/diy/ - Do It Yourself

This thread bought to you by the ADUC7025 Precision Analog Microcontroller. You're analogue, why shouldn't your uC be too?

So this might be more /sqt/ than /ohm/, but if I have a device that requires +5DC and -5VDC, can I run them from separate bench supplies or should I slap together an appropriate reg circuit? It's a device under test/restoration, I have two identical supplies but I can slap together something on breadboard to power it short term if needed.

>>1507029
If your bench supplies are floating, you can bridge two 5V ones to create +-5V relative to the center point

Do you think these generic arduino relays can handle being powered by 12v instead of 5v?

>>1507053
They are, thank thank you anon.

>>1507059
Best to use a resistor to drop the voltage. Use R=V/I to determine the value required, then use P=IxV to determine what wattage resistor you need so it doesn't smoke.

>>1507025

>>1507059
>>1507059
>make a break out board
>call it ``x`` arduino board
The guy that had this idea is genius. But I died inside when I saw a potentiometer board.

What's a good schematic for a distortion pedal?

>>1507082
This was legit good idea. There are many components that require additional components to work with things like raspberry pi or arduino, for example thermocouples, capacitive touch sensors, or the mentioned relay, since you can't power it directly from a digital pin etc.
So instead of ahving to solder 5 different things together and ending up with a mess of wires you have a tiny pcb format.
But some things are pretty stupid breakout wise, like for example mosfet driver boards.. why would you need that?, you can easily drive the mosfet directly from a gpio pin. you can even pull it down from it.

>>1507087
You mean drive the DRIVER from the pin right? Because power mosfets need some current to get going and usually have Von>5v.

>>1507088
no, you can drive mosfets directly from logic pins

>>1507093
obviously i mean logic level mosfets

>I NEED A QUESTION ANSWERED, EE NOOB

I have a CDI box for my motorcycle that might be broken, I attempted to test it and get random results on the resistance of the proper pathways, I'll measure it once and get 0 ohms, another time it'll measure 2 ohms etc (random values not actual values)

Does this mean that it's busted? My multimeter is brand new and I bought it just for this. I'm testing it with no power to it and my multimeter has never been put under any kind of power load.

>>1507025
Anyone know what the connectors on the bottom of this switch are called?

I know there are connectors that can easily plug into them, but I don't know what they are called.

>>1507098
>CDI box
What model? What pins are you measuring?
If you measure at same points and get different values - your measurement technique is shit

>>1507059
what does the 10A 30VDC, 10A 28VDC mean on it? The part number makes me think that it can only handle up to 5VDC

>>1507111
seriously?
it's the rating for the switched voltage and current current and coil voltage

>>1507111

relays quite often come with different coil voltages, so they put the coil voltage in the pn, so you drive this one with 5 v.

listing 28 vdc and 30 vdc is retarded, but maybe they had a big customer with a retarded purchasing dept who kept bugging them "is this ok with 28v, it says 30???" so they listed 28 too. stupider things have happened.

>>1507112
>it's the rating for the switched voltage and current current and coil voltage

put the crack pipe down son

>>1507114
Come on brainlet, apply yourself.
You use the 5V to turn the relay on or off and you can pass 30V DC up to 10A or 125DC up to 10A through that relay. It is really not that complicated at all.

>>1507098
According to this quick google search you can't test CDI boxes with just a DMM https://www.denniskirk.com/learn/how-to-diagnose-motorcycle

>>1507116
I think he was commenting on your retarded sentence structure..

>>1507105
Those are solder lugs.
You're probably thinking pic related, which is available in different sizes. Some of them might even fit to your connector. They're called Faston or Abiko connectors, depending on where you live.

What is a good crimping tool?
I want to buy one from ali but they have so many of them.
I would like some unversal that can do lots of terminals.
I will be using it solely for putiting terminals on wires sized between duponts and normal outlet wires (the bendy ones)

do you think the 9000mah 18650 liions chinks sell will have at least 3000mah?

>>1507156
No. I think if they’re Chinese and have some obnoxious rating like that, they will be <1500mAh

>>1507156
Have a look at some google results on them, plenty have been tested, none come close. I think the lowest I ever saw was a 9900mAh battery that only had 500-odd. I would avoid unless you can get them locally and return them easily when they fail to deliver.

>>1507170
>>1507156
Anon who mentioned the Kobalt drill batteries probably had the best idea. 6x 2000mAh for $20, and I doubt those Kobalt batteries will have used or knockoff cells in there. Even the Harbor Freight stuff has new Samsung sells.

Otherwise Walmart... pic related was $15

>>1507026
based

>>1507153
non-ratcheting: the Molex 63811-1000 hand crimp tool is pretty universal for non-insulated terminals at the smaller end of that range
ratcheting: there are a few tools on ali/ebay with individual crimp dies available. they're all pretty much the same. I've found the SN-01BM to work best with "Dupont" connectors

>>1507156
nope, Pic related

>>1507170
>>1507176
seemed too good to be true

>>1507161
>>1507174
Get the fuck out of here faggot

>>1507176
The fuck? Is that little silver rectangle the actual battery?

That better be cocaine weighing it down.

>>1507178
pls kill yourself tripfag

I would like to power esp32 from a battery.
Is saw it has ams1117 voltage regulator on it so i beeped the 3 legs and one is shorted to ground, one to 3v3 pin and one to VIN pin.
I assume that if i want to save energy i will have to go through the 3v3 pin directly and connect it to a battery. But since it is bypassing the regulator any slight overvoltage will destroy the board.
So i was thinking of getting a small buck converter set to 3v3 and two 18650 liions connected in series feeding the buck.
since the buck isn't wasting energy as heat this should be very efficient and safe right?

>>1507219
LDO regulator with a single nominal 3.7V battery gives 3.3/3.7 = 89% efficiency, you're not gonna beat that with a cheapo buck converter module.

>>1507221
If i read the sheet for the ams1117 correclty it drops at minimum 1.2V which would be only 2.5V left for the board which is not enough

>>1507025
Can I use a transmitter to fry car stereos? People in my subdivision with really loud bass (I'm hearing their music through their car, across 200 yards, and through my house) annoy the shit out of me.
Yeah, I'm aware I'd probably be violating some FCC regulation. They're black, they wouldn't report me even if they knew what the fuck was happening.

>>1507195
I’m workin on it

>20ga wire for relay switch
>12ga wire for power to nails
Sound ok?

welp after some digging i found out that the chinese 18650 liions have a slightly lower capacity than the originally claimed 9900mah, the real capacity is actually pretty close (550mah) but sadly not enough for my needs. I would be perfectly happy with even 2000mah :( god damn it. i will have to buy those overpriced pieces of shit here in my country to actually get a good capacity

>>1507229
Can you get any drill batteries for a decent price?

That Canuck AvE took apart a Chinesium DeWalt knockoff battery that was supposed to be 3.0Ah, and it ended up being slightly used 2000mAh Sony cells so it was actually better than the packaging claimed. Otherwise if you can find a cheaper power tool brand and grab a 2.0Ah+ battery, it will probably have decent cells.

>>1507229
>overpriced
no, your cheap chink shit batteries are underpriced. you want good batteries you have to pay good money.

>>1507223
So get a better LDO.

>>1507237
i buy everything electronics related from china and so far it was all been perfect. i buy arduinos from there for $3 a pop whereas here they cost $19 a pop, and the chink ones gave me zero problems so far
so this is a question of finding a store that carries liions which aren't a pure rip off

>>1507238
i already have 50 small buck convertors (it was min purchase quantity) so i will use those instead

>>1507240
it's easy to make a totally identical rip off of something like an arduino, the cheap components work pretty much identical to the real ones. much harder to do a battery.
you just can't make a battery cheaper and still have it be as good. cheap batteries are always going to have lower capacity and the electrode overlap is going to be shitty too.
go ahead and try cycling those batteries hard, see how long they can go before blowing up.

>>1507244
this is really shitty situation.
this would mean that batteries for my project will cost literally 3 times as much as my entire project

>>1507025
Test

>>1507235
not a bad idea.
maybe from letgo or some shit like that
the problem is that if they were use in a drill they would be cycled alot and probably near dead

>>1507248
I mean if you can get them dirt cheap, maybe you can get 2-3 cells that are still really good and the other are crappy. And the AvE ones ended up all being good cells, but you could tell they were spot welded to something else before.

There’s a company that sells remanufactured Prius batteries that does something similar. They take the failing battery packs and it’s normally only one or a couple cells failing, and the others still have a ton of life left in them. So they test all of the cells, junk the crap ones, and use the good ones for new battery packs.

>>1507248
You could also buy and old lead acid car battery, they are really cheap, and then cut open the casing and pull out the liion ceels.

>>1507252
That’s what I’m trying to do but I think I need a bigger pry bar.

>inb4 this guy >>1507189 gets here

>>1507219
maybe. you'd have to consult the datasheet and see. don't forget to consider quiescent current and the total amount of run time you need

>>1507226
if you have to ask, no, you probably couldn't

>>1507246
many such cases in IoT

>>1507280
the current will be almost nonexistent, basically it is an esp32 with a digital temperature sensors, it will be in deep sleep and wake up once every 60 second, connect to wifi, read temperature, make http GET request with the temperature and go back to sleep
so no heating issues or anything like that
so since most of the time the esp 32 is in deep sleep which is like .5ma or something, then buck would be worth it, since i bet the liner regulator would waste lots of energy during sleep

>>1507280
>if you have to ask, no, you probably couldn't
Nigger, it was a yes or no question, don't be fucking smarmy with me.
Now that I know that answer is, "yes but it requires you to go learn something," I'll go fucking learn something. If it had been, "no that isn't possible," I'd be wasting my goddamn time by learning about it.
I didn't ask you to spoonfeed me the "how" - just to confirm whether my desired end goal was realistic.
Thanks cunt, I'l go learn about radios now.

>>1507226
You need one of these

Meh, it’s not really a spot welder but more of a “make things glow red” machine. I definitely lost some current somewhere in the contraption. I would put my money on the shitty connectors.

Welp, gonna put that project away in the junk bin.

>>1507285
the regulator's quiescent current, I mean, that's what's killing you with the AMS1117, which is not the most efficient linear reg
>i bet the liner regulator would waste lots of energy during sleep
regulators vary widely. I wouldn't bet that without reading individual regulator/converter datasheets

I have a KiCAD question. Component pads in my PCB editor have these rings around them. They seem to be in the front copper layer and disappear when I make "Footprints" or "Pads" invisible.

