
>induction
N-S Field must be parallel to wire-segment.
>skeptical
>when you're wrapping a coil around the rotation axis.
>The magnetic flux density integrated across a coil's area needs to change as a function of time.
coil's area ? any which way? only your way?
Every single parallel axial-wound "series" wire arc-segment goes +NORTH to -SOUTH Change as a function of time with axial-rotation.
>Your coil is symmetric about the axis it's rotating in...
And its axial rotational magnetic N-S field exposure is parallel series to wire "arc-segment" length..... to the axial rotating alternating N-S field poles. .
>(from what I can tell from your simplistic 2D diagrams)
Sorry, my brain reads info + creates a 3D mental visualization construct without even thinking about it.
>saying that if this kind of motor topology was smart we'd see it all the time.
>motor
Alternator.
WE Do, it is exactly equivalent to what you're already doing (see pic).
Equivalent to coil-winding remaining still and rotating the 4 magnets axially, wire always parallel to the axially rotating N-S poles.

>>2408299 help with >>2408316??
Looking at these breadth views here >>2405347 >>2405353 may help visualize 3D render of device.

>>2408299
>motor
He's not making a motor, he's making an alternator.
>induction
Alternating N-S Field must be parallel to wire-segment.
It doesn't matter how you do so, as long as you obey the rule.

>>2408275
And now I've got CSV files, and they're working. The comments/titles/whatever are appearing in really strange locations though, I still don't understand this coordinate system.

>>2407733
>>2407656
How do I know the value of I E ?

realistically how hard is it to DIY a "high power " variable bench supply? let's say an offline SMPS 1 KW, 0-33V 0-33A. is this hard, extremely hard, or basically impossible unless you're an EE with a decade in power engineering? i realize it would be much cheaper and easier to just buy an old xantrex, sorensen, HP, etc. power supply off ebay but i want to learn this myself if it's even remotely possible.

>>2408363
Of course it's possible. The question you should ask yourself is whether you have the time, money, and patience to build it from the ground up. Anything that fails must fail safe or else.

>>2408341
I E = I C + I B
= (ß * I B) + I B
= I B (ß + 1)

Any resources on how to design/build your own custom transformers? Everything I find online is too basic and doesn't really get into the nuances.

>>2408373
Transformer and Inductor Design Handbook

>>2408218
>You'll be measuring possibly very small ceramics remember, so just just relying on the input impedance and capacitance of your microcontroller being negligible likely won't get you far
but that $8 meter project that is popular (but it is not $8 anymore) was originally running on some 8 bit micro and produced decent results measuring down to pf. no idea how they do that. but i am sure it is time constant based. that "arduino tester" is probably a spinoff of that project.

>>2408414
>was originally running on some 8 bit micro

this is faulty logic. the fact that a task can be accomplished by a given 8-bit micro hardly means it can be accomplished by *any* 8-bit micro.

>>2408414
>was originally running on some 8 bit micro

this is faulty logic. the fact that a task can be accomplished by a given 8-bit micro hardly means it can be accomplished by *any* 8-bit micro.
i bet the designers of that LCR meter spent between 8 mins and 8 months to find the bestest cheapest part for the job.

>>2408475
Don't be a nitpicking pickaninny, Anon. He's not optimizing the BOM for mass production.

>>2408477
>optimizing the BOM

you missed the point.
if you have a particular set of features you absolutely must have (like a 14 bit A/D, 1uA standby current, unregulated 3V supply) then you have to be very picky in your parts. it's not ''anything will do as long as it's 8-bit''.

>>2408414
>>2408218
The tutorial I listed does base the design on time constant measurement, and it does show good capacitance measurement down to the pF range. In any case, I'm not designing RF circuits or transmitters, so this is probably good enough for the time being. In audio, the less accuracy the better the sound ^:)

>>2408475
it is not about bitness, but rather the clock speed. all 8 bit micros are fairly slow, probably in the range of 1 to 8Mhz. even 20Mhz clock is probably not good enough to measure time constant in the pf range since ADC speed is even slower than that. so i have no idea how they did it.

>>2408414
>that $8 meter project that is popular
Those things have pretty damn clever code in them from what I remember. If you want to make something functional, trying to understand that design would get you a pretty good end result, with some good education along the way.

>>2408529
16MHz is reasonably fast, for a 10pF cap that's a resistance in the realm of 1kΩ. With 10kΩ you'd get 2 clock cycles for each time constant, with 100kΩ you'd get 20.
>ADC
Gotta use interrupts off the analog comparator instead of the ADC, which is what I assume that circuit is doing. Didn't see interrupts or ACs being discussed in that arduino sketch though, and it uses digitalWrite(), so you can tell it's not going to be very good.

