[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Search] [Home]
Board
Settings Home
/diy/ - Do It Yourself

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • There are 56 posters in this thread.

05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
06/20/16New 4chan Banner Contest with a chance to win a 4chan Pass! See the contest page for details.
[Hide] [Show All]



File: 1558574325614.png (90 KB, 1920x1153)
90 KB
90 KB PNG
all tied up thread: >>1621755

>RULES
0. Electrics ≠ electronics. Appliances/mains stuff to /qtddtot/ or /sqt/. PC assembly to >>>/g/.
1. Do your own homework. Search web first. Re-read all documentation/datasheets related to your components/circuits. THEN ask.
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
jkgamm041
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.
>>
File: LM131VtoF.png (223 KB, 608x890)
223 KB
223 KB PNG
Opening Ritual
>>
File: 1544538547385.png (402 KB, 2064x1461)
402 KB
402 KB PNG
>>1627160
>>
File: IMG_20190610_232748.jpg (429 KB, 1536x2048)
429 KB
429 KB JPG
yesssss
that piece of shit relay is finally working properly now
i had to add a coupler there since the cunt wouldn't work with 3v and i managed to burn the usb power traces, but thank god the vin pin traces were spared so i can still power it via the vin pin so all is gud
but shit man. this is way more pieces that i hoped for. if i want to make a bunch of these plugs for to turn my house devices into iot bois, it will take a shitload of work to put together so many components for so many plugs
thank god for wire wrapping tho, it is making connecting all those pins together a breeze
>>
File: 1529254118564.gif (809 KB, 500x374)
809 KB
809 KB GIF
>>1627179
>that safety cap
>>
so why is this capacitor making the output goofy as fuck rather than smoothing the inverter's output
>>
>>1627179
>shitload of work
and that's why prototype pcbs are awesome
for funsies, post original board to see how much less of this you would have had to put up with had you just hacked the fw

>>1627196
>he doesn't know about resonant circuits and damping yet
>>
>>1627199
I know of them but I wasn't sure if that was going to make an LC tank circuit or not

how then would you go about smoothing the output of an inverter without making an oscillator
>>
>>1627201
usually you just call it a win since power was transferred
you might want to think about regulating the output voltage before worrying too much about wave shape, since true sine wave inverters start to look a bit more like a power amplifier anyway
>>
so all else being equal, an increase in current means a decrease in voltage?
>>
>>1627264
in the context of constant power, yes. was that the context of interest to you?
>>
>>1627268
I can't honestly say. I'm playing with diode clippers.

I see when the breakdown voltage is reached, the diode begins to conduct and thus shunts the current away from Vout. That's also where the waveform of the triangle wave begins to distort.

I'm trying to develop intuition for this. The way I'm thinking of it now is that in order to drive that current through the diode it must "spend" some voltage to do so. So some of the output voltage is redirected towards making the diode conduct, thus depressing the output voltage and distorting the waveform in the process.

Granted, that "spending of voltage" might just be from the voltage required to make the diode conduct and not be an inherent relationship between current and voltage.
>>
>>1627273
>not be an inherent relationship between current and voltage.
>>
>>1627193
>safety cap
what do you mean?

>>1627199
There is no original board, this plug had originally just a humble mechanical switch which i replaces with a relay and all that other ic stuff
>>
File: no.gif (1.18 MB, 300x188)
1.18 MB
1.18 MB GIF
>>1627179
>Touch
No
>>
>>1627499
yes.
it is nice and flat and saves me from having to make an extra hole for an ugly switch
also i hope you aren't on about some reliability meme, i use touch switches in many of my iot devices and i haven't had a single misfire ever
>>
>>1627524
>an ugly switch

not trying to argue, but how do you use that "non-ugly" touch thing if you don't make holes for the three wires, and how is an ordinary 20 cent pushbutton uglier than your touch sensor? Do you have a pic of a finished project that uses a touch sensor? And if >>1627179 is a finished project then maybe you should disregard this entire post.
>>
>>1627532
you don't actually touch touch buttons you touch the casing covering them
for example i can glue the red part under my table and then when i place my finger ON TOP of the table it activates it is really cool, it looks like you are pushing an invisible button or can draw dick with a sharpie there and have a dick button
>>
>>1627541
>i can glue the red part under my table and then when i place my finger ON TOP of the table it activates it is really cool, i

that is amazing, if I understand correctly. You put the sensor on one side of a solid, stiff board, and it detects your finger through the board. Do you have a link or part name, because I'd like to try that.

