[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/diy/ - Do It Yourself

View post   

File: 942 KB, 3840x2400, 1653035179778.jpg [View same] [iqdb] [saucenao] [google]
2799790 No.2799790 [Reply] [Original]

Prev thread at bump limit. I don't often hangout here, forgive me if this isn't perfect.
Prev thread: >>2794153

>I'm new to electronics. Where to get started?
It is an art/science of applying principles to requirements.
Find problem, learn principles, design and verify solution, build, test, post results, repeat.
Read the datasheet.

>OP source:

>Comprehensive list of electronics resources:

>Project ideas:


>Principles (by increasing skill level):
Mims III, Getting Started in Electronics
Geier, How to Diagnose & Fix Everything Electronic
Kybett & Boysen, All New Electronics Self-Teaching Guide
Scherz & Monk, Practical Electronics for Inventors (arguably has minor issues with mains grounding)
Horowitz and Hill, The Art of Electronics

>Recommended software tools:
KiCAD 6+
Logisim Evolution

>Recommended Components/equipment:
eBay/AliExpress sellers, for component assortments/sample kits (caveat emptor)
Local independent electronics distributors

>Most relevant YouTube channels:
Moritz Klein

>microcontroller specific problems?
>I have junk, what do?
Shitcan it
>consumer product support or PC building?
>household/premises wiring?
More rules-driven than engineering, try /qtddtot/ or sparky general first
>antigravity and/or overunity?
Go away

>> No.2799794

Is there a reason the thread copypasta doesn't recommend digikey/mouser/arrow/avnet for partprocurement? Generally a higher price point for the same shit? I'm a bit ignorant to the landscape outside the above, mostly industry experience buying stuff and shipping it fast on company/customer dollar.

>> No.2799799

I need an amplifier circuit that can boost a very small DC signal to around 10-30V while keeping it's shape. The DC signal might be positive or negative and I want the output to also have those two halves. How could that be achieved?
I've made one before for square wave but that had a fixed signal shape and I didn't need it to go in reverse polarity. This one is new for me.

>> No.2799808

A comparator does exactly this: compares two voltages (here input and ground) and outputs positive voltage for pos input and vice versa.

>> No.2799810

Would it work with sine waves as well?

>> No.2799811

Instrumental amplifiers keeping the shape

>> No.2799818

Alright, so the application in question is electrochemical analysis. I've got an equipment that has precise control over voltage and current and can do all sorts of voltammetry and pulse sequences but it can only go so high in terms of voltage. I'd like to amplify that while keeping the shape and accuracy of the signal so what'd measure off of it will be an accurate representation of the chemical processes happening in the electrodes.
From what I can see at a glance from wikipedia it seems an instrumental amplifier would work for that end and the schematics seem simple enough for me to try building one for experimentation.
With that background on the intended use, can you think of any better fit for this purpose?

>> No.2799833

They're still listed among other in the last link. They'd be worth recommending specifically too because how can you not recommend building up orders to get free delivery and have stockpiles of basic components because you added 20 bridge rectifiers and 100 1n4001 etc. that one time.

>> No.2799841

>I don't often hangout here, forgive me if this isn't perfect
Did you read the part at the bottom of the OP that you avoided copy-pasting that said "bake at page 10, post in old thread"???

>> No.2799847

They're included in Octopart. Octopart is an aggregator for the big vendors like Digi-Key and Mouser, and a lot of smaller vendors as well. They also generally have a better parametric search function, allowing you to set a minimum value for the maximum input voltage, for example. Though can get a bit strange when different vendors have different data formats.

The reason I included LCSC separately is because Octopart doesn't scrape all the unique parts on LCSC, it only shows the western parts that LCSC shares with the other vendors. LCSC has a lot of interesting cromulent ICs that you can't get anywhere else, like linear high voltage LED drivers, TP4056s, along with chinky copies of western parts that go under different names, like the GD32F103. Their passives and jellybeans are also generally cheaper, though there is no free shipping threshold. They ship worldwide regardless of location, so if like me you're in a country where Farnell and RS are the only vendors available, you can see why it would be worth not paying $10 each for one of two available SMA connectors. Keep in mind the shipping thresholds though, more expensive part prices might be offset by free shipping.

In no way am I recommending using AliExpress or whatever as a replacement for proper vendors for anything but the most bog-standard parts. Or maybe surplus parts for a game console or laptop. I got out-of-spec TL072s once, and IR2148s that were dead on arrival. Stick to the proper vendors for things that matter.

>> No.2799850

30V is kinda high for electrochem, but sure. An instrumentation amplifier is the sort of thing you want for amplifying a weak signal without incurring offset, but such an amplifier can only provide so much current. Electrochemistry can easily get over 1A, which at 30V is 30W. That's a lot of power for a linear precision amplifier to dissipate. For that kind of power, what you'd end up doing is designing a DC-capable audio-style amplifier stage that runs off a split-rail supply (e.g. ±32V), and using the instrumentation amplifier (or similar op-amp circuit) in a feedback loop in order to make the output perfectly follow the input. It's also an option to go for a switching solution, though with a bipolar supply that can sink and source current I'd advise against it unless you really need high power.

Also the feedback loop of the existing equipment may interfere with this circuit. When it sets 0.1A through an electrochemical cell, it's constantly changing its voltage while measuring current in order to keep it at 0.1A. If it won't measure anything above 0.5A, then outputting 1A wouldn't work, even if your amplifier could take it. You may need to augment the measurement ability of the existing equipment, or even use your own circuit to set the voltage and current independently with a CC/CV amplifier combination.

I'd write down some specs. Maximum voltage, maximum current, accuracy of voltage, accuracy of current, that sort of thing.

>> No.2799940

Max voltage is 30V. It is that high so I can dabble in electrochemical synthesis if I want, might happen and I want to future proof this project. Max current is around 100mA, that should be plenty. Commercial electrochemical reactors have an accuracy of around 0.01mV and 0.1mA, so that's what I'm sort of aiming for.
Sometimes in those processes the voltage isn't the variable controlled, but instead the current. So you set it to something like 4mA per mole and the device gives it as much voltage as it needs to hold that current. My little potentiometer can do that, but at much lower power, I'm not sure how to amplify that. It can reach the currents I need but the problem comes when I start doing synthesis at larger than analytical scales and need higher voltages to maintain that current.

>> No.2800046
File: 161 KB, 920x842, png-transparent-solid-state-relay-electric-potential-difference-electronic-component-alternating-current-others-relay-highdefinition-electric-current.png [View same] [iqdb] [saucenao] [google]

I want to go with relays instead of the mux IC now. Just because I can. It has more Sovl

I have a bunch of old very sovlful looking transparent relays. But they have such an usual footprint that I won't be able to assemble more of those PCBs - 5 pcs is the minimal quanitity

Please recommend me some common, easy to get (ideally on chinkexpress) DPDT relay.
Only requirement is it should have a nice lout *klack* and be PCB mountable.
(EXTRA DOUBLE TRIPLE) bonus point if it is transparent

Also can I use an NGATE MOSFET to switch them? Most of the time I see BJT trannies being used.

>> No.2800055

Almost all microwave ovens have two of these, they’re as common as… microwave ovens in dumpsters. I haven’t bought a relay in years.
You can de-case them and make little acrylic boxes for them and light them with leds like those tube amps. They will work fine without the case, I have bigger relays that never had cases.
You can go full “clack” if you buy a relay that electricians call a “contactor” for some hillbilly reason.

>> No.2800062
File: 443 KB, 1199x829, 4EFCE4C3-BD4C-496A-AEB3-994BFBAB5D2A.jpg [View same] [iqdb] [saucenao] [google]

Get an opa541, it’s 10 amps.
It costs like $10 too, so I don’t know why you’d ever need much else for that kind of thing.

>> No.2800063

Oh, and I forgot to mention those lazy fuckers use this shit in missiles to control servos and whatnot instead of designing something with 3055’s.
Although, at $10, I can’t blame thme.

>> No.2800082

recc me some klickity klack relays

>> No.2800084
File: 99 KB, 720x941, 1703888555592-0.jpg [View same] [iqdb] [saucenao] [google]

I found those HK19F on ali. Dirt cheap shit, I hope this magically doesn't transate to the longevity...
Guess I'll dremel the cap of and glue a plexi plate or something on

>> No.2800097

China makes better relays because they don’t give a fuck about environmental laws. U.S. probably can’t even make a mercury wetted contact. China can.

>> No.2800106

In my country that turns into 100, plus importation and taxes and it's almost 300. Any cheaper option? Can I make one with discrete components or would that be too complex?

>> No.2800116

Just want to put this out there. Thank you for giving better advice, by whatever margin, than /g/ on actual technological discussion. Threads like this >>>/g/100495470 are a regular occurrence there.

>> No.2800153
File: 622 KB, 1014x621, 431D9497-E468-4DC0-8438-BAC7A7C06299.jpg [View same] [iqdb] [saucenao] [google]

You can absolutely make a good op amp with discrete components, and doing so with even the cheapest transistors tends to really boost the current available at the output.

>> No.2800157

>Max voltage is 30V. It is that high so I can dabble in electrochemical synthesis if I want
Aren't electrochemical potentials like no greater than 5V differential? Maybe if you get some really concentrated reagents or do molten salt stuff the nernst equation can push them past 8V but idk man. 30V sounds overkill unless you're dumping 3/4 of that voltage across the series resistance of your system. Running at an overly high voltage in that kind of situation sounds like a recipe for unwanted side reactions, especially dissolving your "inert" anode. Though if you have a glassy-carbon anode I'll retract my last statement.
>Max current is around 100mA
100mA at 1V with a 31V power input means 3W of power dissipation worst case, you'll either want a power package monolithic amplifier or a BJT output stage. Combining a monolithic audio amplifier (LM1875, LM3876, etc.) with a jellybean op-amp is an idea, see EEVblog's vid on combining opamps.

>I'm not sure how to amplify that.
Does your existing piece of equipment go up to 100mA? If not, you could just amplify the current output by converting it to a voltage and back with some op-amps and resistors, or even modify the existing current sense resistor. But then you'd have the issue where you have to use a conversion factor on the current-per-mole.
It's a pain, but you may want to try to modify the firmware of your existing equipment, or build an entirely new setup that has its own display and control mechanism.

I was using Panasonic DW2 and JW2 relays recently, the DW2 is pretty small and I think even perfboard compatible.
For larger but transparent ones, try these:

Y-you're not from Brazil, are you? Go to sound-au.com for some discrete op-amp designs.

That's a bait thread. At least compare this to their gnu/linux general, but most of their threads are too fast to get good advice anyhow.

>> No.2800160

I plan to try this with TIP31 and TIP32, maybe KSC1845/KSA992 too or op amp or both. For first ever I'll be content unless it shorts supply and lights the speakers on fire. That wouldn't even be so bad I can test with the speakers picked up from by trash collection area.

>> No.2800165

Yeah, but I also want to be able to do electrosynthesis so I want something beefier.

>> No.2800172

>Also can I use an NGATE MOSFET to switch them?
Yes but there's no real benefit to using a more sensitive MOSFET when some 2n3904 saturation goes brrr on any logic level

>> No.2800209

There are a few benefits to using a FET for switching a relay:
>lower power dissipation - means potentially more reliable, and can be in a smaller package
>doesn't need a series resistor if you use a logic FET instead of a power FET
>less load on the signal source - could be a problem with a dozen relays
>insulating channel between the channel and gate means less likely to fry your MCU in event of failure
Though if you're sticking to THT parts your selections are limited. And by that I mean you have literally one common logic FET, the 2N7000, and it's a bit shit. The resistor point is also moot if you use a logic transistor (i.e. a transistor with resistors built in) or even a logic transistor array like the ULN2003. It's a darlington array so not only do you not need resistors, but you don't need to worry about input current either. Plus it comes with freewheel diodes built-in. It (as well as similar arrays) are the obvious parts to use for relay driving, as they minimise the required components. There are MOSFET versions of the ULN2003N that might beat it for current capability and thermal characteristics, but they're less common, and probably not found in DIP packages.

The main downside of using FET-based relay drivers is ESD sensitivity, but when they're soldered into a circuit with a CMOS output driving their input, this really isn't a concern. FETs might be more prone to damage by inductive spikes than are BJTs, but either way you should be using a freewheel diode.

>> No.2800238

I bought a bunch of HF electronics, but guess what

they are all stolen by some faux angel.

I regrette moi, but I want to just buy more of them.

I'm still wondering if you girls will just disconnect everything 'once it works', just cuz the voices tell you.

I'm really wondering why DigiKey is so expensive. Mouser doesn't seem to have anything. The best way to get parts is still to buy the completed product, and break it into pieces.

I have some "lifepo4" but they're DED. I don't know, phosphate explodes you know?
But maybe the iron mountain... the lithium the lithium...

I think it sucks. Whatever I do.

Anyway, I could powder out of one of these cells, and the result would be -- that I spread them way above the house... like at 800 meters?... The lithium might be 'detoxification' for all the mountains, but I'm not sure, I'm not sure..

>> No.2800241


where's the app?

where's the "I don't care where I shop" app?

>> No.2800288

>mercury wetted contact
So I should NOT open them, I guess??

>> No.2800290

Regardless of userbase, imageboards tend to become shit as soon as they get faster. Add the masses of bots and retarded underage normalfags and you have the cesspit that is current /g/.
hopefully /diy/ will always stay /comfy/

>> No.2800307
File: 128 KB, 541x436, file.png [View same] [iqdb] [saucenao] [google]

aaaah fucking retard chinks and their pozzed bogshit website

>> No.2800309

take your meds

That's why i'm in favour of a "tree chan", where the faster board categories just get split into smaller slower boards. A general thread that's been sufficiently fast for enough time becomes a new board itself. Zoom out and you see the entire hierarchical tree. The mascot would be a leprechaun.

