/ohm/ - Electronics General - pp amplitude edition

[Anonymous No.1555499]

tell this one to your valentine: "are you AM? Because you modulate my pp amplitude"

>RULES
0. Electrics ≠ electronics. Mains wiring goes to /qtddtot/ or /sqt/. PC assembly to >>>/g/.
1. Do your own homework. Re-read all documentation/datasheets related to your components/circuits, and do an honest web search, before asking.
2. Pics > 1000 words. Post relevant schematic/picture/sketch/9001.5 hours in MS Paint with all part numbers/values/etc. when asking for help. Focus/lighting counts.
3. Read posts fully. Solve more problems than you create.
4. /ohm/ is an anonymous, non-smoking general

>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

>Project ideas:
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/

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

>Design/verification tools:
LTSpice
falstad.com/circuit/circuitjs.html
NI Multisim
CircuitLab
iCircuit for Macs
KiCAD (pcb layout software, v5+ recommended)

>Components/equipment:
Mouser, Digi-Key, Arrow, Newark, LCSC (global)
RS Components (Europe)
eBay/AliExpress sellers, especially good for component assortments/sample kits (caveat emptor)
Local independent electronics distributors
ladyada.net/library/procure/hobbyist.html

>Related YouTube channels:
mjlorton
paceworldwide
eevblog
EcProjects
greatscottlab
Photonvids
sdgelectronics

BigClive

>Li+/LiPo batteries
Read this exemplary resource first: https://www.robotshop.com/media/files/pdf/hyperion-g5-50c-3s-1100mah-lipo-battery-User-Guide.pdf
>I have junk, what do?
Take it to the recycler.

[Anonymous No.1555501]

>>1551656
old

[Anonymous No.1555506]

>>1555499
This thread's digits brought to you by the good old 555, and the Maxim MAX5499 dual digital potentiometer.

[Anonymous No.1555508]

>>1555506
Also the TI TL499A, for when you want a linear regulator and a boost converter in one chip because why not.

[B.P.P. No.1555533]

Yesterday was National Battery Day. Happy belated.

HF has a new meter! Throw out yer Fluke and grab the Ames at a fraction of the cost! They claim this is more of a bench meter made for fegs like you compared to the previous top meter that was advertised as a sparky and tech meter.

[Anonymous No.1555536]

>>1555533
never had a multimeter but I really want one with a clamp

[Anonymous No.1555559]

>>1555533
>

>>1555536
I'd get a clamp-meter seperately to a DMM such that you can measure voltage and current at the same time. The AC/DC UNI-Ts don't look awful for the price.

[Anonymous No.1555565]

>>1555536
>>1555559
I dunno why they even bother with the meter part. Everyone just hangs 'em off their nipples anyway.

[Anonymous No.1555567]

>>1555559
oh man I just missed that get

[B.P.P. No.1555575]

>>1555559
+1 about getting both. AC-DC clamp meter is nice, but you can’t read current through the leads with most of them so you’re SOL for lower current measurements (like <5.0A).

That Uni-T seems to be the most affordable DC clamp meter from a brand that anybody would touch, but it’s small and has a low cap for most measurements. I think that Ames brand has a couple DC clamp meters for like $70-$120 but I’m weary of buying anything electronic from HF for >$25 because of the 90 day warranty, but those meters are far better quality from the old Cen-Tech trash. One of those higher end Ames clamps has an inrush setting which looks interesting.

I think the cheapest Klein clamp that will read DC is $100-$120, that’s probably a solid choice. My Southwire was $80 but only reads 400A and people said that was low for starters on lots of vehicles, however that’s probably the only thing the average guy will go near that’s pulling so much curent. Maybe a 600A or 1000A meter was worth the extra $20-$40, but I have been content with the 400A model.

>>1555536
First meter should probably be a regular one with fused inputs. Fuckin with the little electronic gadgets will keep you <10A anyway.

[Anonymous No.1555577]

>>1555575
>pic
I too like spot-welding my wrench to my battery terminals on my car/motorcycle

[B.P.P. No.1555581]

>>1555577
Since the microwave is all the way in the kitchen, I use that trick to reheat my coffee. Short the wrench for a few seconds and then drop it in the half full mug that has gone cool. Made my wife tea like that once as well.

[Anonymous No.1555597]

>>1555575
I wouldn't touch a DMM from harbour freight anyhow, if you're going to spend less than $200 you can get better value from the usual suspects, and if you're going to spend more you might as well get a Fluke or similar name-brand. EEVblog has DMM reviews from all sorts of brands and there's a very comprehensive spreadsheet of DMM specs on the EEVblog forum.

I've got a superior HIOKI Japmeter, and the levels of protection on it are top tier (ceramic sand-filled fuse, isolation slots, etc.) which I got because where I live Flukes are only sold at a fairly high markup, and shipping from elsewhere would have a similar effect on the price. If you do get a cheap meter, I'd take it apart before turning it on, just to check for dry soldering, dirt on the PCB, and possible sources of mechanical strain. A big thing to check is how the banana sockets are mounted to the PCB, as pulling and pushing on those could easily cause the traces nearby to work harden and break in a year or so. An easy to replace fuse and battery are also things to watch for, but apart from that the spreadsheet will contain basically any information you could care about, from battery life to different ranges and modes, even obscure features like inductance measurement if you hunt for it.

http://www.eevblog.com/forum/testgear/multimeter-spreadsheet/

[Anonymous No.1555600]

>>1555597
Oh just to note, the DMMs in the spreadsheet are only those good enough to be both autoranging and true RMS AFAIK, but in general they're a good source of DMM eduction for buying any DMM, be it one on the spreadsheet or not.

[Anonymous No.1555614]

>tfw no gf
At least I have you guys

[Anonymous No.1555618]

>>1555614
amen

[Anonymous No.1555652]

>>1555575
Have the Uni-T 210E, cheap as, does DC clamp amps down to 1mA resolution and up to 100A, having it so small is a nice feature IMO.

[Anonymous No.1555663]

how this different than arduino pls

[Anonymous No.1555670]

>>1555663
How is what different from arduino?

[Anonymous No.1555705]

>>1555499
Anyone can help with H bridge control? I want to read a value from a encoder/pot and control a DC motor speed and direction (well two pots). But how would I go to control it with only 1 pwm pin and the smallest ammount of digital pins?

[Anonymous No.1555708]

>>1555705
Use a gate driver IC. Or if you don't need discrete FETs (low current) use an H bridge IC

[Anonymous No.1555714]

>>1555708
The H bridge and the drivers are ready, I'm asking about control schemes, not gate driving.

[Anonymous No.1555715]

>>1555714
What's your driver?

[Anonymous No.1555717]

>>1555715
Discrete driver, with enable pins on each half bridge.

[Anonymous No.1555719]

>>1555717
You're essentially asking how to interface with it. I have no idea how to do that from "it's a gate driver." You need to say what the model is so we can look up the datasheet.

[Anonymous No.1555720]

>>1555717
>>1555715
>>1555714
No wonder people mistake this for arduino general. Fucking retard.
http://modularcircuits.tantosonline.com/blog/articles/h-bridge-secrets/h-bridge-control/

[Anonymous No.1555730]

>>1555705
Left leg = PWM AND direction, right leg = PWM AND NOT direction

[Anonymous No.1555735]

I'm mounting an extra light on a tractor, is there any downside to using the frame as the ground and only running a single 24V wire to the light? Or should I run a proper 2-wire cable, with 24V and dedicated ground?

[Anonymous No.1555737]

>>1555730
>>1555720
thanks. Furter info is welcome, for me and other posters I think.

[Anonymous No.1555738]

>>1555735
>Ground=safety measure, you don't get shocked to death if you touch exposed metal on the lamp
>return wire = the one that goes on the negative of the battery (also called ground)

[Anonymous No.1555739]

>>1555738
It's a vehicle, it only has one kind of ground, the one that goes to the negative of the battery. The metal frame is directly connected to the negative of the battery and it's pretty thick all the way, thinnest point is probably a couple M8 screws.

[Anonymous No.1555746]

>>1555739
So it'll work m8.

[Anonymous No.1555748]

Explain this bullshit
>32 GB microSD card: $5
>32 kB microcontroller: Also $5
Why are microcontrollers not made with more internal storage when it's so cheap nowadays?
Also, the chip in pic related, ESP32, can be had for like $7 but the processor is 160 MHz, 10 times more than typical Arduinos. Why are these much slower and barely cheaper µC's still so popular?

[Anonymous No.1555757]

>>1555748
uC are made to be micro-controllers, they have to do a/d conversion, read and write, PWM, pulse some leds, DSP, I/O stuff. memory cards are made to be memory cards, they have a i/o interface, some logic and a whole lot of memory.
>esp32 and arduinos
because you use what you need. You only use very fast things when you need very fast things. Most stuff only need some nameless 0.03 cent chinese uC

[Anonymous No.1555761]

are the tiny fakeduinos supposed to be disposable?
it's useful for lots of things but it's essential for only one

[Anonymous No.1555764]

>>1555761
They are open source hardware, the cheapest o ne from ali work just fine.

What logic function/IC I should buy to
>pass a clock signal
>switch that signal between point A and B
>only switch during rising clock edges

[Anonymous No.1555772]

>>1555757
I'll try to make my first question more clear: Why not put, say, 1 MB of storage in basic microcontrollers, so programmers wouldn't have to worry about running out of program space? Based on the price of memory cards, it seems like it would cost next to nothing to do so.

[Anonymous No.1555773]

>>1555764
solved it

[Anonymous No.1555774]

>>1555772
Silicon die space.

[Anonymous No.1555783]

>>1555748
> Why are microcontrollers not made with more internal storage when it's so cheap nowadays?
Are you talking about RAM or flash? Because the latter is much cheaper than the former. The reason they don't put more flash on the uC is because that forces people to pay for it whether they want it or not.

> Also, the chip in pic related, ESP32, can be had for like $7 but the processor is 160 MHz, 10 times more than typical Arduinos. Why are these much slower and barely cheaper µC's still so popular?
The uCs in Arduinos are a lot cheaper than $7 if you're just buying the chip; particularly if you're buying them by the million. There are plenty of applications where you don't need performance, and a smaller, slower uC will draw less current (which is important if you want it to run for months on a couple of AAA cells).

[Anonymous No.1555784]

Is there a way to set up Eagle's autorouter so that traces on the top plane will have a fixed width and none of them will connect at the component pads? A customer wants to make a board single sided because he thinks it will be so much cheaper but I don't think it's possible to route everything by hand

[Anonymous No.1555786]

>>1555772
first, that 32GB of flash isn't guaranteed to be 100% good and isn't as finely addressable as MCU flash. it is not desirable to cut such corners for high speed program flash
second, plenty of other people can code in a restricted environment, why can't you?
third, they do make them, you can buy as big a flash as you like for your development work
fourth, knock yourself out https://hackaday.com/2013/12/29/hacking-sd-card-flash-memory-controllers/

>>1555783
you can tell the /pol/ escapees easily by their belief that their personal ues case is more legitimate than others' use cases

[Anonymous No.1555795]

why is it that this filter circuit fails to attenuate the signal at all?
I'll sweep the pot and nothing happens.

[Anonymous No.1555796]

>>1555795
How do you do these?
Are they actual simulations?
I really need something like that for autisms sake.

[Anonymous No.1555798]

>>1555796
this is an absurdly useful website/javascript applet called "Falstad"
http://www.falstad.com/circuit/circuitjs.html

it's 100% free, but I would be willing to pay a good chunk of change for a high-performance, offline, desktop version programmed in something other than js.

I know people like ltspice and shit, but compared to just this little applet, for learning electronics and experimenting, spice fucking sucks.

[Anonymous No.1555802]

>>1555798
Huh, I should send that to my old teacher.
Also it looks like the full source is available for the java version, have you tried that?

[Anonymous No.1555804]

>>1555802
Holy shit, no I have not.
I wish it were in C but I'm not going to complain.

[Anonymous No.1555809]

>>1555795
because you haven't hooked up the wiper of the pot

[Anonymous No.1555816]

>>1555809
thanks, I'm retarded.

Related question: why does this high-pass filter only affect the current (yellow) on the "out?" wire, and doesn't do anything to the voltage/frequency on the "out" wire?

[Anonymous No.1555821]

>>1555816
an RC filter works kinda like a frequency-dependent voltage divider. the proper output of that circuit is between the cap and the wiper of the pot
1. because there's no voltage on the "out?" wire, only current to ground. E=IR, if R=0 then E=0
2. the formula for the corner frequency of an RC filter is f=1/(2*pi*R*C) where f is in hertz, R is in ohms and C is in farads. the formula for the reactance (oversimplifying, the ac "resistance") is X=1/(2*pi*f*C). (corollary: at the corner frequency, X = R, so the output is exactly half the voltage of the input)
the resistance of the pot isn't being shown, but at 1500Hz the reactance of that cap is about 1.06 ohms, which is probably why you're not seeing a whole lot of effect

[Anonymous No.1555825]

how come this srr uses amperage to turn it on but voltage to turn it off?
I'm sure there's a good reason they designed it to be that way, I'm just not sure what it is

[Anonymous No.1555831]

>>1555825
>this ssr
What ssr?
It needs some minimum current (and voltage) to operate the LED for the optoisolator and related electronics. If it's a mosfet ssr there's gate charging and discharge circuitry.
Below some threshold voltage there will be a cutoff where the SSR shuts down quickly.
These limits provide hysterisis and a guarantee that the ssr can switch on/off quickly without wasting heat in a linear region.