In the footprint editor on the left, the rings aren't there.

Anyone know what's going? I'd love to get rid of them.

>>1507317
Never used KiCAD, but the size suggests they're solder mask expansions. Check the pad settings for solder mask or displayed layers.

>>1507320
Found it. It's an indicator of "Net Pad Clearance."

Any ideas how to disable that? Besides setting the clearance rule to 0

>>1507340
Got it, it's in display options

>>1507177
>Get the fuck out of here faggot

>>1507177
>>1507345
I prefer Pic related

>>1507351
At 75, I'm too uninformed geekwise to know where to find the mod/pack you're using.

>>1507355
https://www.4chan-x.net/
shit's cash, if you know what I mean
I haven't tried but you can probably use the plain old tripcode filter with just a single exclamation point to mute ALL trips

Is it acceptable to use a single (schottky?) diode to make a higher voltage rail than your silicon rectifier bridge such that I can drive an N-channel MOSFET on the high-side? Rectified mains btw.

>>1507355
4chan X browser plugin, allows you to draw little pictures in the quick reply window, among many other things. Hotkeys for spoiler, math, etc. tags, hotkeys to hide or show all images on a page, change date/time formats, it has a very extensive settings page.

>>1507286
It's possible, but it will require a few hundred dollars of RF equipment, if not a few thousand. You'll have to send them RF at an audio frequency (10kHz would work) and have it get picked up by the amplifier, meaning you'd need to induce at least a few hundred mV across a tiny wire. The electric field strength needed to pull that off would require a large amount of transmitting power even 5m away, so trying to do so at 200m would require a directional antenna and even more power. Not to mention you couldn't use a self resonant antenna at such a low frequency. Other options are frying their electronics with an EMP (not an option) or somehow trying to match their class-D amp's PWM frequency, which might not do anything at all. But as far as FCC regulations, I don't think you'd be in trouble at all, since I don't believe bands below 100kHz or so are regulated.

>>1507367
>Is it acceptable to use a single (schottky?) diode to make a higher voltage rail
No. Draw all the diodes inside the bridge visible and you should see why. On the other hand, you can add a capacitor and another diode to make a kinda-sorta doubler.
You need a protection diode for the mosfet gate, too.

>>1507285
A good LDO will waste less when the ESP's in deep sleep than a good buck.

>>1507378
I still don't really see the problem

>>1507391
that's amazing

>>1507397
what?

>>1507362
thanks!

>>1507391
Look at D1 and D5, you're not gonna get anything across C2. Simulate it and see.

>>1507419
Oh yeah the other diode drop is on the -ve rail. Still I can use a schottky diode, is there an issue with that?

>>1507367
y tho, just move that bitch to the low side and have more gate drive than you can handle

>>1507391
see if you can find the hidden C-W voltage doubler in there
protip: you can't

>>1507422
I want to make a buck converter, so having a high-side switch would mean my output voltage is referenced to the neutral/earth line.

Or would using a low-side and flipping the live and neutral lines be recommended instead?

>>1507424
Oh it wouldn't be referenced to 0V anyway since it's a full-bridge rectifier.

What is a good control scheme to cool a led without wasting power when it isn`t very hot (It has a heatsink, but I`ll add a fan because sometimes it really gets toasty). Bang-Bang control is adequate? I`ll proably go analog because It seems like a simple thing

>>1507443
Add some hysteresis

>>1507124
The generic term would be 'quick disconnect'.

>>1507156
LOL no. Read the datasheet from the specific manufacturer carefully, I'll bet that 'rating' is based on some ridiculously low current draw, like you'd expect from a 3V lithium coin cell like a CR2302.

>>1507221
>>1507223
In a battery-powered design, look at what your typical current draw will be to determine if you want to use an LDO or a switching regulator; low current draw, use an LDO, medium or high current draw, use a switcher. There's even some buck/boost types that include an LDO and will switch to that automatically when appropriate, to cover the middle ground. Linear Tech, I think?

>>1507226
>Can I use a transmitter to fry car stereos?
Not really. But if you have access to a CB radio with a linear amplifier (i.e. kilowatt or better; highly illegal!) and an directional antenna, you can overload their car stereo, and some diode junction in the audio section of it will act as an AM detector, so everything you say into the mic will come out their speakers. Can be rather unsettling if they don't know what's going on, and if you're 'creative' about what you say when you do this.

>>1507308
Try using some carbon rods as electrodes instead of those nails.

>>1507367
>>1507391
If a few tenths of a volt are really going to make a difference, then make a full-wave bridge out of four Schottky diodes and be done with it.

>>1507506
can confirm, my home stereo used to pick up transmissions from trucks on the highway next to the house. the speaker wires were about the right length to be a half-wave antenna. It got detected in the output stage, the volume control had no effect.

>>1507507
It worked with the nails when I tested it with wire straight to the battery. The relay is rated at 40A, maybe that is the weak link. Or one of the connections. I did 12ga wire with everything but the switch, and I think it was 14ga or 16ga I used to test it.

Dude in the video used copper nails, I think they were larger than the little copper plated ones I bought but once again they worked in the test and I was thinking about filing the point a tiny bit.

Carbon rod sounds interesting. This may be a retarded question, but could I use the lead from a regular #2 pencil?

>>1507519
Simpsons did it!

>>1507510
I'd only need a "half-bridge rectifier" since I wouldn't need the two diodes that establish a negative rail. Also please just make all your quotes in the same post and dump the trip.

>>1507507
>>1507525
Hmm this guy did a different yet simple setup. He uses a DC power supply (mine is sitting in my guest bedroom and I’m trying not to open it til Xmas), 3 big capacitors in series, and some graphite rods for the electrodes.

Probably not as fast as the big battery but still a cool idea and he claims he can spot weld copper with it (which I’m assuming is more difficult than the nickel strips).

I wanna load test this car battery somehow and see what kind of balls she had left in her. Seems to be taking and holding a charge better than I expected, but my little 4A charger is great for the 10Ah scooter battery and slow as steam coming off a cold turd with a full size one. She sure did blow some holes in the nickel strips with the test run and that was at 11.9V.

I need a fuckin purpose before I get the motivation to perfect the thing. Any ideas on an easy fix to control the output for different materials?

>>1507495
>>1507498
>>1507503
>>1507506
> gets a trip just to rile up the board even more.
Another trip to filter. Great to see you have the community at heart you cock lord.

Is it safe to step down pc fan voltage from 12v to 5v with just a resistor? I want to reduce rpm by half

>>1507620
You can filter all trips automatically so you experience can be 100% trip free

>>1507654
Do you prefer to filter the entire reply chain to a tripfag's post too or just the actual post the tripfag made.

>>1507620
>>1507654
>>1507656
>the tripcode is the problem
Delusional

>>1507659
Nope, but it's a symptom that you can filter to solve the problem

>>1507665
>>1507659
The true chemo would be to implement a silent permaban for anyone who attempts to post with a trip

>>1507659
see
>>1506401
Now leave it off dipshit.

>>1507667
They are not without their place, but to assume a mechanism to prevent thread hijacking and ongoing verification for works in progress is for turning the board into your personal blog is the actual delusion. Even fucking bunkerbro, the one guy on /diy/ who actually warrants a trip, manages to refrain.

>>1507659
see
>>1506401
Now leave it off dipshit.

>>1507667
They are not without their place, but to assume a mechanism to prevent thread hijacking and ongoing verification for works in progress is for turning the board into your personal blog is the actual delusion and arrogant in the extreme. Even fucking bunkerbro, the one guy on /diy/ who actually warrants a trip, manages to refrain.

>>1507652
Even if 5V is enough for running it, your fan might not start.
For an extra of kludge, you can put a large capacitor in parallel with the resistor to give the fan an initial start pulse.

>>1507674
if i connect the fan to 5v it runs just fine, just don't want to waste a buck convertor to step down 12v if i can use a resistor.

>>1507675
Fan pulls more current starting than it does running at steady speed, if you size the resistor to get 5V steady state it'll be getting less than that when trying to start, might not be enough.

>>1507681
Fuck. Why can't anything ever be just simple for once. Will 500uf cap be enough?

I fitered trip* in comments
Liberation

Ok so I know little about circuits and I have a poor idea of what Im doing but here I go.

Say the decoder output table (y3 y2 y1 y0) is :
HHHL
HHLH
HLHH
LHHH

Which is the same as the encoder input table.
Is the (pull up?) resistor on the left the right way to make the horizontal lines go high when a button on that line is not pressed? Or should each line have it's own resistor?

>>1507059
Answer? Yes, for a few quick seconds. Otherwise, use a 100 ohm 1 watt resistor

>>1507746
Each line should have at least it's own resistor. Otherwise once you pull any line low, all other lines will go down as well since all 4 lines are connected.

Though personally I would never use this configuration

>>1507682
get a slower fan
or use Pic related

>>1507059
no, add a small voltage regulator

>>1507771
>voltage regulator
>for relay
kek

>>1507792
point-of-load regulation is a legitimate technique in power design. no doubt the device you're sitting in front of right now has half a dozen of them
begone lolduino thot

>>1507763
Thanks frien.
I did this in order to use the least pins possible on the microcontroller it will eventually be attached to, but what do you mean by you'd never use this?
Too unwieldy?

>>1507798
Just to prevent keypress ghosting I'd add diodes, other than that it's just mostly preference how you do it

how do i balance charge on my liions without spending hunderds on a balancer?

>>1507850
Series right? Have a few difference comparators. You'd need a separate voltage stage for each battery, but start there. That's only if you're optimizing. If you're not, just put each battery at 4.1-4.2 volts above the previous negative terminal. Don't think normal regulators sink current so have fun, you're going to run into issues if you want to charge it all at once. In series you could switch which battery - is "ground", so you only need a 4.1-4.2 supply and if each battery is charged enough, switch to the next one.

>>1507797
>legitimate technique
What does that have to do with anything?
Anon's point is a simple resister would do and a VR is overkill for this application

>thot
Ah, faulty reasoning explained

>>1507088
Depends on the frequency, but you can couple them with a transistor/other mosfet/not gate if it's a problem. Around 100KHz you'd need ~.05-.15A depending on the mosfet. A circuit I worked on, I was pushing hardware to its limits. I would've needed 1-1.5A for 1MHz, and the goal was 1.5MHz at highest speed. I didn't know power drivers existed before that.