>> No.2800316
File: 99 KB, 640x640, 97c10a80b0dfa1881e7f43ee647c69c7559294ad95590ad4a09ed1f640be1946.jpg [View same] [iqdb] [saucenao] [google]

weirdly I can still log in over the browser from my phone and just placed an order this way.
Wondering if I got banned because I opened 2 disputes in the last 2 months or so getting scammed over empty epoxy blocks larping as ICs
That would be really invigorating since I've been using that bogshit scamsite for ~4 years now and propably have over 1k orders

>> No.2800322
File: 23 KB, 804x588, Bildschirmfoto_2024-05-16_10-36-17.jpg [View same] [iqdb] [saucenao] [google]

why do they always use pnp trannies to switch relays?
can I also use a mosfet?

>> No.2800324

Anon, I need a project for my final year humiliation ritual. What do?

>> No.2800336
File: 3.36 MB, 4000x3000, DSC_2511.jpg [View same] [iqdb] [saucenao] [google]

Doesn't look good, does it?

>> No.2800337


>> No.2800338

you generally want at least one side of your large electromechanicals to be tied to ground
picrel is floating

>> No.2800340

so PNP?

>> No.2800341

Anon, can you tell me what's going on with the output on the tranny? >>2800336

>> No.2800342

1) It's NPN, not PNP in your picture.
2) "BC109": obviously that's an old circuit or a circuit drawn by an old person. Back then the commonly available MOSFET needed 10V to switch on completely. TTL had less then 5V, so bipolar transitors were used.
3) Yes, you can!

>> No.2800344
File: 71 KB, 1522x410, Bildschirmfoto_2024-05-16_13-50-23.jpg [View same] [iqdb] [saucenao] [google]

I also want to use it to switch a 1khz pwm signal.
But it doesn't really work. >>2800336
It says 10ns turn on time in the datasheet https://www.farnell.com/datasheets/694801.pdf

>> No.2800345

The osci output is very jittery and inconsistent
it has those weird spikes

>> No.2800367

>you generally want at least one side of your large electromechanicals to be tied to ground

Why? And when you say "at least one side", are there situations where you ground both sides?

>> No.2800374
File: 2.27 MB, 4000x3000, DSC_2513.jpg [View same] [iqdb] [saucenao] [google]

BC33 good 'nuff

>> No.2800378
File: 1.96 MB, 4000x3000, DSC_2515.jpg [View same] [iqdb] [saucenao] [google]

Fried another tarduino plussing in my laptop while osci/psu was connected

>> No.2800380

>Fried another tarduino

>> No.2800381

Nakasashi sex with all the JK assembling the PCB

>> No.2800384

>go absolutely insane trying to fix a circuit made of opamps
>output slams to rail regardless
>give up
>last opamp on the breadboard
>make a simple buffer
>1kHz sine clipped
>swap opamp
>it werks
I may be retarded,

>> No.2800387

Ground loop or whats that called

>> No.2800410

So how hard is it to read from an analog sensor and control a BLE device according to some function of sensor data with an ESP32? I have very little embedded experience, but know how to program in python and C#.

>> No.2800419

Depending on how quick your uptake, moderately. Can use Arduino, Reading ADC is hello world level exercise. Connecting BLE likely is boilerplate you can copy from examples, after that the controlling part difficulty should depend on the device.

>> No.2800421
File: 271 KB, 1019x862, retardproject.png [View same] [iqdb] [saucenao] [google]

Beginner here trying to get a stepper motor going, any idea why I'm not getting a proper read trying to set the current limiter on an A4988? I've got everything hooked up properly, sleep and reset is bridged, the motor hasn't been hooked up once so I don't think I've fried it, but trying to use a screwdriver just shows up as ~0.3V on the multimeter, and as I spin the screw it stays there until it randomly drops to 0 before another turn brings it back to 0.3. I just don't know what to try anymore.

>> No.2800468

> is it live, or is it memorex?
> am I retarded, or is it zheng scamming me?

>> No.2800470

I can’t tell what your mulitmeter is set to, AC or DC volts, but I think the stepper motor controller is going to try and give out pulses, so it won’t necessarily read correctly on DC or AC, although AC is more likely to be closer.

>> No.2800490

>zheng scamming me
To be fair my current PDIP stock comes from Digikey so if the opamp was shot it was probably on me.

>> No.2800584

Best brands/stores for high voltage DC power supplies (around 300V 10A) Google just brings up aliexpress crap.

Ideally also has some sort of remote control ability.

>> No.2800587

Meant to include a question mark somewhere in there lol

>> No.2800590

You realize that’s gonna be more than double what you can pull out of a standard north american outlet eh?

>> No.2800596

Gotcha, thanks.

If you think I don't know what I'm doing, you'd be right.

>> No.2800599
File: 135 KB, 1323x659, bk power supply.jpg [View same] [iqdb] [saucenao] [google]

>Best brands/stores for high voltage DC power

if you use digikey's parametric search, you'll find plenty of contenders
unfortunately, high power comes at high prices
very high, cheech and chong levels of high

>> No.2800606
File: 34 KB, 640x271, arc flash.jpg [View same] [iqdb] [saucenao] [google]

33 9V batteries in series

>> No.2800656

Rectified 220-240V mains is in the realm of 310-340VDC, you'll be able to pull 7A continuously from that without tripping a breaker, higher for shorter periods if you keep mind of your breaker's trip time curve. Assuming you have a lot of bulk of output filter capacitance. If this kind of voltage is fine for you, it's going to be at least an order of magnitude cheaper than a switching supply capable of the same power. Though the output isn't isolated but solidly mains referenced, so neither the positive nor negative output will be safe to touch. And tying an end to ground would trip your RCD. Maybe you can isolate it by buying a (2nd hand) 2400VA isolation transformer, though that would probably double the cost.

A rough calculation:
>I = C*dV/dt
>C = I / (∆V * f)
>f = 120Hz (halve this for a single-diode rectifier)
>∆V = 10V
>I = 10A
>C = 8mF (double this for a single-diode rectifier)

American houses have some 240V outlets IIRC, be it for an oven or electric water heater or whatever. Microwaves maybe.

What's it for?

>> No.2800665
File: 2.91 MB, 4000x3000, DSC_2519.jpg [View same] [iqdb] [saucenao] [google]

What does it mean by that?
Now I get the same jittering again

>> No.2800675
File: 3.12 MB, 4000x3000, DSC_2521.jpg [View same] [iqdb] [saucenao] [google]

Guess its the breadboard

>> No.2800676


>> No.2800701

any issues with this circuit?
the trannies are controlled by a tarduino

>> No.2800702
File: 90 KB, 730x870, retard.jpg [View same] [iqdb] [saucenao] [google]

hurrdurr forgot to attach the picture >>2800701

>> No.2800707

looks fine to me
though using one of those 10-step telephone contactors would be cooler

>> No.2800709

got myself a 450V shock from a tube amp last night. would not recommend

>> No.2800721

I'm using the cheapest chink relays I could find at 50ct a piece https://youtu.be/WVRW3949kdA

>> No.2800748

Pls help. Why does it do this?

>> No.2800775

>Why does it do this?

we might hazard a guess if we knew what ''it'' is
dont be mysterious: show schematic, show what place you're probing, note all DC voltages
anyway, wild guess is that you forgot the diode across the relay and you're getting large negative kickback voltages

>> No.2800808
File: 214 KB, 348x468, 8c03204e-1aef-4029-8f73-38cc0b0b8197.png [View same] [iqdb] [saucenao] [google]

No, that's the PWM brightness control for the LCD backlight. The BC337 has a base resistor of around 5.8k, driven by an arduino.
No clue if I should have PWM anywhere near an analog audio path in the first place ...
Wont have access to the shit for next week anyways. The relays and the other shit should have arrived by then (hopefully).
Next trouble is I'll have to wrap my head around the software stuff, especially this SPI fuckery and get the LCD and digital pots (for baxandall eq) which both use SPI to somehow play along as the tarduino only has one SPI interface as far as I'm aware.

>> No.2800831

>PWM anywhere near an analog audio path
Near is a loose term but I'd put all digital things on separate PCB in the final version.
>SPI fuckery and get the LCD and digital pots (for baxandall eq) which both use SPI to somehow play along
Shouln't be a problem, SPI can have multiple slave devices.

>> No.2800943

Does anyone here have experience with working on synths? Someone is offering an untested Yamaha DX-7 near me for $150; I was thinking of making a lowball offer and seeing if they take it. The thing is opened in the pictures, comparing it to online photos everything seems to be there. If I decide to go through with the purchase, what are some common issues that these have that I should take a look at? I know about the battery requiring replacement, and everything else I see online seems like the standard "sliders and knobs need deoxit" type of issues.

>> No.2800990

LEDs take some time to turn on and off. You’re basically using an LED like a pull-up resistor across its own capacitance, and it’s effective resistance gets really high at low voltages. If the backlight itself looks fine, who cares what the waveform looks like. If not, try lowering the frequency.

>SPI can have multiple slave devices
Yeah but most MCUs can’t have multiple hardware SS/CS lines for some reason. Gotta either use seperate SPI pins or not bang that CS line.

If boards aren’t cracked or have fried chips on them, it’s probably fine. Even cracked plastic you can replace with a 3D printer.

>> No.2801011

I didn’t know anything about synthesizers, but reading up on the DX7 for a few hours… it’s is an amazing and iconic piece of technology from the early 80s… almost an alien technology. It’s practically an arbitrary waveform generator.
I’d love to have one, broken or no, just to mess around with.

>> No.2801019

>it’s probably fine

a 40yo piece of electronic gear used by degenerate musicians is likely gonna be as healthy as an 80yo junkie hooker
like mom

>> No.2801034

I meant more that it’s probably repairable for <$30 and a half dozen hours.

>> No.2801043

Checked. If I had to guess the hillbilly reason is because they make and break contact at multiple points simultaneously. There’s probably relays like that but most I’ve seen are a one channel deal.

I’m seconding the definite purpose contactors. Not the cheapest but definitely the most clack for your stack?
Usually 120V coil but I’ve seen 24VAC.

Go watch paulmcwhorters stepper vid or something. A picture of your actual breadboard would have been more helpful to figure out how you wired it wrong. It doesn’t go with just plus and minus like a regular motor. Steppers take plus,minus and a control wire with a square wave to tell it what to do.

This is DIY and I want pics of a series-parallel transformer configuration and the biggest diodes you have ( or a bunch of small ones parallelled) Preferably with all the components zip tied to a 2x4 with the photo labeled “circuitboard”.

This will work too but the xtra xfmrs will look cool and isolate from mains.

>> No.2801061

It bites pretty good but unfortunately I've been bit too often and got used to it...not good. Taught me to be more careful, can't do electronics as a corpse. What did you do?

>> No.2801106

This might give you some ideas:
If you want something specific, maybe a metal detector with phase discrimination? I'd like to make a large format one myself, the size of a pallet that would be wheeled around. The larger the coils are, the deeper they should be able to detect. Maybe you can even have an array of transmit and receive coils to increase the effective depth, might be a good use of some magnetics simulating software. Huygens Optics has a neat video on how metal detectors work.

If you want something more specific, please elaborate.

>> No.2801179

>cnc-ing a pcb
>notice that Y axis stopped working, the screw that holds the screw rod & the stepper motor loosened and it is no longer moving on Y axis
>the damage is not that bad, I noticed it early
>turn CNC sideways to screw things back
>3 out of 4 screws that holds X stepper motor falls off
Fuck, did vibrations loosened them or what? It is amazing only 1 screw was holding X stepper motor

>> No.2801210

Ghosts did it. The derelict-dead dread spring washers and the essence of threadlocking.

>> No.2801237

recently bought a 0-30V, 0-3A dual output power supply that can be wired up series or parallel. i was thinking, how badly would 60V DC shock me if i were to touch it?

>> No.2801240
File: 246 KB, 1000x1414, 6d2bfa55af4eeb8a6d02bba4201da5321a2090f6.jpg [View same] [iqdb] [saucenao] [google]

I bought around 20 TDA1521 stereo amp chips just because they are selling for peanuts and now I don't know what to do with it.
Thought about buying PCBs and selling them as DIY amp kits.
Any experiences selling your own DIY kits? Is there any demand for something like this?

>> No.2801246

Probably fine as long as you don't have a wound where you touch it. Or try to lick it

>> No.2801255
File: 1.40 MB, 1440x1920, shitfuck.jpg [View same] [iqdb] [saucenao] [google]

Heya fellas.
Is there a simple way to remove any epoxy or glue that's been used to hold down a CPU?
I'm wanting to resolder in a new CPU in my laptop but on my first attempt I learned that the chip is glued to the board at the corners, making desoldering a huge pain in the neck.
Anyone got experience and wisdom to share? Thanks.
Pic related; the result of my first attempt.

>> No.2801261

1. Heat the epoxy with a shitty iron and scrape it up with a toothpick, or use a little acetone and let it sit.
2. You don't need flux to remove chips.

>> No.2801280

a solvent + toothbrush. i would start with 91-99% IPA and see if that dissolves it. if not move on to other things. keep in mind that harder solvents like acetone can dissolve plastics.