[Anonymous No.1555833]

>>1555831
>These limits provide hysterisis[sic]
How so?

[Anonymous No.1555838]

>>1555833
Schmitt trigger input

[Anonymous No.1555841]

>>1555838
Actually come to think of it I don't know if that's a general property of SSR inputs.
Anyway I maintain the point that ssrs are designed to switch on / off quickly like mechanical relays, and not dwell in a region that wastes heat. At least the ones I've looked at.

[Anonymous No.1555843]

>>1555748
>Why are microcontrollers not made with more internal storage
How much do you want? Some ucs have a couple megs of flash.

[Anonymous No.1555845]

>>1555825
"how much voltage do I need to turn it on" and "how much current can I allow through before it doesn't turn off" are rarely pertinent concerns. they'll usually have plenty of voltage and the ability to prevent current flow in one direction above a certain voltage. they'll want to know whether the driver can safely supply the LED enough current to turn it definitely on and whether the driver will have a low enough voltage in the off-state to turn it definitely off without leakage

[Anonymous No.1555846]

>>1555841
>quickly like mechanical relays
that settles the case

[Anonymous No.1555849]

>>1555816
You're feeding in noise? Filters cannot remove signals that are already at the noise floor of your system. They only work on signals above the noise floor. There are ways to lower your noise floor with DSP assuming your electronics and external factors won't introduce noise above the lowered noise floor but that's all way beyond the scope of this post.

[Anonymous No.1555852]

>>1555849

[Anonymous No.1555853]

>>1555852
Okay, well just because you don't understand it doesn't mean I'm not right.

[Anonymous No.1555868]

>>1555773
Well, tomorrow I'll buy the one of those dual flip flop ICs and finish this, it seems to transition with only a small discontinuity. Not simulating parasitics, so no resistances and a lot of ringing.

[Anonymous No.1555870]

>>1555821
thanks for the detailed answer, I'm playing with the formulas now

>>1555853
for the record that anime poster isn't me.
I'm trying to learn signal filtering based on my knowledge and intuition from music production.

With my music I *can* filter noise, so I guess that's why I'm confused for it not working.

[Anonymous No.1555876]

>>1555870
sweet, it's working now for a triangle wave. The knee frequency formula helped a lot.

[Anonymous No.1555880]

>>1555870
>>1555876
You *need* to learn about the laplace transform and fourier series if your goal is that.

[Anonymous No.1555883]

>>1555880
I know about fourier series and the fourier transform from a very high level; I know how they work and what they do, but I'm still doing my pre-engineering courses and am only in Calc II, so I haven't actually had any formal training on them

the Laplace Transform I'll be honest, I don't know much about. I'll start reading about it now.

[Anonymous No.1555884]

>>1555876
cheers

>>1555868
the glitch is caused by the DFF flipping after part of the positive clock pulse has already been sent through the switch. if you put an inverter on the clock input, the switch will flip while the clock is low and be in position in plenty of time for the next rising edge

[Anonymous No.1555888]

>>1555748
There's actually a MCU inside a SD card.

That's actually an excellent question

[Anonymous No.1555928]

I have a motor and led that are in parallel. The motor draws 100ma at 10V from a 5V source and the led draws a max of 900ma from a 3V input. The motor is attached to a switch and whenever the led is on when I turn the switch on theres a brief drop in voltage I believe, the brightness drops by like 80% then returns to 90% brightness with the motor active. I don't really mind that the current or voltage drops when the motor is turned on but how do I prevent that quick drop? It looks kind of unprofessional. Do I just stick a capacitor parallel to the led that will slowly drain when I turn the motor on?

[Anonymous No.1555930]

>>1555798
>spice sucks
Get good

[Anonymous No.1555931]

>>1555928
you could stop using your shitty USB power source and get a real bench supply already

>>1555930
meh, LTspice's UI leaves much to be desired

[Anonymous No.1555932]

>>1555928
If the motor turn on can sag the voltage, a capacitor alone might not be enough because it'll suck down the charge from the capacitor too.
Add series diode to your (capacitor || led) if a capacitor alone doesn't work.

[Anonymous No.1555940]

>>1555928
use a motor soft start circuit. turn on time should maybe be 500ms. diode is optional.

[Anonymous No.1555946]

>>1555884
Thanks man!

[Anonymous No.1555947]

>>1555932
Adding a 10uf cap didn't do anything and I don't have enough space to add a diode so I'll drop that idea
>>1555940
Interesting, I'll try this. Thanks

[Anonymous No.1555967]

>>1555940
>diode is optional
I don't think so

[Anonymous No.1555971]

>>1555967
The one above 'R' just allows the cap to discharge quicker no?

[Anonymous No.1555978]

>>1555971
oh that one

[B.P.P. No.1555984]

>>1555614
Meh, the pussy and back scratches are nice but I’m about to spend $2k between V-Day and her birthday in a couple weeks and I would love to buy a scope and a bunch of little electronical shit with my tax return instead of a Kate Spade bag and a couple days at Disney.

>>1555652
Is that mA through the leads or the clamp? If it’s the clamp, how accurate is it at those low levels? Everybody seems to say that DC clamps like to bounce around a bit until you get them settled just right and preferably away from interference.

[Anonymous No.1555990]

Is there a inverter and level shifter circuit? Preferably with discrete components

[Anonymous No.1556002]

>>1555984
mA through the clamp, doesn't have any current measurement through the leads. With drift and compass effect about +/-2-5mA usable accuracy,

[Anonymous No.1556020]

>>1555990
Yes, several ways to do this
Quick and dirty way with discrete components: n-mosfet, and a resistor voltage divider at the drain

[Anonymous No.1556022]

>>1555714
>I'm asking about control schemes
You mean like Phase/Enable vs PWMA/PWMB?

[Anonymous No.1556175]

>>1555748
We really need to step up how much memory embedded things have. I would instantly buy a printer/wifi dongle/anything IF IT CAME WITH THE DAMN DRIVERS. There's literally no reason they can't have a standard interface, so if you insert a device, and it's in a not-ready state, that you can't switch it over to a mass storage device with drivers and/or compile instructions. There is literally no reason in 2019 that every device couldn't have a 128MB chip slapped on it, for even the most egregious offenders like HP.

[Anonymous No.1556180]

For not-autism-tier use, what's the point of high-end multimeters? I've had numerous cheap ones, and they all give out consistent measurements. Only the cheapest ones had a resistance of about .5-1 ohm on the leads for resistance checking, but I could just replace the connectors.

>>1555786
Memory controllers have already supported virtual blocks for years. It started on mechanical drives because to have 10^12 * 8 bits perfectly done (for only 1 terabyte) is next to impossible. Even SSD's actively do things with TRIM. Without even hardware support, certain filesystems can be written in non-contiguous sections., so even if you had huge chunks of drive screwed up, you could re-allocate software-side. A 100MB storage would have far less errors than a 1TB drive though.

>>1555798
An interesting note on Falstad, it's not that good. Would you be interested in a C/C++ SDL-compatible circuit tester? All it would take is some controls for delta-t, graphic, initial conditions, imaginary numbers for a basic implementation. I've done a little messing around with a rendering library (Irrlicht) that has file load/save features, menus, text boxes, etc because it's meant for games and 3D. Wouldn't be terribly hard to extend it, including virtual chips.

>>1555888
I know someone who did an arduino project where they slapped together a memory card, USB storage block and charger, and the arduino to read files off the chip and play them according to buttons. May not have been an arduino, but they make SD breakouts for mini boards.

[Anonymous No.1556182]

I noticed adding a cap from a mosfet gate to ground decreased my minimum led current which is great but how? My idle circuit current without the cap is 11ma and when I turn the led on with my pot switch it jumps to 120ma but with the cap my idle circuit current is 15ma and theres no current jump when I turn on the pot. Is it because the extra cap is lowering the duty cycle of the pwm to the gate?

[Anonymous No.1556189]

>>1556180
>For not-autism-tier use, what's the point of high-end multimeters?
Years ago, I got a Fluke not merely because of accurate measurements (although accuracy is important) but rather because it was so much nicer to use than all the chink garbage at the time.
Had a backlight, auto shutoff, lots of useful bells and whistles, good ergonomics, fast autoranging, and the nicest of all - the measurement reading on the screen updated way faster than the display on any chink meter I've used. It also had a bar on the lcd that responded very quickly to transient events so you could tell if something was happening faster than the measurement reading could update.

[Anonymous No.1556192]

>>1556182
The cap is sinking/source current, which just slows your transition times. If you're getting a dimming effect you've basically just made an RC filter that's preventing full turn-on.

It's important to realize the gate _is_ a capacitor, so any reduction in current or capacitance in parallel will decrease your transition times. We don't know your frequency, but with a uF cap and a frequency over a few KHz I could see an issue if your driver isn't outputting that much current (like 20mA for an op-amp, or even less for a digital out pin).

If your "driver" also is on-high but off-floating, your only discharge is leakage. A mosfets normal capacitance is really low, and so is the leakage. A normal capacitor might have a lower capacitance/leakage ratio, so you might be influencing the on/off times that way. Are you sure your driving thing can sink current? Some outputs only sink 1/10th their source capacity I think.

[Anonymous No.1556195]

>>1556189
Makes you think, with cheapish gizmo oscilloscopes like the DSO series, why nobody has slapped together a multimeter+oscilloscope. Oscilloscopes aren't meant for autoranging, but I have a decent one, (DSO112, not the A) that has a touchscreen and only a few non-critical flaws: it has a live USB mode but isn't supported by anything, and the refresh rate and touch response drop at certain timing intervals for some reason. The 112A might be supported for live mode.

I don't consider auto-off a feature because I hate the bastard turning off if I'm doing measurements longer than ten minutes, and if I wanted to measure a varying circuit I'd use an oscilloscope, not a meter.

Some car people really like the DSO112s because they have built-in 50V tolerance, are light, rechargeable, and only need the touch interface instead of 30 switches like other models.

[Anonymous No.1556200]

>>1556175
Kinda a recipe for chink printers to include malware

[Anonymous No.1556209]

>>1556195
>why nobody has slapped together a multimeter+oscilloscope
Fluke did that too

[Anonymous No.1556221]

>>1556200
...literally drivers these days from HP, Dell, etc include (((analytics))) software. I'd much rather have a one-time built-in driver, even if it's a blob, than an always-on driver that self-updates for the latest and greatest spyware. Either way, for non-free software, you have to install a blob. It would make it far easier for consumers if you could just put the drivers on the damn thing. Reputable companies could even offer a guarantee that it works out-of-box forever, instead of having to sift through a site which may need a 200 megabyte download that breaks later. HP has been known to intentionally provide firmware updates that render their printers inoperable. I have one sitting in my room right now that they fucked over the ink and print speed settings in firmware, so you have to do the slowest print if you want black to be black, otherwise it comes out gray or literally nothing is printed on the draft setting. I know it's the firmware because open-source drivers have the same issue, which only happened a few years after the printers came out (I'd never fucking update printer firmware, it shouldn't).

>>1556209
They probably charged $2000 for it

[Anonymous No.1556246]

i have the pinout for this chip but it doesn't tell you which way round it's supposed to be :/

[Anonymous No.1556253]

>>1556022
Just how to wire the switches so I use less pins, only one pwm output and don't burn anything. I kinda solved it here, but if you have any sugestions I'm happy to accept>>1555868

[Anonymous No.1556254]

>>1556246
LIttle ball on the left tells you where pin 1 is

[Anonymous No.1556256]

>>1556175
After this post, I'm convinced we deserve skynet.

[Anonymous No.1556258]

>>1556254
you're a peach

[Anonymous No.1556276]

>>1556258

[Anonymous No.1556305]

>>1556180
tell me how many wait states and random pauses you're willing to allow for that level of indirection and be outvoted by the billions of chips bought by people who aren't willing to pay for or tolerate any
also read https://en.wikipedia.org/wiki/Memory_hierarchy and either or both of the computer engineering texts by Hennessy & Patterson
bottom line, it's a silly idea and you can do it yourself in code if you really care that much

[Anonymous No.1556309]

>>1556305
You talking about memory allocation? It's literally built-in to every single hard drive. Platter drives have at least 1-3% reallocatable sectors, and use them for manufacturing defects. SSD's have their entire block system virtualizable with TRIM, because one of the best benefits of TRIM is wear levelling. If that's what you're complaining about, you are so horribly misinformed. On the software level, files are written in block sizes, and the OS handles the fragmentation. GNU+Linux tries to avoid it with automatic defragmenting, and Windows started scheduling defragmentation cycles since what, Vista or 7?

[Anonymous No.1556326]

>>1556309
no, I'm referring to that all those levels of indirection consume time and silicon, that chip test time costs about 5 cents per second, that there are least one or two OSes intermediating all that happy horseshit, and that you're too fucking oblivious to recognize that MCU users want determinism and predictability
stop being /pol/ and assuming that anyone who doesn't agree with your use case is misinformed. go buy a pi and leave us alone

[Hopefully the right Thread to (...) No.1556327]

>>1555499
Well the people on /adv/ are not very helpful, which is why I'm asking you guys here.
I'm keeping it short. 2 Big companies offered me an apprenticeship, one as an Industrial Electrician (Energy supplying company) and the other one as a Software Developer.
>How is the job market for Industrial Electricians?
>What is more important a University Degree (E.E) or work experience?
>What about a career, still worthwhile, or will I end up as a standart worker on a power plant?
>Will Robots take my Job?