>>1507864
Does anybody want to spoonfeed me on mosfets? Was trying to read the Wiki and stuff but a little over my head. Are they basically like a relay? Except it’s not 0% or 100%, it can be anywhere in the middle based on the input voltage? If that’s the case, then what is going in/out of the 3 pins? And can they handle high current like a relay?

They seem to be in everything I open up.

>>1507854
No, they are all in parallel since i want to increase capacity not voltage

>>1507881
>Does anybody want to spoonfeed me on mosfets?
Not really but ok
>Are they basically like a relay?
That's one use for them, but important differences are that they don't isolate the control from the signal, they don't require any significant current to drive, and only work in one polarity. If you try to drive them in the other polarity they'll just act as diodes no matter what you do with the gate voltage.
>Except it’s not 0% or 100%, it can be anywhere in the middle based on the input voltage?
Yeah
>If that’s the case, then what is going in/out of the 3 pins?
For N-type, which are the important ones: drain, gate, and source. Drain to source is the signal path and acts like a resistor when the FET is on. Gate-source voltage changes the resistance. It's a bit more complicated as neither of those effects is linear with Vgs or Vds over the whole operating region.
>And can they handle high current like a relay?
Depends on the FET and how well it's heat sinked.

>>1507881
>>1507886
And take off your trip, fucker. Why would you WANT to be known as the guy who doesn't understand FETs every time you post?

>>1507886
>>1507887
Thx bruh

And see, I’m learning! Your little beppu is growing up!

>>1507887
So trying to read the FET wiki now... Source would be like 30 on this relay? Drain like 87 and gate like 86?

Took of the trip for this one post since somebody was actually helpful on the ol’ Taiwanese spearfishing forum.

>>1507897
I'll tell you if you promise to leave it off for all future posts

>>1507881
it's a space future triode.
normally the electrons sit in a pool at the bottom doing nothing, but when you suck some electrons out the middle of the tube, electrons from the bottom fizz up to the top.

>>1507881
>>1507881
>picking a trip when you don`t even know the basics or made 0 things to make the thread better AND ask questions
die

>a symptom that you can filter
filter trip* in comments
no more Borg posts

>>1507898
What if I leave the trip off, but every post includes a pic with Newports in the shot?

>>1507925
That's not bad

If I'm reading your picture right 30 is drain, 86 is gate, 85 is source

>>1507921
Hi there!
You seem to have made a bit of a mistake in your post. Luckily, the users of 4channel are always willing to help you clear this problem right up! You appear to have used a tripcode when posting, but your identity has nothing at all to do with the conversation! Whoops! You should always remember to stop using your tripcode when the thread it was used for is gone, unless another one is started! Posting with a tripcode when it isn't necessary is poor form. You should always try to post anonymously, unless your identity is absolutely vital to the post that you're making!

Now, there's no need to thank me - I'm just doing my bit to help you get used to the anonymous image-board culture!

I've got a couple questions.
What's the point of the arrow on the trim pots? There are some of them that aren't marked, what kinds of resistors are those?
What effect do you get when using a capacitor inline instead of across?

>>1507954
Resistor with arrow coming in the middle - variable potentiometer
Resistor with only two terminals - regular resistor (fixed resistance)

Capacitor inline blocks DC and allows AC to pass (generic usage).
There are more specialized uses but at the moment no need to know.

For audio stuff capacitor inline blocks any bias DC voltage that is present and only allows AC signal (audio) to pass since it is what you want and what generates audio in speakers.

>>1507954
The arrow represents the tap on the pot. If you measure the resistance between both non-arrowed terminals on a pot you'll get the max resistance value of the pot. If you leave one side of the pot floating and measure from the connected terminal to the tap can get any resistance from 0 to the maximum value of the pot depending on how far you turn it.

Capacitor inline forms a high pass filter which is typically used as a DC block though depending on the size of capacitor it can be used to filter out low frequency AC too. Usually a DC block though to prevent DC from the bias networks backfeeding into your signal source and into your load. This is typically known as an AC coupling capacitor. Capacitor across forms a low pass filter which takes out high frequency noise. Also referred to as a decoupling capacitor because it shunts high frequency noise to ground due to its very low impedance at high frequencies. Additionally these caps can serve a dual role, they also store some energy and can quickly supply it if there's a sudden spike in load power consumption this helps keep the voltage on the DC rail constant where it otherwise might have dropped a bit and then recovered.

>>1507957
>>1507958
Thanks.
One more thing, how would you go about powering that circuit?
Oh also, what's a good place to snag the relevant Op Amp?

>>1507965
Since schematic seems to be using 9V, should be good idea to stick to that voltage. External supply would probably be best option.

Component stores are mentioned in >>1507025

Anyone know of a good way to clean a motherboard? I'm envisioning some kind of lightly pressurized isopropyl alcohol hose/nozzle. maybe in a bath with a filter to recycle the alcohol?

>>1507977
>Check the store in the OP
>Find the chip
>See price
>Heart skips a beat
>Realize that commas and points are switched
Oh thank god, sure didn't help that they were showing one more decimal than you'd expect

>>1507988
I always clean electronics with isopropanol (99%) and a soft bristle brush.
A pressurized spray system is overcomplicated, I spray the alcohol out of a regular spray bottle. that way you can spray with exactly the force you want, and adjust the spray size.

>>1507086
Take a look at the Big Muff Pi anon:
http://www.bigmuffpage.com/Big_Muff_Pi_versions_schematics_part1.html

has anyone here ever tried using a R2R ladder+uC to generate sines and ramps in the absence of a signal generator? If I use only parallel 1% resistors to get 2r and r the error gets cut down in half, maybe it`s possible to make a 10bit dac?

>>1508053
8bit resolution percentage is 0.4%, 10bit is 0.1%.

"For a 10-bit converter, even using 0.1% precision resistors would not guarantee monotonicity of output."

https://en.wikipedia.org/wiki/Resistor_ladder#Accuracy_of_R%E2%80%932R_resistor_ladders

>>1507667
>>1507671
How about make it impossible to post with a trip in more than one (bump-limit not reached) thread at a time? Make it take like an hour for your trip to get verified for each thread? Make it a bannable offence to have more than one tripcode? (you know who you are)

>>1507992
What do they call it in europe, a decimal comma?

>>1508086
The tripcode conforms to the rules.
Instructions are at http://4channel.org/faq#trip
The asylum is not run by a few triggered inmates.

>>1508119
>The tripcode conforms to the rules.

correct. the sad but funny thing is that the smartest board on 4chan (?) never learned the basic rule, DO NOT FEED THE TROLLS.

bepis rules this board because it is full of morons and he is somewhat creative in his trolling.

>>1508119
The point of tripcodes are for verification of someone's identity as being the same as the one who posted earlier without requiring any form of IRL identification, be it a name or distinctive piece of furniture or scar. It's there in the first place to prevent people from faking posts by a known anon (like bunkerbro or any other person that has anons interested in their updates across threads) which is simple to do since we're on a website that's anonymous by default.

Not only do we have people using tripcodes when nobody is interested enough in their updates to want to avoid posts by an imposter, but people using (what I assume to be) their own name as their tripcode. This is out of line of the original purpose of tripcodes and an obnoxious waste of time for all involved. Having to go out of your way to filter every useless tripfag when there may be an occasional person who does deserve a tripcode is tiresome, especially since the ones we do see often display subpar ability.

The suggestions I made would keep tripcodes able for use for their original purpose while disincentivising them for other uses. Extremely low-quality posts are against the rules, after all.

>>1508125
fuck off, even bunker anon does not use a tripcode. how about you start doing /diy/ instead of being a fag?

>>1508128
Bunker anon doesn't need to because he always posts sick OC. If it were an anon reporting from North Korea who can't post images (since he doesn't have a camera/doesn't want to get caught) the use of tripcodes would be perfectly justified. And if bunkerbro did use a tripcode I don't think any of us would fault him for it.

Is my maths wrong? My sim is giving me a different result to my calculation.

>>1508125
That’s my old wallet bro, that’s how you know it’s an impostor

>>1508126
>>1508086
>he actually took the time to type all of this out

I rode the short bus to high school. It was more bumpy than the regular bus. There were only like 4 of us on there and for Xmas the driver gave us all Bugles chips.

>>1508240
please go back to using a tripcode

>>1508174
I1 and I2 don't depend only on V1 and V2. Review you circuit analysis and algebra books

>>1508243
Fuck i thought so. For such a simple circuit it has such a complicated transfer function.

>>1508242
The grass is always greener, amirite?

>>1508243
>>1508244
Oh never mind, I forgot all about Millman's theorem.

So I'm testing a variable duty-cycle buck converter, and I've hit the snag that I though I would. Namely, the feedback voltage is smaller than the voltage set by the potentiometer, so the differential amplifier is feeding the comparator a voltage within the bounds of the triangle wave, but the resultant duty cycle isn't enough to get the output voltage up to the desired level. Since the output voltage doesn't change, the feedback voltage doesn't change, so the duty cycle doesn't change, and this sag / undesirable stead-state would happen to some extent regardless of what load I put on.

Pic related, but it's kinda awful. Oscillator voltages, feedback voltages, and potentiometer voltages are all set to vary between 2V and 3V, which I assumed would give me plenty of room compared to larger ranges. If the output voltage is higher than what I set it to be then the comparator should shut off completely, and if it's 340V too low then the comparator should turn on 100% of the time. Some funky low value resistors are there just for ensuring the transistors turn on quick enough.

>>1507925
Then you'd get banned for avatarfagging. So yes, please do.

>>1508285
any proportional feedback loop maintains a little bit of error to keep it pushing in the needed direction. try adding more gain to your error amp (U1?). to start, you might want to see the duty cycle swing from 0% to 100% as the feedback voltage varies from 99% to 101% of the reference voltage, maybe even tighter

>>1508285
cont.
I assume it's a problem typically avoided by simply turning that sag as tiny as possible (<100mV or so), since eliminating it entirely, while possible, would require vastly different circuitry. So I think I'll need to change the gain of the differential amplifier (currently at unity) to be much higher, which will probably require its ground bias (voltage source V4) to change too. Changing the inductance would also probably help.