>> No.2801292
File: 2.02 MB, 737x9364, how to be like matsumae ohana.jpg [View same] [iqdb] [saucenao] [google]

Just do it (yourself). Amplifiers are common DIY kit though, just getting noticed might be difficult without having something special(optional) or aggressive ads on Amazon or such. Just speculating though, I haven't sold anything I've made or made anything I'd presume to sell yet.

If they don't sell you'll still have 20 TDA1521 for something else.

>> No.2801293
File: 357 KB, 744x1052, a9560273fb94c1e262d8fca54ed71c68.jpg [View same] [iqdb] [saucenao] [google]

It won't, it's a goddamn worksite to remove. Desoldering four op amps from junk Creative 2.1 computer speakers took me something like an hour or pliers, heat from soldering iron, liberally applying craft brew PRF IPA(didn't really do much). Soaking in acetone beforehand sound worth trying, or maybe let it cook in oven so the epoxy is softened throughout before even attempting to remove it.

>> No.2801294

Hello frens, sorry if this is a noob question but any good recommendations for books that provide electronic projects in order from very beginner to somewhat more advanced? I learned alot of the theory already but I wanna start getting practical experience (lots of it).

The projects that interest me the most I've looked up online and they're very intimidating and advanced, I don't trust myself with all of that until I learn how to do easier stuff

>> No.2801297
File: 130 KB, 1238x1150, QFS_QMS_elevated-standoff.jpg [View same] [iqdb] [saucenao] [google]

I think I'll sell it with my digitally controlled eq that I'm currently working on. I created seperate eq and amp boards and want make them stackable like pcrel.

>> No.2801303

Anywhere that can convert a circuit design into a formula for Kirchhoff or some other equation?

>> No.2801323

doing your own homework and learning this stuff is not pointless anon, I know first year EE is not the most exciting, but you have to learn it to be able to move on to better stuff.

>> No.2801359
File: 135 KB, 695x927, file.jpg [View same] [iqdb] [saucenao] [google]

lol, I either forgot to 0 Z at some point or added an extra 0

I should get some large sacrificial wood boards. All I had was 2 small pieces and I was not able to secure it properly. Some parts were flexing and the copper at those parts were not properly etched all the way through. I tried going a second time and ruined it

>> No.2801380
File: 717 KB, 2190x2064, hmm.jpg [View same] [iqdb] [saucenao] [google]

These days we just send a tiktok of our homework to chatgpt and it does it for us ain't it cool to live in the future

>> No.2801383
File: 12 KB, 307x206, snufkin huh.jpg [View same] [iqdb] [saucenao] [google]

i love that image too much.
need an anime about cute anime girls who get shrunken in order to build pcbs

>> No.2801385

Remember that manga series that taught some high school level stuff, including one of electricity? An anime like that would be really cool.

>> No.2801396

>Remember that manga series that taught some high school level stuff
i do not, got link?

>> No.2801397

Look for "the manga guide to [thing]". There's one on electricity, physics, biochem, all sorts of stuff.
It's an manga episode and then a small chapter elaborating on the topics, then another manga episode and so on and so fourth. Very cute but obviously no college level stuff.

>> No.2801430

I crave the ham radio anime, the one I posted a synopsis for a year ago. Cute girls doing stupid shit innawoods and exploring the electromagnetic spectrum. What could be more comfy?

>> No.2801436

What anime? I didn't knew it existed but not I need it.

>> No.2801475

It doesn’t exist, that’s why I have to write it. I am bound to this destiny.

>> No.2801480

Then do it before I eventually kill myself either willingly or by accident (most likely by accident).

>> No.2801512

There used to be a "read this before you try anything" link for lithium ion batteries in the OP. Anyone know what IOM talking about?

>> No.2801530

>early 20s
>"lets make X today!"
>make the circuit on a breadboard or etch using a permanent marker and copper clad
>hodgepodge a simple firmware for the MCU in the circuit, looks simple but actually quite clever and ingenious code
>by late night of the same day the project is done and working, there are bugs but it still works for the most part
>make a shitty enclosure out of cardboard
>sleep happy knowing I had fun and did fun stuff
>best of all, I never even considered not doing something because it may be too difficult or too much of a pain

>late 20s
>"I want to make X! Let me plan to make X in the next 2 months"
>do nothing the entire week
>weekend comes, make detailed plans about how I will make the project
>make a detailed schematic in kicad to eventually make a PCB out of (from a paid service ofcourse)
>spend a few weekends slowly working on a "robust" design
>get a PCB made, still have issues in it (usually abandon project at this point)
>start writing firmware
>firmware is "modular", very stable, probably could get MISRA compliance, yet is incredibly boring, tedious to write and appears to have no intelligence or ingenuity put into it (no clever or hacky code)
>fix all the bugs over a few weekends
>start working on the "mechanical design" in CAD
>spend a few weekends making the perfect, low cost of manufacturing enclosure out of 3D printed ABS or PLA or whatever
>order it from a service, get the part a week later
>project is done
>just kidding, I gave up on the design phase of the project because everything is "too hard" or "too expensive" or "I can't get it robust enough in the given timeframe"

Get a job they said. Your hobby will always remain your hobby they said.

>> No.2801532

It got moved to elsewhere in the OP github. And by that I mean I forgot to add it anywhere and it's only in the "oldOP.txt" file. Here:
It's a bit of an odd document, as it specifically refers to Saft's own lithium cells, so it may be hard for noobs to apply it to other cells. Especially high current RC lipos, which will have quite different maximum current ratings for charge and discharge. It's probably worth looking for a better existing document to direct people to, or I could try writing one myself.

this hurts me
sometimes i can still get really into projects and spit them out quick
but man i burn out when it doesn't go well

>> No.2801540

i found it: >>2476213

>> No.2801542
File: 395 KB, 667x374, 4chan OLD NEW.png [View same] [iqdb] [saucenao] [google]

i miss the days of cardboard enclosures too.

>> No.2801560

this is why I refuse to use kicad or pcbs for anything

>> No.2801569

Still have to validate whatever chatGPT says, I find every 1 in 5 or so answers have an error in/

>> No.2801641

i had this chinky light-up christmas stocking thing that ran on AA batteries. i liked it but i hated how it ran on AAs and it didn't have under voltage cut-off, so it would kill my rechageable batteries if i wasn't paying attention and ran it for too long.

i spent like a month designing a two PCBs: one for a stm32g0 mcu + human interface + debug interface + 18650 charging and bms; and another PCB for a boost LED driver and a dozen RGB LEDs with individually PWM dimmable channels. it was going to have RTC functionality so i could program it to turn on and off at certain times and ability to control brightness by altering PWM duty cycle.

i then spent another month designing a 3D printable enclosure that could fit 2S 18650s, both PCBs, cutouts for USB charging and push buttons for human interface, etc. in such a way that it would bolt up to the acrylic sheet from the chinky christmas stocking.

got the PCBs made, built the boost converter/RGB LED PCB first. ran way too fucking hot because the data sheet lied about boost converter efficiency at low input voltages, like 60% at 3.7V input. in an autistic fit i just threw it all in a closet and forgot about it.

>> No.2801672

Yep. But soon, because continued use of chatgpt, and the belief that it is “smart” soon it will be.
If I found someone using chatgpt to complete homework assignments, I would still pass them with a C grade rather than initiate any kind of proceedings against them.
They paid tuition, after all.

>> No.2801674

> kills rechargable batteries
Get some NiCd, they’re tolerant of discharge.
EBL in china is reputable. You can’t make NiCd in the US because of the EPA, so stop worrying about it.

>> No.2801699
File: 1.46 MB, 4000x3000, 1000000198.jpg [View same] [iqdb] [saucenao] [google]

is this safe

>> No.2801710

It's an ungrounded socket, so you aren't protected against catastrophic failures like broken insulation coming in contact with an appliance's metal housing. The surge protector is very old, so the MOVs and whatever else is inside are probably dead.
Buy a new one; they're like $20.

>> No.2801715

not my problem, I'm helping some move out...just thought it was kind of odd

>> No.2801717

Yeah, they're weird. Traps for idiots, mostly.

>> No.2801721

omg that's so unsafe. it wins the unsafe award for unsafetylike conduct.

>> No.2801722

>american electrics
no its not safe

>> No.2801723

i'm reporting this to the consumer safety bureau right now. this is appalling.

>> No.2801736

>in an autistic fit i just threw it all in a closet and forgot about it.
Its the same with me, the projects which I ironically make the mistake of doing "properly" often end up nowhere, meanwhile one-off, prototypes and shit that I whip up in minutes often work well and stay with me for a long time.
I have no idea what's the reason behind this phenomenon

Just today I quickly whipped up a keychain thing with an attiny13 which beeps every 20 minutes. The keychain is attached to my grandfather's water bottle so that he remembers to drink water. This project took me like 2 hours and it all turned out well.

>> No.2801780
File: 120 KB, 800x627, eldar gods 07.jpg [View same] [iqdb] [saucenao] [google]

>keychain is attached to my grandfather's water bottle
Pretty sure forcing gramps to constantly drink water constitutes eldar abuse. And the ground glass, meant to speed things along, is no excuse.

>> No.2801795

He turns into a prune without water and gets UTI

>> No.2801802

When you get old it's easy to forget to eat or drink and you.dont feel hungry or thirsty

>> No.2801806

Put one of those shock collars for dogs on him and set phaser to stun every 15 minutes. Maybe he'll remember better.

>> No.2801854
File: 7 KB, 1013x595, FishFinder Diagram.png [View same] [iqdb] [saucenao] [google]

I have a fish finder that powers on when set in it's cradle and I want that to also power on a cheap ethernet network switch.

Is there risk of drawing too much amperage through the fish finder in this circuit?

If this creates noise can I add something, like a diode and capacitor, between the fish finder and buck converter to stop it?

>> No.2801883

New idea: headphone detection for a DAP/PMP/phone/DAC/amp. When you plug in your headphones, the device runs a low amplitude frequency sweep through them while measuring the current to uniquely characterise the headphones. If it’s a new pair, it will ask the user to calibrate their sensitivity by listening to a test wave, such that maximum and minimum volume levels are sensible. It builds up an internal database of each pair of headphones, so whenever you plug in a known pair it will apply the right sensitivity gain to a PGA. You could also set EQs for each pair. It would be nice if you could have an online database it can check against too. You’d probably do a frequency sweep only of the high-end, so it takes less time.

Draw a diagram showing the power wires going to the fish-finder’s input too. You might be able to tap a wire off it, but I’d probably want to use a MOSFET.

>> No.2801911

Better idea: put the headphone cups together (as they would be on your head with little to no airflow) and use the other headphone as a dynamic microphone to sense the output of the other, and build the real in-use characteristics.

>> No.2801915

That DC/DC converter is likely to limit the current to unusably low values.
You’re better off to put a low value resistor in series with the fish finder’s DC inputs, and a comparator to sense when it’s “in use” or not when the voltage across the resistor is sufficient.

Also, if open up the fish finder, there’s a good chance it already has a 5 V on board somewheres.

>> No.2801933
File: 97 KB, 1518x403, Yaesu Microphone Bias Power.png [View same] [iqdb] [saucenao] [google]

So I'm wiring a setup to get bias power to an electret microphone while using a radio not really designed for it, so I need a dc blocking cap on the input, all I have on the shelf are little electrolytics. The cap used in the diagram isn't polarized. So I should put the positive towards the voltage right?

>> No.2801936

also now that I've done more reading there seems to be dc offset and input impedance might be an issue

>> No.2801942

A: doesn’t work for IEMs
B: wouldn’t the frequency response as a speaker and as a microphone be different? Especially for hybrid units. Also my bad intuition is telling me it would cancel out to provide a flat response on the other side.

Depends on if the DC DC converter is in series or parallel with the fish finder. Worst case it’s just turning on because it gets contact through the two pins on the charging dock. But it could have a reed switch or something similar that tells it to provide power only when the device is inside it, to prevent water sitting in the dock and corroding the electrodes. Or maybe it even charges wirelessly like a toothbrush.

Do you have any luck using a comparator on an unamplified current sense resistor? Even then it might not work due to low current draw when fully charged. Opening up the fish finder is sketchy too, since it presumably needs to be waterproof.

>> No.2801948

>So I should put the positive towards the voltage right?

typically yes, but not necessarily so
a smart operator (aka a smart-ass) would do one of two things
- measure DC voltage between the 2 points and orient the cap accordingly
- use two polarized caps in series of the same value, connecting + to +, and using the -'s as the endpoints of the cap

>> No.2801949

well the radio is providing the 5v and the resistor is there to ball park it to around 3.3v for the element so I would think all the dc voltage would be on that side since the other side is a microphone input

>> No.2801950

what situations would a high resistance meter, capable of measuring up to 20 petaohms, come in handy for.

all i can think of is testing motor windings, transformers, or things like that

>> No.2801951


>> No.2801952

Where did you see ground glass?

>> No.2801953

>I would think

i'd measure and be sure
or i'd take precautions and thus be assured

>> No.2801955

completely false. 120V AC can be rectified to 170V positive and 170V negative, for a total voltage of 340V DC

>> No.2801956

>see ground glass?