Pic related is a quick comparision i made - both companies are located in Germany. Thank you very much in advance Anons

[Anonymous No.1556342]

>>1556326
>testing the chip
You don't need to test memory, kind of the point. They don't test cpus, gpus, memory cards, mice... I have a hard time thinking of anything that is tested. 2KiB program autists can have their 5 cent chip, I'd gladly pay 50c if it meant we could have 32 megs.

[Anonymous No.1556348]

>>1556327
Industrial Electrician would be more reliable, and there's probably a union, but you might get bored. It's going to be a lot of physical work, going to places. Most of the design should be handled by actual Engineers. There is a path from worker to Engineer, but it'd be hard if you started a life (kids) before you got an Engineering license or didn't have a supportive spouse.

The programming job is high-risk. Every year they outsource jobs or insource people, but if you can get a good social group going and/or government clearance, you're set. Programming is 99% reputation, don't believe anyone who tells you it's knowledge or skills. Not physically stressful, but you'll be overworked and produce shitty code under bad management, and everyone will know it but nobody will saying anything.

[Anonymous No.1556376]

>>1556348
Obviously you don't want to get replaced at some point, which is my biggest fear in the IT Market. Simply by being not good enough, or doing something that someone else would be better at. It feels like the Skill gap is far greater in IT than in a blue collar field, mostly because the competitions is higher.

Apparently my title after the IT apprenticeship would be "Dynamic AX Developer", which sounds fancy, but also like mainstream office BS
Thank you for your insight Anon

[Anonymous No.1556386]

>>1556327
Electrician anytime. You can learn to code anytime but the experience you'll get in a industry is amazing. People make this big noise about "technology" and IT jobs, but it's not all that buzz to be honest

[Anonymous No.1556389]

>>1556342
Are you retarded? Can you think of any aplication you need 32 mb of memory in a uC? What makes you think you are right when the chinese hivemind or people who design, sell and build ICs for a living don't seem to care about that? If you reached the point where you need 32mb of memory in a microcontroller it means you are doing something wrong, you probably should use a custom job, fpga or just get a fucking computer.

[Anonymous No.1556391]

>>1556386
And getting good techs and engineers is getting more valuable today, where everyone seems to think that covering bad designs and installations with a android app is a good idea. And I'm talking from the freaking 3rd world, no idea of how is it where you live.

[Anonymous No.1556394]

>>1556389
And also, 32mb of cpu takes a lot of space in a die. Space that could be used to do things a uC should do.

[Anonymous No.1556396]

>>1556394
Fuck, forgot my pic. The big blob is memory, the tiny square is the interface and management.

[Anonymous No.1556401]

>>1556396
Atmega328P, most of it is also some kind of memory or another.

[Anonymous No.1556406]

>>1556401
Going from 32kb to 32mb is a huge increase in area as you can see. Apparently only you had the need for a 32mb program memory.

[Anonymous No.1556416]

>>1556389
Records. Configurations. Large programs with many paths, even if the cpu will only be processing a subset. Basically you wouldn't need to use hyper-turbo autist limited optimized libraries and worry about blowing your memory if you go outside it.

>>1556394
>>1556396
The memory would be storage, not CPU memory. CPU could still have like a 1KB cache or whatever piss you want a tiny chip. Just basically a tiny-ass SoC which only has USB support at best. Making the chip a little bigger would make it easier to integrate a clock, DAC/ADCs, outputs which can source and sink current (even a 555 can do 200mA). Even if you turned a 5c chip into a $1 chip, it'd save so much time and effort to not have to add in a bunch of bullshit that everyone adds anyway. Time is money, and if you spent only an extra hour for 100 units, you just increased the cost of each unit by 10c at minimum wage 0 overhead cost.

[Anonymous No.1556423]

>>1556309
>You talking about memory allocation?
>>1556342
>You don't need to test memory
>>1556416
>Records. Configurations. Large programs with many paths
>CPU could still have like a 1KB cache or whatever piss you want a tiny chip

It is clear by now that you're completely ignorant of how any of this works. Your mental model of how all this works is just retarded and wrong.

[Anonymous No.1556429]

>>1556342
>They don't test cpus, gpus, memory cards, mice
>they don't test cpus

[Anonymous No.1556473]

>>1555499

Okay, can anyone tell me where I fucked up?
I was trying to recreate a board from an old magazine and as far as I can tell, I did the schematic and board correctly (the pots and power transistor are to be wired out of the board, hence the pads.)
But Eagle still complains about there being an airwire at N$8 and wants me to route it, except that's not how it works on the original PCB.
Please don't mind the fat routes, I'm a pleb sharpie mask nigger.

[Anonymous No.1556519]

>>1556473
you mean, other than by using EAGLE?
N$8, between LED1 and R3, looks pretty well routed to me, you could try highlighting N$8 on schematic and board and seeing what it thinks is there, also ripping it up and routing it again to see if it then recognizes it

[Anonymous No.1556523]

>>1556519
Meant N$18, sorry.
Eagle insists both sides of R15 should be on the same route, which doesn't seem right to me.

[Anonymous No.1556540]

>>1556523
user error

[Anonymous No.1556543]

>>1556540
Oh god thank you so much, I knew I made a retarded mistake somewhere but I'm too braindead right now to spot it.

[Anonymous No.1556583]

>>1556416
You must be some brainlet that coudn't learn assembly and now wants your favorite uC to run your 2 terabyte pyton code.
>records
Literally everything that needs some sort of static record or config that doesn't fit the main memory goes into a eeprom, and guess what? It works just fucking fine. No plane is going to crash because you take calibration data from I2C instead of chip memory,
>5c chip into a 1 dollar one
Literally a 20x increase in price. Are you retarded? Open up digikey/mouser and see to what decimal case they put their prices in.

[Anonymous No.1556585]

>>1556583
In case you want to overcome your mental retardation, you could try using a 2mb eeprom
>https://www.st.com/en/memories/m24m02-dr.html

[Anonymous No.1556593]

>>1556585
Oh, it's 32mb, no problem
>https://www.microchip.com/paramchartsearch/chart.aspx?branchID=7130106

[Anonymous No.1556601]

meanwhile, those web tools for power systems design are kinda neat. with stock USB charger transformers going for about 25 cents each, and a few firms on ali selling custom wound flyback transformers in 10-piece sample lots starting at $6.50 each, it's almost too easy to roll your own offline power supply

>>1556593
no, it's 32 megabytes
https://www.aliexpress.com/wholesale?catId=0&SearchText=256mbit+flash
whoops, looks like it's $2-3 each in 10-piece lots. surely it'll be cheaper and faster if we try to cram it onto a microcontroller die which likely requires very different process parameters

[Anonymous No.1556614]

>>1556601
or, you can buy your own core and bobbin and follow the detailed instructions
>inb4 that's not a web tool
yeah I downloaded the non-web version because power integrations is especially niggered and locked me out of my account for some reason. that's fine, their chips are cheap on the gray market on alibay

[Anonymous No.1556629]

I have to fix an old motherboard which has exploded capacitors. What is the correct way to desolder and solder onto it without fucking it up?
I don't have much experience soldering onto PCBs and every time I tried I fucked up by destroying the round contacts around the holes.
Also since I'm going to buy capacitors I wanted to know what else is a good idea to have around. I might want to get a bit more into making circuits and stuff.

[Anonymous No.1556633]

>>1556629
step 1: hot air to heat the area and the internal layers all at once
step 2: a solder sucker to clean out the holes after you've removed the part

[Anonymous No.1556647]

>circuit porn
fully organic, no mosfet

[Anonymous No.1556739]

I'm not sure if this is the right place to ask and it's a bit autistic but here goes.
I haven't been satisfied with my various cellphone chargers over the years, I've used: 2008 Nokia 5V 0.5A, 2011 Huawei 5V 0.6A, 2014 Moto 5V/12V QC, 2016 ZTE QC, 2017 Apple 5V 1A. My issues with them have been mostly coil whine and some voltage instability (devices flicker and stutter).
So I'm wondering if I could build my own reliable/clean/efficient 5V 2A basic power supply. It seems like I could start from scratch with a custom SMPS or transformer? on a PCB (fun and customizable), or some kind of integrated module (quick and reliable) and some connectors and packaging.
Finally I am new to electronics and don't want to electrocute myself or burn the house down. Thoughts?

[Anonymous No.1556749]

>>1556739
you can make a better, more stable, quieter, and simpler supply than all of those chargers without too much trouble. the reason is that those are optimized for cost and size above all else, so they use a switchmode topology. you don't care about that so you can just make a linear power supply as pictured. smps can make noise in the 10kHz+ range (generally when they're poorly designed, i'm surprised the apple does) but mains frequency transformers for linear regulators can still make a 120Hz hum. it's a much more pleasant noise but you can eliminate it too by selecting the right transformer. i don't remember what load cases cause the hum though.

you'll have to google how to wire the data pins on the usb port here as phones use them to detect charger current capability. some phones use cancerous proprietary detection schemes that i don't know about.

[Anonymous No.1556756]

>>1556739
Buy a name-brand 2A/10W charger you shitlord, at 0.6A or 1A it's no wonder you're getting those issues. I've been using the same Apple 2A charger daily for nearly 9 years or so and haven't had a single issue with it, no whine, no overheating, no device flickering.

[Anonymous No.1556757]

Not sure if this is the right general, but I am having a pain in the ass issue with an inverter overheating. On sight we have 5 of the same model of machine, with identical inverters. One of the 5 always over heats, to the point I just jammed an air wand in the housing in order to keep it running. The enviroment is very filthy, so I made some filters for all the intakes, which has drastically increased up time for all the devices, save the busted one.

Any idea's how I can keep the inverter running cooler for less money than running an $5/day air leak over it?

[Anonymous No.1556760]

>>1555499
>"are you AM?"
Because you are noisy and your carrier zero crossing is equidistant

[Anonymous No.1556764]

>>1556757
First of all, take it apart and use a toothbrush and some IPA to give it a scrub down. The heat sinks might be covered in dust or dirt and the fan might not be running due to shit in its bearings. Also check for any other components that look damaged in some way, but since it's still working fine apart from the overheating I suspect it's a dirt issue. Secondly, does your load even need an inverter instead of a HVDC source from a boost converter? If it runs with a SMPS or brushed universal motor, the answer is no.

[Anonymous No.1556767]

>>1556764
Will other beer work? I'm not a fan of IPA.

[Anonymous No.1556779]

>>1556767
It has to be something that evaporates easily at room temperature and without leaving residue behind, so no.

[Anonymous No.1556781]

>>1556767
>>1556779
Oh but a damp cloth should be fine if you're careful not to turn it on for a while after.

[Anonymous No.1556794]

do resistors cause a net increase in voltage of a section of a circuit due to cutting current? Or does it "go away" because of heat dissipation

[Anonymous No.1556799]

>>1556767
you gotta use either IPA or vodka, anything else just gums it up worse

>>1556749
>Finally I am new to electronics
>Thoughts?
don't work with mains power (yet)

>>1556794
E = I*R
so no, if anything, well-behaved loads will do the opposite

[Anonymous No.1556803]

>>1556767
also UNDER NO CIRCUMSTANCES do you use stouts (gums it up worst of all) or tequila (causes devices to lose time and wake up later in a bathtub full of ice with a capacitor missing)

[Anonymous No.1556808]

Aside from a trigger, what circuits I can use to take a slow transitioning signal and turn it into a sharper one? Something that takes less space than an IC.

[Anonymous No.1556809]

>>1556808
Logic level mosfet and some passives? Can't be picky about where the transition happens though.

[Anonymous No.1556833]

>>1556192
Thanks. What do you mean by transition time, would that be the duty cycle? I am getting a beautiful dimming effect, something I was striving for and randomly experimenting, achieved. Without the cap its a linear dim which means once I turn it on and the 100+ma floods into the led the change in brightness is noticeable only at the beginning of the turn.

I'm working with 23.5khz and the strange thing is, it only works with a 1kuf cap. I tried using a 100uf, 470uf and 3.3kuf cap and all of them seem to only turn the led on once it reaches ~20ma. I don't understand exactly why the 1kuf is special in this case where the led turns on immediately when I turn my pot switch on. I suppose theres some mathematical reason for this but I don't know circuitry enough to figure out why only that cap value works.

My driver is a basic 555 in astable mode so I'm pretty sure its sinking the current

[Anonymous No.1556839]

>>1556833
>transition time
I think anon is talking about edge time, the time it takes the voltage to reach 90% if rising or 10% if falling.

[Anonymous No.1556855]

>>1556808
74LVC1G17

[Anonymous No.1556871]

>>1556808
Is surface mount not an option for you? Those can get pretty small.

[Anonymous No.1556887]

So my brother gave me an old kit of electronics he had from university, it's got a whole lot of different resistors, capacitors, a breadboard, LEDs, and a bunch of different chips:
1x NE555P
2x CD4009UBE
2x CD4013BE
1x MC14001UBCP
1x CD4011UBE
1x CD4030BE
2x MC14081BCP
1x CD4511BE

Are there any projects a total noob might be able to do to make use of these components?

[Anonymous No.1556893]

Possibly a dumb question, but how bad exactly are chink solder wires, really? If I get something marked 63-37 leaded, 2% flux, 0.5mm diameter, what could really go wrong with it?
Asking because despite being warned on various boards to not cheap out on my solder wire, I'm in bongland and amazon only has chinkshit (except suitably marked up, as usual). Or an immeasurably vast choice of various lead free options, fuck that. And I don't know of any physical shops nearby that are likely to stock it - though if you guys tell me to still keep away from chinks, finding one is going to be my next option.