>>1508292
>you might want to see the duty cycle swing from 0% to 100% as the feedback voltage varies from 99% to 101% of the reference voltage
Ah, that does send me in somewhat the right direction, that's far less arbitrary than what I was thinking. But I'll need to do a fair bit of math to figure out how to alter the gain of the amp for that to happen. Thanks a bunch!

>>1508292
>>1508295
>a fair bit of math
Oh all I had to do was increase the gain a whole lot because all the other factors cancelled. Neat!

How does everyone build your projects?

So far I've just been using perfboard and making connections with spare wire and bending component legs. It's a pain in the ass to solder though.

>>1508326
Depends. If I'm just testing shit breadboard, I use mostly through hole parts because of this since SMD doesn't easily lend itself to prototyping without a metric fuck ton of breakout boards lying around and at $1-4 a board those costs add up when you need a lot of them.

For more permanent but not too serious stuff I use protoboard like in your pic. The ones that have buses like on the breadboard are nice but they're relatively expensive so I don't have many on hand. The cheap ones with holes are okay, I usually make connections with component legs. People say to use solder bridges to connect pads but I can never seem to do this right which is not helped by the soldermask.

For any project I want to look nice or that doesn't lend itself well to breadboard due to high voltages or parasitics or whatever I'll get a PCB made.

I wanna try out Manhattan construction but I don't have any plain copper boards and they aren't cheap either.

>>1508326
I don't :(

>>1508326

Tab wire, cause i like to step shit up and make things harder then they need to be. looks cool though hay.

>>1508326
With lots of hot glue

>>1508326
>>1508373
Forgot pic

>>1508374

>mfw hiromoot adds a new rule against newport posting

>>1508380
Back to Camel Wides I guess

>>1507881
Even though you're a tripcode user, the physical model of a mosfet makes them way easier to understand: they have an npn/pnp set-up like a bjt, but the middle channel is non-doped normally. There's a capacitor made against this middle channel by having the gate connection and an insulator. If you build up static on the gate, you displace charge on the other, making a conductive path. More static charge, greater current capacity. The opposite long side doesn't have the gate connection, so there's non-linear 'body' problems. It gets worse because as you push higher voltage on D/S, the ends of the channel pick up a static effect, so the channel shortens and becomes more conductive. mosfets are actually non-directional devices, but we build them in a way which makes them better "one-way".

>>1508174
That's a Y connection. You could transfer it to a delta to figure out the current going from V1->V2 and then use IR to figure out the voltage drop through the resistors, to get V3.

I am considering using a mosfet instead of a relay. But since mosfet works with some electron magic fuckery instead of airgapping shit like a relay, it surely has a lekage current.
How long would it take for a closed mosfet to drain an AA battery connected to it?

>>1508424
It would drain it from the source input?

Don’t power tools all use mosfets in the switches? They don’t seem to drain the batteries too bad.

>>1508426
Imagine being this determined to cause drama in a community. Why the fuck is anyone still responding to this faggot?

>>1508424
It depends on mosfet, voltage and temperature. Your fet's datasheet tells the worst case value, but you can expect the leakage current to be significantly less than 1uA at room temperature and at voltages relevant to AAs. Worst case values can be one-two orders of magnitude higher.
Your typical AA battery has a capacity of 2500mAh or so and it would take couple of hundred years to drain it at 1uA current and around 3 years at 100uA.

>>1508427
I’m just holding up my end of the deal >>1507948

probably a retarded question, but why does my body have a decent connection to ground potential even when standing on e.g. concrete, which is a very bad conductor? is it merely because of the size of the area i'm standing on which makes the resistance low? if i take a look at the formula for the resistance of a wire R=rho*l/A, it would make sense

>>1508437
>why does my body have a decent connection to ground potential even when standing on e.g. concrete
Where does this assumption come from? Some electrical safety stuff? Wet concrete can be conductive enough for it to matter.

>>1508440
do you mean the assumption that you have a decent ground connection or that concrete is a bad conductor?

>>1508440

>tfw doing electrical work barefoot on wet concrete

>>1508446
The first one.

>>1508452
i did assume that because if you for example touch a 230V line while standing on the floor, a considerable current flows through you. which must mean that the current flows to ground, no? so for that to happen, there has to be a somewhat decent connection

>>1508456

If current flows to any appreciable degree you are in trouble. when you grab a hot mains and it tingles or even makes you twitch a bit, it's just capacitive current and is very small.

>>1508457
>it's just capacitive current and is very small
well but still there must be a conncetion to ground, no?

I'm looking to buy a bunch of cheap, simple, and, safe ~100W 12V power supplies, y'know the kind where there isn't any exposed mains wiring for idiots and children to fry themselves on. So far i'm thinking of ordering good ol' simple laptop power bricks from Aliexpress, any other suggestions? They need to be cheap so i can keep the costs down but at the very least kinda safe so i wont get into trouble when some idiot hurts themself after buying my product.

>>1508462
>well but still there must be a conncetion to ground, no?

through a tiny capacitor (high impedance) and the relatively large resistance of your body and probably the floor as well. The floor might also be largely capacitive.

all of which is why you can touch the mains with one part of your body and just tingle.

>>1508464
>simple laptop power bricks

no need to specify laptop; it won't hurt but it won't help either.

>>1508471
I'm not buying them for laptops.
All i need is 12V and around 100W in a small, safe, and cheap package.

>>1508477
So just search for whatever 12V 8A-10A power supply and find the cheapest.

Alibaba if you really want to buy in bulk

>>1508484
>Alibaba if you really want to buy in bulk
Oh i had completely forgotten about that website
I live in the EU, could it be cheaper to order from amazon.de for example than from china in bulk due to import taxes? Any experience on this?

>>1508326
for quick and dirty proofs of concept I use le ebin solderless breadboard and my minimal stock of through-hole components
for somewhat pernanent stuff that I'm in a hurry to build, usually it's a combination of cut copper tape and gouged board, which is good down to SOT-23 sizes. Pic related, a balanced antenna amp and bias tee, "free" Manhattan islands in lower right. sometimes I will use a .100" protoboard and my dwindling stock of THT components
for permanent and/or complex projects where I'm willing to wait a minute, I order prototype pcbs from china

>>1508329
>$1-4 a board
>plain copper boards and they aren't cheap either.
if you order direct from chink they're like 5-20 cents each
plain single-sided boards are like a dollar each. use nibblers or (alternatively) tin snips to make islands. or, if you can find insulated copper tape, try that, maybe with a bit of Kapton tape for extra layering and jumpers
also, pro tip: SOT-23 can fit onto an L-shaped section of a .100" protoboard

>>1508489
probably not, if they paid the import tax too

>>1508326
Breadboard, perfboard or I draw and etch a PCB. I don`t have a laser printer so I have to do it manually.

>>1508518
Greetings, fellow magic marker user.

>>1508715
Thoughts?

>>1508513
>>1508518
If there was a PCB 3d printer equivalent, would you buy it?

What if it could only do single-sided boards?

>>1508803
There is, you can use a CNC router to carve away the copper where you don't want traces. Alternatively, you could probably use a low-power CNC laser to burn/ablate away an etch resist, or simply use a "2D printer" to print ink (either with an inkjet on an acetate transparency or with a laser printer on the photoresist sheet directly) where you want to avoid exposing photoresist. Theoretically, a violet laser (GaN or doubled GaAs) would be usable for exposing photoresist without printing a template to obscure it, but you'd have to tune the power correctly to avoid exposing other parts of the PCB.

>>1507910
Best explain I seen in a while, thanks for laugh anon

>>1507988
If it got wet while it was powered, it's done for.

>>1508920
I'm guessing it's just covered in that sticky sort of dust that a PCB accumulates after years in a room with another person.

>>1508942
fans and speakers don't like baths. maybe just take it outside and hit it with canned air with extension straw, you might be surprised

>>1508920
H2O isn't conductive you genius

>>1508985

>>1507988
> lightly pressurized
Pussy. I drench my shit in windex, let it sit for a minute or two then hit it with a pressure washer. Pushes all the dust out of card slots and blasts all the shit off the surface. Follow up with a methylated spirit bath to displace all the water and make it dry faster. Have not had a single board, vintage or modern, fail, but just make sure you remove any batteries and don't get closer than 100mm or so otherwise you may blast stickers off.

>>1508988
Pure H20 is not conductive. The shit that comes out of your tap that is full of mineral and other contamination, is. >>1508985 is still a potato for making such a dumb unqualified statement though.

>>1509001
See also.

>>1509001
I assumed it was bait, since the assumption is that if it got wet while it was powered on, it probably wasn't due to spilling distilled water on it. Also it's H2O not H20, sweaty.

>>1508828
I CNC'd my own boards for several years but mostly get them printed now that those services have become so fast and cheap. CNCing the traces for chips smaller than SOP was almost impossible for my $1000 machine and things like solder masks, silkscreens, and through hole plating is also really nice.

>>1509021
>$1000 machine
Got a link? Considering getting my own machine, though PCB "routing" would be a secondary use. I wonder if a pick'n'place head for a generic 3-axis CNC machine would be viable, along with a solder paste extruder? At that point, doing a solder mask and silkscreen with some sort of felt-tip pen attachment wouldn't be inconceivable, though certainly a waste of time compared to offshore PCB production.

>>1509022
I use a genaric 4060 model from AliExpress with a better cutting head and a homebrew set of rotary encoders to track position. The base models keep track of position with dead reckoning so if the motors ever jam or slip it doesn't know and ruins everything. It took a lot of experimenting and practice to get consistently good results; don't expect to just plug it in and go like an inkjet printer.

>>1509032
How rigid is it, could I cut aluminium with patience enough?

>>1508803
My plan is to make a laser engraver to engrave pcb or photo coating, also the normal alternative is cnc.

>>1509054
The plan is to use DVD carriages, I`ve tried using the lasers but they are so tiny and they cut off their legs so soldering them is impossible.

Is it better to drill one sided pcb from the copper side rather than from the top side?

>>1509061
Assuming you're hand drilling, is there any other way?
> can visually check you are centered on the pad.
> no risk of the pad lifting as the drill comes through.

>>1509065
>> no risk of the pad lifting as the drill comes through.

I agree with your first point about centering on the pad, but your second point seems backwards, as the drill will try to lift the copper if the copper side is on top, and if you have some wood or something on the far side there will be no tendency to push the pad down.