- logical deduction
- S. Holmes, and A. Christie

>> No.2801957
File: 54 KB, 612x632, 886d20b72cdb7cdc82fc1e699668895c3c8e075fabb38b81c12e437f9025492d_1.jpg [View same] [iqdb] [saucenao] [google]

definitely fit the smart ass moniker

>> No.2801958

you can rectify each half of the AC wave separately and get double your numbers

>> No.2801962
File: 855 KB, 2560x1920, PXL_20240517_141602920.jpg [View same] [iqdb] [saucenao] [google]


>> No.2801963

>total voltage of 340V DC

that's peak-to-peak power which is fake and gay
real power is measure in RMS by convention and by law

>> No.2801964
File: 126 KB, 1500x1000, f-1587fc-03b-1500x1000.jpg [View same] [iqdb] [saucenao] [google]

generally motor testing is done with a megohm meter that can test at hundreds of volts

>> No.2801965

you dont measure DC power by rms.

to turn 120V AC to 340V DC, you use a half wave rectifier on each half of the AC to charge capacitor, one positive 170V, the other negative 170V. this gives you a 340V DC supply.

this is how filament LED bulb power supplies work.

>> No.2801966

but 1000V is a better test right?
the motors i used to test would routinely measure in teraohms

>> No.2801967
File: 21 KB, 994x853, Bridge-voltage-doubler-circuit.png [View same] [iqdb] [saucenao] [google]

I think this is the circuit im thinking of

>> No.2801968

>charge capacitor

utter nonsense
sure you can boost voltage and smooth it out but, at the end of the day, the power out cant be greater than power in
the diode/caps dont magically multiply power by 2.8x

>> No.2801972

threw some alligator clips around and tested it live
works fine, little noisy through 3 ft of test clip cable but that's to be expected

>> No.2801971

>voltage is a measure of power
sir are you retarded.

>> No.2801976

>>voltage is a measure of power

that's not what i'm saying
that's i'm arguing against
pls keep up

>> No.2801977

Ah, you made it up.

>> No.2801979

-illogical delusion

>> No.2801980

Sorry, the original statement below made me think it was referring to voltage
>You realize that’s gonna be more than double what you can pull out of a standard north american outlet eh?
since 300V 10A isnt what id call "more than double" what a US outlet can do.

but yeah, anon should be looking at like a simple 300V 5A circuit maybe at best

>> No.2801987
File: 114 KB, 1007x1027, autists often dont get it.jpg [View same] [iqdb] [saucenao] [google]

>-illogical delusion

>> No.2801994

>300V 10A isnt what id call "more than double" what a US outlet can do
that is trivially equal to 3000W
a US outlet can do 1200W
sounds like more than double to me

>> No.2801997

US outlets do 1800W but arent generally permitted to do over 1500W sustained.
So to me, it sounds like somewhere in the range of less than double to exactly double.

>> No.2801999

and thats at 15A. some circuits are rated for 20A, or 2400W

>> No.2802031

Ah, I see. Thanks for the resource anyway, outdated as it may be. I'll take anything I can get.

>> No.2802041
File: 259 KB, 883x614, 84702F26-6E1E-4AAF-B776-F04D7E0CC45D.jpg [View same] [iqdb] [saucenao] [google]

> IEM and cancel
Don’t know what an IEM is, i wasn’t in the iraq war. Improvised explosive munition? We used to test the surface microphones with primacord back in the day.

They cup’s responses effectively multiply, if you think about it, very small diameter ultrasonic transducers have the best power transfer with identically sized receivers.

The best pair of phones I ever had was this brown realistic pair in picrel.
Searching the internet now, it looks like they might have been made by koss and re-branded.
The cups were like a PVC material, and when you pull them off you could feel the vacuum break.
You didn’t need any of that active noise-cancelling voodoo.
In my opinion, those around ear headphones were the best. Of course, back then, almost all headphones were around ear. There is a lot of this “on ear” horseshit floating around nowadays. You’re getting scammed, they’re just saving 5 cents in plastic.
Anyway, they’re gone now, and I never appreciated them until I had to replace them with modern bullshit. You don’t know what you’ve got till it’s gone.

>> No.2802043

There was never any doubt you're on the spectrum.

>> No.2802046

I was including losses and reasonable safety margins in whatever DC conversion scheme was going to be used.

It’s generally not a good idea to run your outlets at a sustained 1500 watts. Very few things are rated for that I find, and you don’t know what else could be on that circuit. Well, I do because I wired my own house, but most people probably don’t.
You can pull a lot more than 1800W for very short periods of time. e.g. the lights dim when I start my 1 HP radial arm saw since it presents almost a dead short when not moving.
Either way, you generally want a safety margin.
I’d power this on two separate circuits (two plugs) at 150 volts DC at 10 A each and put the two DC supplies in series to get your 300 V.

But then, again, there are lots of people who design things using nothing but the ABSOLUTE MAXIMUM RATINGS table nowadays, so maybe I’m not updated on the latest design trends and eveyone is using superconducting lk-99 wires with no losses.

>> No.2802174
File: 429 KB, 1079x534, 1000014444.jpg [View same] [iqdb] [saucenao] [google]

I'm sure there was an episode of magic school bus that did that.

>> No.2802183

Carrots, cheese, dentures, a tiny horned demon and a hippie witch bus driver. Heavy drug use.

>> No.2802252
File: 409 KB, 960x960, 1706619280734.jpg [View same] [iqdb] [saucenao] [google]

Guys, how can we test for mains polarity in Europe, if plugs/sockets are reversible; does it even make sense to talk about it, in Europe?

>> No.2802271

>how can we test for mains polarity in Europe
Multimeter. Set to AC, remove a probe and touch the terminals with the remaining one. The live should read somewhere between 10-40V due to the capacitance of the probe wires. Neutral should give little to no voltage like this.
>does it even make sense to talk about it, in Europe
Yes because not everyone uses L+N circuit breakers. If you connect the neutral to the breakers instead of the live then the live remains on all the time and you can shock yourself when working on outlets or switches.
For outlets it doesn't make much difference which way around you connect them.
For light switches it matters if you use one of those touch activated ones. Those must be connected to the live or they will not work.

>> No.2802281

thanks, device in picrel should test all the possibilities, plus ground, for a few bucks, but my question is in the sense that, since everything is reversible, you can just be plugging either end of a device's circuitry to either live or neutral wire; does that have any implications for either the device or you as an user, here? or is that just a safety concern re. mains line, as you mentioned?

>> No.2802290

btw, how can I check if I have l+n breakers; would those things tell me that?

>> No.2802297

>does that have any implications for either the device or you as an user, here?
No. It's called alternating current for a reason.
The live alternates between a positive voltage and a negative voltage. The devices don't care where the live is, they just care about the potential difference. If the voltage is positive on the live and the device is plugged in the "wrong" way then it sees a negative voltage and positive when the live is negative. It's the same sine wave, just flipped.

No, you go to the breaker panel and look at them. Regular breakers have 2 screw terminals, one top one bottom. L+N breakers have 4.

>> No.2802305

How do I get started with embedded/arduino/etc? I feel a bit overwhelmed by all the stuff it seems like I need to learn to even start. All I really know how to do so far is program in c/c++.

>> No.2802330

ok, i asked that because on some devices, when i touch an outer metal part, i feel current passing, but then if i reverse the plug I'll no longer feel it; other times I'll still feel it, such as with my laptop (which doesn't have a ground connection), and all I've been told about that it that it's ok, not to worry...

>> No.2802333

Can you tell just from picrel which are those (would have to remove the front plate to check inside)? those are for the lighting and non-grounded sockets, btw

>> No.2802342

>I feel a bit overwhelmed

go here and follow your nose
from n00b to DoA (Doctor of Arduino) in 3 months

>> No.2802352
File: 1.36 MB, 3521x2641, 1701938923523.jpg [View same] [iqdb] [saucenao] [google]


>> No.2802369

Just start experimenting bro

>> No.2802431

The arduino hardware and software (or things like it) is by far the easiest way to get into programming microcontrollers, but it’s also a dumbed down partial layer of abstraction that you may not like. I’d start out there to get at least some benchmark of what you’re able to do, but consider moving afield. If you already use VScode, consider using PlatformIO instead. If you’re comfortable with makefiles and command line tools, I’d say jump to using tools like avr-gcc and AVRdude directly.

It’s odd that you’d feel it only in one orientation. Usually capacitive coupling is equal from both live and neutral, sounds like somehow it’s quite asymmetric. Dead Y-cap, maybe. I’d consider requiring it to ground that metal.

>> No.2802460
File: 20 KB, 1100x976, fishfinder network switch.png [View same] [iqdb] [saucenao] [google]

Haven't figured it out yet, too dumb. I'm youtubing comparators to better understand, learning about OP amps.

Pretty sure the unit just makes contacts with the pins to complete the circuit. There are buttons to turn it on or it can be set to auto turn on when connected.

Is this something I could accomplish with a mosfet if it is just completing the circuit?

>> No.2802464

If the cradle lights up or something, you could even use a photodiode to sense it and activate the 5 V supply through similar means to you schematic there (which is coming along nicely)

If you are worried about the transistor/mosfet you could always use a small relay. What’s the coil maintenance current on a relay, like 20 mA?

>> No.2802469
File: 49 KB, 852x337, roller switch and magnetic switch.jpg [View same] [iqdb] [saucenao] [google]

>you could always use a small relay

a relay has 2 elements, a control signal and and an output signal
you can simplify that to 1 element
- a roller switch that's activated when the unit is inserted, which pushes down on the arm
- a magnetic switch which operates similarly but doesnt require touching, just proximity

>> No.2802472
File: 355 KB, 1463x971, fishfinder with relay.png [View same] [iqdb] [saucenao] [google]

I thought about using a relay like this, only because I've used them in the past, but not exactly sure how to set that up or how much power I can draw through the trigger side.

With something like this, would the current just ignore the trigger and flow through the DC+ to DC- because there is less resistance?

Also, if use a regular 4 prong automotive relay I'm not sure how much current I can draw across the trigger pin.

I honestly could throw another toggle switch on the control panel, just being stubborn.

>> No.2802473

You’ll probably want to measure the current coming out of the dock and using that to turn on a MOSFET. That means putting the current through a current sense resistor, and comparing the voltage across it to some threshold voltage (as set by a voltage divider). Comparators turn on when their + voltage is higher than their - voltage.
It may work just to measure voltage at one of the charge dock pins, depending on the charger it may float high when it’s not connected to the fish finder and you’ll be able to quite easily detect its drop. A MOSFET kinda acts like a comparator if you don’t mind some variance, but I’d probably want to use a proper comparator. Make sure its input voltage range is wide enough.

Using a BJT like your pic might work, but it will probably interfere with the charging process and not charge it fully. Anything else in series other than a very low value resistance will probably be even worse for charging. Maybe a low resistance low voltage relay coil might be ok…

>> No.2802475

Appreciate the feedback. The fish finder doesn't have a battery that gets charged, just runs off the boat battery. Would I be comparing it to 0 then?

>> No.2802477

Wait then why would you ever unplug it?

>> No.2802478

Just so I can take it inside for security.

I think this route will work, just tested having DC+ and IN hooked up like this and it powered up fine on a 5v relay, will just have to get a 12v one.

I'm sure there are far more efficient, cheaper and simpler solutions but I'm dumb and this works. Thanks everyone for trying to help.

>> No.2802509

>I think this route will work, just tested having DC+ and IN hooked up like this and it powered up fine on a 5v relay
Yeah but is the fish finder getting enough power? Very little current will be flowing through that thing, maybe in the ballpark of 0.01A. If that "2A" means "two amperes of current", then whatever you put between the battery and the fish finder should have an effective resistance less than 1Ω; it should drop no more than 2V or so. I'd probably want to use something in the realm of 0.1Ω or smaller. A BJT base-emitter junction (as in >>2802460) will likely work since that drops only 0.7V, but you'll need to find a BJT that can handle 2A of base current, which is hard to find. The TIP41 can do it though.

>> No.2802512

oops, /mcg/ died

>> No.2802528

Can tarduinos do anything right

>> No.2802572

why doesn't /mcg/ use a mcu to keep itself alive

>> No.2802588

It reached bump limit, retards.

>> No.2802594

So use a mcg to make a new thread, genius.

>> No.2802722
File: 2.23 MB, 2309x4028, IMG_1924.jpg [View same] [iqdb] [saucenao] [google]

I picked up the synth for $100. It appears that besides some cosmetic issues, the only thing that was wrong with it was that it needed a new battery. A message popped up on the screen saying to check it. Measuring it confirmed that it was dead. I removed the battery and since the original was soldered to the board I printed a CR2032 holder, clipped out the old battery, soldered in the holder and stuck a battery that I had on hand into it. Upon the second powerup, it displayed information about the ROM that was on it. Messing around with the menus and keys, it seems that they all work fine. Now I need to figure out how to work the thing. There’s some software that’s available on GitHub called dex or something like that which allows you to program it using a USB MIDI interface which might make it a bit easier.

>> No.2802730
File: 153 KB, 1920x1080, 1712660791008337.jpg [View same] [iqdb] [saucenao] [google]

what is this tool called?

>> No.2802737

I’m pretty sure it’s an underfill removal tool.

>> No.2802741
File: 23 KB, 569x348, 2280s-32-6_front[1].jpg [View same] [iqdb] [saucenao] [google]

What's the point of precision bench power supplies if they are only precise at the PCB level and any type of interconnects and especially external cables will result in a perceivable voltage drop making their specs irrelevant?
For example
>Accuracy ± (0.02% + 3 mV)
But a 3ft AWG cable would result in a voltage drop of about 5mV for 1A. Make it 5A and you are out of spec by an order of magnitude. Conversely, to keep the voltage drop to 2mV at 5A you'd need a massive 1AWG cable! That's just not realistic.

>> No.2802743

thanks! I had no idea such a thing (underfill) even existed.
looks like a useful tool for other kinds of shit too.

>> No.2802746

Pretty sure they're calibrated to account for the voltage drop in their original cables.