[Anonymous No.1556894]

>>1556887
there's not a ton you can do with random logic, but with those chips you could make a variety of single digit counters. i think you have enough to make a timer that counts down from X minutes and then blinks the segment rapidly, with a button to reset. try that. the 4511 drives the leds, the 4013s can be arranged as a down counter, and the 555 can provide a pulse every minute. the other chips help make the down counter, blink, or glue things together.

[Anonymous No.1556903]

>>1556871
>>1556855
>0.7mm x 0.7mm SMD IC package
>wipes_sweat.jpg

[Anonymous No.1556912]

>>1556893
The metal is fine, the flux may be shitty.

[Anonymous No.1556932]

>>1556401
I could never interpret these die photos.

[Anonymous No.1556956]

>>1556912
Nope, cheap chink solders are made of whatever reclaimed crap they could find.
>>1556893
I recommend Kaina with blue label (the red one is said to be worse).
See this review. It's in Russian, but you still can watch the gifs that compare different solders to this one.
https://mysku.ru/blog/aliexpress/56840.html

[Anonymous No.1557005]

>smt components
>pen with conductive ink
Is this how people prototype/breadboard with smt?
I can't imagine breakout boarding every individual smt component for a prototype

[Anonymous No.1557011]

>>1557005
dead bug. that pen is a useless gimmick

[Anonymous No.1557020]

Do you remember a guy askinga bout optocoupler speed and load resistor that was too dumb to read the datasheet? Well that was me.

Anyways, even with proper load the pc87 rise/fall time can't go lower than a couple microseconds. BUT
I've built and tested this circuit, and managed to get it down to 100ns (in a breadboard, so when soldered it'll be less). What it does is diminish the changes in the coupler Vce and that helps with the miller capacitance. I'm posting this because 4 pin optocouplers are quite cheap and some of you might be put off by how slow they might be.

[Anonymous No.1557022]

>>1557020
Top graph is the pnp output, lower one is the Vce.

[Anonymous No.1557023]

>>1557022
PC817* jesus fuck

[Anonymous No.1557035]

>>1557005
personally, I use gouged board techniques and don't go any smaller than 0603 or SOT23
>can't imagine
it happens, Pic related

>>1557020
intredasting. you might want to try reading that datasheet again, particularly the absolute maximum ratings section, particularly the collector current rating of 50mA

[Anonymous No.1557039]

>>1557035
oh, I've never seen those kinds of breakout boards before. That's neat, a lot more practical than I thought

[Anonymous No.1557040]

>>1557020
If you care about speed, photodiodes are faster than phototransistors.

[Anonymous No.1557053]

this is probably a stupid question but that's what this place is for
I have leds wired in series to my optocouplers, will this impede their efficacy?
the optos efficacy I mean, would it be smarter to have the leds parallel? or would it make no difference

[Anonymous No.1557063]

>>1557053
Series is fine if you've got sufficient voltage available and want a similar current in each

[Anonymous No.1557066]

>>1557005
larger SMD stuff can be used with some creative placement and cutting on veroboard. For more complex stuff, getting a PCB made has never been cheaper.

[Anonymous No.1557067]

>>1557066

[Anonymous No.1557080]

>>1557035
I did kek, the 100 ohm resistor is for the simulations, in the real circuit I think it was 330

[Anonymous No.1557086]

>>1556912
>>1556956
Alright, thanks.
Related question, does 60/40 have any advantages (mechanical strength, durability, I dunno) over 63/37, or is it just historical and "good enough" therefore slightly more common despite being strictly inferior?
Also any opinions on meme alloys like 62/36/2 with silver? Not that I'm gonna be able to get my hands on some, but I'm curious if anyone has any experience with those they can comment on.

[Anonymous No.1557165]

>>1557086
I'm thinking of putting pressure sensors around one of the columns of my bike and doing some FEM of the column to use them as an anemometer. Any guesses if it'll work?

[Anonymous No.1557167]

I'm thinking of putting pressure sensors around one of the columns of my bike and doing some FEM of the column to use them as an anemometer. Any guesses if it'll work?

[Anonymous No.1557256]

spent the day breadboarding this ungodly mess. it's a 16 bit binary to decimal converter. Switch on the LEDs corresponding to the bit, and it converts it to decimal. completely useless, but i think it's cool. Once I'm done debugging I'm gonna start designing the PCB

[Anonymous No.1557387]

>>1555931
>meh, LTspice's UI leaves much to be desired
it's not for normies and it's all the better for it
this is a tool for engineers. we don't want it looking pretty, we just want to get work done fast

[Anonymous No.1557433]

In a transistor voltage follower, how important is the load resistor?Why shouldn't I just use a huge value to aavoid heat?

[Anonymous No.1557454]

>>1555499
Need some help bros...
I'd like to measure vibrations of a small motor to keep track of it and also to compare vibrations of various small motors.

Does anyone know what the best vibration/acceleration sensor is for Arduino or RaspberryPI that could measure this reliably?

Thanks!

[Anonymous No.1557456]

>>1557454
MEMs sensors are the cheapest, piezoelectric are the most accurate.

T. ISO Class III Vibration Analyst

[Anonymous No.1557457]

>>1557256
nice clock, achmed

[Anonymous No.1557476]

>>1557456
>MEMs sensors are the cheapest, piezoelectric are the most accurate.
>T. ISO Class III Vibration Analyst
holy crap! that's awesome anon. do you have any specific recommendations on which chip/board I should look for? I'd like to get 2-3 sensor boards and give them a test and see which of them would work out the best.

also, have any recommendations on how to analyze vibrations? like what kinds of tools/libraries I could use to compare two motors?

Thanks so much!

[Anonymous No.1557481]

>>1557476
not the other poster but look up fast fourier transform. It’s for looking at what frequencies comprise a signal.

[Anonymous No.1557488]

>>1557481
Thanks anon! Yes, FFTs... that's what I managed to google so far. I still remember FFTs from school but didn't know you could use them for this. I'm still not clear as to how analyze two signals qualitatively and say "X is better than Y". Will do more reading... thanks again. Appreciate it.

[Anonymous No.1557501]

>>1557488
>I'm still not clear as to how analyze two signals qualitatively and say "X is better than Y".
Make spectrograms and this will become more clear. You'd record a spectrogram of a new motor with new bearings, running unloaded and also at your expected loads.
When "something's wrong" like a bad bearing, or the load becomes unbalanced, or whatever, the spectrogram will appear unlike the original good spectrograms, in that frequencies will show up that didn't show up before.
Processes and machinery health is monitored this way, I think.

[Anonymous No.1557507]

>>1557433
that load resistor sets your output impedance. it's good to set it high for that reason, and also to keep the input impedance high, but there are other considerations which might call for a lower value (cable impedance, output current, etc.)

>>1557387
they could certainly make it faster to set up a voltage source

[Anonymous No.1557511]

>>1557507
F2, vol, enter, left click, right click, number, enter

Show me a program where that's significantly easier.

[Anonymous No.1557521]

Can you use the Fourier transform instead of the Z or Laplace transform to find the transfer function?

[Anonymous No.1557531]

>>1557521
Sure, you can use any linear transform if it makes the math easier.

[Anonymous No.1557545]

>>1557488
I personally use Python and the fft modules, it makes it pretty easy

[Anonymous No.1557685]

I have a couple of transformers similar to pic related that I got off an old power supply, and can't find any data sheets online. How would you suggest i determine how they're wired up?
I was gonna measure the reistances between each terminal and assume they correlate to the relative amounts of wingdings, don't know if that's smart or not.

[Anonymous No.1557695]

>>1557685
Resistance won't correlate to number of turns well since the higher turn count coils will mostly be thinner wire too.

Measuring inductance will give a better approximation.

[Anonymous No.1557701]

>>1557695
or just measure the voltage ratio

measuring the resistance will at least give you an idea of the arrangement of the taps

[Anonymous No.1557712]

>tfw you design circuits by trial and error and messing around with resistor values on falstad instead of doing calculations

[Anonymous No.1557714]

>>1557712
sometimes it's easier to do that than to do the math. and whatever's easier is better if it gets the same results. fuck captcha holy shit.

[Anonymous No.1557726]

>>1557685
I did something like that yesterday. The resistance tells you what is connected to which pin and lets you draw the diagram. If you can measure L (or compare voltages) you have the turn ratios. I haven't found a way to get the real number of turns other than winding an additional (known) number on the coil to estimate the AL of the core and go from there. Not all transformers have a spare winding space for that. With open coils this works well but it is still an estimate because the additional (outer) winding is not so well coupled magnetically.

[Anonymous No.1557739]

Usually even when you make calculations you end up messing arround with a prototype.

[Anonymous No.1557741]

hate to admit this but I'm having trouble translating my breadboard to a circuit board
are there any tools to help me make it good?

[Anonymous No.1557744]

in an RF antenna, if the R is a capacitor, what is the F?

[Anonymous No.1557746]

>>1557741
The tool is called circuit diagram.

[Anonymous No.1557808]

Car stereos and other electronics often use a knob you can turn to adjust volume o scroll through a menu
What is that component called?

[Anonymous No.1557809]

>>1557808
Trimpot

[Anonymous No.1557811]

>>1557808
Rotary encoder

[Anonymous No.1557812]

>>1557811
Oh shoot, I had stopped reading at volume

[Anonymous No.1557855]

>>1557812
and you were still right. software-intermediated volume is far from uncommon these days in consumer electronics

>>1557685
typically you're not gonna find any datasheets because they tend to be custom-wound for a particular customer application
>measure the reistances between each terminal
they often use different gauges of wire. that's not going to work very well
the best way is probably to pump a 1V sine wave into one winding and measure the voltage at the other(s). also, for these sorts of transformers, which terminals are start resp. finish is also important
but, you won't be able to make very good use of these without a short course in ac-dc converters. if you try to hook them directly to mains voltage without appropriate driver circuitry, they'll burn out

[Anonymous No.1557858]

>>1557712
I always do that. Calculations never work in reality anyway.

[Anonymous No.1557938]

can't figure out how to reconcile two intersecting ground rails on this circuit board
I was always awful at those pipe games

[Anonymous No.1557960]

is there some easy way to detect DC overcurrent fast enough to not destroy my source?

I was thinking about using an op amp somehow to trigger a relay if the current is higher than a value i set with a potentiometer but i am too brainlet to actually design the actual circuit ;_;
And ideal thing would be to limit the actual current but that is basically fairy magic to me at this point.

[Anonymous No.1557973]

>>1557695
>>1557726
>>1557855
thanks for your help boys

[Anonymous No.1557975]

>>1557960
I may have no idea what the fuck I'm talking about, but how about using a transistor to limit the current?

[Anonymous No.1557978]

>>1557960
Use a fuse?

[Anonymous No.1557980]

>>1557855
>software-intermediated volume is far from uncommon these days in consumer electronics
Yea but I'm >>1557809
Good luck configuring that one to work.

[Anonymous No.1557981]

>>1557960
What's the source? You can't protect a transistor from over-current with a relay or fuse; they're too slow.

[Anonymous No.1557985]

>>1557981
the source is an old computer PSU
i don't want to use fuses because they are too slow and aren't easily resetable

[Anonymous No.1557986]

I want to wire up buttons 0 through 9 on to this rotary selector.

When i click it to 1 it should press the 1 key and then stop pressing it, not keep it held down.

I can only find this switch >pic related

But that doesnt do what i need it to. What google search term should i use to find the kind of rotary switch im looking for?

[Anonymous No.1558009]

>>1557960
>>1557985
something like this?

[Anonymous No.1558014]

>>1557986
they don't exist as stock items
you'll need to use extra circuitry, whose exact nature depends on the bigger project

[Anonymous No.1558023]

>>1557960
>>1557985
art of electronics, see the subchapter on circuit protection and learn the various magics. What you want is in there.
One problem with ultra-fast overcurrent protection is that transient (brief, unsustained) overcurrent may be allowable and you wouldn't want your circuit to cut power for cases like that.

[Anonymous No.1558024]

>>1557960
>>1558009
fairy magic explained: Rshunt and Rbias are a voltage divider that set Qpower's Vgs to some nice saturated fully-on value. Rset develops a voltage proportional to load current. when that voltage reaches about 0.65V, Qshunt turns on and starts stealing current from Rshunt, reducing Vgs and pulling Qpower out of saturation
but if you're using an old computer PSU, those generally have current limiting built in already, and without affecting the voltage to the load. if anything, you would use this circuit to protect your load from the PSU
also I fucked up the schematic because it was upside down in my brain, the correct Rbias is
Vgs * Rshunt / (Rshunt + Rbias)

[Anonymous No.1558026]

>>1557986
if such a thing exists it would be called a momentary rotary switch.

>>1558014
probably some monostable circuit

[Anonymous No.1558028]

I was halfway this veroboard when I realized there would not be enough space. I have no regrets.

[Anonymous No.1558032]

>>1558024
that seems simple enough, i will try to put together something like that and if it doesn't work i will just roll the dice and buy a pack of fuses for the psu

[Anonymous No.1558047]

I have a job interview coming for a company that needs a HVPS system. I don't know how high the voltage is or how much power it'll need to source. I'm thinking I should brush up on HV transformers, CW bridges, and switching regulators. Are there any other important topologies I should consider?