>>1509067
I dunno, sure, a drill has a twist, but that's to pull swarf out once it has been removed by the cutting edge. I guess making sure you're using a good, sharp drill is the key. It's not a problem I've ever had at least.

Anyone have experience with this shit?
About a year before I started in my current job, the previous engineer decided to replace all the lights in the factory with "high efficiency fluorescent tubes".
They bought about 50 extra tubes.
Fast forward to now, all spares are gone and lots of bulbs are blown.
Turns out they're a plastic tube with a shorter than normal fluorescent tube inside and built in ballast (pic related). So turns out they pulled all the existing ballasts out of the fittings so I can't use normal fluorescent tubes. I could go to LED tubes but would have to rewire all the fittings so live and neutral were on the same end.
The Chink company they bought the tubes off have dropped off the face of the planet and any lighting suppliers I've asked have never seen anything like these tubes.
The ballasts are still good but I can't fit a standard tube into the fitting with the ballasts on it.
Anyone ever see these things before or have any idea where they can be bought?

>>1509105
Looks like you got chinked.
Replace all lights with led light bulbs and never look back.

Sensing circuit with lm324

>>1508985
Rain water through a basement window sure is conductive

>>1509125

Did Karly Kloss take up electronics?

>>1509129
*Brainlet.png*
I don't get it, is that because it's beautiful?

>>1509135

It looks like something a girl would do, if she were more interested in the artistic appeal than how it functions.

Is there a reason for that monstrosity looking like a jungle gym?

>>1509125
Wow, i bet FM radio signal on that thing is amazing

>>1509136
We don't use breadboards here, so any prototyping is done 'jungle-gym' style (I like that term more than freeform, I'm adopting it)

>>1509119
Looks like I'll have to.
There's a few hundred fittings to be rewired so

>>1509144
Be happy you have work son.
If there was nothing to repair you would get fired because nobody will pay you to do nothing.

>>1509125
uh, what's it sense?

>>1509150
yet the insurance industry still exists

>>1509150
Also make sure the leds are powered by smooth dc and not AC, nobody want to look at a blinking led all day

>>1509150
I have proper work to be doing anyway.
Don't know why I keep getting roped into this handyman bullshit

>>1509054
Laser engravers just rough up the surface of metal. They don't cut through it like you need for PCBs. Even powerful laser cutters can't do PCBs unless you coat the copper with acid paste that activates when the laser heats it.

>>1509197
that's the same stuff they use on fine-rule production pc boards, isn't it? tell me more about this photo-activated etchant

>>1509192
>proper work
no such thing, any work that needs to be done is proper work
in your case that is anything that involves moving electric charge from point A to point B, including but not limited to digging ditches to put cables in.

>>1509197
anything transparent or reflective like metals is near impossible to laser for obvious reasons
not to mention the high melting point
drill cutting that shit will always be faster, better and less smelly

/ohm/, I recently built myself a nice sturdy desk out of 2 sheets of 4x8x3/4" plywood with 45s cut out to meet in a corner.

I am looking to pimp my workstation out to help it help me better.

digging around on amzn I found some high temp ESD mat in various sizes I can lay on it. I'm considering buying more ESD stuff and having a proper pc/electronics workstation in my office. snap rings, snap tool, dummy plug with only ground connect, ground bus for hooking up desk mat, floor mat, wrist/ankle straps. (also thinking of having this on top for easy access for grounding other things.
are there good brands or stores for this stuff? amzn is spotty on products in this category.

i already have some panavises and jaws for them, though im always open to new ideas.

I have a soldering iron, but not a fancy station with a temp controller. recommendations?

also seeking recommendations for solder fume air filtration. currently office is in a basement, do i want some charcoal/hepa setup?

cont.

>>1509210 cont.
I have also been digging around on amzn for clamp mount arms to get things off the desk where I can.
led lamp arm
https://www.amazon.com/gp/product/B079BCYXSQ/ref=ox_sc_saved_title_3?smid=A1BN6ZFNGSN7IU&psc=1
monitor arm
https://www.amazon.com/gp/product/B01NH0HTM5/ref=ox_sc_saved_title_6?smid=A3W3STWCK1O3NR&psc=1
phone arm
https://www.amazon.com/gp/product/B079QY6RFQ/ref=ox_sc_saved_title_5?smid=A2OLS0OEBUDXH7&psc=1
laptop arm
https://www.amazon.com/gp/product/B00W58D4KI/ref=ox_sc_saved_title_4?smid=A3686AOLDPXXOD&psc=1
microphone boom
https://www.amazon.com/gp/product/B07H9VRPHJ/ref=ox_sc_saved_title_1?smid=A13Q9HWV9EQ56Z&psc=1
5diopter led light (i only found 3d and 5d, should a 5d cover all needs or is there a use for a 2ndary @ 3d?)
https://www.amazon.com/gp/product/B07B4L5NVJ/ref=ox_sc_saved_title_2?smid=A3U81HV8YX54FG&psc=1

im not shilling anything, like i said i dug, in each category, this appeard to be good choices for each, but i would like opinions.

>>1509212 cont
fuck it heres some more products im looking at equipping my desk with.
tv arm
https://www.amazon.com/gp/product/B07GFRP5HT/ref=ox_sc_saved_title_7?smid=A3W3STWCK1O3NR&psc=1
surge protectors
https://www.amazon.com/gp/product/B00019OWWW/ref=ox_sc_saved_title_8?smid=ATVPDKIKX0DER&psc=1
https://www.amazon.com/gp/product/B00019OXF8/ref=ox_sc_saved_title_9?smid=ATVPDKIKX0DER&psc=1
https://www.amazon.com/gp/product/B00EUB7C8I/ref=ox_sc_saved_title_10?smid=A3NZR5PY384PKK&psc=1
https://www.amazon.com/gp/product/B01MT3SRIJ/ref=ox_sc_saved_title_1?smid=ASF0S1GFIHF5V&psc=1
esd mat
https://www.amazon.com/gp/product/B00O0Q3ITU/ref=ox_sc_saved_title_2?smid=A2TAQNLPPQ8TIQ&psc=1

im not shilling anything. like i said i dug in each category, this appeard to be good choices for each. but i would like opinions.

>>1509216
Nigger just get a bigger desk.

>>1509238
nigger im not building a bigger room.

>>1509239
nigger add shelves to your wall

>>1509239
nigger this is all consumer electronics shit, take it to >>>/g/

>>1509141
>We don't use breadboards here
Why would you do that to yourself?

>>1507659
Tell you what: Get a petition together for Japmoot to remove the ability to use names or tripcodes, and I guess I'll have little choice. Meanwhile it's 100% legal, you can't get me banned for it, so how about you stop being a gigantic crybaby about it and just cope? Seriously I come in here and give good information/advice, and this is what you complain about? And by the way, how dare you attempt to infringe on my freedom of speech and freedom of expression? It's my CHOICE to be who I am, and it's not up to you or any other user of this site to decide who I am or how I present myself. So how about you back the hell off?

>>1509271
nigger i done did.

>>1509281
nigger shit.

>>1509292
not the anon youre talking to, but i just lumped you in with bepis due to your stupid high horse bullshit. ditch the trip, tripfaggot.
no one cares who you are on here, only you.

>>1509129
>>1509135
>>1509136

>>1509292
Do you know why avatarfagging is against the rules here?

>>1509125
>Sensing circuit
what does it sense?
does it make sense?

>>1509290
must be the Newports

>>1509321
>what does it sense?
anything

>>1509292
This has to be a troll. No one could be this flat out delusional.

>>1509317
>Do you know why avatarfagging is against the rules here?

it's also one of the "rules" that is never enforced, like the one against macro image replies.

and I admit, I don't understand why it's against the rules, unless it's for someone changing one pixel so they can post the same image multiple times in a thread.

>>1508240
>There were only like 4 of us on there and for Xmas the driver gave us all Bugles chips.

kek

>>1509352
I think because it's attention-whoring that goes against the concept of an anonymous imageboard. Going by the rules for which you can report someone, avatarfagging and signature use are put in the same tab, and there doesn't appear to be anything that image macros would fall under aside from "this post is extremely low quality", which is enforced occasionally, typically with regards to posts in broken english asking a very simplistic question which could just as easily be chalked up to the underage rule being enforced.

In any case, I can't see any difference between tripfagging (or posting images with carcinogen-sticks) and "signature use".

>>1509361
>"this post is extremely low quality",

if they tried to enforce that rule 4chan would implode

>>1509364
It's a relative scale, I suspect that both we and the janitors have been desensitised to a great deal of "extremely low quality" posts.

>>1509197
well I could just tape a permanent marker in it and it would draw pcbs faster than me anyway.
Also
Anyone ever found uses for BLDC motors that are used in cd drivers to spin the disk? I got some laying arround, and I`ve managed to find uses for the steppers, but the bldcs seem worthless.

>>1509410
I find permanent markers are pretty shitty when it comes to sticking to copper in an oxidising bath, nail polish is the way for me. Piping some sort of nail-lacquer into a thin (teflon) nozzle would probably work on a CNC machine.

As far as BLDCs go, how about a fan to blow away soldering fumes, a solenoid winding setup (with appropriate gearing), diy centrifuge, a mini power tool like a saw to cut PCBs, or a hendheld sander/dremel tool?

>>1509438
>diy centrifuge, a mini power tool like a saw to cut PCBs, or a hendheld sander/dremel tool?
thought o fthat but I`m stuck in the mechanical part, I have no idea on how to attach anything to the disk spinning head that houses te magnet.

>>1509448
You'd probably want to fasten something around the external rotor, be it a hollowed-out cogwheel or just a piece of haphazardly reamed steel tube that you can screw things on to. Friction-fit it well enough and the kind of torque those things can put out shouldn't be able to cause it to slip, and there's always loctite if it is an issue. If you're going to gear it down then getting an appropriate cogwheel and hollowing out the middle would cut out one of your steps. It may also be possible to simply screw an arbor of some kind through the top surface of the motor, but then you might start to face an offset torque.

t. never done anything like it

>>1509455
I`ll try your suggestions but sorry man, they don`t make much sense

>>1509448
>stuck in the mechanical part
move along to /diy/sqt

I have been using regular raspberry relays for a while (the blue ones) and they could be triggered directly by the raspberry's logical 3.3V
But now i need to use a bunch of latching relays for a battery powered project.
I only managed to find an array marked for arduino, but they are listed as 12V.
Wouldn't that mean that even arduino wouldn't be able to trigger them with 5V?
https://www.aliexpress.com/item/Retail-Self-Lock-Relay-DC-12V-6-Channel-Relay-Module-with-optocoupler-High-Low-Level-Trigger/32920431357.htm
It even says the trigger voltage needs a resistor, that doesn't seem as a logical trigger voltage at all, it seems like you are supplying current directly to the coild, so i would have to use a bunch of transistors to switch the relays from an external power source. Is this correct?