>> No.2802762

Nice get.
I have some yamaha keyboards, but no synth.
Surprising considering my penchant for 80s techno music.

>> No.2802763

The sort of people who are buying these are smart and know what they're plugging in to the front panel and compensate for that in many different ways. Please be more creative.

>> No.2802775
File: 129 KB, 500x375, Keithley 2280S_rearview_NB_0.png [View same] [iqdb] [saucenao] [google]

>result in a perceivable voltage drop making their specs irrelevant?

if you peek at its ass side, you'll see sense wires which can compensate for the error.

>> No.2802778

Looks like an xacto knife handle with a piece of sheet metal stuck in it. Sounds like a diy project.

Nope, it’s €120 underfill manipulation separation and extraction workstation accessory.

>> No.2802783

Oh can they do kelvin sensing? It kinda scares me to use kelvin sensing from a PSU, since if one of the voltage sensing wires gets broken the output could shoot way up in voltage. Maybe they sense the impedance of the sense wires or something.

>> No.2802823
File: 102 KB, 1592x902, huh.png [View same] [iqdb] [saucenao] [google]

A cc/cv op-amp circuit for regulating power to a low-current electrochemical cell. Any thoughts?

I think maybe if I invert U1B and U1C, and invert or bypass U1D, the output polarity would be the same (still negative feedback) but I'd be able to put diodes across U1B/U1C to stop them hitting the rails. Like a normal ideal diode circuit.

>> No.2802836

how hard is it finding device schematics? like laptop motherboards.

>> No.2802853

near impossible. manufacturers don't like releasing them, so only the chinks have it, and the only way you're getting those are from them is thru shady sources.
>have junk, what do?

>> No.2802860

Motherboards are not that complicated. Dense, and 16 layer, but not complicated. 90% of it is the CPU connecting to pins on the accompanying chipsets, ram, superio, etc. many times it’s not possible to get complete specs on the chips themselves, because there’s no real need (pins A0 through A63 connect to pins B0 through B63 on the other black box chip).

The remainder are probably the things that go wrong, like the DC/DC converters and battery charge, USB drivers, etc. due to high voltages, high currents, water damage, and static discharge.
For those, look at the app notes for the chips and that will usually be the exact circuit used on the motherboard. If not, very similar.

> impossible
High-end motherboards, like apple have been leaked, but generally it’s impossible with low-end mobos… mostly because nobody cares enough to fix them, and it will change every year.

I find the best thing to have is a working device where you can trace through voltages on the bad one, comparing it to the good one (without the need to really understand or have the schematic)

>> No.2802865

Googled 555 formula (granted bad wording, but still)
>Starting to invest in mutual funds at age 25
>Increasing contributions by 5% annually for 30 years
>Retiring at age 55

>> No.2802874

>finding device schematics? like laptops

you can get paid software that pulls droves of intimate service data for your laptop from the internet
it's made by third parties, and they're careful to keep pirates away

>> No.2802881

Depends what manufacturer and model. Apple? Yeah, sure. Lenovo? Maybe if it's a Thinkpad. Just a random HP envy x382whatever from 2011? Good luck.

>> No.2802895

Any ideas for a cheap actuator that can drain a vessel of its ferric chloride? The vessel will be a custom job made from PVC pipe and cement, I hear that stuff can handle FeCl well. A ball valve with a motor coupled to it is doable, but a bit clunky if I can avoid it. Not sure about anything that uses a rubber seal. I want good fluid mixing so a siphon method is not ideal.

I was thinking of using some clear heat-shrink tubing and pressing it flat with an RC servo. The clear stuff is more flexible and seems to have PVC in it, so I hope it could be sealed to the rest of the PVC tank using that cement stuff, and that it's chemically resistant.
Any other thoughts?

>> No.2802896

Get elecdroid for your phone

>> No.2802902

>look for 25MHz crystal oscillator
>find a renesas part on farnell
>only part in the local warehouse
>look at its datasheet
>describes pinout for 4-pin and 6-pin package
>the one being sold is the 5x3.2mm package
>nothing tells me how many pins the 5x3.2mm package has
>link in datasheet just leads to renesas's home page
>find oscillator's page on their website
>look at other documents
>see package information
>no description of different packages
>see i can download eecad files
>click through without knowing which package's files i'll get
>it asks me for an email address so i can manually request which files i want
i'm gonna choose a different brand, i can wait

>> No.2802907

Regular water spigots used for the portable and potable water jugs used in camping are often 100% plastic. There is also spigots on these big-ass laundry detergent containers so you don’t have to lift a 6 liter container every day. They have a push-button to open them.

>> No.2802912

But what kind of motor would I need to use to actuate one of those? I'm trying to automate it based off time / voltage / whatever, such that I don't corrode my electrodes by leaving it too long.

>> No.2802917

Are you the electrochemical cell/cv cc guy?
The camping spigot with a geared down servo seems to be the best bet.
I wouldn’t be too obsessed with the type of plastic, I think my ferric chloride came in clear-ish polyethylene bottles with polypropylene caps.
I’m sure it’s not “rubber” either, silicone and viton are pretty resistant.
In my etching set-up, I use aquarium stuff, like those blue oxygenating fish bubblers to circulate it and some air pumps, which is another great place to investigate for valves and whatnot. Also aquarium heaters, warm FeCl works better than cold.

>> No.2802918

Oh, by the way, if you leave a lot of FeCl out in the open air for a while, inside your house, everything metal in your house will rust. Everything. Even stainless steel.

>> No.2802938
File: 2.11 MB, 3689x2076, PLATINUM.jpg [View same] [iqdb] [saucenao] [google]

>Are you the electrochemical cell/cv cc guy?
Yeah, I'm making a device to reduce the copper back out of the FeCl and oxidise the Fe ions again. Need to run it real slow to not oxidise my platinum anode, even platinum doesn't last in acidic chlorinating environments. I could upgrade to iridium or glassy carbon or boron doped diamond to solve that, but I'm not made of money. I've done it before with a platinum wire and it seemed to not dissolve the wire, so I hope it will be fine with this platinum plated anode I bought from aliexpress where the platinum is in quotation-marks, pic related.

The idea is to measure the voltage and/or current of my system, and when it goes above a certain level, I'll drain the FeCl out of the reactor tank and into a bottle. I'll have a vibration motor to periodically shake any climbing dendrites back down onto the copper, and also use it to shake off any droplets stuck to the cathode after draining. The cathode will be copper, so leaving it in the solution after I stop the current would corrode it and use up the etchant again. I'll also add a battery so it doesn't dissolve the cathode in the event of a power cut, and drains the solution before it goes flat.

Though there's also the option of just leaving the electrolysis going permanently until I manually decant it, no auto-valve needed. I don't know if that's more likely to dissolve the anode or cathode, I think the only side-reaction will be iron being reduced again at the cathode and re-oxidised at the anode, but with very low concentrations of Fe(II) at the anode I'm not 100% certain that the anode will remain intact. With any luck the cathode starts reducing Fe(III) before that happens? I might ask /sci/ but I'm not expecting anything. Maybe I try this first, if it doesn't work I'm only down one $15 anode.

Air bubbling might not be desirable, dissolved oxygen will increase the pH as it oxidises the Fe ions.

>> No.2802949


>> No.2802953

it does not fill me with confidence

>> No.2802991

any idea about the elon musk's brain reading chip? chinese knockoffs soon? and an open source API? is that something we could DIY soon?

>> No.2802997

It’s probably going to be released with the roadster.
I can’t wait for the home-kit.
I signed some people up for the pre-testing phase.

>> No.2803008

There is one or two “forever” regenerating solutions involving electrolysis and, I believe, copper sulphate. Maybe one was nurd rage, and/or the comment section of a nurd rage video, it was a few years back.
Ferric chloride is used to iron plate soldering iron tips. Perhaps you could investigate using iron electrodes, although that is not intuitive. You might have to reverse the flow of electricity in this scheme. Then you would only need mostly an acid I won’t name here.
Most plating factories just raise or dip (via slow conveyer belt) the electrodes in the “vat” of electrolyte rather than draining it, and trying to do wet works.

>> No.2803020

>crystal oscillator
Crystal oscillators (that are literally just a quartz crystal in an enclosure) will either have 2 or 4 pins. VCO/TCXO/VCTCXO and things like these will have more pins but these are not crystal oscillators, they are way more complicated ICs.
>renesas part on farnell
I checked, these are " Low Phase Noise
Quartz-based PLL Oscillators". Do you specifically need a fancy oscillator like this tho? Or just any 25MHz clock source, because if not, just buy one of these: MCSJK-7U-25.00-12-30-60-B-30

>> No.2803060

Something something Cartman something something V-Chip.

>> No.2803084

So I got a sensor that outputs from 0.5 volts to 4.5 volts, how can I scale this to 0 to 3.3 volts?

>> No.2803089

>something we could diy?
Oh fuck no, you need your own semiconductor fab for that. I don't think you can even do the surgery to put one in without a very expensive robot.

>> No.2803135

Voltage divider.

>> No.2803139

I know a guy who works out of an old dentist's office in the mall. He's even a real doctor. He just lost his license temporarily over some kind of paperwork bullshit. He'll fix you right up. Trust me.

>> No.2803146

He's that guy who stole someone's skeleton, isn't he?

>> No.2803149

Won't work. I need the 0.5 to map to 0.

>> No.2803154

seems like you'd be better off with magnetic "needles" imo, driving probes into brain tissue doesn't sound great

>> No.2803158

> won’t
Will. You need a negative voltage relative to 0 V.
If this is a battery powered thing, it’s trivial, add a 1.5 V cell somewhere. Or, even if it’s not!

We just had a discussion of negative voltage generators ^^^ up there somewhere, where you need an oscillator (single transistor will do) an electrolytic cap, a signal diode, and a zener diode. If you have ferrite cores and some enamel wire, you can make a joule thief-like circuit to generate the negative voltage as well.

You could also scale it with a single supply rail-to-rail op amp if you’re into those kinds of things.

>> No.2803163

I remember there was an “omni” magazine article about this in the 70s, chips hooked up to your brain. I think it was to help with memory.
Anyway, it’s just another scam with stolen and already failed sci-fi idea for this generation’s gullible and ignorant investors.

>> No.2803168

the laptop is a low end Asus; typically those devices are heaters, or other very somple, low cost electrical apparatus

>> No.2803170

oh, btw, one place that happens is on an escalator at a local shopping mall (the metal frame under the rubbery handle). I've reported it to a security guard; that was months ago...

>> No.2803178

Didn't Cartman also have an anal probe installed by the Martians? It was long time ago so it was probably analog. A modern device could combine all these bleeding edge features. Imagine your anal probe is also your brain!

>> No.2803195

You can't interface with single neurons this way. For fine control it's necessary to read and write from single neurons. Invasive neural interfaces are the only way.
And how is it a scam? And how has it failed? Monkeys have had simpler versions of the tech in their heads for more than a decade and it still works. Sure, neuralink is not very well managed I'll admit that. It's one thing to kill monkeys, it's another to kill monkeys and not have learned anything from it

>> No.2803205
File: 322 KB, 1280x720, 72ad11d0-a0e3-43f6-8c06-f7a3fbc64bda.jpg [View same] [iqdb] [saucenao] [google]

Don't know if that belongs here, but I think I got banned from AliExpress for some retarded reason.
Can I somehow make an appeal via email? - I can't access (((their))) bogshit AI chatbot that usually handles that.
With normalfags catching on Ali has gotten quite unbearable anyways. So I might just buy off Tabao with an agent or whatever if it turns out they banned my 6 yo account that I purchased dozens of chinky electronics components with.

>> No.2803207

neither do those shitty probes lol

>> No.2803209

(They didn't explicitely tell me I am banned)

>> No.2803214

Stop sending your dick pics to random sellers

>> No.2803253

you pay you go

>> No.2803257

I think they use copper chloride instead of sulfate, wikipedia tells me the spent Cu2SO4 decomposes back into Cu metal and CuS2O4 in water. Though apparently it's used for etching iron and zinc and aluminium, so idk.
Cupric chloride is no easier to regenerate than ferric chloride, though it has the added risk that pulling copper out of solution can continue to deplete it and eventually turn it all into HCl. Well maybe you can pull the iron out of FeCl too. Either way, electrolysing HCl is a recipe for chloroplatinic acid. The forever etchants usually use H2O2 or bubbling air to regenerate it, both methods increase the pH and slow the etchant. I could also then add acid to counteract it, but that just serves to dilute it further. I'd rather have a simple plug and play solution. If I can't get it working with chlorides, I'll jump to bromides. It's a good idea I promise.

>Most plating factories just raise or dip (via slow conveyer belt) the electrodes in the “vat” of electrolyte rather than draining it
Definitely worth considering, but I think draining the liquid is easier for me, if I can get the squished tube to seal well. Moving the rack of electrodes means having them on their own structure, and properly aligned with the vat. Even bolting a DC motor to a ball valve would be easier, I'll only have a couple of inches of pressure to worry about. Anyhow, I'll try the non-draining method first.

I wonder what trying to etch in an ultrasonic cleaner would be like? Real fast?

>> No.2803258

>Do you specifically need a fancy oscillator like this tho?
as i said, that was the only part in the local warehouse for quick and cheap shipping
>will either have 2 or 4 pins
the same part has both a xo version that only needed 4 pins and vcxo version that required 6 pins, and the pin diagrams showed nc pins too, so i could not tell for certain which one i was getting.
all i wanted was a "3.2mm by 5.0mm package has 4 pins" written somewhere

now i'm just going for the txc-7c oscillator, since it's the only one in stock and also has a part built into kicad. if all manufacturers use the same packages and pinouts (or if i can import symbols and footprints without faffing about) then i could go for something cheaper, but who cares if its like 5% of the bom cost instead of 2%.