I have an rgb led that says it has 3v front voltage for green and blue and 2v for red.

I assumed 3.6v would barely work since 3v fron voltage but even with high resistances (330 ohm) it was pretty bright. Then i tried 2.4v and it still works despite being under front voltage.

Are they lying?

>>2408713
>would higher temperatures force a greater percentage of solvent out
Which colours? I'd expect the red to light for sure, and probably the green too since they're usually closer to 2.5V than 3V, but not the blue. At least not a very bright blue.

If you're lucky they may have voltage/current curves for each LED on the datasheet.

>>2408718
Here green light, i connected 2.4volt and i am getting 200mV drop across resistor. That means i have 2.2v front voltage on the led right?

It is drawing only 0.6ma with that resistor (330ohm) and still pretty bright. This led is magic

>>2408718
>>2408723
Blue is pretty dim, voltage drop across resistor is 0.6v which leaves 1.8v to led. I dont get it

>>2408732

>>2408732
>>2408734
Is there a regulator IC integrated with the LED?

>>2408739
It is just a led

>>2408723
>>2408732
>>2408734
Forward voltage isn't a threshold, it's the voltage value of of an exponential curve at full forward current when under standard conditions. Pic related.

>>2408751
Thanks. I thought it was constant, some kind of threshold as you said

>>2408614
>10pF cap that's a resistance in the realm of 1kΩ
that's tc=100 picoseconds. atmega32 max ADC speed is 1Mhz and the max sampling speed to get the 10bit resolution is 15KsPs and you need at least 2 points per sample, so it is 7KsPs? no way you can measure anything even close to pf with these speeds.

could probably measure 2t or 3t if the ADC/comparator resolution is good enough. or could measure the time difference between the rising and falling edges at the tc voltage levels.

how am i supposed to figure out the resistors i need to use with a transistor? i'm trying to use a transistor in a simple push-button circuit to light up a red LED. i've got a bunch of PN2222 transistors and a good selection of resistors, i just can't figure out how to go from 5v in to the transistor to an LED without blowing anything. also i plan on having a push button if that makes any difference

https://www.researchgate.net/publication/332049933_Capacitance_Measurements_System_Using_RC_Circuit
this seems hilarious. maybe not.
this design measures down to whopping 1nF with a huge resistor on the order of 100Mom LOL
and they measure 0.5Vcc threshold for some reason, which is less than the TC 1-1/e, to make their life harder?

>>2408781
Yeah that's why you don't use the ADC. You use the dedicated internal analog comparator hardware connected to interrupts on a 16MHz timer.

>>2408799
The current you want to turn the LED on is presumably 20mA, so select the resistor such that V_LED + V_CE(@20mA) + V_resistor(@20mA) = V_CC = 5V. The value of V_CE may be ignorable, it's usually pretty low like 0.2V or so, gets lower with transistor saturation. Some datasheets have a curve for this, but not usually on small-signal transistors. Of course, V_resistor(@20mA) = 20mA * R. Assuming 0.2V from emitter to collector, and 2V for a red LED, we get (5V - 0.2V - 2V)/20mA = R_LED = 140Ω.

Now consider the minimum hFE of a PN2222, a glance at the datasheet suggests 50. Then divide your collector current by that minimum hFE and you get 400µA. Now you need a base resistor, such that V_BE + V_resistor(@400µA) = V_CC = 5V. Naturally, V_resistor(@400µA) = 400µA * R, and V_BE = 0.7V, so it follows that (5V - 0.7V)/400µA = R_base ≈ 10kΩ.

So we have the base resistor as 10kΩ and the LED resistor as something around 140Ω. 150Ω is the next highest standard value, so I'd go with that.
Adjust those equations depending on the current through the LED, the forward voltage of the LED (changes slightly with current), the minimum hFE of the transistor, etc.

>>2408829
What are they measuring time with, a stopwatch? And how low leakage does their instrumentation equipment have to be for 100M to ever make sense?

>>2408844
jesus christ, all that to turn on an LED?

>>2408844
> You use the dedicated internal analog comparator hardware connected to interrupts on a 16MHz timer.
Or if you have a better 32 bit MCU, in addition to being faster, you could use ADC with DMA, or ADC with an interrupt on a voltage threshold detection.
>What are they measuring time with, a stopwatch?