>>1509628
It does say "requires 12V to trigger" in the listing description, though since they say to put a ~6kΩ resistor in series with the coil chances are you can power it with less. What are you switching with it anyhow? If it's DC under 20A or so then MOSFETs would be a better choice. Your use of an "arduino" or "raspberry pi" branded hardware is less than optimal in the first place, even if it's knockoff shit for significantly less than the stupid markups we get from real sellers. Any reason you can't use this:
https://www.aliexpress.com/item/anal-beads/32902847358.html
Or this:
https://www.aliexpress.com/item/baby-car-mouth-vagina/32905958885.html

If you can read the fuzzy images or otherwise get the part number through inference of the legit boards then you'll be able to look up the relays' datasheet, but it likely won't do any good since the power rail of the thing is set to run on a higher voltage, with the correct resistors for the driving transistors to run on 12V. Try and run it on 3.3V or 5V and the LDO won't even function, let alone the relays.

>>1509639
I need to switch 230V AC so fets are not an option and ssrs are overpriced shit
i use raspberrys and arduinos for all my projects for one simple reason: programming them is some next level shit simple
i can literally put anything together nad have it work reliably in 30 minutes where using other ICs would take literally a week to complete the same task with little extra benefit apart from saving some cents and miliamps

the second relays you posted are not latching, the first ones look good though.
I mean 12v is not really a problem, i can just use fets or transistors to amplify the digital output with an old 12V phone charger adapter to trigger the relays, it's just that it's extra work where some chink store is maybe selling latching arrays that work directly with logic signals

>>1509646
>ssrs are overpriced shit
Correct, latching relays are probably your best bet then. My question wasn't why you're using arduinos and such, but why you're searching for products specially made to work with them. Any old 5V triggered relay will work with a 5V AVR. Arduino/Rpi boards are a good start when it comes to pre-built modules/boards, but searching solely for them will cause you to miss out on products aimed at the "ESP32", "IOT", "micropython", etc. crowds, which is why I'd advise simply searching for "latching relay board" or "latching relay module". How many do you need and what kind of current are you putting through them anyhow? If it's low current then TRIACs might be worth considering.

While I was looking for dedicated TRIAC boards (there aren't any) I stumbled across a UVB sensor with a little quartz-window IC on it, which makes me wonder whether you could use a quartz-window EPROM as a UV sensor.

>>1509653
I search for those keywords because it means the relays and sensors and shit i find is guaranteed to work with 3V - 5V logical level without any added circuitry

I need about 10 relays, they will be in the basements turning on heaters around the house.
I wanted to buy once of those nice 12 channel relay boards chinks make where you connect power for all the coils once and then you are left witch 12 logical pins to control each relay so only 14 wires total and that is it for 12 relays which is awesome.
I will maybe just ditch the battery and use the regular relays, since like i said chinks make them in 12 channel arrays and they cost 3x less than latches

>>1509660
Or, you can search for "5V coil relay" you dumb faggot.

>>1509664
not only are you wrong, but also a bully

is there anything you don't fail at?

>>1509653
>triac
>low current

>>1509669
/ohm/ is getting very retarded, I think the tripfags scared away the decent people

>>1509653
>which makes me wonder whether you could use a quartz-window EPROM as a UV sensor.

>>1508985
at high enough voltages anything is conductive.
>>tfw using a material with less conductivity than deionized water as a high voltage transmission line

>>1509671
But the tripfag doesn’t post on this thread anymore so it must be the one whiney ass anon who will shitpost for hours after a one post by a trip.

>>1509707
> only one person
Who manages to be on the board across all timezones, every day of the week. How delusional can you be?

>>1509710
>How delusional can you be?

whether it's just you, or 1000 autistic spergloards like you, the problem is that you never learned the first rule of the internet: if you don't feed a troll, it will go elsewhere. Bepis/Newport is the most successful troll I've seen in ages, and you are the reason why.

>>1509710
He must be really bored.

It literally is mostly one person. He fell off for a couple months and it was nice and quiet. And then a couple weeks ago he came back and sits there clicking every thread to search for tripcodes and post the same three pictures. It’s like 12hrs a day of nonstop bitching and moaning, then anon eats his tendies and gets some sleep, then comes back for another day of his main hobby.

>>1509719
it`s because /diy/ is usually a board where people don`t troll much. It and /out/ are the best boards.

>>1509653
>>1509689
Answer: yes.
If you connect a voltmeter across just about any two pins on a quartz window IC you'll see a voltage appear

>>1509724
>If you connect a voltmeter across just about any two pins on a quartz window IC you'll see a voltage appear

I assume you mean in the presence of UV. Is there a predictable relationship that applies to all quartz window Eproms?

>>1509725
no you retard the UV is just to clear the chip memory. Just buy a photodiode/transistor sensible to the band you are looking for.

>>1509723
>/out/ is the best board
And bepis has been posting there for years. Only difference is the lack of some cunt who samefags and posts 5 replies bitching every time he sees a name or tripcode

>>1509727
>no you retard

lol. I think moot refused to put thread specific IDs on boards like this just to create confusion. I posted the skeptical girl up there >>1509689 , and then >>1509724 seemed to imply that you could use an EPROM to detect UV, so I asked for him to elaborate >>1509725


this post is probably too long for you to follow, so feel free to flame moar.

>>1509728
Except /diy/ has never had tripfags. Ever. Look around. None. There was Mr T, but, if his posts were anything to go by, he was literally insane and prolly got locked up. End result, zero tripfags.

We like it that way. Why do you think none of us use them? And now you're showing your disdain for the board by avatarfagging? It's not trolling, nothing about it is clever, its just one fuck head pissing in the pool for the sake of being a disruptive cunt. Kill yourself.

>>1509736
>It's not trolling, nothing about it is clever, its just one fuck head pissing in the pool for the sake of being a disruptive cunt. Kill yourself.

do you even know what "trolling" means? Your response is what a troll craves. You have been trolled totally. Get it now? I doubt it; you're the kind of thickheaded moron that trolls seek out.

>need to draw the schematic and layout for a demoboard for an educational robotics software
>software in question was made to control a motor using a DPDT relay and a tip122, so the control scheme is the one in the left
>the board needs to be as small as possible, so I'm going with a L9110s H-bridge
>turns out the control signals produced by the software are incompatible with any h-bridge design out there

What do I do? Is there a cheap motor driver that will take PWM and DIR signals rather than make me control the h-bridge inputs directly?

>>1509777
Also, what if use logic gates to transform

A B => IN1 IN2
1 1 => 0 1
0 1 => 1 0
1 0 => 0 0
0 0 => 0 0

How do I find out which gates I need?

>>1509781

Figured it out:

A B => IN1 IN2
0 0 => 0 0
0 1 => 1 0
1 0 => 0 0
1 1 => 0 1

IN1 = NOT(A) AND (B)
IN2 = A AND B

Now, do I make logic gates out of transistors or what?

>>1509785
>he doesn't know about the 74xx1G logic gates
a couple of Pic related should do

Is multisim broken inconsistent trash or is it just me?

>>1509811
It's trash. Only good for the most basic of simulations. SPICE stuff usually is (Google up Bob Pease's rants on it-boomertastic).

>>1509805
It would still be missing a NOT gate

Or do I use the NAND gate? Like B NAND A

>>1509815
BTW, he told me to simply stuff the relay in the board rather than use the motor driver ic

>>1509815
you use the AND gate with one inverted input (fig. 6)

>>1509814
I've spent two hours trying to get a simple switch working. The same circuit will perform COMPELTELY DIFFERENTLY if I replace one switch with the exact same switch from the parts list

I can't even explain how broken transistors seem to be because their behavior is completely inconsistent. I'm 99% sure they just have random number generators connected to each pin.

>>1509821
I figured that out. That IC is really just 6 pins?

>>1509824
Yes it is, but it's 20 cents each. The relay is only really going to be "cheaper" because the client already has thousands of them in stock, otherwise the cheapest option would be to use discrete transistor logic gates, and the most space-saving would be to use a couple of that magic logic gate, instead of using a $2 motor driver ic. Thanks for the tip, buddy.

>>1509301
>>1509317
>>1509326
Stay mad. Meanwhile I'll do what I want and IDGAF.

what's a good book to read about power supplies?

>>1509880
care to be a bit more specific other than "power supplies" ?

>>1509880
if you don't need colorful history, read all the manufacturers' app notes you can get your hands on

I have a bandgap reference in a project that's 1.25V. If I want a lower reference voltage can I just divide that down with resistors and still maintain accuracy? Obviously 1% or better and metal film for the good tempco. Would a constant current source + resistor be better?

>>1509880
Switching Power Supply Design 3rd Edition by Abraham I. Pressman.

>>1509895
it's a shunt reference, right? I = E/R, your divider will just divert *some* current from the shunt

>>1509899
Like a TL431 or Zener? No. It's a bandgap reference like pic related

>>1509885
The only power supplies that I know how to diy are transformer+linear regulator, so I want to know more
>>1509897
thanks this seems perfect
>>1509892
will do

>>1509910
>will do

this isn't an in-depth app note, but it's an excellent starting point related to the design found in many current laptop power supplies:

http://file.yizimg.com/332467/2010080308450435.pdf

http://www.micro-bridge.com/data/CRpowtech/PT4201E.pdf

>>1509908
sure, why not. the op-amp output is very low-Z and shouldn't be disturbed by any reasonably sized voltage divider

I am getting some strange behaviour from my esp32. Sometimes it just stops working.
I have it connected to my pc via usb so i watched the serial output and from time to time i get a brownout error.
It usually works but sometimes not, i assume that the brownout happens during wifi connection sometimes since it needs short bursts of energy.
So i was thinking about connecting a capacitor between the 3v3 pin and the ground pin (3v3 pin is connected to the output of the onboard voltage regulator which has it's input connected to micro usb port on the esp32)

Do you think that should work?
What capacity should i use?