>an escalator
deadly escalators, nemesis of the chinese

>> No.2803267

that was scary shit, right there...

>> No.2803281

or mill an aluminum pallet?

>> No.2803302

>76 pin QFN
>1cm by 0.6cm
don't you hate it when you plan a project and buy parts without realising how small they really are?

>> No.2803329
File: 208 KB, 612x408, 1685942991889447.png [View same] [iqdb] [saucenao] [google]

I am not 100% sure this belongs in DIY but I thought I would ask in the electronics thread.

Has anyone here ever repasted a graphics card and/or replaced thermal pads? I am hearing enormous hype--suspiciously enormous--for various "thermal putty" products that are touted as replacements for thermal pads. They sound very good in theory, as you just put a blob atop the VRAM chip, or whatever, and reassemble the card, and the heatsink mashes it down to get a perfect gap-filling fit. Proper fit with thermal pads is, of course, not something you can depend on. If the manufacturing tolerances of the graphics card, the PCB, the height of the tops of the chip dies after they have been soldered the dimensions of the heatsink, etc., etc., etc., were always dependably perfect, there would be no need for thermal pads in the first place, because everything would just fit and nothing would ever need more than the tiniest dab of thermal paste. The factories usually use thermal pads of fragile mushy material anyway, but the higher heat conductivity of the big brand-name thermal pads that are touted by so many of the tech Youtubers comes at a cost. They're mostly relatively rigid, stiff material. Trying to crank all the screws down and force everything into tighter contact has been known to crack circuit boards with some brands, even when they are supposedly the "right" thickness according to the graphics card manufacturer's spec sheet. Thermally conductive (for some values of the word "conductive," I know that they are typically orders of magnitude inferior to copper or aluminum for heat conduction) putty as a gap-filler seems like such a clever idea that I am suspicious. Why haven't we always been using it? Why can you only get most of the most highly hyped brands on Aliexpress or Wish? Image of guy applying putty probably not related. Probably.

>> No.2803330

You're right, I don't know what I was thinking.

Thanks for sharing the TIP41.

>> No.2803336
File: 277 KB, 965x824, 3B59FB3C-36AC-403A-AD30-6B30C1842A0A.jpg [View same] [iqdb] [saucenao] [google]

> draining
I somehow forgot to mention peristaltic pumps. I have a few of these I collected from junked inkjet printers. It kind of “massages” the fluid through the continuous tube with these little wheels on the rotor. Dirt cheap. Easy to control. Instead of cutting a hole, just pump it out into the container.

I’m glad someone has the balls to do stuff like this., and I’m not the only one that re nickel plates my probes after sharpening them and other stuff like that.

> etching iron, zinc, aluminum
I was going to experiment with aluminum pcbs, etch them, and there is this “aluminum welding rod” alloy… alumaweld which seems to be able to wet aluminum surfaces like solder.

>> No.2803354

I've heard that some thermal pads have phase-change materials built in that travel throughout their thickness, like a bunch of microscopic heat pipes. At least that's what I saw being advertised when I was researching thermal transfer materials for a motor driver's transistors. My guess is that an isotropic putty will be worse than that.
Be the chad you were meant to be, CNC machine copper chunks to fill all but the tiniest gaps between the ICs and heat-sinks, also machine supporting blocks to clamp against the far side of the PCB to prevent flexing. Go full chad and ditch the thermal interface materials altogether, just lap the surfaces finely and ring them together.

>I collected from junked inkjet printers
Oh I didn't know printers used them, got rid of my old printer half a year ago without scavenging anything from it.
Peristaltic pumps are what gave me the idea for just having a little RC servo or gear-motor pressing against a flexible tube in the first place. Though in my case there's no need to make it with multiple lobes in order to be able to push uphill, all I'd need would be a single lobe. I guess the silicone tubing would be pretty resistant to FeCl, but I don't think I have anything like it, hence why I was thinking about using heat-shrink-tubing. Maybe I can buy it locally from an aquarium or hardware store though.

>wet aluminum surfaces like solder
Applied science did a video on ultrasonic soldering, bet it would look nicer with some kind of flux or inert gas, not sure if it would be feasible for small-scale or surface-mount stuff though. Also it might break bond-wires inside ICs.

>> No.2803370
File: 55 KB, 640x360, continuous ink system.jpg [View same] [iqdb] [saucenao] [google]

>>I collected from junked inkjet printers
>Oh I didn't know printers used them

they dont, usually
you'll find them in industrial printers that operate like 12 hours a day
often called a continuous ink system
coz we all need continuous advertising in our mailboxes or else capitalism implodes

>> No.2803374

tried to buy one of those for my printer once
the chinese vendor was kind enough to tell me my epson printer won't take reusable cartridges without buying a bunch of disposable microchips. i guess they keep track of the serial number of each cart and how much ink has been drawn from each one
i hate printers, i'd rather use a plotter

>> No.2803711

>1 Amp = 10 micromoles per second
Is this right? Is electrochem really that slow?

>> No.2803724

> silicone tubing
It’s *everywhere* here since the government banned those thin one-time-use plastic drinking straws after watching a video where a sea turtle caught one in it’s nose.

You know when you fix your kids’ 3.5 mm headphone jack for the 10th time, and they complain about the looks of the heat shrink job over the jack? I’m using re-usable silicone straw bits to cover it now, and as strain reliefs. It looks great, and pretty tough stuff. And heat resistant.

>> No.2803725

I think older models would have them more frequently than newer ones. I find stacks of them by dumpsters. Where they belong, for sure!
I think the pumps came out of Cannons, but I can’t remember.
The newer ones have almost nothing in them compared to the older ones…. It’s hilarious.

>> No.2803726

>1mol = N_A of particles = 6.022E23
>1C = 6.24E10 electron charges
>1A = 1C/s
>1A = 6.24E10 electron charges / second
>1A = 6.24E10 electron charges / ( second * N_A)
>1A = 103.6E-15mol/s
looks waay lower than 10µmol/s, but maybe i goofed in the math

>> No.2803754

the charge of an electron is 1.6e-19, that second line is off

>> No.2803764

Goofy story behind it, so I'll greentext it.

>buddy is getting divorced
>but not just yet
>his separated wife installed the stupid automatic door lock wrong
>it starts beeping at him
>he takes it apart
>strained connector she fed through wrong side of the piece of shit has broken off
>he tried to glue it back together
>after dicking with it for multiple hours, he calls me and asks if I can solder the connector back on
After about half an hour and some solvent, couldn't get the connector off, so it's time to make a shitty bodge job connector.
I luckily HAVE the exact dupont ribbon cable around, and while it's not the same connector, I can solder down the ribbon cable to the board, hot snot it in place, and then run out the cable to where it has to meet away from the board and connect to the other cable.

My question is based on the fact that my kit has dupont headers and connectors (the receptacle is like this https://www.digikey.com/en/products/detail/amphenol-icc-(fci)/65039-027ELF/4272526)..
How should I make this abortion? Ribbon from board to header pins WITH heatshrink, or should I just solder to the headers and hot snot that, too (aerospace mode), or is heat shrink/vinyl tape the move? And then I'll just crimp the other bitches to shove them into the right hole on the 1x10 receptacle side, right?
Sadly, I can't just solder this whole bitch together because the cable has to go through the door during installation.

>> No.2803765

Ah, fuck, forgot why the greentext is important.
>he'd buy a new one, but thinks his soon to be ex wife will be like OMFG YOU'RE CHANGING THE LOCKS ON ME?

>> No.2803787

you're right, the answer is 10.36E-6mol/s, am retard
also yeah electrochem is pretty slow, that's why you need stupid high surface area for reacting at any decent scale, like in those meme HHO gas generators

>> No.2803809

Using the header looks better but solder+snot would be more robust. But steer clear of vinyl tape it's cursed item that brings with it not only bad connections but bad connections with sticky tape residue.

>> No.2803854
File: 117 KB, 1280x960, P1070537.jpg [View same] [iqdb] [saucenao] [google]

Can anyone recommend a cheap voltage standard accurate to at least a millivolt?
I have an opportunity to buy pic related for $50. However, its made in the 60s, uses 2 vacuum tubes, and likely isnt going to just work without some work. It is also noisy beyond the millivolt range. however, its only $50.

Should I get it, or is there a better option for $50+work or maybe like $150 good to go. also curious what there might be thats accurate or at least stable beyond a millivolt

>> No.2803884

I get that just for the fallout4-motif workshop.
I got a cheap 10.0 voltage reference from china.
You’d think there would be some common “natural” voltage references that you could build.
How accurate is a brand new zinc carbon cell, for example… maybe it’s 1.6352 V in all cases.
I assume if you build some kind of electrochemical device, it will produce a known exact voltage for that cell chemistry, independent of other variables.

>> No.2803936

ok I got it. luckily theres a pretty fair amount of info on restoring these.

>> No.2803997

Temperature variation is gonna bite you, though not as much as the thermals of a hot cathode tube. I think they used to use mercury-based electrochemical cells as voltage references. You could probably get decent accuracy measuring the potential across a pair of quality reference electrodes and keeping them at a fixed temperature, but that seems like it would be more expensive and less repeatable than an LTZ1000.

>> No.2804023

Lol, it’s basically a zener diode and resistor dividers. I could have thought of that.

>> No.2804028

Thats just a single voltage though right? I should probably get one of those too, and when ever I calibrate this one I can make sure it matches the LTZ1000 at 7V or whatever.

The vid I posted above only calibrated to 100.0V. But this should (theoretically) be accurate to, for example, 7.000V

>Temperature variation is gonna bite you
I suppose I could maybe run some thermocouples internally to get an idea of temps.

>> No.2804037

whatevs, it looks cool, and has a bunch of 1% resistors and a pot (I guess) that allows 1 mV adjustment. I always wanted a voltage standard, and as meh as this one may be, for $50 I think its ok.

>> No.2804039

If built today, it would have a quartz window showing off the tubes illuminated by LEDs.

>> No.2804065

>people put shitty carbon-film pots or trimpots on their precision voltage references
do not do this

Follow along with Marco Reps as the builds an LTZ1000 reference board. Again.

>> No.2804071

not trying to be smart or anything, but how much does an LTZ1000, with supporting circuitry and PCB, cost? Is an LTZ1000 good to use as a voltage reference from like 1V to 100V?

Sure I dont actually need completely variable from like 1V to 100V, but I think 1V, 2.5V, 3.3V, 5V, 12V... and so on would be nice. Guessing the LTZ1000 is good for all this with the right circuitry

>> No.2804079

$75-100 per IC, plus another $20-50 for good metal film resistors, maybe $200 all up if you're not adding your own ovenized enclosure atop it all.
>Is an LTZ1000 good to use as a voltage reference from like 1V to 100V
It's a fixed 10V reference. You can use precision resistors to divide it as you need, but generally you use such a reference to calibrate the ADC inside your measurement equipment (or even have one inside your benchtop multimeter). In that case, 10V would be the full-scale-range of the ADC, and the meter itself would have the dividing resistor arrays and programmable-gain-amplifier built-in for measuring different voltage ranges. With a fixed voltage precision voltage reference combined with a fixed precision resistance reference, you can perform a full calibration of most meters.

FYI I was trying to be snarky with my suggestion of the LTZ1000. While it is a really nice piece of kit, it's completely superfluous unless you have a 6+ digit multimeter that requires that kind of accuracy. There are cheaper voltage reference ICs out there that might be more suitable for your applications, check digi-key and filter by tolerance or tempco or whatever. But even if you make your own reference, you'll want to get it initially calibrated by an instrumentation house somewhere. Buying a pre-calibrated one may well be a more cost effective solution unless you're as autistic as Marco Reps and want like 12 of the things and to measure their voltages relative to one another for long-term error tracking. But I wouldn't trust a chinese calibration house for a reference you get via free shipping.

>> No.2804080

>how much does an LTZ1000, with supporting circuitry and PCB, cost?

yeah, $700 is insane
when i needed a 2.000V reference i wired a 9V battery to a 7805, and used a 20-turn pot to adjust it to 2V using a good calibrated meter
the meter becomes the reference
any tuning can be done in situ at the time the reference is needed
using just a fingernail, which you can buy anywhere

>> No.2804092

>a good calibrated meter
when was your meter last calibrated? i look up places in my country and only see "contact us" or "inquire for quote" and they want a business email, so i assume it costs more than my $250 hioki meter. other places exclusively take flukes. none are in my town either.
getting my crt scope calibrated by someone is also probably a pipe-dream, so buying or building a reference that will stay good for 10 years might be the way to go.

how hard can a josephson standard be to diy anyhow?

>> No.2804094

>it's completely superfluous unless you have a 6+ digit multimeter
i have a 5 1/2 that I would like to make sure is reasonably accurate

>> No.2804150

do multi-channel analog cmos switch ICs (dg444) ever have a failure mode where there is current leakage between channels? trying to troubleshoot a fairly complex PCB. physically pulling this IC fixes the problem but it's also upstream of a ton of other shit.

>> No.2804244

Negative voltage so can be the coil induction when witch off (which is inverse to power voltage). Maybe your diodes are to slow to high in impedance and/or bad wiring. Decoupling is always good use a 30 Volt zener some inductance ... On 25V with analog I power relais with nominal voltage and reduce it to holding voltage to minimize em-energy.