Yes, it is a very crude approach.
They pull the pin high and start the counter
and wait until ADC reaches 0.5max and then stop the counter
digitalWrite(7,HIGH);
t1 = micros();
while(analogRead(1) < 2048){}//ADC=2048 equal to 0,5VS
t2 = micros();
digitalWrite(7,LOW);

seems a horrible approach overall.
But I think setting the Vthreshold to 0.5VDD which corresponds to a nice 2048 ADC makes sense. It is easier to divide by ln(0.5) in the formula, than deal with a precise voltage level 1-1/e = 0.63212055882 for the sake of an easier math.


>>2408852
why do you need a transistor to turn on a LED in the first place?

>>2408854
>why do you need a transistor to turn on a LED in the first place?
i think we both know i have no idea what i'm doing

>>2408857
Post circuit, not meme frog

>>2408857
transistor is a digital switch. you said you also had a push button. so that's two switches. i mean if you just want to experiment with a transistor used as a switch that is fine. depends on what you want to achieve.

>>2408857
also please posting frogs, since my wife occasionally walks by and she may think i am posting on a white supremacist website since she read on fb that it's a white supremacist frog.

>>2408858
it's already been taken apart. this was mostly to compensate for the fact all my buttons seemed to be normally closed circuit, when it turns out that all i had to do was turn the sumbitch 180 degrees and put the wires in a different place
>>2408859
i have 3 boxes of chinese parts and i intend to use all of them

>>2408862
*please stop

>>2408799
Transistor? >>2407986

>>2408868

i was looking at that diagram, and then it changed in front of my eyes.
never been spooked by a GIF before.

>>2408852
An easy approximation is to say that the base resistor is 50 times larger than the LED resistor. Make that 75-100 times larger and you're probably still safe, considering the LED drops some of the voltage. Get the LED voltage from a normal LED calculator or your own head.

>>2408868
>1k base resistor
>current ratio of 10
Kinda overkill mate.
>LED forward voltage decreases but is nonzero at zero current
what?

In school they taught me that current flows from negative to positive. Once I looked at some schematics, they always had the current flowing from positive to negative. I was very confused!

>>2408953
Conventional vs electron current, there's probably a wikipedia article on it. Technically electrons flow from negative to positive (inb4 holes as charge carriers) but in practice that's a level of abstraction that just isn't useful. We treat the electron as negative, say conventional current flows positive to negative, and don't ask any questions. The only times the specific charge carrier matters is when talking about the internals of semiconductor systems or vacuum systems or maybe also thermal noise. At which point you're deep enough into it that you should be confident enough to get mixed up it the polarity thing.

I mean I'd love to teach people to use tau instead of pi when doing math and physics, but the moment they step into the real world and see pi everywhere they'll ultimately be worse off than had they just used pi in the first place.

>>2408316
>Alternator
This makes sense.

hey um... so this is gonna be a real dumb question, but for some reason this confuses me. If the voltage in the house is 120V or 110V, how does plugging in a fan (for example) end up using like 12V or something?

>>2408968
start on this
https://en.wikipedia.org/wiki/Electric_power
volts isn't the only unit relevant to electricity
see also
https://en.wikipedia.org/wiki/Ohm's_law

>>2408981
Okay, let me try to ask this in a different way.
If i'm trying to calculate how many 12V fans i can plug into a 15A breaker circuit, i want to know how much current the fan uses. So what i want to know is, when dividing wattage by voltage, am i using the 12V or the 120V?

I'm just confused on the voltage of the appliance vs the actual house voltage. Sorry again, i know this is probably a dumb question

>>2408371
Thanks a lot!

>>2408983
Oh you were talking about 12V computer fans, not fans that you literally plug into the 120V wall socket? The power supply in your computer steps down 120VAC to 12VDC. If your fan draws 200mA of current at 12V, that's 2.4W of power through P=V*I. Assuming the power supply is 100% efficient, this will draw the same 2.4W of power from the wall socket, which at 120V is 20mA of current. The difference between AC and DC isn't particularly important here, but it can be confusing.
A simpler example would be using a normal AC transformer to turn 120VAC to 12VAC and putting a resistor across it.

Any wall wart or power brick or what have you is designed to convert wall voltage to some more convenient lower voltage while keeping an efficiency close to 100%. In reality we see efficiencies around the 90-95% mark, so the wall socket will draw a bit more power from it compared to what the load consumes, the extra power is turned into heat inside the PSU.

>>2408983
>>2408997
So you need to figure out how much current a 12V fan takes (they usually draw more current while starting than when up to speed). If you assume 2.4W for each fan, that's 750 fans total. That would be inordinately expensive.