>>1509822
Complete shit. If the circuit is simple enough just build it!

>>1509822

>>1509926

Bob is getting on my fucking nerves. He needs to take a chill pill before I snap.

>>1509923
I have to submit multisim simulations for this project

>>1509940
simulators are crutches for the weak minded
a true designer simulates the circuit in his mind and takes any readings he needs from that

>>1509940
Maybe you just suck. I use multisim regularly and aside from the occasional convergence error it behaves consistently enough for me.

>tried making this
>https://www.youtube.com/watch?v=gTt_YBzJ_Dk
>cheap fuck so I don`t buy epoxy
>mix hot glue with charcoal on a pan covered with foil
>press on metal grid
>make one capacitor
>stores 1-2V before bubbling
>after discharge there is a 0.6-0.9V open circuit potential
>can push 50ma in short circuit
>vinegar and salt electrolyte
I wonder what is making this battery, maybe the steel of the grid is reacting with air, dunno

>>1509980
Also because I only pressed with my hands, and did not use that metalic sponge thing to improve connections, the resistance of each cell is like 1kohm. Top kek

>>1509822
I`ve designed my linear power supply in multisim, it seems you suck.

>>1509669
They've got what, 2 NPN voltage drops in them? So maybe 2V or so worst case scenario. If it's one of those common TO-92 TRIACs then I would't put want than 600mW being dissipated, which equates to 300mA or so. I'd normally call that a low current, though I suppose when it comes to mains it's definitely not small, especially for 240V systems.

And if it's in a power package then there's nothing wrong with using a TRIAC for much higher powers. But as I understand you'd need a bit of peripheral circuitry to drive them such that they remain isolated.

>>1509918
That picture is of a fixed duty-cycle, no? From what I've seen the more advanced isolated switching supplies use a dedicated feedback winding in the flyback.

>>1509940
Post a picture so we can see what you might be doing wrong? Spice has it's ups and downs but for the most part it gives a fairly faithful representation of what you're trying to achieve.

Decided to check out pyrography, to jewy to buy a pen when I have some gay shaped tips I never use anyway. I don't plan on using it for soldering again but out of curiosity how fucked will a tip be after being kept at about 430°C for dozens of minutes?

>>1510025
If you don't jab it in a wet sponge and thermal shock it to death, you should be fairly safe.

>>1509929
He's dead bro

>>1510002
nigger, triacs are the last resort in power electronics, they are used from suicide showers to HVDC converters.

>>1510045
I thought leaving a tip untinned for extended periods of time would oxidise it to fuck? If not then I use a copper sponge rather than a wet one so I should be fine then.

>>1509926
>Bob Pease's rants
top kek, he seems autistic but fun

>>1510058
Just add a little no-clean flux on the tip every now and then

>>1509777
Stepper drivers tend to use this exact control scheme. I'm sure someone makes the same thing for DC motors.

>>1510075
he took his craft seriously but he was very personable
Bob Widlar had a lot more fun tho

How do i simulate 5ohm resistance with just a wire?

>>1510179
What do you mean "simulate"? IRL or in a simulator? All real wires have resistance, but common wires will have a pretty low resistance per metre, 5Ω would require quite a lot of wire.

>>1510179
what else do you have

>>1510182
>>1510183
I need to test if 5V power supply can handle 1A without voltage drop and i don't have a heat dissipating 5 ohm resistor, i have the small blue ones, but i don't really want to waste 5 1ohms just to test this

>>1510187
Do you have an op-amp and a MOSFET?

>>1510189
yes

>>1510196
Then you can probably make a variable load using the MOSFET (or more than one) as the load-bearing element, alongside a current sense resistor with a low enough value that it can handle the 1A without getting warm (0.1Ω or smaller).

>>1510199
Why not just lm317 it in that case?
you wont even need an op amp just a single pot

>>1510210
>lm317
oh yeah

>>1510210
Can you post the diagram?

>>1510223
just google lm317 constant current
this isn't exactly revolutionary

>>1510224
but some of the power with that is dissipated within the set resistor

>>1509922
Cap will just cause more brownouts.
It will supply juice for the wifi burst, but then will eat even more current than the wifi to fill back up
So you will just delay the brownout by a bit

>>1510224
But your 'single pot' solution would be, given the power in question.
>>1510225
That's why I asked for a diagram.

I'm considering building an rc car from basics as an engineering project and my robotics fundamentals are lacking. I have a 130 size motor (unused in a school project that just hooked batteries directly to motor) I want to use with arduino but don't know how to shop for any of this shit - ESC or IC motor controller, how to pick a steering servo motor, find some cheapass radio transmitter/receiver or try bluetooth via android.... Can anyone point me in the right direction or suggest a path of least resistance? I'm looking for fundamentals to forge forward into more design - a bigger crawler with a clutch or maybe a plane or drone.

I'm trying to build a lamp flasher circuit that can switch off during the day and start flashing in the night. The quiescent current should be as low as possible during "off" state. This circuit will deliver pulses into the |ON/OFF pin of an LM2575 IC which drives a small light bulb.

The CD4093 quad NAND seemed perfect for the task so I decided to make a prototype with it. It has wide operating voltage range suitable for battery use (3..15 V) and very small current consumption (<1 µA).

I breadboarded and measured the circuit in pic related with a 5 V supply. It consumes a constant 0,43 mA of current during daylight. In dark the consumption fluctuates between 0,90 and 1,20 mA. The consumption of the LM2575 is not included in this measurement. According to the data sheet its standby consumption is around 80 µA.

Do you think there's anything to optimize in this design? If anything, I will probably choose a much larger value resistor R104 + slightly smaller capacitor C101 for a slower flashing cycle. Also the resistor divider formed by R101, RV101 & R102 always pulls certain amount of current depending on the trimmer pot setting and the light falling on the LDR. Would be good if the divider could be somehow switched off during the day.

>>1510227
fuck.
then how do i limit cap charging speed but unlimit discharging speed?

>>1510279
You can't without losing voltage. Normally you do that with a resistor and a diode.
>i assume that..
Did you verify your assumption? Did you disconnect the wifi and experienced stable operation?

>>1510289
>i assume that..
no frien, i am a man of science, so i added a 5s delay in code before and after the wifi burst is fired which increased the decrease of the brownouts greatly, confirming my assumption that adding a delay will give more time for the tiny little shitty onbaord caps to charge and get ready to be milked by the wifi

if i can solve it this way, i will try to wire power directly to the 3v3 rail which bypasses the voltage regulator, that way it will eat up an extra buck converter but meh, it's worth the stability

>>1510267
Do you have enough 400mA fuses? The 2575 needs about 9V in to get 6V out, but you only have about 7V and maybe less after a while. The initial cold resistance of the 2W lamp is much lower than the nominal 18Ω, which means the voltage starved 2575 will draw a lot of current if it starts up at all. Can you measure the cold resistance? I would be interested to compare it to the 8% I found for my 12V/10W halogen lamps.

>>1510210
0.1 < 1.2

>>1510264
try the lolduino thread

>>1510267
>LDR
use a photodiode
>LM257x standby current
use a low-Rds(on) switch in series with its Vin, or better yet, for a lamp, put it in series with Vout
>logic inputs floating
ground them
>logic inputs at all
use a microcontroller

>>1510295
>about 50% more power available to the ESP32 instead of causing the heat death of the universe
unironically that is a good idea and solves two problems with one stroke

>>1510350
>initial cold resistance
I totally did not take that into account. I measured 2 Ω from the lamp.

>>1510392
>use a photodiode
Very good suggestion. I will try using some superbright leds for sensing the light level.

>>1510392
>>logic inputs floating
>ground them
I grounded the unused CD4093 inputs and in light its current consumption decreased to 0,23 mA. and in darkness the consumption fluctuates between 0,08..0,18 mA. I'd say that's a decent improvement.

Still not sure what I should use for driving the lamp. I was thinking to go with an LM2575 but as >>1510350 mentioned two 18650 Lithium batteries in series might not provide enough voltage for it. If I use a PNP transistor or P-channel FET instead for driving the lamp, I would need to have a current-limit circuit to protect the lamp the battery voltage.

>>1510469
the XL4005, also commonly available as a module, has a 0.6V dropout and can run at a 100% duty cycle so should still deliver plenty of power when the battery drops low. shutdown current is on par with the LM2575. on the other hand, maybe you should prefer the load to stop running entirely in case the battery reaches 3.2V/cell

>>1510433
>I measured 2 Ω
Thank you. To let it flash you would need a short start-up current of 6V/2Ω=3A, but the ON current is only 2W/6V=(1/3)A. If you limit the current to that value the lamp will come up slowly, no flashing. If you have a PSU that can set voltage and current, set it to 6V and (1/3)A and see how the lamp reacts when you connect it.

>>1507367
>allows you to draw little pictures in the quick reply window
mind explaining how you do that?
I'm using 4chanX but I never saw that.

I didn't find /hard/ so I'm asking here.

I'm working with an STM32 and want to use a timer interrupt. My goal is to have function that starts the timer for n milli- or microseconds (n is a parameter) and then executes another function on the interrupt. Then the timer should stop.

Is it a bad idea to keep enabling and disabling the timer (using TIM_Cmd) to accomplish this? Basically all I want to is execute a function later, but still do stuff in the meantime.

>>1510974
>Is it a bad idea to keep enabling and disabling the timer (using TIM_Cmd) to accomplish this?
why would it be? just consider all the cases that are applicable to your situation e.g. can the start function be called while a delay is in progress and what should happen if so; the start function and its interplay with this/other interrupts, and so on

>>1511073
I was wondering about performance issues and if there was a better way to do it, since I will be using it a lot (the main purpose will be to check for silent gaps between incoming messages).

>>1511103
>check for
does the action need to be performed immediately at timeout? could you use the SysTick timer, copy the value to a static var upon message receipt, and subtract now-time from that-time to get the interval?