[Anonymous No.1558056]

>>1558047
Thyristors and Triac shit too. The HVDC conversion near here, in itaipu uses xboxhueg thyristors

[Anonymous No.1558059]

>>1558028
Nice zoo.
I love the giraffe

[Anonymous No.1558062]

>>1557744
>

[Anonymous No.1558075]

>>1558028

[Anonymous No.1558079]

Why are microcontroller datasheets so intimidating

[Anonymous No.1558080]

>>1558079
because they're packed full of peripherals that you don't need. 90% of the datasheet is just scrolling through peripheral registers seeing what you need to write to make them go away. the most important parts are the memory map, pinouts, and maybe the instruction set.

[Anonymous No.1558081]

>>1558079
>What do you need the controller for
>thing
>go datasheet
>1.how to turn it on
>2.how i config
>3.how i thing

[Anonymous No.1558082]

>>1558059
>>1558075
thanks people

[Anonymous No.1558090]

>>1558079
Keep reading and learning and all those mystery functions will make sense and you'll see how truly useful microcontrollers are with all their jam-packed peripherals.

Just for laughs, check out the latest generation xeon datasheets and i86-64 architecture manuals, all those combined are over 5000 pages. It makes a microcontroller datasheet look like a light read.

[Anonymous No.1558140]

>>1558079
It's like climbing a mountain, look tough from the bottom, but if you take it step by step you just die of lack of oxygen because you didn't bring any gear and you're a dumb fuck and this is a stupid hobby

[Anonymous No.1558144]

This is now a cross-chan YLYL.

https://55chan.org/b/res/26415313.html

[Anonymous No.1558145]

>>1558144

[Anonymous No.1558153]

>>1558145
oy vey

[Anonymous No.1558155]

>>1558153
kek

[Anonymous No.1558163]

>>1558144
What is this supposed to mean?

[Anonymous No.1558168]

>>1558163
Solder paste (not flux).

[Anonymous No.1558174]

>>1558144
>paid 200 reals for this, did I get ripped off?

[Anonymous No.1558187]

>>1558163
Rosin flux, if you for flux in brasil you'll get that expensive surface mount syringe thing

[Anonymous No.1558195]

Do the standard transformer "n1/n2 = V1/V2" rules apply if your coils don't fit the "long thin solenoid" approximation? I've got 100m of wire and wrapping this around the cylinder it would give me 400 turns, and I need V1/V2 = 400, so my primary would just be a single turn.

[Anonymous No.1558200]

>>1558195
What sort of core are you using?

[Anonymous No.1558220]

>>1558200
Air. I'm relying on the stacking wire insulation of the transformer's wire in order to get sufficient voltage, If I make overlapping windings or use a ferromagnetic core I'll get breakdown. Similar to a tesla coil's transformer.

[Anonymous No.1558222]

>>1558220
ye, for air core the n1/n2 ratio won't really apply since leakage inductance will be significant relative to mutual inductance.

[Anonymous No.1558224]

where can I learn how to make my own power supplies that are 90%+ efficient?

my goal is to make a custom power supply for my vertical farming rig that has 3.3v 10a, 5v 10a, 12v 40a, and 24v 40a outputs off a 110/220 input.

Ive been rigging server power supplies with 12v 75A rails to get the job done but I need something a little more efficient since I intend to scale this. The 3.3/5v are for raspberry pis and microcontrollers. The 12 is for pumps, motors, and solenoids mostly, and the 24 is for a 660W LED array (constant current desired).

[Anonymous No.1558228]

Best way to strip copper off PCBs after you've taken all components off? I want to re use them, and crate my own connections with wired and solder and whatnot.
Scraping it with a knife takes too much time.

[Anonymous No.1558234]

>>1558224
The hell are you using all those voltages for? Sounds like there's a better option somewhere.

[Anonymous No.1558236]

>>1558187
Ask for ** jesus fug

>>1558195
You need a magnetic return path. (The iron/ferrite thing has to have/be a hole)

[Anonymous No.1558237]

>>1558224
pic related is a decent book.
also check out LTpowerCAD, from analog/lt, might be worth a look. and their design guides and app notes. they do a lot of power supply stuff.

[Anonymous No.1558238]

>>1558236
tesla coil transformers don't do that and I can't afford to do that

[Anonymous No.1558239]

>>1558236
Magnetic flux**
>>1558228
That is a very very very bad idea, even here on the third world plated copper plates and vero board is somewhat cheap. But you could use Ferric chloride, or throw them in a pot with vinegar. Assuming they have no masks.

I usually reuse them to make casings and plataforms for other projects.

[Anonymous No.1558242]

>>1558238
They transform very little power and have high frequency. If you want any sort of usable current you need a flux return path.

Think of this as putting a 10M ohm resistor on a 5v source vs putting a 100 Ohm resistor on the same source. The 10M = air, the 100 Ohm = ferrite/iron.

What you are doing is to couple two inductors, and that is not a transformer per se.

[Anonymous No.1558244]

>>1558242
I would argue that any set of coupled inductors is an autotransformer, but that's semantics. It's impossible for me to use a conductive medium as my core as I'll be dealing with hundreds of kV. It's a budget build so having some expensive high-voltage insulation rated ferrite probably isn't feasible. I'll be running it as a flyback transformer at at least a 100kHz.

[Anonymous No.1558254]

>>1558244
The whole point of transformer cores are to be NON conductive.
>auto transformer
Need a galvanic connection, they are just one coil desu.
>inductor
>stores energy in a magnetic field
>transformer
>transfers energy from one coil to the other
What you are saying is
>how I do x without any means to do x
>you are doing it wrong
>no I am right, how I x
Are you sure you want to mess arround with high voltage?

Anyways, good luck.

[Anonymous No.1558262]

>>1558234
3.3v is for Raspberry Pi and various sensor breakouts.

5v is for arduino and most other breakouts not using 3.3v.

12v is for pumps, motors, solenoids, heating elements, and TEC modules.

24v is for 360 LEDs running off a PWM controller at its max supported voltage.

>>1558237
thanks anon. Good leads here I'll look for a copy.

[Anonymous No.1558264]

>E.E. is only available work that pays where i live
>Guessimanenginerd.jpg
Where do i start? I've tinkered since I was 5 but I haven't built anything
I'm looking for things to build to expand my practical knowledge
What should I buy in terms of tools to get started? Have loads of scrap in my drawers

[Anonymous No.1558284]

>>1558224
>33w + 50w + 480w + 960w
>90% efficient
the most efficient power converter is the one that's not there. why are your heating elements and mechanicals not running at mains voltage wherever practical?
90% is a very tall order. a single switcher to handle all of that seems like a really bad idea, with detrimental effects on the quality and accuracy of your output voltages. if I were putting together a system like this, I'd put the lights on their own off-line CC supply with a current/brightness input, add a 12V@50A switcher (off-the-shelf, probably) for what pumps etc. I couldn't run from the mains, hook that bitch up to some copper bus bar, and use smaller buck converters from the 12V bus to run controllers and sensors as required, an amp or four at a time. it's a step up from two dozen wall warts polluting the space, to be sure
if you still go to roll your own, you might want to have a look at some of TI's switching power supply controllers and their app notes to see if they have something to make life any easier for you, and also think about Pic related

>>1558244
any chance of using your primary as the core, encapsulated inside however much paraffin you like?

>>1558228
>I have junk, what do?
Take it to the recycler.

[Anonymous No.1558300]

>>1558254
Coupled inductors != autotransformer. Air-core transformers work fine for RF applications, and yes they are called air-core transformers for a reason. Ferrites are insulating, but they're basically a bunch of pieces of metal seperated by insulating epoxy/ceramic, hence will have a lower breakdown voltage than straight epoxy/ceramic. If they have an insulation voltage rating significantly less than that of air, that's basically a no-go. I'm working under the assumption that a non-specialised ferrite will not fulfil this condition, but I could be wrong.

Anyhow, since I'm boosting up to 200kV maximum I'll need a ~8cm long ferrite at minimum, and the product of the diameter and length of the rod needs to be at least 6400mm^2 due to needing to wrap all the wire around it in only one layer. In other words, the volume of the ferrite would have to be pretty damn large or it would be pretty damn long. I've seen those beefy toroids in switching PSUs that Dave takes apart, but without one I doubt I'd easily be able to use this core. So basically, even if they're better insulators than diamond they're infeasible to use. I think it's correct to assume that a thin core a few cm from the walls of the coil will not suffice.

Now I could try some sort of fancy core winding technique to get more layers on the core without compromising its insulation rating, but I'm already winding 100m of the stuff so that's a pass from me. I only want ~1µA of output current, or 200mW of coupled power, and I'd be fine with even a tenth of that, though I can hardly find data on the output power of a tesla coil with a similar-sized core.

[Anonymous No.1558336]

I've been told traces on iPhone PCBs are fractions of mils. Is that true? I thought no one really went below 2 mils, and that's already extreme.

[Anonymous No.1558346]

Anyone here knowledgeable about motion control/encoders? If so:

How do I know how many counts per revolution I need for an encoder for my application?

I need to be able to accurately capture a 1000 Hz signal on an input side running from 1-133 rev/sec (60-8000 RPM) and a 1000 Hz signal on an output side with gear ratio 1:100, so speeds are .01-1.33 rev/sec (.6 - 80 RPM).


Here is the math I have but am not sure if it is correct:

For a 1000 Hz signal, it takes .001 sec for one period. But if I want to accurately capture this and accurately obtain velocity from this position data, I need 100 counts (is this a good assumption? what would be a better value?). Hence, I need 100/.001 = 100,000 cts/sec. Now, if I divide this by my rev/sec, I'll get cts/rev which is what encoders are rated at.
Assuming input is running at 10 rev/sec, I need 100,000/10 = 10,000 cts/rev.
Assuming output is running at .1 rev/sec, I need 1,000,000 cts/rev.


I strongly suspect my math is incorrect however because the best encoders that meet the speed requirement of my application are 1024 cts/rev for input and 72,000 cts/rev for output. Would these cts/rev actually be good enough for my application then?

For context, I need to obtain the Bode plot of a gear drive and hence plan to give sinusoidal inputs to my motor. I do not expect this the bandwidth of my gear drive to be more than 1000 Hz, but since I am not sure what it's exact number is, I would like my design to be capable of accurately measuring a 1000 Hz signal. This sinusoid will be in terms of velocity which is why I want to obtain velocity data.
Also, I know it may seem odd that I am using position first and then getting velocity rather than getting velocity directly, but my client would like this system to be capable of positional control.

[Anonymous No.1558361]

>>1558346
You can get velocity from anything more than two counts per revolution.

[Anonymous No.1558365]

>>1558346
>Also, I know it may seem odd that I am using position first and then getting velocity rather than getting velocity directly, but my client would like this system to be capable of positional control.
The odd thing is needing 1000Hz readings of angular velocity for the entire range all the way down to 0.01r/sec
You also don't need 100 counts to determine angular velocity, you can do that with the period between pulses, or an average of the last few pulse timings, or calculate by however many pulses happened in the last millisecond.
In the worst case (0.01r/sec) with a 72000ppr encoder you'll only get 0.72 pulses per millisecond, so some samples won't have a new angular velocity to report.

[Anonymous No.1558366]

>>1558361
True, but wouldn't only using two be subject to a lot of noise?

[Anonymous No.1558376]

Not sure if I should ask here instead of making a separate thread. Basically I want to buy a cheap (sub $100) chinese laser engraver, there are different brands, but they are mostly the same. The XY mechanism is just repurposed from CD/DVD/bluray drives, whereas the laser is generally a 405nm one, probably from bluray drives. Now they are advertised to be in the 1000mw to 2000mw range, which is absolute bullshit. Has some anon tested the actual optical power output of those lasers, or know exactly what kind of laser diode they mount? I suspect a lot of engravers actually mount 200-300mw diodes in various states of overdrive. I actually don't care much about power as long they can engrave wood and cardboard. Frankly the results I've seen are quite good, but the reliability of the machines is questionable. I'd prefer to buy a "1000mw" machine over a "1500mw" one if the 1000mw one is more reliable. Does any anon here have any experience with them?

[Anonymous No.1558377]

>>1558376
Why is that bullshit? The actual laser module looks decently sized though I don't exactly have a good idea of scale. You're right to be somewhat skeptical, but not critically so.

To measure its optical power output (so you can send the data back to alibay and get it free by the dispute process) you could do a simple calorimetry by shining the laser through a window onto something very dark like carbon, which should be immersed in a known quantity of water inside a well insulated container. Shine the laser in for ~1 minute and measure the temperature difference of the water, then use the heat capacity of water to figure out the power input. To decide how much water and what length of time you use, pick values such that the optical power input dwarfs the losses due to imperfect insulation, and that the water increases its temperature an amount that isn't too close to boiling (I wouldn't exceed 60°C) but isn't too small such that the measurement error via your thermometer is too great.

[Anonymous No.1558378]

>>1558377
>Why is that bullshit? The actual laser module looks decently sized though I don't exactly have a good idea of scale. You're right to be somewhat skeptical, but not critically so.

Well, the only time I could find data on the actual diode in the machine it was a 200-300mw optical power diode and it was sold as a 1500mw machine. Probably they count power as power dissipated by the diode, not optical power. They are chinese what do you expect?

[Anonymous No.1558381]

>>1555536
Clam meters do not allow measuring current through the cables
i learned the hard way after my overpriced clamp arrived

[Anonymous No.1558391]

>>1558381
You have to measure the individual live or return cable, both is a net 0A.