For example, if I want to have 5 of them, how close together can they be if they are on the other side of a 3/4" thick board?
>>
>>1627547
just search for touch buttons, they are used everywhere these days
in fact you can make one yourself with just a piece of foil wrap and a wire
>>
>>1627554
>in fact you can make one yourself with just a piece of foil wrap and a wire

do you have an answer to the other part, where I'd like to have five or more of them on the other side of a 3/4" board? Also, how would I detect the signal from a piece of foil wrap and a wire located 3/4" from my finger? I assume it's capacitive, but can a weak capacitor like that really be used very far from your finger?
>>
>>1627561
I think they measure the impedance from the plate to "ground", but I might be wrong and they measure the fluctuating mains-frequency electric field coupled from your body instead. In either case, your body is a pretty good antenna for coupling to your surroundings.

A cursory glance at a product page tells me they're good to 5mm or so, but I guess you could multiply that by the relative permittivity of the relevant material. Some plywoods show they might have a relative permittivity up to 8, but MDFs look to be only good to 1.2 or so. But this also is quite dependant on frequency. A screw screwed most of the way into the board from the other side would certainly help though.
>>
>>1627561
>>1627554
>>1627547
>>1627580
https://en.wikipedia.org/wiki/Capacitive_sensing
https://www.youtube.com/watch?v=mWR9Q_pTagw
>>
File: hey.jpg (14 KB, 240x240)
14 KB
14 KB JPG
>>1627561
Be the first to make it work.
>>
>>1627429
safety capacitor across the primary/secondary. apparently, instead of using a proper certified Y cap they just put a greencap in there. kek

>>1627547
>they are on the other side of a 3/4" thick board?
they work much better with only a couple mm of veneer separating them from their finger. 1/r^2 you know

>>1627561
>fucking touch switches, how do they work
http://ww1.microchip.com/downloads/en/AppNotes/00001334B.pdf

>>1627602
>>>/pol/ tryhard
>>
>>1627727
>fucking touch switches, how do they work

excellent reference. so they are wonderful, but rather challenging.
>>
File: 1558372349275.jpg (156 KB, 900x1600)
156 KB
156 KB JPG
hi im a brainlet and have no idea what i am doing other than a few college digital design courses i took. what can i do with parts of my old mp3 pic related
>>
>>1627842
read OP
>I have junk, what do?
Take it to the recycler.
>>
how do I become the recycler
>>
>>1627867
just a recycler or The Recycler?
>>
Has anyone ever found anything easter egg-y in a datasheet or elsewhere unexpected?
>>
>>1627931
closest I found was in a computer book's index, where there was a fake entry in the index for "daemon" or "demon" or something on page 666 (which wasn't in the book)
>>
>>1627842
The lipo (and its protection board) looks good, the screen may be usable if you can desolder that flat flex and find a datasheet for it, and some of the components on the PCB might be worth salvaging depending on your skillset and such. The main IC might be reprogrammable, but I can't see it being worth doing. I suspect the IC shown there isn't the main IC but the flash memory IC, so arguably you could replace it with a higher capacity IC to modernise it, possibly install rockbox on it if supported.