>>1511106
That sounds like it could be a good idea, but I'm wondering how I would handle timer overflow. My current idea is to use a "state" variable and the timeout action would check if the current state was "waiting", and if so change it back to "idle". Another interrupt is triggered when a new part of a message has arrived, which does something depending on the current state. So if it is still "waiting" it means that timeout hasn't occurred and that the timer needs to be reset or disabled. That's the gist of it but I'm not sure if it will work out in practice because I don't have much experience with programming on this level. Your suggestion does seem like it would be easier to debug and reason about.

>>1510494
>>1510605
I ended up using an LM 317 to provide constant current. It gives a nice gradual turn-on of the lamp, excellent for my application (a miniature lighthouse). An IRF 630 works as an on/off switch.
Simple, cheap and inefficient but has zero quiescent current consumption when the lamp is off.

Now, the next challenge is to plan some scheme to recharge the Li-Ion cells...

>>1510974
>I didn't find /hard/
Oh yeah, that existed. There's an arduino thread up though, not that I'd particularly recommend you go there.

>>1511134
Use a dedicated chip for the Li-ions, like a TP4056. Not sure the preferred way to provide overcurrent and overdischarge protection however.

>>1511143
Why what's wrong with that thread?

>>1511165
Nothing that wrong, just that it isn't likely to have people more experienced with microcontrollers than the established /ohm/.

>>1511129
>handle timer overflow
are you sure you really need to? overflow happens only once every 49 days or so. if you use unsigned arithmetic you would need to worry about it only if you anticipate more than 49 days between commands
anyway, if you really do, it's trivial to write your own tick implementation that uses 64-bit ints thanks to the magic of weak references, see Drivers/STM32*_HAL_Driver/Src/stm32*_hal.c
and
http://www.micromouseonline.com/2016/02/02/systick-configuration-made-easy-on-the-stm32/
if I may ask again, in a slightly different way: do you need an actual transition to an idle state for anything outside the message processing code, or is it enough to use the time-since-last-part in the receive-part routine as a factor in the waiting state's next-state decision?

>>1511134
here I am adding some considerations for outdoor projects that may or may not apply to you, feel free to ignore if not
there are TP5100 modules that will charge 2S, even 3S. you can probably power them from a 12V solar cell. bonus: you can use the solar cell voltage to enable/disable the circuit and lose the LDR, however, you will also need to somehow regulate battery temperature, and you will need to select balanced cells, or use new cells
https://www.aliexpress.com/item/TP5100-1S-2S-3-7V-7-4V-Lithium-Lipo-Li-ion-18650-Benis-Charger-Charging-2Benis/32910466083.html
in any case you can trivially find 2S protection boards
you may also need to ensure that the capacity of your cells is adequate to run your lighthouse over those long Michigan winter nights, especially given the temperature restrictions of Li+ cells. you might want to add some temperature-controlled, gentle resistive heat to the battery somehow
your circuit seems to draw about 25mA average over a blinking cycle, so round up to 30mA to account for the CdS cell in the very worst case, and divide by a conservative 2000mAh for a lifetime of 66 hours in the ON condition

>>1511144
>>1511171
I plan to make a tiny lighthouse just for decoration/indoor use. Maybe I should consider using a different power supply than batteries, maybe a small 9 V wall cube that can deliver 250 mA or something like that.

>>1511214
If indoor then a wall-wart is likely a better option. I kinda wish houses had a 5V rail by default. You wouldn't pull that much current through them, at least not enough to warrant thicker wires than are in the walls for AC already.

>>1511134
The 317 is not required and drops too much. You can use the FET as both switch and limiter. The optional grey parts allow a delayed current limit.

>>1511252
Low voltage lines are becoming common in new homes with all led lighting, aren't they?

>>1511305
Not to my knowledge.

With low-voltage DC, the problem with the much higher currents isn't just the maximum current rating of the cables, but also the resistance. A 1V drop at 120V is negligible, at 5V it's too much.

There might be a viable case for e.g. 48/50V DC. That would reduce the cost of the PSUs for low-power applications (e.g. lighting, and devices which are powered from wall warts) without needing to use thick bus bars.

But then you have a chicken-and-egg problem. No-one's going to make devices which work off 50V DC when such systems are uncommon, and no-one is going to install such systems until there's a use for them.

>>1511305
Efficient LED lighting (i.e. no dropper resistors) needs to be constant-current. You can't have a constant-current DC rail in your house since the required current would change as you turned lights on and off, meaning you need a seperate PSU for each light or set of lights. At that point it's easier to simply have each PSU being an AC to constant-current DC converter, as opposed to having one large ~48VDC PSU that feeds a series of constant-current drivers.

At the moment I believe we're still at the point of having each individual LED light having its own power supply instead of a power supply for each set of lights on the same switch, but that's not a particularly difficult change. Just have the power supply itself either in the attic with an array of constant current rails that lead to each DC bulb, or just put the lot in series, since a fair few 24V bulbs can run off buck-converted mains. But then you have to standardise the current that each bulb takes such that replacing a dead 24V bulb with a 12V bulb won't ruin things. I use the term "bulb" loosely. Having all the bulbs in series makes the power supply far easier to make, but if one bulb dies the lot will die (unless you have a few "balancing" lines).

Having such a power supply would also mean you could give it IOT functionality and control the current being fed to your bulbs through your phone with ease, or just with a wire or two going from the switch to the PSU. With the multiple rail method, you could even tune the current to each bulb for more even lighting.

>>1511345
>OT functionality and control the current being fed to your bulbs through your phone with ease,
fuck off zuckerborg

>>1511449
Just saying, dimming by controlling the PSU directly is a much easier way to dim your LEDs than worrying about whether you're getting dimmable bulbs.

>make breadboard circuit
>30ish components
>ok lets go to perfboard/make a pcb
>either misolder something or design something wrong in the pcb in ways that would be too much trouble to repair
Did you suck ass when you were starting in electronics? I sure do

>>1511451
I avoid perfboard even for simple prototypes, it's simply awful to solder onto and I frequently delaminate pads (it's the shitty FR2 stuff). I've enough nail polish, etchant, and copper-clad lying about that I've no issue with etching myself a PCB every now and then.

>table so covered in wires, meters, hardware and shit that i don't even have a space for a glass of water
>clean it all up
>same exact thing a week later
i am starting to think my table is haunted

>>1511468
Same thing here, apartment broke student fag that has no space for bench.
https://www.youtube.com/watch?v=t513IZ5V9Nk

Do you think this heatsink can handle this 230v 50w led without the fan connected?

>>1511479
Eh, maybe.

>>1511481
I will leave it on for couple of hours and see if it melts.
Worst case scenario i can use a small ac dc transformed to power the fan but it would just be an extra ballast to log around

>>1511483
70 deg on the led after 5 mins of runtime... guess i better do the fan just in case

>>1511497
Poor thermal contact? If so, fan will not help much.

>>1511479
>>1511497
you're using thermal paste, right?

>>1511524
of course

QUESTION: Does anyone use ZenitPCB to design circuit boards?

I want a few 40mm x 111mm two layer boards from a china manufacturer.
I need to create the file to place the order.
The board is quite simple with few traces and only one component pad location.
The rest is just mounting holes and plated through holes for attaching wires.
I can only get as far as the board outline. I can't figure out how to place holes and traces.
I've looked for instructions online and in the included 'help' file but they all start by opening a file with a semi-complete board and the jew-tube videos just show pointing and clicking with no audio to explain what's going on.

I'm aware this isn't the best program to use but I'm on < 1mbs DSL so downloading a 1GB file to install KiCad isn't an option.

I'm sure I'm just being stupid about this but I've tried for a couple of days before asking so, no bully please...

>>1511538
>ZenitPCB
How you even find this shit is beyond me.
Just download eagle and be happy or write to zenit whatever support

>>1511538
>I'm on < 1mbs DSL so downloading a 1GB file to install KiCad isn't an option.
just start the download when you go to sleep. turn off everything else on your network for bonus speed

Do you need to ground the USB-B connector shield when designing a PCB that will have them? I designed a PCB, had it made in China, soldered everything and it works fine, but I'm wondering whether I should ground the little tab thingies on the USB-B connectors or if they should be left alone. How do professionally made products deal with them?

BTW, I'm not going to etch a single board until I get my high-precision mini-CNC built, the pcb from the china house was simply perfect

>>1511538
I second the Eagle recommendation. It's very small, very easy to use (also to screw up) and it generates the Gerber files on its own. Eagle > Altium

>>1511615
The tabs are normally soldered to GND, for electrical and mechanical reasons.

So turns out that connecting a pc fan to 12v makes it spin really loud. What resistance should i use to slow it down to say about 60%?

>>1511640
Use a potentiometer + LM317 chip to make an adjustable speed fan.
Or an NE555 + potentiometer + transistor for adjusting speed by PWM.

>>1511678
That would be too needlessly complicated, i just want a fixed 60% speed on the fan.
I calculated the resistance i need is about 47ohms, i just hope those small blue chink resistors can handle dumping .5W

>>1511679
Use a few big diodes in series to drop the voltage if you don't mind heating up the diodes

>>1511682
how much are diodes dropping? i have some of the standard chink ones that are black with a grey stripe

>>1511468
the steady state of workbenches is full

>>1511497
stop using kapton tape you fucking mong

>>1511538
>1GB
that should take you perhaps 3 hours or so. it's truly worth it
also
>1GB
pressing X to doubt

>>1511615
Henry Ott says yes

>>1511684
nominally 650mV each

>>1511693
>>1GB
>pressing X to doubt

Well I need some help.
I said I'd help a group that's designing rockets (not commercially) and was asked to look into antenna theory and design.
I read a lot about it but frankly it seems really complicated because I'm hardly aware where I should start.
The goal is to maximise the range of the antenna without data loss, all while keepint weight and complexity of the telemetry within the rocket low. (the ground unit can be bigger).

I thought that generally speaking a directional antenna is better than a normal mono or dipole. I basically want a high efficiency and am okay with a low bandwith.

budget is private so somewhat limited, then again antennas tend to not be all that expensive.

The things I have no idea about:
1. Will a less efficient antenna in the rocket (dipole w/ reflector etc) bottleneck? Or can I use a bad antenna to send and use a very efficient one to recieve the (worse) signal?
2. Does a Yagi-Uda antenna make sense? Or do I want to use a dipole antenna with a reflector?
3. where do I look at a practical guide? I'm interested in application, less in theory (unless I need it to understand obv)


Essentially what do I need to look into?

Thanks for helping a anon out!