[Anonymous No.1558395]

>>1558391
I mean the the multimeter wires for measuring voltage

[Anonymous No.1558397]

>>1558395
oh

Why would you want to do that, assuming you have an AC/DC clamp meter? Why would you not own a seperate DMM for the µA range?

[Anonymous No.1558475]

>>1558366
Only if the angular velocity is changing very rapidly

[Anonymous No.1558500]

I need a help from a big strong brainy man.
I have this part
https://www.aliexpress.com/item/10Pcs-KSD301-85-Degrees-Celsius-85-C-Normal-Close-NC-Temperature-Controlled-Switch-Thermostat-250V-10A/32827167060.html
Basically from what i understand it is a heater with a thermostat that turns it self off if it reaches over 85C

It says on it 250V 10A, i mean that is A LOT of current for such high voltage.

But it is installed into a food heater that operates on 12V DC power, so that would mean that if you operate it from 12V it would consume fucking 200A and that just can't be right.
I tried measuring resistance ot the in and out contacts and it shows either 0 or no connection, which would be about right since with such massive amperage it has to be short tier tiny resistance.

I am too afraid to connect it to my shitty little DC power source which can only pump out about 30A

So how in the fuck are you supposed to power this tiny thing with just 12V DC?

[Anonymous No.1558503]

>>1558500

Man you're totally confused. Forget everything you think you know about this part.
This is a switch, and it opens once heat it to 85 degrees. That's all it is.

It does not need any power to operate, since it is just a fucking switch.
It can take 10A at any voltage from 0-250V, but not more. With 30A you will probably melt the shit out of it.

[Anonymous No.1558505]

>>1558500

This is not a heater, it's just a thermostat. The 250V and 10A is the maximum switching capability (those are typical values often seen on switches)

[Anonymous No.1558511]

>>1558505
wait a minute, i think i see it now, if you flip wires in a peltier thing does it start cooling and heating on opposite sides?

[Anonymous No.1558531]

>>1558511
Yes, but I don't see how that's related.

[Anonymous No.1558534]

>>1558531
well it's a fridge that also heats food and since i only found the peltier modules and no heat spiral i assume this was the heating unit
but this makes sense since they can use the peltiers to also heat the food reversing the polarity

[Anonymous No.1558570]

>>1558534
>well it's a fridge that also heats food and since i only found the peltier modules and no heat spiral i assume this was the heating unit
Are you trying to repair something or figure out how something works?
>well it's a fridge that also heats food
Never heard of such a contraption, need pics and a brand/model #

[Anonymous No.1558652]

Opinions on Uni-T products?

[Anonymous No.1558666]

I'm interested in designing a PCI card that has a raspberry pi compute module mounted onto it, so that it could be connected to via SSH over USB. Has someone made one of these already? I'm a first year EE student, any recommended places to look at stuff like this?

[Anonymous No.1558671]

>>1558666
>PCI
Make it PCIe or go home. You can't even find a new mobo with PCI slots now.

[Anonymous No.1558678]

how do I go about automating a potentiometer so it goes high to low in rapid succession constantly

[Anonymous No.1558680]

>>1558666
PCIe for power and a integrated ethernet card configured to ssh into the pi, perhaps.

[Anonymous No.1558687]

>>1558678
replace it with a digital pot
Pic related would be most friendly to simply attach a couple of 555 timers: a fast one to the INC pin, a slow one to the U/~D pin. tie ~CS low

>>1558666
>stuff like this
hackaday.io?

[Anonymous No.1558688]

>>1558666
>via SSH over USB
Why bother with all the rigors of interfacing to PCIe if you're just going to connect it via USB serially anyway? Just connect it via normal USB instead. Motherboards have so many USB ports compared to PCIe slots anyway, and you don't need the PCIe bandwidth of even an x1 slot.

[Anonymous No.1558695]

>>1558688

[Anonymous No.1558697]

>>1558678
What are you using it for? most of the time you think of using a digital pot, you don't really need one.

[Anonymous No.1558699]

>>1558695
source of pic

[Anonymous No.1558702]

>>1558678
use a motorized potentiometer with a nice big knob

pure sex

[Anonymous No.1558703]

>>1558699
https://tools.ietf.org/html/rfc1925

[Anonymous No.1558715]

>>1558702
i do have some small dc motors that I took out of an xbox controller, but it's a bit impractical

[Anonymous No.1558721]

>>1558702
Do they sell those ones on ali? Some with a built in motor and a digital input so you can hook it up to say arduino and send it command to set it self to some value, but also be able to set it by hand?

[Anonymous No.1558723]

>>1558703
>https://tools.ietf.org/html/rfc1925
thonks

[Anonymous No.1558730]

>>1558721
I have never seen a digital one.

But it's pretty easy to implement something like that with an Arduino and an H-bridge. You just need to turn on the motor in the right direction until you read the desired value on the potentiometer.

[Anonymous No.1558774]

>>1558678
>high to low in rapid succession
but never low to high

[Anonymous No.1558815]

lord deliver me from this fresh hell

[Anonymous No.1558827]

>>1558815
she looks dicker'd mate

[Anonymous No.1558829]

What kind of connector is this? It's not RJ11 and has an offset tab thing

[Anonymous No.1558831]

>>1558827
can't decide yet if it's right fucked or if I'm the one that's fucked.

[Anonymous No.1558832]

>>1558829
rtfd

[Anonymous No.1558848]

>>1558815
>mfw I discovered lead free solder simply grows nickel hairs because fuck why not, no one knows why or how to fix it
>they break free and short stuff out

[Anonymous No.1558850]

>>1558829
it's RJ11 with an offset tab, of course

>>1558848
tin
you fix it by controlling your soldering process or slapping some conformal coating on your assembly

[Anonymous No.1558853]

>>1558832
No such d to rtf

>>1558848
Man this is so nice, now I get to continue using poisonous LEAD solder

[Anonymous No.1558854]

>>1557726
>With open coils this works well
It does indeed. When you remove the shrink you can just wind on and finally arrive at the original number of turns as well as the AL of the core. The calculation is left to the experts though..

[Anonymous No.1558858]

>>1558848
>lead free solder simply grows nickel hairs
Anon, that's not solder...

[Anonymous No.1558868]

>>1558853
It's not that easy to get intoxicated by metallic mercury and lead.
>https://www.ncbi.nlm.nih.gov/pubmed/8355325
I don't think it stays in your body forever and ever like the bioavailable salts and organic compounds.

Anyways, I think europe should have studied more the myriad of problems the abrupt change would cause before going "OMG LEAD" and banning it even for hobby use. And IIRC correctly there were no problems of lead FROM solder ever causing problems.

[Anonymous No.1558869]

>>1558868
And if with planned obsolescense and lead solder consumer stuff was being thrown in massive ammound, with shitty unreliable lead-free solder the rate of which they are discarded increased..

[Anonymous No.1558871]

>>1558848
no solder here mate, just aluminum and gold welded to each other.

[Anonymous No.1558873]

>>1558871
how that even happens? (I really want to know)

[Anonymous No.1558874]

>>1558868
>banning it even for hobby use
have they yet? it's my impression that the worst of the problems with lead-free solder are to do with oxidation and slag, which a healthy dose of flux would fix

[Anonymous No.1558885]

>>1558873
I'm not actually sure. but the aluminum forms a complex with the gold kind of like an amalgam.

[Anonymous No.1558886]

>>1558874
Atleast in britain they did.
The soldering process is diferent, the flux they use is not really safe for the workers, the process involves higher temperatures (some plants had board sagging after the bath because of the heat), it uses way more tin (tin is rare and the mining is no enviro friendly either). There are a bunch of problems.
Most health issues from led in PCBs and electronics comes from places where they burn the electronics to get the metal. (That makes the metalic lead and mercury react with other shit and be released in the smoke and dust)

[Anonymous No.1558887]

>>1558874
if they ever attempt to ban it im stashing 1km of solder in my garage

[Anonymous No.1558910]

>>1558850
>it's RJ11 with an offset tab, of course
This is weird enough already, but they use a cable that has one of the plugs facing up and the other one facing down so that the pins on the sensor/microcontroller boards can have the same order. Is this the default for telephones? When you buy a RJ11 cable, is one of the ends the reverse of the other?

[Anonymous No.1558929]

>>1558829
Lego used to use offset RJ11 for NXT sensors/motors/etc,had i2c+power + motor drive on top and RS485 on one.

[Anonymous No.1558932]

>>1558910
>When you buy a RJ11 cable, is one of the ends the reverse of the other?
Kinda, if it's a flat cable they'll be opposite sides up so that pin 1 is pin 1 on both ends.

[Anonymous No.1558934]

>>1558929
hmm, not related, Lego used offset to the other side.

[Anonymous No.1558955]

>>1558929
Makeblock uses a similar plug, but they put 2 digital pins from the arduino on 2 pins, gnd and vcc in 2 and I2C in another pair. Not a bad approach.

[Anonymous No.1559025]

>shopping hard on price on alibay
do you want pulls? because that's how you get pulls

[Anonymous No.1559026]

I want to make a charger for a supercap system amounting 250F/5.4V with a 12V DC source.
I have several XL6009 switching buck-boost chink modules in my toolbox and I figured out I can give them some use now. They are not too powerful though, they can output only 1.5A.
What would be the best way to charge them? To use the 1.25V reference voltage to make it behave like a constant current supply or any other method?

[Anonymous No.1559031]

>>1559026
attempting to use boost modules to reduce voltage will not end well. use a buck module instead if your Vin is > than your Vcap
fortunately, the modification detailed in Pic related works just about the same way on either boost or buck module

[Anonymous No.1559070]

>>1559026
Keep in mind that those supercaps will look like an absolute dead short when you start charging them. Your charging circuit has to be capable of going into current-limit instead of just shutting down (or blowing up). Some switcher circuits can't handle that.

[Anonymous No.1559127]

What is this called?

[Anonymous No.1559128]

>>1559127
nRST is pulled low when DTR high and RTS low, GPIO0 is pulled low when they're the other way around.

Presumably on an ESP8266 board? It's so the USB to serial can automatically reset the chip to the serial bootloader for programming.

[Anonymous No.1559129]

>>1559128
yes! thanks so much

Is there a name for this type of circuit?

[Anonymous No.1559178]

>>1555499
Does anyone here have experience with simavr. I have been trying to run this very simple program compiled using avr-gcc:
avr-gcc -mmcu=attiny13 main.c -o compiled

and then I try to simulate it using:
simavr -m attiny13 -f 8000000 -t compiled

Then I get the following output:
Loaded 344 .text at address 0x0
Loaded 0 .data

Then nothing happens and I need to kill the program.

[Anonymous No.1559189]

>overspend on a 60° 100K potentiometer to repair a broken throttle controller
>manage to overheat it soldering and ruin the tracks
FUCK

[Anonymous No.1559215]

I can hardly believe how fast chink post is.

It only took 4 months for these babies to arrive.

Too bad i actually forgot i ordered them and why i ordered them by now.
I am not even sure what they are, they are relays of some sort

[Anonymous No.1559228]

>>1559215
>you'll never be a stupid american that has so much disposable income that his hobby is buying stuff

[Anonymous No.1559262]

how I reduce the switching noise from a H-bridge, also rec me some cheap 1A, 24V max H bridge IC or a half bridge one (or something near those specs)

[Anonymous No.1559286]

>>1559215
I remembered what i wanted to do.

So basically here is the idea.
I have a solar invertor that makes 230V and regular outlet that makes 230V and they musn't touch each other at all.

So i am going to power the coil in this dpdt relay through 230AC outlet and then connect live and neutral to the two of the 4 poles this thing has and the invertor 230AC to the other two poles so that if power goes out in the house, the coil turns off and switches power to the inventor 230V almost instantly.
Since most devices like computers and shit have caps in them to smooth over power flukes this switch should be fast enough to prevent any reboots and such when the relay contacts are moving

but man, looking at those contacts they are fucking cloooose together as fuck, i am really afraid of arcing issues, if the outlet power ever touches the invertor it's fucking game over for anything connected to it

[Anonymous No.1559287]

>>1559262
No idea, made my own.
Pic is a H bridge with discrete drivers and a pwm driver I scrunged up to test stuff.
Rise and fall times are on the region of 200-300ns.
Now I'll make the board that'll house the controller and the pwm switching so I can run AC stuff and modulate signals.

>voltage:
5-20V
>current:
It's limited by the PFETs I think. They handle 48A, while the NFETs handle 100+A . I guess it'll be limited by heating.
>control scheme
Only diagonal FETs are active at the same time, only one PWM signal is used. Total digital pins used will be 4. (ground, supply, pwm, multiplex)

Isolated inputs.

[Anonymous No.1559289]

>>1559287
The rise time is probably lower, as the rise time function of my scope is affected by jitter, so most readings will never go below 200ns

[Anonymous No.1559303]

>>1559286
Fug. I will have to rig up some servo that basically moves two pieces of metal to flip between the sources to make sure there is a big ass gap between them so no danger of an arc is present. That is the only cheap way i know how to make sure the dicks will never touch without buying an over priced piece of shit automatic switcher.

[Anonymous No.1559306]

>>1558874
> have they yet?
No. The RoHS (Restriction of Hazardous Substances) directive applies to devices "placed on the market" in the EU. The concern is primarily about toxins ending up in the air and water after disposal. There are also restrictions on disposal (Waste Electrical and Electronic Equipment directive), but it's taken for granted that these will never achieve 100% compliance, particularly for consumer products.