>>1627931
What about that "write only memory" datasheet posted a while ago?
>>
File: 1542622500107.jpg (39 KB, 421x636)
39 KB
39 KB JPG
>>1627955
I'll see your Signetics 25120 datasheet and raise you the Umac 606 phantasatron
http://www.tubecollectors.org/archives/606.pdf
>>
File: 1557802324697.jpg (25 KB, 243x269)
25 KB
25 KB JPG
>>1628041
the rare National Semiconductor LM0901A1411090451C "Polish Operational Amplifier", which "may be substituted directly for paper weights and fish lures". includes characteristic graphs, mostly characteristic of the comparatively un-butthurt workplace culture of the 1970s
http://idea2ic.com/FUN_DOCUMENTS/PolishOpAmp.pdf
>>
File: 1530081473123.jpg (70 KB, 1280x720)
70 KB
70 KB JPG
Noob here, can I use relatively thick audio cable instead of dedicated 6 amp cable?
>>
>>1628066
For power or audio? Or some other signal? What voltage? Frequency?
>>
>>1628049
>If a short length of 300-ohm twin lead is not available, use two short lengths of 150-ohm twin-lead
http://steamer.mindfart.com/
>>
>>1628082
>If a short length of 150-ohm twin lead is not available, use two short lengths of 300-ohm twin-lead in parallel.
works
>>
>>1627727
>safety capacitor across the primary/secondary.
Not him, but why would you want to add a short between the two coils? With ac it acts as a short and would defeat a purpose of the transformer?
>>
>>1628066
yes it's all copper inside so it doesn't matter (unless you got it from chinks in which case its aluminum)
the only difference is that audio and data cables have shielding around them
>>
File: 1535359549016.png (279 KB, 1062x1375)
279 KB
279 KB PNG
>>1628106
that cap is kept small to keep touch currents non-dangerous
anyway its main purpose is to keep the entire output section and cable from becoming an antenna. it could be grounded instead or in addition, and if there were a ground lead, almost certainly would be
>>
>>1628078
220v, 50hz. Going to power two 30-45w light-bulbs.
>>1628110
It's actually both, aluminum for one polarity and copper for another. Don't ask me about it.
>>
>>1628124
>aluminium
In audio cable? Sure it's not tinned copper?
>>
>>1628126
I'm not but it's felt like a different, less flexible metal when I bend the tips around pins.
>>
>>1628128
>less flexible
Pretty sure aluminium would be more flexible. It's very difficult to solder onto aluminium, so that should give it away.
Without knowing the distance between the conductors it's hard to tell whether it's safe to put 220VAC across them. Half an amp should be fine if you described it as thick wire, but knowing the conductor diameter/AWG number would be good too.
>>
File: 121g.jpg (11 KB, 324x450)
11 KB
11 KB JPG
>>1627931

https://www.oreillyauto.com/flux-capacitor
>>
>>1628234
Jaycar (aus/nz overpriced electronic components store) has one of these on its online catalog for $10,000 or so.
>>
>>1628234
The line at the end tells me some retard tried to buy it
>>
Is my train of thought correct?
For reasons, the pure battery voltage of my lipo can be connected to the board. A DCDC converter that shares the ground with the battery supplies the entire board.
I used a pnp mosfet to prevent short circuits by reverse voltage protection. If I were to connect the mosfets ground to the boards ground and I connected the pure battery voltage I would create a short, right?
Consequently, if I didn't connect the mosfets ground to the boards ground and I connected the pure battery voltage cables the right way. Would the OPamps noninverting gate be powered by the pure battery voltage. Because they share a ground at the battery's boost converter pcb.
>>
>>1628338
>I connected the pure battery voltage in reverse*** I would create a short, right?
fixed
>>
>>1628338

Just for fun, use "pure battery" and "battery" in a sentence, so I can get an idea what you mean by pure battery.
>>
>>1628344
I thought that would have been more confusing but I guess you are right
>>
Has anyone used the capsense lib for arduino and measure how much delay there is in the detection. I am hoping it is on the magnitude of 1ms but I think Ive read that it is more like 5ms.
>>
>>1628338
Well, theoretically, your "pure battery" ground is shared with your DC-DC boost ground, which is your circuit ground.