You also need to get a power supply that can step down the wall power to 12V and supply enough current. Which, considering 1800W is such a large amount, would need to be a 150A power supply. This would also be inordinately expensive.

Good luck!

>>2408299
>Magnetic induction works by the poles going past the coil, alternating between N and S.
>poles going past the coil?
N-S poles line up temporarily to induce a voltage across series wire-segments, then its "removed" completely for that stored energy to be released..
Then, same is done in the opposite S-N direction.
This is an intermittent circuit, makes no attempt at full duty-cycle per rotation.
this is a simple relaxation circuit.

>>2408862
>also please posting frogs
ok
>she read on fb that it's a white supremacist frog
Apu is green. I think your wife might be (is) mentally retarded. Sorry about your luck, Chief. Til death. lmao

Can anyone recommend a budget ESR meter for an amateur ohmie? I read that ESR is the only good way to identify faulty caps while they’re still in circuit, and all I have is a cheapo DMM.

I built a guitar pickup. I used a piece of really thick wire slathered in nail barnish, really thin wire salvaged from a real pickup, a tiny toroid salvaged from a cfl ballast, a slice of copper tubing cut open and flattened into a plate, two rectangular neo magnets stacked together, some scrap plywood, masking tape, superglue and solder. It works! But it has very low output. Maybe with twice the turns on the secondary it would come closer to a typical single coil solenoid pickup output. I have no idea how many turns I used. I could probably squeeze in some more if I can manage not snapping the tiny wire.

>>2409329
What are you using for poles?

>>2409329
Use steel for the plate, and steel screws or cylindrical neodymium magnets for the poles.

>>2408316
>>2408173
>>2409207
yeh, not gonna work.
You can't excite others to fix what they don't yet realize is broken..
As lame ass boring as it will be, you will have to come up with the solution yourself.
Reinvent it from scratch.
Don't ask me how, I'm as pissed off about it as you.
just stop larping, its wasting time.

Yes, it sux.

Where did you guys learn actual practical info?
I'm a graduated EE who has been looking for work and I've gone through all these books on circuit theory, did all the projects in Sedra, and I'm getting thrown for a loop in interview questions.
Some examples:

>You're on a desert island surrounded by mines. How would you make a buck converter?
>Explain how power is delivered to your home and how it ties into buck converter design. How are electrical standards in California different from say, Colorado?
>What wire gauge would you use in high frequency signal use-cases?
>Pick the specific alloy you'd use for corrosive wire design. (He gave me the answer, alloy 400, then asked why it was used.)
>Explain the differences between Type 304, type 304 L, type 347, and type 430 wires.

I haven't even fucking heard of those wire types. This was for a CAD job.

>>2409385
>You're on a desert island surrounded by mines. How would you make a buck converter?
What was your response?
>inb4 ligma

>>2409385
>>You're on a desert island surrounded by mines. How would you make a buck converter?
should have walked out then and there

>>2409385
>>You're on a desert island surrounded by mines. How would you make a buck converter?
Teorethically doable but it would take you a long time to even make the necessary tools. What kind of mines though?

>>2409442
>What kind of mines
Buck converter mines

>>2409385
>How would you make a buck converter?
Step 1: Build a boat.

>>2409341
There are no poles
>>2409345
Why a steel plate? Why have polepieces? If you can have the magnet very close to the strings, why bother?
With this little project I was trying to replicate the "Alumitone" pickup made by Lace which uses no polepieces, and two coils in one contiguous aluminum frame with reverse magnet polarity. For the transformer I think next time I'll use a bigger toroid or chuck something together from the laminated core of a broken transformer, and lots more windings.

>>2409385
What, like landmines? Take your cellphone apart, use the speaker as an inductor, and manually switch it to and from the battery positive and negative. Very low frequency synchronous buck.
For the 2nd question, I imagine it’s asking about how AC power is stepped down to LVDC frequently, and how noise or voltage drift specs of mains that may impact SMPS design, which may change between states.
Wire gauge questions, idk, there’s probably a decent rule of thumb. For high frequency signals maybe you’d want litz wire.
How are alloys relevant to an EE? Leave that shit to the mechies.

>>2409471
>If you can have the magnet very close to the strings, why bother?
To adjust pickup height for individual strings. Makes a big difference to tone.

I "built" the cap meter. It works but I'm sure it works a lot better if you know the exact values of the internal capacitor and internal pullup resistor used by the Arduino. However, all I need is a general idea of what the caps are. Some of them are a whopping 22% below the nominal value. For example, I have a pair of 22n caps that might be older than me, and the measured value is 17.2n for each. Not even newer but lower value caps measure that much below nominal value. I saw it recommended in the comments of the tutorial page to wait a few minutes to get a more accurate reading.