[Anonymous No.1559322]

Lads,

how to build a delay audio effect without using tape or a IC? I want what the ICs do but building it myself.
Is it possible or feasible?

Thanks

[Anonymous No.1559331]

Is there a way to calculate a DC motors shaft RPMs that is powered by a 12V PWM signal?
I have data sheet that has a chart where you can read off the RPMs based on the input current but that's only for 12 volts input.
I need it so I can confirm my speed sensors values.
Alternatively I would use a stepper motor but I don't have one right now.

[Anonymous No.1559332]

>>1559322

[Anonymous No.1559335]

>>1559332
I'm sure that's reverb, not delay.

[Anonymous No.1559337]

>>1559332
is that a primitive delay line or something?

[Anonymous No.1559341]

>>1559331
Usually that is given in the datasheet or by whoever made the motor. The only way I see you doing that is with a stepper motor. You can put a tiny resistor from ground to motor -, measure the average current and get a rough estimate for your chart.

[Anonymous No.1559343]

>>1559341
The resistor is for measuring current. Measure the voltage across it, use ohms law to get the current. Pic the lowest value you can that still gives you a decent measurement and does not interfere with the motor too much. Usually 1 or .1 ohm works ok.

[Anonymous No.1559351]

>>1559343
I guess that wouldn't be accurate enough, thanks though

[Anonymous No.1559353]

>>1559351
Currents in inductive stuff is continuous, and shunt resistance measurements are as precise as your resistance value. But estimating speed from current is not usually a good idea because it's dependency of mechanical load. Usually it's done with hall effect sensors or a regular tachometer

[Anonymous No.1559355]

>>1559331
The motor current is equal to the applied voltage minus the back-EMF, divided by the winding resistance. The back-EMF is proportional to rotation speed.

If you plot a graph of current versus speed, you should get a straight line with a negative slope. The Y-axis intercept is the stall current (zero speed = zero back-EMF), which is the applied voltage divided by the winding resistance. The X-axis intercept is the no-load speed, i.e. the speed at which the back EMF is equal to the supply voltage.

From those two, you can determine the winding resistance and the motor's Kv or Ke (Kv is in RPM/volt, Ke in volts/RPM; sometimes radians/sec may be used instead of RPM). That should allow you to compute the current-vs-speed line for other voltages. For PWM, the current will vary with the applied voltage.

[Anonymous No.1559368]

>>1555499
this popped up in my autoplay and thought i would pass it along; I think I've seen the video but never noticed this fuckin part lmao, so bizarre.
https://youtu.be/-E_zUJ-SKRE?t=30

On a related note, you guys ever watch the old Nat Semi-Pease videos. Some of them are typical corporate shill but some are pretty interesting. kind of ahead of their time with the online 'show' i wish they still made them

[Anonymous No.1559369]

>>1559368
well the yt link was supposed to point to 0:30 but it's not working for me

[Anonymous No.1559375]

>>1559178
what program?

>>1559286
yeah, you shouldn't have chinked on that. I think you're looking for something like this?
https://picclick.com/Ge-Relay-Contactor-600V-20A-Mod-1-291850741743.html

>>1559322
>I have no idea what the fuck I'm doing
>make up a new class of design for me
if you have to ask, the answer is no

>>1559369
works for me, did you try rebooting it?

[Anonymous No.1559421]

>>1559332
>>1559335
>>1559337
'spring reverb'. Not anything like a delay.

[Anonymous No.1559453]

>>1559322

[Anonymous No.1559454]

>>1559375
>>make up a new class of design for me
i'm asking if that's possible or how i would go about it

[Anonymous No.1559480]

>>1559454
you start by understanding how the existing or previous solutions work, at the basic level. you might start by reading the datasheet for any of the ICs commonly used in delays (as archived by affectionate amateurs) which typically contains a block diagram of the IC. wikipedia is a surprisingly good place to get information on engineering and technical topics. you gotta know your history, mane
then you consider what it would take for a discrete implementation. since you probably don't want to match and solder 2000 MOSFETs, I think we can write off the bucket brigade device as infeasible
since you don't want to do mag tape, you'll have to find some other means of time-shifting a signal, perhaps with the help of mother google. as acoustic means go, air is one of the slowest media through which you could propagate and time-shift a signal, to allow for a reasonably sized build. it was good enough for the famous Leslie speaker, it's good enough for us
now let's put it in a container. you'll want something that won't lose much sonic energy, so a hard surface. at the same time, you might (or might not) want to limit reverberation off the walls of the container. you could use PVC pipe, with a speaker and amp at one end. you might (or might not) acoustically treat the inside of the pipe end caps with one of many sheet or coating treatments. shelf liner seems like a good choice
at the other end you would add a microphone, perhaps on a stick with a bushing of some sort for delay time adjustment, or a mechanical flanging effect if you feel so bold. feed that into your audio chain and/or speaker amp as you like

[Anonymous No.1559490]

>>1559480
nice, thanks a lot

[Anonymous No.1559512]

Can someone explain to me how a current mode amplifier works? One like the LM359

[Anonymous No.1559530]

>>1559512
shit's wack
http://www.ti.com/lit/an/snoa653/snoa653.pdf

[Anonymous No.1559535]

>>1559512
if you're good i might post some pics from a NatSemi handbook
actually you can probably just google it LM3900 app note they have a solid 6-7 pages on how revolutionary they thought it would be.

From my understanding it's one big current mirror. Unlike an VM op-amp, Zin is pretty low and the input current into V+ is mirrored (through the feedback resistors of your choice) on the V- pin. They can be pretty versatile but if you're not working with very high BW (maybe audio) then you're not missing out

[Anonymous No.1559540]

>>1558224
as a small protip/fyi the atmega can run down to 1.8V I do believe but not sure what other ICs you have

[Anonymous No.1559560]

>>1559530
>>1559535
Thanks! It's datasheet says the unitary gain BW is at 400Mhz, does this mean one can make oscilators at near that frequency?

[Anonymous No.1559571]

>>1559560
no, but high tens of MHz might be possible

[Anonymous No.1559577]

>>1559560
You can make silicon MEMS oscillators can be made to oscillate at more than 1GHz.

[Anonymous No.1559596]

>>1559189
I use to see that all the time with people building guitars. Apparently it's common to solder the ground wire onto the metal case of the pot, which requires getting the whole thing hot and often kills it.
I don't know of any way to unfuck it when that happens.
could you use a 500k regular pot? 60 degrees of rotation on that should give you about 100k. you could trim that value with external resistors.

[Anonymous No.1559599]

>>1559596
Why not crimp the ground to an eye connector and simply use it as a washer on the pot's threads?

[Anonymous No.1559627]

>>1559599
fuck if I know, but that's what we ended up doing.

[Anonymous No.1559628]

>>1559577
Yeah I'll just make one here with screwdrivers and drills. I was asking about using a current amplifier to do so top kek
>>1559571
Nice

[Anonymous No.1559633]

>>1559599
1. they often use the case of the pot as a convenient junction point for ground
2. the threads aren't necessarily connected to the case (not that it matters as long as the electronics compartment has grounding foil)

[Anonymous No.1559635]

>>1559628
you should probably just use a DDS or some other clock generator instead. more precise, can still be FMed (if the right chip is used), probably lower power, and the parasitics of an RC or LC design will not work in your favor

[Anonymous No.1559646]

>>1559633
>junction point
Ah, well in that case I'd just use a piece of solid-core copper wire.

[Anonymous No.1559676]

>>1559646
should work, as long as it's mechanically secured. you don't want the wiring moving around inside the electronics cavity, shorting against signal wires and/or fraying and breaking

[Anonymous No.1559806]

So I'm pretty retarded but want to make a test set
Potentially a huge test set, like 100 pair of wires to test. Each pair should be able to be tested with respect to ground and each other. Most important tests to me are resistance and capacitance.

Buying a 100 of those diode tester all in one kits or whatever sounds expensive

Buying 100or 200 relays also sounds super expensive.

Is there a smarter way to test all these wires?

[Anonymous No.1559836]

>>1559633
>they often use the case of the pot as a convenient junction point for ground
If we're talking about a very low-level, high-impedance signal like from guitar pickups, then grounding the potentiometer case might also be done to minimize EMF pickup as much as possible.

[Anonymous No.1559838]

>>1559577
>1GHz
Usually if you want frequencies that high, you use a PLL as a multiplier to a very stable reference oscillator, that's how modern microprocessors and chipsets do it.

[Anonymous No.1559840]

>>1559838
If anyone was interested:
https://www.digikey.com/product-detail/en/texas-instruments/LMX2594RHAT/296-46560-1-ND/7056949
Up to 15GHz frequency synthesis

[Anonymous No.1559852]

>>1559596
>could you use a 500k regular pot? 60 degrees of rotation on that should give you about 100k. you could trim that value with external resistors.
You could but it's not going to be as accurate in practice and will lack solid min and max limits, which is kind of a deal breaker for a control device pot. You can actually calibrate around a different resistance range without resorting to external circuitry but it makes a big difference for the full range to be compressed within 60° and hit a wall immediately outside of that.

[Anonymous No.1559858]

>>1559806
CPLD, FPGA, MCU, etc.
Multiplexing transistor arrays with a bunch of shift registers sounds like the right idea.

Though to measure every one of 101 points (including ground) to every other point, that would be 5151 measurements.

[Anonymous No.1559919]

hello /ohm/

I bought a cheap old guitar amp from the 80ies (a Marshall JCM800 model 2204). I got it really cheap, way below usual prices, so it does not basically not matter what I have to invest in new parts. Transformers, tubes and chassis are in a very good condition.

First of all: the amp work seems to work. There is sound, it is loud, maybe bit noisy.
However, studying the schematics (pic related), I found that mine has a strange power supply. The setup for this amp should be: 50u+50u reservoir cap
to outputtransformer and choke, choke to to 50u cap, powertubes and dropping resistor, dropping resistor to 50u cap, phase inverter and another dropping resistor, that one to another 50u cap, second preamp tube and a last dropping resistor, and form that last dropping resistor to a last 50u cap and the first tube. (see schematic)

But mine is strange: 50u+50u cap to OT and choke, choke to 50u+50u, powertubes and dropping resistor, dropping resistor to 50u, phase inverter tube and a dropping resistor, and that dropping resistor goes to a last 50u cap which supplies both V1 and V2.

I figure that this basically raises plate voltage on the first tubes while lowering the overall filtering of their power supply.

Are there any other side effects of this? Why would anyone change it to this configuration? Does this change its overdrive characterists? (in a negative way, which means LESS overdrive, as that is what this amp was built for: overdriving the preamp)

Should I convert it back to original specs? I guess this change was made when the amp was recapped, but for what reason? Is this some kind of common modification to tube amps?

I am really new to tube amps, any help is highly appreciated.

[Anonymous No.1559935]

How reliable are proximity sensors like this?
Am i guaranteed they won't trigger by themselves in a month or two of running?

[Anonymous No.1560038]

>magic smoke

[Anonymous No.1560060]

>>1559935
the only way to know for sure is to verify that in your own environment. that said, I've had one running for a year or two and it hasn't false-triggered yet

>>1559919
congrats on your find!
>more capacitance
just at a guess, it might sound a bit less saggy than stock. whether to convert it back is your call to make. you could certainly tack on some switches in series or parallel as appropriate, and evaluate for yourself, bearing in mind 400V and all that

[Anonymous No.1560063]

How do ferrite beads work? I know what they do, but how putting a high u core around the wire will lower noise? Does it behave like a one turn inductor?

[Anonymous No.1560066]

>>1560063
basically, yes

[Anonymous No.1560068]

>>1559858
Every combination isnt necessary, the resistance from one pair to another is usually okay, or at least not part of this testing.

I work for the phone company and they explicitly sold or trashed all of these pair tagging units 15 years ago so they would be less competitive with the cable and cellphone companies.

It would make my work so much easier I wonder how cheaply I could diy it. This is like 2000$ used, 5000$ new

Its main purpose is to make identifying pairs very quick. You clip on in the splice with another unit and it communicates back what pair, 1-100 you clipped on so then you can put it on the right connector.

[Anonymous No.1560069]

I have a switch I bought from aliexpress that has a protective cover, kinda like you see in aircraft and is also LED lit when in the on position.
What I want to do with it is to interrupt a circuit while also having the LED on. and having the LED off while the circuit is closed.
The switch has 3 connection points: a positive, rated for up to 12v, a ground for the LED and another point for whatever is being switched.

[Anonymous No.1560072]

>>1560068
Analog switches? With some smart connections you can make that with under 10 parallel units.

[Anonymous No.1560073]

>>1560072
>>1560072
I mean, assuming you have a way to connect all those pairs. I have no idea of how phone shit works, but wiring 100 terminals would quickly ensue in suicide for me

[Anonymous No.1560074]

>>1560073
>>1560072
And they come in 16 pin DIP. Jesus fucking hell, phoneposting makes me press the post before I finish saying everything. There are probably some modern god how many channel mux/demux surface mount chips if that's your thing. I'd check mouser to see their part names.

[Anonymous No.1560076]

>>1560060
>I've had one running for a year or two and it hasn't false-triggered yet
Do you use them bare neck or hide them under something? I see they activate at most at about 3mm distance from finger which is pretty small so hiding them under a 2cm thick piece of wood ain't happening it seems

[Anonymous No.1560079]

>>1560076
Not him but they are capacitive sensors, so distance isn't really their thing

[Anonymous No.1560083]

>>1560079
then what sort of device do need i can glue to say on the inside of a closet door and it will be able to detect my finger through 2cm of wood?
i was able to find no such sensor on chinkmarkets so far

[Anonymous No.1560086]

>>1560083
It's called a button with a 2cm wire.