>> No.2804254
File: 33 KB, 1086x1035, Capture.png [View same] [iqdb] [saucenao] [google]

i have an ohmmeter on pins 2 and 3 of RV1. i rotate the potentiometer through all 10 turns and i see it go from ~1 ohm at one extreme to ~1.2k ohms at the other extreme.

maybe 3-4 turns from ~1.2k i see the resistance peak at ~1.6k.

how does this make any sense?

>> No.2804263

Don’t use an ohmmeter on powered devices

>> No.2804270

it's unpowered when i measure it.

>> No.2804281

> etch thousands of josephson junctions
I’m pretty sure you can use ordinary semiconductor bandgaps, and/or solar cells somehow. Just hold it at 20 ℃ or so and use low current. Even dave jones was sending single photons through a led a few years back on his bench.

Almost nobody needs more than two decimal digits of accuracy.
It’s the same with a machinist’s paradox: to build a car you only need to know pi to 2 decimal places.

Marco reps needs help. He’d think he needs to know Pi to 10 decimal places to 3D print a replacement plastic wheel for his tonka truck out of PLA even though it starts melting at 60 ℃ and has horrible expansion coefficients. It’s a fantasy.

Imagine, in a typical production circuit, all the solder blob sizes, components with wide tolerances, wire lengths and dimension differences, induced voltages and noise, capacitive and electrostatic charges, humidity, altitude, microphonics, thermal conductivity coefficients, wildly varying ESRs, different batches of components, varying copper and plating thicknesses, things with PTCs and NTCs, etc, etc, ad nauseam…

A lot of that you can’t control in the circuit itself, it literally changes .1% when you breathe on it. You just hope most of that shit cancels out in the final analysis.

>> No.2804286

> cmos
> touched the pins
> carried my chip in my polyester pants’ pocket and went to the discotheque wearing crepe rubber shoes on a rayon carpet
> during winter
> only bought one chip

When ordering components, especially as something generally useful as the dg series (I’m in the cd camp though… more common imho) always order more than one.

I learned this from my dad, when he took me to the hardware store to get nuts or bolts for some project. He’s always get at least one more than he needed, sometimes two… and put the extras in the parts drawer.

The reason I always advocate using the most bog-standard components instead of fancy ones is they’re useful in more situations, and digikey now requires $100 for free shipping. Either that, or risk chinese clones.

>> No.2804296

Where does "bog standard" come from? Is the word bog even necessary in the context? What does a bog have to do with it anyway? These are questions that have been burning in my mind for 27 years.

>> No.2804356

Hello guys,i want to enter the PLC programming world and need advice on where to start.

I already know control systems, some electrical stuff and some knowledge on how to read a technical drawings.

i you want to

>> No.2804365

well better choice posting here than /mcg/
are you in America or euro?
In the US Rockwell/Allen Bradley is king. Euros are majority Siemens.

>> No.2804376

Not unless there’s a fault that jumped through the insulating gates of FETs. Check the datasheet, but I’m fairly sure there’s no drain-to-source connections between control and signal lines, ignoring the power rails’ body connections.
When you take the IC out of the circuit, are you replacing it with wire jumpers that manually make the same connection that causes the problem?

Is Vout connected to anything? Have you verified the markings on R6? If R6 is a lot lower, or if there’s resistance to ground on the output, you’d get nontrivial parallel resistances resulting in a lower resistance than you’d expect.

> Almost nobody needs more than two decimal digits of accuracy
I’d like closer to three, just for the assurance that the 1% tolerance of their components I work with are more significant than any systematic error from my meter. Also my scope has terrible DC offset issues that I need to fix. I want to calibrate them since it’s been like 7 years just for that assurance, and there doesn’t seem to be an easy and comparatively cheap method since I only want 4 digit accuracy.

Marco has fun and makes YouTube bucks, living the dream if you ask me. Some people collect stamps or coins, he collects ppms. It is not a means to some esoteric end, it is a hobby unto itself.

>> No.2804400

not /ohm/, fuck off

>> No.2804414

you're just a bowl of sunshine aren't you
PLCs make money using electricity, your little Arduino or rasppi don't

>> No.2804419

>Arduino or rasppi
Well that isn't /ohm/ either. PLCs and MCUs and SBCs fall under the blanket of the (really slow) /mcg/, though I've seldom seen anyone with any PLC expertise over there, you're more likely to find people who can program FPGAs. It's probably better off if you look at tradie threads, sparky threads, or the SQT. Even /g/ might have some PLC expertise. This general is for people designing and building circuitry, it's a seperate layer of abstraction from programming.

>> No.2804420

/mcg/ doesn't do PLCs either
there was /ind/ which was Industrial General but it was short lived

>> No.2804432

I actually do see PLC threads here, but usually they're their own thing. Sort of like how sparky/trade shit isn't /ohm/ either.

>> No.2804454

Has anyone here managed to make a set of miniature spy cams? Like can you connect them to a network like regular cctv cams without WiFi? Do you have to risk audio with these miniature cams? I thought about getting a spy pen but it seems retarded and way too low quality

>> No.2804571

Just use your phone like everyone else

>> No.2804575

I should have used “common as dirt” which is clearer for an international audience, yeah a lot of phrases don’t make much sense anymore.
It’s probably a british term, like “motorized rollingham” means “car” there.

We used to call capacitors “condensers” back in the day, and before metric happened. So you’d see “mfd” which would be interpreted differently today using metric prefixes.

>> No.2804576

/g/entooman here, I have a mouse problem and I want to make an automatic pantry door closer. The one at ground level about takes a kick to the corner to actually close, so I would either need to fix that or put a motor down there to ratchet it shut. I also want a delay between door movement and close.

What electronics could I use to achieve this? I can solder but I haven't worked with microchips since high school (5 years ago)

>> No.2804599
File: 3.30 MB, 800x600, output.webm [View same] [iqdb] [saucenao] [google]

I am making a remote for my media server, today I printed the PCB and soldered it all. Only a few fuck ups, but it works

Basically it allows me to connect to a server and control it or run scripts. I have various scripts on my media (and other stuff) server and I want to be able to control it via this remote

>> No.2804601
File: 3.84 MB, 1264x1806, file.png [View same] [iqdb] [saucenao] [google]

Here how it looks like, the wire in middle, between the keys, is me forgetting routing one of the pads. And there will be 2 more key displays on top later

>> No.2804671

I am in latin america, i have access to siemens, schneider electric and some other chinese brands to work with.

>> No.2804689

Fix the door getting stuck, put a spring on it. Use a damper to make it slow.

>> No.2804724

Im testing some analog ics and I want to get a stable 1.8-3.3v. I could use the mcus voltage but something tells me thats a source of noise in the circuit. Should I get a low noise voltage regulator or will my cheap bench psu work. Or am I overthinking it and I should just keep using the mcus 3.3v line.

>> No.2804753

Is a red pitaya useful to have?
My dad pulled one out of his storage and gave it to me, seems pretty cool?

>> No.2804778

Personally I'd replace those soldered vias with riveted vias. But I guess if you fold over the wires before soldering they won't just fall out when you reflow other stuff on the pad.
Also deburr those holes for the top three switch pegs. And route it nicer next time, there's some senseless curves and turns here and there.

Is it for a power rail, or just a reference? if it's a power rail, some people would just put a small SMD choke (with following cap) on the digital power line, which is generally fine. Others would step it down from a higher rail (e.g. 5V) using a common LDO, but LDOs have pretty shitty ripple rejection at higher frequencies, they're only really good at ~120Hz ripple rejection from a linear transformer PSU. If you do want this kind of linear regulation, you'd be better off putting a capacitance multiplier behind an LDO, or even replace the LDO with a discrete BJT regulator. Using a decently sized cap on a zener-BJT regulator that uses a TL431 instead of a zener works real well, maybe add a choke there too. Generally this is overkill unless you're stepping down a really noisy rail, but I'd definitely do at least this if I were making a hi-fi microphone ADC circuit. Or a preamp for measuring noise.

>> No.2804849

I am >>2804601

I don't have riveted vias. It sounds like they could cause connectivity issues with my lousy diy setup or are they good? The pins I used were pretty tight fit so they didn't fall off

Those 3 holes were manually drilled and I forgot to deburr. Other holes didn't need deburring because I made them in my cnc. Those 3 are larger and I didn't have the bit

And for the routes, I tried to put distance between lines and try to wiggle things a lot but at some point I probably said "fuck it" and left those weird turns

>> No.2804906

So I just got a Arduino CNC Machine Kit and I have yet to start building. What else besides the Arduino CNC Kit (i.e. the Arduino, the "step up" controller, and the motors) do I need to start milling me out something "good"?

>> No.2804926

Define kit, do you just have the electronics?
If yes, you're gonna need a frame, some sort of aluminium beams is probably fine.
How are you gonna drive the axes? belts? screws? how are you gonna mount those? How will you connect the motors to it?
How are you gonna hold the tool? how are you gonna drive said tool?
Do you have a power supply?

>> No.2804957

What's your favourite colour of bog-standard aluminium? I'll just boil some watour in me electric kettle and plug me fish and chips into the 220V mains as I watch the BBC and await your answour. Cheerio, chap.

>> No.2804960

Ah yes a fellow european chap.
We're not like those barbarians across the pond with their aloominoom.
Pleasure speaking with such an esteemed gentleman such as yourself.
Chip chip cheerio, god save the king.

- Nigel Thornberry

>> No.2805007

Oh in thia case I'm just driving an LED panel, no coils involved. It is on a breadboard, so indeed can be bad wiring.
So by decoupling you mean put a zener in parallel or?

>> No.2805039

i kept wondering why the little shitbox i made kept reporting it had low battery charge even though i charged its lipo fully via tp4056 module so i finally checked the bat lotage with a multimeter and found out it was only 3.5V so i was like what the shit, the tp4056 always shown blue led to indicate it was fully charging but then i realized the USB wall wart is supplying 5V and i have polarity protection diode on the +, so the tp4056 was getting like 4.2V and it means when the battery charged to about 3.5V the TP voltage output was too low to keep charging so the charge current fell under 10ma so it probably assumed it's fully charged and turned the blue led on making me think it was fully charged.
What a fucking motherfucker of a bug.
But then what to do? If i ditch the polarity protection diode i can power the tp4056 directly from a 5V usb wall wart, but i don¨t wanna.. since i accidentally reverse the polarity of the charging cables all the time.

>> No.2805049

5V relay with diode on the coil
If the polarity is correct, diode conducts and relay closes to allow 5V further down to the devices.
Probably the simplest option that gives no voltage drop to the stuff you're powering.

>> No.2805059
File: 32 KB, 700x250, dc_transformer.jpg [View same] [iqdb] [saucenao] [google]

Sorry for the dumb question, I am a newbie. Would a raspberry pico consume more power if it is powered using 2 batteries instead of 3. I am guessingg yes but I wanted to confirm. The transformer converts input voltage to 5v and if input voltage is lower, it will draw more amper from it? Or I am wrong and there is more to this and it will pull the same amper no matter what

>> No.2805077
File: 187 KB, 810x1380, b.jpg [View same] [iqdb] [saucenao] [google]

The buck boost converter will have a higher initial efficiency with just two AA batteries in series. But if you put three in series the converter may be more efficient on average, and you can probably discharge the batteries a bit more. On the other hand, batteries might not be consistent and if you're more likely to get a dud that ruins performance if you use three.

>> No.2805138

I think the rivets would be fine, though I haven’t tried them. They also arguably want a dedicated punch to stop them collapsing. I just used solid core copper wire and a punch to make small flush rivets that can even be in an SMT pad, though it’s a pain to do.

Use a P-FET instead of a diode.

>> No.2805167

Yes, I currently only have the electronics. I am thinking of driving the axes with belts - seems "up my alley" and for the other questions - I'll learn as I go. Also, no, I don't have a "power supply".

>> No.2805168

Also, what type of power supply do you recommend?

>> No.2805186

I’d recommend getting one of those inkjet printer/scanner power supplies, I think they have two voltages, one of them is around 30 V usually. They drive some motors in the printer, so maybe that’s a good match for what you need.
My local thrift store has a dozen of them. Most common one has a purple end-connector with 3 pins.

>> No.2805213

I mocked up a circuit with a comparator and current sense resistor on a bread board and it worked with the fish finder.

Took me forever to wrap my head around the comparator, kept trying to make a follower voltage in error. Thanks for the input, that was the way to go.

>> No.2805228

Depends on what voltage your electronics want. If it's similar to a 3D printer with TMC2209s or whatever, that's about 24V. Check datasheets to see what the acceptable ranges are, the TMC2209 for example can handle up to 29VDC, and as low as 4.75V, but the lower the voltage the less torque you'll get. 24V supplies aren't as common as 12V or 5V, you might find some at a thrift store but don't hold your breath. I'm running two old ATX PSUs in series (I had to remove the ground reference on one of them) to power a 3D printer, though the major power drains aren't the motor drivers but the filament heater and bed heater, and even then it's vastly overkill. In your case it will probably be the spindle motor that eats the most power, hopefully it runs at the same voltage as the steppers, unless it's a mains spindle.

Whatever draws the most power is usually what you want to set your primary power supply to output, the rest of the power rails can be derived from DC-to-DC converters or even linear regulators if the power demand is low enough. Multi-rail supplies do exist, but they're generally less common and more expensive than single rail supplies. With the exception of used ATX supplies, they're great. If you buy new, get a name brand Mean Well or similar, they make some dual-rail supplies so browse their catalogue.

Cool stuff. Post a schematic.