I don't see why you put a mosfet on the "pure battery" ground.
>>
>>1628421
try the arduino general?
>>
>>1628423
Well if the Gate of the mosfet was connected to the circuit ground and the cables leading from the battery to the circuit were falsely connected (Battery voltage directly connected to circuit ground) there would be a short circuit

The battery is supposed to be removeable for charging so you would have to reconnect those cables everytime you connect its DC-DC converter to the circuit via USB ports.

So to prevent a short circuit I used the p-channel mosfet and connected its gate to the battery ground
This reverse voltage protection is not included in your schematic because I didn't go too much into detail regarding that
>>
>>1628338
>reverse voltage protection
Jeri Ellsworth did a vlog on reverse-polarity protection. you should find it and watch it
>>
File: muh precious.jpg (369 KB, 1000x419)
369 KB
369 KB JPG
The guy who taught me how to solder used to work as a cellphone repair tech. He had this amazing white flux that allowed him to beautifully rework microscopic components by hand and with no mess. He wouldn't tell me the brand, citing his NDA. Said his shop bought it in bulk. He was willing to part with this 1/10th of a CC because he was such a good pal.

Does anyone have any clue what this stuff is? I haven't even used it because when it's gone it's gone.
>>
File: 1545509563088.png (25 KB, 574x357)
25 KB
25 KB PNG
>>1628442
also, your reverse polarity protection isn't protecting the dc-dc converter which almost certainly will release its magic smoke. instead just use a large-ish resistor between the battery and the voltage detector input. the comp's bias current is too small to worry about (if you chose your comparator correctly) so a large resistor (100k) won't cause significant error. in fact, as a general principle, a diff-pair (such as used in op amp/comparator inputs) should see the same resistance on both inputs for best balance and minimum offset, so you should add this resistor anyway, equal to the effective resistance to your voltage source on the other input. in the case of a simple voltage divider it's just Ra||Rb
in case of reverse connection, your dc-dc and maybe some stuff on the load side of it will be destroyed (but, thanks to the resistors, exceeding the common-mode range on the opamp/comparator or exceeding the input current rating won't be the direct cause of catastrophe)
anyway chink boost modules use a fair to large amount of current even with no load so don't expect more than a fortnight of operation from full charge
>>
>>1628338
I failed to decipher both the painting and the prose.
>>
>>1628457
Probably this stuff
https://store.rossmanngroup.com/index.php/amtech-nc-559-v2-30-cc-16160.html
>>
>>1628442
>>1628458
good video
The DC-DC converter has a built in short circuit protection, I suppose. good tip though
but ultimately the p-channel mosfet would take care of it too

>>1628461
Let me try again with this slightly altered schematic.
here, the mosfet's gate is connected to ground. If the battery voltages were reversed like in this picture, the battery voltage would short out every component on the board and probably destroy the dc-dc converter aswell
so my question is if it's a good idea to not connect the mosfet gate to the circuit ground
>>
File: 1551656368327.jpg (106 KB, 541x348)
106 KB
106 KB JPG
>>1628468
Thanks for the identification and store link, Anon. It's cheaper there than Amazon by several dollars. And Amazon is FULL of chink knockoff fluxes of every variety.

Just bought a 30cc syringe. If it's not the same I'm sure it's a damn sight better than the rosin-based poop I need to throw away.
>>
File: 1553544427659.png (31 KB, 574x357)
31 KB
31 KB PNG
>>1628457
that's what flux should be expected to do tho
the YouTubers who work with leaded solders seem to like *genuine* Amtech RMA-223 and NC-559 fluxes (as opposed to the vaseline available on alibay). those who work with lead-free have some no-clean options which are not quite as easy to find in hobbyist quantities
I've been using SRA's TF5000 tacky no-clean flux with the occasional coat of SRA's 99-20 RMA liquid when dealing with age-oxidized component leads. both are supposed to be safe to leave on the board, but the solder wire I'm using produces heaps of white powdery residue and burn marks. as a matter of caution and workmanship I wash those assemblies where possible by shaking in isopropyl alcohol for a few minutes