>>2409360
The electrical yield is still orders of magnitude less than what the new motor drives. The 2 sections combined looks like a flagpole with most volume unused.

A radically different geometry is needed.

>>2409680
>radically different geometry
There is no alternate to how it is wired.
Also, the old way makes a 3phase output easier, and will always be proffered by industry.
The only solution to our flag-pole problem is a wider diameter.

>>2404526
rolling for a project

>measuring open-circuit ripple voltage on cheap AC-DC SMPS modules
>getting ripple frequencies on the order of 30-600Hz
>pic related
I imagine this is just occasional switchings in some sort of intermittent mode, while the actual switching frequency under load will be far higher. Are common wall-warts fixed duty-cycle? I'd have thought that more sophisticated ones would vary the duty-cycle, causing their no-load ripple to be the same frequency as their normal operation, and the resultant ripple to be similar if not smaller than when under load.

Note that I have only tested shitty cheap SMPS modules, I'll check more expensive ones tomorrow. Guess I need to also test them under a variety of load currents, for which I may need some sort of variable load. By which I mean a length of nichrome in a bucket of water.

>>2409689
>radically different geometry
I'm not saying your unused-volume solution is lazy, bruh...
but, could it possibly look any gayer!!

do you guys also 3d print? I feel like I need cool covers etc for my current/future projects but 3d printing seems a bit too much, what would be the alternative?

>>2404526
So I have this weird Li-ion behaviour in that it reads ~3.4v, but applying voltage with my constant current supply it only draws around 300mA instead of going into CC mode and immediately goes up to the applied voltage.

I've had this several times with Li-ion batteries and have just thrown them out in the past because I assume they can't be revived when in that state.
I just wonder what caused them to fail in that weird way. They haven't been deep discharged or anything and other batteries in that pack charge and discharge just fine, having most of the specified capacity left.

>>2409927
>I feel like I need cool covers

diy'ers dont into aesthetics.
just functionality.
for all aesthetics questions, kindly visit /gay/

Got info on making legit diy capacitors for use in pcb's? I want to use common materials and get capacitances in uncommon/hard to find values in the 15pF to 1000 nF range. I'll be working with voltages below 50v and frequencies in the audio range.

>>2409972
Wood is ultimate aesthetic tho

>>2409972
Did you glue knobs and switches onto wood blocks with nothing inside? lmao
>Hey kid, wanna buy a blocky-talky?

>>2404526
RF spectrum analyzers down-sample the gigafuck frequency to something more manageable before sampling. How does the spectrum analyzer knows how much it need to downsample to avoid alias or to not fuck up things by downsampling too much?

>>2409983
I made a variable capacitor out of cardboard and plastic wrap for a radio receiver once, it was the size of my arm and about ~10-.1nF.

>>2405385
>Filtering
consider:
>transformers and transmission lines
>very fine steps
>>2405353
>>2405347
why does every schizo idea is so similarly presented?

>>2410017
>schizo
its a hand held generator, you KMART coomsumer.

>>2410025
like these chinesium flashlights?

>>2410030
Doubt that thing would power a house, bruh.

>>2410033
you can only power a house if you have enough power to power a house dude, hand-held or not. The only handheld 5 hp + motors I know of are hidraulic.

>>2410011
RF black magic

>>2410037
>I know of
exactly.
converting the turnforce to electricity isn't going so great >>2409360.

i'm drawing up a simple interlocking logic gate. the input (on the left side) is pulled down already. do i need pulldown resistors on the NAND gate output and the AND gate output as well, or is this unnecessary?

>>2410086

>>2410086
>is this unnecessary?

very very unnecessary.
logic gates are designed to feed each other without any external stimulus or lubrication.

>>2409397
I would reply with a joke
Buck state: broken
and we would laugh heartily and move on

>>2409385
>I'm a graduated EE
So you must have a lot of theoretical knowledge but has no chance to apply it during the interviews since they ask you bullshit questions? Not sure if this is bait but sad if true.

>>2408862
This is definitely the most white supremacist electronics community that I know of

>>2409385
>How are electrical standards in California different from say, Colorado?
115V in california, 110V in colorado due to high altitudes. I'd also talk about the standards in Texas. If you catch my drift.

>>2410011
Not downsampling, downconverting with a mixer
>>2410044
dclet cope

>>2409558
what's the significance of landmines in this story?

>>2409972
those boxes look fucking awesome.
how do you make one? it looks like a solid box though. is the inside hollow that was carved out or it is an actual box with the walls?