[Anonymous No.1560087]

What do I do if I have a pair of 4 ohm speakers and a driver board rated for 6-32 ohm speakers, pic related.
>inb4 put them in series and use only one channel

[Anonymous No.1560089]

>>1560086
that looks gross and is not waterproof and shit

[Anonymous No.1560090]

>>1560089
Why do you have a underwater closet?

[Anonymous No.1560091]

>>1560074
That looks pretty good.

Wiring them up is the second problem. The frame on the right of the pic is an industry standard and you can get cables in 25 pair groups that you just clamp on. Not sure where to get them though.

[Anonymous No.1560092]

>>1560090
I want this not only for hidden furniture buttons but also other projects

[Anonymous No.1560094]

>>1560076
bare electrode side against ~2mm ABS

>>1560083
well, nobody does that so nobody's making a chip for it. you need to bend your project to the available materials. cut a cavity in the back side of the door or something

[Anonymous No.1560098]

>>1560074
CD4067 16:1 mux/demux
also in DIP btw

[Anonymous No.1560101]

>>1560094
>well, nobody does that
i refuse to believe that, having working painted on button on your stuff that actually works is way too cool for nobody to do it

[Anonymous No.1560102]

>>1560101
Connect a metal thing to the sensor and to the end of the wood thing.
The way people do it is by masking the sensor, near the surface. embedding it in stuff.
hmmmmmmmmmmmmmmmmmmmmm
I think I'll make a new area of electronics, I'll call it embedded systems

[Anonymous No.1560104]

>>1560102

[Anonymous No.1560132]

>>1560101
alas, lad, the capacitance equation contains the inverse of the distance between the two "plates" as a term, but does not contain a "cool" term
touch sensors routinely measure capacitance variations on the order of a couple of picofarads. by increasing the distance to 10x you decrease the capacitance change to 1/10, which is some fraction of a picofarad, which is tiny compared to the capacitance of the wires you want to run between the MCU and the sense area, which is part of why you've been failing for months
the touch sensor board does help in decreasing the length of delicate signals (which is always a good idea) but you still aren't going to see 2cm performance out of it
>It is an art/science of applying principles to requirements
so you don't have much choice but to get the wire closer to the finger to be sensed, which means thinking mechanically. side-drilling, back-carving, use a circle cutter, back-drilling blind holes for screws. you got plenty of options

[Anonymous No.1560191]

>>1560069
bls respond

[Anonymous No.1560195]

>>1560191
did you try connecting LED- to the load

[Anonymous No.1560212]

>>1560195
Yes, the LED lights up when the switch is on and the current passes through the switch just fine..
Isn't there some sort of way to invert the switch's "state"? Kinda like an NOT logic gate?

[Anonymous No.1560215]

>>1560212
Unless theressomethinf in the documentation you probably need to open the switch and change the LEDs connections

[Anonymous No.1560217]

>>1560215
stfu donny

>>1560212
Pic related

[Anonymous No.1560223]

>>1560212
sigh, fine

[Anonymous No.1560231]

>>1560217
Thanks, that was what I was looking for.
Also stupid question, can I use just any resistor before the LED? I want to connect this to a car's electrical system so I don't want this shit to catch fire.

[Anonymous No.1560232]

>>1560231
the 1k in the drawing is probably already built into the switch, for about 10mA forward current. check the documentation/packaging. or, just put it there for safety and replace it with a wire if the LED doesn't appear bright enough

[Anonymous No.1560234]

>>1560217
>>1560223
how did you get that load symbol?

[Anonymous No.1560235]

Is anyone familiar with simulink?
I'd like to measure to time that passed between a rising and a falling edge but I don't know how.

[Anonymous No.1560237]

Threads almost dead so excuse me if I double post. Building a temp control for injection molder and here are some schematics I redrew from a shit ''precious plastics'' one.

problem is my PID controller is missing the NO1 slot and I really have no idea where to reroute this now. I don't really know where to go from here because my /ohm/ knowledge is limited.

Pic related

[Anonymous No.1560256]

>>1560234
lamp in outputs

[Anonymous No.1560261]

>>1560191
never touch a data sheet
never post a diagram

[Anonymous No.1560270]

>>1560237
read datasheet?
https://www.gmelectronica.com.ar/datasheets/B2E-en.pdf

[Anonymous No.1560273]

>>1560217
'leakage' current to the load through the LED and resistor

[Anonymous No.1560290]

>>1560217
Not the guy that started this but how exactly is this supposed to toggle the current for your load? The circuit can only toggle power to the LED, the load receives power continuously no matter the state of the switch.

[Anonymous No.1560309]

>>1559070
Exactly, I'm trying to either mess with the feedback network to make it a current source of 1.5A or add a hefty, low resistor in series.
I'm considering seriously leaving just the space to get myself a LM2596 buck converter, design the PCB by myself and, in the meanwhile just charge the caps using just some resistors in series.

[Anonymous No.1560312]

>>1560290
>load receives power continuously
...through the LED. if you have a low resistance load like a big lamp or motor, it will pass that tiny current and not do anything.
if it's something electronic like a microcontroller it might behave unpredictably when the switch is open.

[Anonymous No.1560320]

>>1559322
Performing delays using analog, discrete components is basically impossible.
Why not making it simple and buy a microcontroller with a good ADC and DAC (16-bit), some SRAM chip and program it?
It's small, clean and has no mechanical parts. The thing is that you will need some firmware programming skills and basic DSP knowledge, but it's a nice challenge.

[Anonymous No.1560321]

Can test pads (the round exposed copper contacts with no hole) be used as fiducials for pick and place?

[Anonymous No.1560323]

>>1560312
Or you could do something like pic related and escape all uncertainties.
Circled in red is OP's switch. Underlined in blue is a powered switch that opens when powered.

[Anonymous No.1560326]

>>1560217
What would be the method to inverting that behaviour, a single NPN?

[Anonymous No.1560332]

>>1560326
Something among the lines of a latching relay I guess.

[Anonymous No.1560345]

>>1560332
ew no

[Anonymous No.1560362]

Easiest way to generate triangle (not sawtooth) waves without opamps? (I might need them at more than 1MHz, which is the limit for most amps.
I'm ok with discrete transistor madness

[Anonymous No.1560375]

Can I drive a loudspeaker (8 ohm) with a mosfet circuit designed to drive LED strips with a PWM signal?

[Anonymous No.1560392]

>>1560362
Is this for PWM purposes? A 2-transistor based relaxation oscillator would be a simple option, but I'm unsure about how that topology would work at high frequencies.

>>1560375
I think you'd want a half-bridge or a totem pole.

[Anonymous No.1560395]

>>1560323
it's only uncertain if you don't know exactly what kind of load it's connected to. there are fuseholders with indicator lights that work the same way, and I'd use them on a motor but not on an outlet.

[Anonymous No.1560447]

>>1560362
i just came up with this. it's a switched current mirror. it's just a sketch, real parts will need emitter degeneration and source/sink matching and trimming.

the left 4 transistors are the current source/sink mirror. the middle two and diodes switch between sourcing and sinking. the divider biases the triangle to midway between the rails. the cap integrates the current to get a triangle.

[Anonymous No.1560450]

So I'm a noob and I need help getting started on this project. I want to make a rudimentary 3d printer out of a disassembled printer and arduino. It has 3 stepper motors. I'm curious if I can utilize the old PCBs? Where do I start in learning that? Embedded programming? Pls

[Anonymous No.1560459]

>>1560447
Do you actually need an accurate triangle wave? Because usually a constant-voltage + resistor + capacitor ramp is close enough.

[Anonymous No.1560461]

>>1560459
i don't think anyone actually needs an accurate triangle wave

[Anonymous No.1560542]

>>1560237
>pid controller
>timer
pay attention, or put the tools down and find someone to do this for you

[Anonymous No.1560544]

>>1560450
>I have junk, what do?
Take it to the recycler.
seriously tho, there are arduino generals and 3d printer generals. either would be more appropriate to your request

[Anonymous No.1560557]

>>1560459
>constant-voltage
constant CURRENT. Otherwise you get a curve.

[Anonymous No.1560559]

>>1560557
I meant what I said, that's why I said "close enough". 555 and comparator-based relaxation oscillators both produce those curvy triangle waves and I see them used for PWM purposes all the time.

[Anonymous No.1560564]

>>1560060
I did a quick test by adding the missing dropping resistor back and running V1 and V2 from different capacitors (aka original schematic): Could not hear a bit of a difference.

I also temporary added a so called post-phase-inverter master volume with gator clips after the decoupling caps of the phase inverter. Surprise, surprise, at full volume (pre-amp and master set to 10, while lowering loudnes with the temporary master volume) something is wrong with the amp. The output transformer makes a hearable noise, a high frequency kind of buzz that also comes through the speakers. Also notes and chords that are played suddenly die or have a big drop volume after some seconds. That thing is bugged. I really don't like to go to deep in this
>bearing in mind 400V and all that
and the fact that I have very little experience with this kind of amps.

Shit

[Anonymous No.1560567]

/synthdiy/ wannabe here, in the new thread let’s discuss the book The Art of Electronics

[Anonymous No.1560569]

>>1560564
the difference would be audible on hard attacks, but if you can't hear the difference may as well leave it
>high frequency buzz
>suddenly die or have a big drop in volume
usually means too much plate current. whatever you were adding was probably wrong. what's wrong with the master volume that's there anyway?

[Anonymous No.1560575]

>>1560569
>what's wrong with the master volume that's there anyway?
The master that exists is before the phase inverter. One cannot overdrive it without the insane volume level of a 50w amp. By adding the post-phase-inverter master I can overdrive it. It is fairly common to add these to marshall and fender style amps. I planned to do this as I was aware that the "master" on the JCM800 amps is usually worthless. I had played one before, I know that you get way less overdrive when dialing them back that way.

>usually means too much plate current. whatever you were adding was probably wrong.
I already did an A/B test with and without the addition but the buzz from the transformer (!) stays. It sounds similar to this video https://www.youtube.com/watch?v=zQXNpyp1nF4

>too much plate current
But how is that related to the filter caps?

[Anonymous No.1560578]

>>1560575
my bad, I was thinking of noise from the tube itself. disregard my post

[Anonymous No.1560579]

>>1560567
can someone please recommend me some books similar to Art of Electronics? I am a retard looking to become an expert in /synthdiy/

[Anonymous No.1560583]

>>1560578
No problem. However, does the sound not indicate that there is at least a mild problem with the outputtransformer? And why the sudden drop in volume? Whenever I hold a chord and the notes decay, the distortion reduces like it should and then it basically drops into nowhere.

[Anonymous No.1560588]

>>1560132
fuck guess i have no choice.
I can use that kind of round saw drill to make nice circular hole but how the fuck do i remove the middle if i don't want to cut the whole way through?

[Anonymous No.1560618]

>>1560447
>>1560392
thanks friends. I think I'll try the curvy rc oscilators.

[Anonymous No.1560624]

>>1560618
And yes, it'll be for a pwm thing. I made a 555 one for testing motors and stuff but the designs with that single pot isn't that good. I had forgotten about a super simple one also using single one, in the art of electronics. (The right one doesn't have a completely constant frequency due the small resistor.)

[Anonymous No.1560627]

>>1560624
My main goal was to make a voltage controlled one. I guess the pot will have to do.

[Anonymous No.1560718]

>>1560624
>a super simple one
Maybe the one in the middle?
None uses pin 7 to oscillate but keeps it for better things.
The lesser known applications are worth a closer look, pin 5 e.g.

[Anonymous No.1560730]

is it alright to use a 5% tolerance over a 20% tolerance cap? This is for an audio amplifier

[Anonymous No.1560734]

>>1560730
Of course it is.

[Anonymous No.1560736]

>>1560734
thanks I looked it up before coming here but everyone else gave such vague answers

[Anonymous No.1560740]

>>1560736
as long as the dielectric type is the same and voltage rating is equal or better, sure. it takes a bit of care to change to a different dielectric type

>>1560583
hmm, well all I got now is generalities. how are your voltages looking vs the notations on the schematic, particularly around the PI and the output stage? is bias correct? are your power tubes fresh? is the output selector set correctly for the load? (you are connecting a load, right?)

[Anonymous No.1560757]

>>1560542
its the same thing

[Anonymous No.1560767]

Explain pull up resistors using only 1 (one) sentence.

[Anonymous No.1560806]

>>1560767
they weakly pull a node, which is (temporarily) not driven and whose logic state is therefore undefined, to a defined state which can be overridden by a strong drive to either state

>>1560757
2/8 felt a small rustle in my jimmies

[Anonymous No.1560824]

ATTENTION:
=-=-=-=-=-=-=

NEW THREAD: >>1560823

[Anonymous No.1560963]

>>1560450
Start by researching types of motors, and motor drivers. You can buy driver modules cheap from china to drive pretty much any type of motor if you know what you have and how to wire it.
If you really want to use the motor drivers that are already on the PCBs, start by googling the numbers on all of the chips on the board so you can learn what they are. chances are there will be a motor driver chip you might be able to wire into.
Do not try to write your arduino code from scratch, work from an existing project. there are at least a few arduino 3D printers already.