>> No.2805234

interesting solution, but relays are big and bulks, altho they are pretty much zero voltage drop

how would the circuit look like with a pfet? since with a diode it's great it's literally just a single component plugged in series with the wire

>> No.2805243
File: 20 KB, 474x271, extremely common circuit.jpg [View same] [iqdb] [saucenao] [google]

did you google image search "pfet reverse polarity protection"?

>> No.2805269

Nice to know some of that was useful.
A lot of circuit breakers use comparators on a current transformer to pop it on over current.

>> No.2805275

damn that looks really simple, why the fuck have been using diodes for polarity protection all these years like a fucking retard, i hate myself

>> No.2805280

what the fuck? i was wrong
i hooked up 6V directly to the input of tp4056 board and it's still showing the battery as fully charged despite the battery only being at 3.5V, how is this possible?

>> No.2805281

are you fucking kidding me? took the lipo out, hooked up 4.2V directly to it, still 0 current flow, what in the shit

>> No.2805283

never mind i had a bad contact, it's charging now when hooked up directly to the battery
so back to the mystery why the fuck is 4056 undercharging it

>> No.2805286

>showing the battery as fully charged despite the battery only being at 3.5V
>why the fuck is 4056 undercharging it
0.7V drop across your protection diode?

>> No.2805290

>>why the fuck is 4056 undercharging it
>0.7V drop across your protection diode?
no, i solved that by bumping the voltage to 6V so 4056 is now getting 5V on the input

>> No.2805294

Try shorting BAT- to the USB connector's ground with a jumper wire for 2 seconds, then plug in 5V.

>> No.2805334

Hey, thanks for the suggestions for power supplies and for my particular project I want it to not be a 3D Printer but instead be a CNC "Milling Machine" (i.e. it can be equipped with tools needed to shape metal pieces into parts e.g. say turning some metal stock into a door knob for instance or even other general "milling" tasks).

>> No.2805364
File: 3.42 MB, 4000x3000, DSC_2562.jpg [View same] [iqdb] [saucenao] [google]

Looks like I finally found an honest seller not sending me empty epoxy blocks. They were quite expensive at 3.8€ per chip though

>> No.2805367
File: 3.39 MB, 4000x3000, DSC_2563.jpg [View same] [iqdb] [saucenao] [google]

These look rancid. But I expected nothing else from 20ct per chip. As long as they work
Don't want to imagine the poor chink slaves who had to desolder them

>> No.2805381
File: 1.10 MB, 4000x3000, DSC_2565.jpg [View same] [iqdb] [saucenao] [google]

Its like christmas :^)

>> No.2805382

>Don't want to imagine the poor chink slaves who had to desolder them
Shouldn't they have short legs if they were soldered somewhere?
Hard to imagine someone using them as is

>> No.2805386

I suppose they would need some headroom to be mounted to a heatsink.

>> No.2805432

Usually you can get away with smaller power supplies for the axis motors because they are so geared down, and possibly use worm drives.
Some of the DIY cnc people use trim routers connected to mains power separately to drive the actual spindle.
Of course you’re going to be cncing wood (or maybe hard, friable plastics) at 20,000 rpm. My router is “variable” but only slightly, it still likes being a high speed.

>> No.2805443

> desoldering is slavery
If I told you how much time, over the last 4 or 5 decades, I’ve spent desoldering stuff, you’d be horrified.
One thing I’ve learned, is that generally you want to leave things on the boards until you need it, but I don’t have that kind of space, it would look like sanford and son’s backyard.
Like yesterday, I needed a 7809 for a fluke battery replacement adapter. Yep, I’ve got one!
A lot of stuff I got back in the 60s and 70s so it’s “original”
You can still get (kinda sorta) original parts from their lineal descendants, like you want to get your 2N3055 from onsemi because it seems likely they took over those fabs from Motorola…And they’re definitely better now than the original RCA ones.
The reason you see so many fakes (and the 2N3055 is no different) is because the creation of the die is separate from the packaging and encapsulation.
So there is all these little factories that encapsulate chips from bulk wafers all over the place, they simply run at night and put through some random tiny 5 A transistor die from a written-off wafer that failed testing instead of the 15 A die for the original authentic part, and slap them in a TO-3 case nd ship them to audiophiles in a class a amplifier kit, and sometimes it more or less works for a while.
I think half the guys here struggling with their circuit wondering why it doesn’t work or only kind of works is because of fake components. I know it costs extra, or it’s extra work, but try and buy legit parts at least initially until you can call bullshit, or salvage from old equipment… usually you want the rev 0 boards because you know the next rev is going to cost-reduced.

>> No.2805445

How to eliminate ground loops, preferably in a low cost and /diy/ way? First time having trouble with them, I get noise in audio from USB soundcard to headphone amplifier and through HDMI->TV->Amplifier. Headphone amp has ground lift switch that eliminates them, as does powering it through an isolation transformer. For TV, using optical connection would certainly eliminate the loop but that'd need another DAC because NAD C320 obviously doesn't have Toslink. Could I create isolated ground for computer and audio equipment by driving a grounding rod deep enough to the ground outside, running wires inside and wiring outlets with L/N from normal supply and each with separate ground wire to the isolated ground?

Unboxing /ohm/-orders is joy, and even more like christmas when you pick up something that comes from a weird sender name from a random country and have no idea what it is because shipping took three months.

>> No.2805447

Forgot to mention, already tried powering everything from the same outlet but didn't help.

>> No.2805467
File: 7 KB, 209x113, cuts.png [View same] [iqdb] [saucenao] [google]

What do the little cuts in the line mean? Inverted?

>> No.2805553

By the ‘00’ and ‘01’ labels, I can deduce it means ‘bus’ as it usually does, without the ‘2’ bus callout indicated because it’s obvious.
So, 2-line bus, which is the most degenerate kind.

>> No.2805556

I’m not convinced it’s an actual ‘ground loop’—those are very specific things but I hear a lot of noise about them (ha ha, get it? Noise? Nvm.)

There is not one secret trick to get rid of the noise, but you can try shielding (shielded cable, ferromagnetic cases, etc) or if it’s coming from the mains (possibly ground) you might want to try filters.
The isolation transformer not only eliminates any literal direct connection (although ground can be passed directly through) on hot and neutral, it also has a HF filtering effect.
Is this a class D amplifier by any chance?
Usually HF just goes everywhere and it might interact with the class D frequency to bring it down into the audible range.
You might want some low pass filters. Some power bars have them.

>> No.2805561
File: 3.22 MB, 5504x3096, 1711354949217560.jpg [View same] [iqdb] [saucenao] [google]

I've never soldered anything in my life, but I have a cheap Chinese soldering kit
Somehow, I've seemingly snapped off the red wire on the left of this board. It's a PS5 controller.

I assume I just solder this back on and it will be work, right? How easy would this, just this, be to do? Would I need to use extra solder wire or just sort of hover my iron between the red iron and silver blob of a connection point?
Any help is welcome

>> No.2805563
File: 242 KB, 1062x764, derf.jpg [View same] [iqdb] [saucenao] [google]

I posted a few threads back about solutions for logging AC power consumption, anyway I ordered this 10 USD module and it seems to work pretty well. The graphs show power and current for an aquarium LED lamp with variable brightness/colours. You just query various registers over I2C to get voltage, current, power etc.
Honourable mention to the anon who suggested setting up a webcam to read the LCD of a kill-a-watt meter, but I think this will be a bit more robust in the long term.

>> No.2805565

Absolute minimum instructions assuming you just have a working soldering iron: Remove a bit of the insulation from the broken red wire at the end. Heat up the blob of solder until it turns in to a liquid, then push the end of the wire in, wait a second, then remove the soldering iron.
There are tons of other tutorials online that will go in to more detail.

>> No.2805566

You’re one of those guys that spaz-out on the controller during the boss fight, right? Stop doing that.
Anyway, yes. Get a cheap 27 watt (approx) soldering iron, and some rosin-core flux solder. Maybe there’s a kit with a little spool of it.
Strip off 1/8” (2mm) and tin the end of the wire.
Do not nick the conductor while doing this.
Then place the tinned wire on top of the pcb blob and pres. It will melt everything together and hold it for a sec or 2 until it cools.
The hard part is re-assembling those fuckers without loosing the springs.
Now, for good measure, you might want to drop a dollop of silicone or something on top of those solder connections ensuring you catch the wire insulation in the blob to act as a strain relief when the controller hits the wall because you rage quit.

>> No.2805569
File: 1.68 MB, 4000x3000, PXL_20220907_093214685_1.jpg [View same] [iqdb] [saucenao] [google]

you can definitely fix that
here's a bodge wire I had to put in a GameCube controller after it slid off my couch and broke the trigger button trace somewhere

>> No.2805570

> webcam
That was me. I think that discussion degenerated into a real vs apparent power and computing real-time transcendental functions bonanza.
Your solution looks good, although the power looks odd, as if it’s actually a capacitive drop-off.
Which is the module? I can’t tell from the pic.

>> No.2805578

>remove a bit of the insulation from the broken red wire
How would I do that, and is it absolutely necessary? I'm not asking in a "i don't want to do it kind of way", more of a I'm just curious how this works kind of way
>blob of solder
is that the blob that's already on the board?
thanks for the tips

Nah, My controller started drifting, I sorted it out before by just cleaning the potentiometer discs, but this time as I opened my controller I noticed that wire was snapped. I don't know how it happened, probably my fault but I'm confused about it. I've opened and closed it plenty of times. This time I was a little tired but I just don't recall any moment I snapped the wire.
>tin the end of the wire.
>Do not nick the confuctor while doing this
Any more on this? I don't want to screw this up, not sure what I could do wrong.
I already have 2 cheap soldering irons but I don't know if they're good. They came with solder wire and stuff. I'm not sure I have a wire stripper this thin so that might be a hurdle.

thanks. i think this is just a vibration motor so nothing major. but i feel like i should at least try sort it. i mean, it's unlikely to completely mess this up and ruin my whole controller, right? worst case scenario i guess is i ruin the red wire completely and can never reconnect it

>> No.2805587
File: 11 KB, 275x288, Screenshot 2024-05-28 214304.jpg [View same] [iqdb] [saucenao] [google]

It's this product from "M5 Stack":
I might be using the wrong power measurement, it reports "active power" and "apparent power", but I can believe the LED power consumption would be a bit strange because it's got its own power supply downstream of the logging unit.
I had to cut the connector cable because they seem to use their own proprietary "HY2.0" connector which doesn't seem to correspond to any normal standard, I'm using Sparkfun qwiic connectors to connect it to my Pi.

>> No.2805590

in a humorous twist, i have now snapped the red wire on the other side too.
i guess this is my fault for deciding i would sort this before i head to sleep. it's now almost 3am and this has gone great.

>> No.2805597

Well I was just assuming it had similar parts to a 3D printer because you didn’t mention otherwise. If you want something more detailed then tell us what motors and motor drivers you have. I searched “arduino CNC kit” and just got a bunch of different tutorials.
The eternal machining general may have more expertise on CNC mill design.

Isolation transformers on the audio lines. Maybe grounding everything even better could fix it, if the source of the noise is an ungrounded SMPS. I blame the TV. Mock up a cable that connects to a piece of metal on the TV (mounting screw, USB shield, etc.) and plug it into the ground hole of a socket.

You might be able to solder through the insulation, but it will be awful and smell a lot, and probably make a bad joint.

>> No.2805604

>You might be able to solder through the insulation, but it will be awful and smell a lot, and probably make a bad joint.
In the absence of a wire stripper, what else can I use? Will I just need to buy a dedicated one? As you can probably guess, I'm not the best with this stuff. I have some small pliers but even they seem too big for the tiny controller wires, which I seem to be continually snapping off at an alarming rate now.

>> No.2805625

>I've seemingly snapped off the red wire
dont feel bad. its probably aluminum wire which breaks after bending just a little bit. pretty common on shit thats built as cheap as possible. but yeah just strip the end of the wire, melt the solder, and shove it in. highly recommend using flux on both the solder blob and the wire

>> No.2805629

>In the absence of a wire stripper, what else can I use?
Xacto or sharp knife. you can probably just use your fingernails too. i personally hate the following method, but i know people who would just take a lighter to the tip and burn a bit of insulation off the tip.

>> No.2805649

Just use a craft knife, wire cutters, or even scissors. You just have to be careful not to cut all the way through the wire.

Aluminium wire doesn’t wet to solder at all, it might sometimes be used for crimped connections but probably not for anything soldered. Maybe they can tin plate it, but it’s Sony not Shenzhen.

>> No.2805656

>Aluminium wire doesn’t wet to solder at all
I think its sometimes copper plated or something.
maybe it is copper wiring, or a shitty grade of copper. but I have encountered wires like that breaking all the time from fatigue. I thought copper didnt do that but maybe im wrong.

>> No.2805662

Um, does this circuit actually work with a battery? The gate-source voltage is the voltage across the battery itself, so the FET will always be on so long as the battery voltage is greater than the threshold voltage. It can still source a bunch of current back into a reversed power adapter until it's discharged below this point.

Not sure if this is still the case when there's a TP4056 between a lithium cell and the PMOS diode circuit, depends on if current can flow back through the IC or if the input voltage can drop to zero.

Copper definitely work-hardens and breaks, especially thicker strands. Really fine strands don't usually have this issue, but even then soldering them to a blob of solder without any strain relief or wicking of the solder back up under the insulation is a quick way to end up with broken strands. Tinned copper usually wicks solder up under the insulation, which adds some strain relief, which explains why I usually have this issue with non-tinned (especially somewhat oxidised) copper.