>>1628475
load short-circuit ≠ line reverse-polarity. review the absolute maximum ratings for the converter IC and you will probably see that the input voltage must never go negative
>ultimately the p-channel mosfet would take care of it too
it does cost an order of magnitude more than a resistor and is better suited to power protection than signal protection where a plain resistor won't work
... so, instead consider using a very-low-Rds(on) PMOS in line with the battery so that it also protects the dc-dc from reverse-polarity faults (Pic related, red). the lower the Rds(on) the less load-related voltage drop you will see in your voltage measurement under normal conditions. the AO3401 is an excellent jellybean PMOS for low-voltage switching if you can work with SOT-23s
>>
>>1628499
Hmm if I did it that way like in your schematic then the 5V of the DC-DC converter would be on the comparators input and not the lipo battery voltage which would be around 4.2V - 3.4V
Thanks but I'm already using the NX2301P which has similar specs
>>
>>1628528
>5V of the DC-DC converter would be on the comparators input
the comparator's upper input is connected to the battery through the PMOS, same as before. I've just placed the boost input "after" it instead of "before" it, so to speak. note the X signifying a broken connection
the dc-dc converter is a boost converter, correct?
cool, the AO3401 was just a suggestion
>>
File: 1551539552629.jpg (64 KB, 1024x798)
64 KB
64 KB JPG
>prototype pcb got into production 2 days ago
>check the manufacturer's design rules again
>minimum distance from trace to ground plane is 8 mil
>have 6 mil on my board
is it already over? this is my first pcb and I don't know if I fucked up and wasted 40 bucks in the process
>>
>>1628533
>I've just placed the boost input "after" it instead of "before" it, so to speak.
I realized that now. Wouldn't that overshadow the 4.2V-3.7V from the battery though?

>the dc-dc converter is a boost converter, correct?
I think it is. Though it's hard to tell since the markings on the ICs are barely visible.
>>
>>1628535
My opinion is that you should tie that knot again.
>>
>>1628535
>$40
>for an 8 mil design rule
should have gone with JLC
anyway you're not screwed, but you might want to pick up some pen knife blades

>>1628542
>overshadow
not sure what you mean. the boost just takes buckets of current from the input and throws them to the output. the input voltage is undisturbed (assuming an ideal source). there will be a small perturbation due to change in current demand from the dc-dc and its load. you could (and should, for reasons mentioned before) put a resistor on that input, and maybe add a capacitor so that brief load transients won't cause the comparator to change state based on spikes
>>
>>1628569
Oh damn I failed to see that the input of the dc-dc converter was connected infront of the opamp's input.
But since my dc-dc converter and battery are one unit I would have to hack it apart sadly. Thanks for the info.
>>
>>1628569
for that money im getting 3 boards and the option to revise my board which I did before but its too late at this point
also i want to support local manufacturers
ill definetely try that
>>
>>1628535
you're fine unless some idiot fucks with your gerbers. i've ordered 4/4 trace/space from houses that list 6/6 minimum. it means you're more likely to get a bad board but that's not really a big deal when you get 3-5 per order.
>>
>>1628535
The usual cheapo china production houses do test for continuity through the tracks so you're probably fine for open circuits, may need to inspect closely for shorts.
>>
File: 1533254068047.jpg (54 KB, 788x699)
54 KB
54 KB JPG
>>1628499
Thanks for the flux info.

>those who work with lead-free have some no-clean options which are not quite as easy to find in hobbyist quantities

Any idea what those are?
>>
>>1628678
I was thinking 100g pots, but it looks like they do 30g syringes now, you serious hobbyist you https://www.kester.com/products/product/rf741-rework-flux
iirc this is the flux you see in action on the jkgamm041 youtube channel
>>
>>1628678
Don't bother friend, I was a lead-free fanboy until I realised basically every product out there was tailored for use with 60/40. The "joint gets weaker if you mix solders" is a meme, and the fumes contain no lead, so you should go to 60/40. Or 67/33 or whatever the cool kids are using. It's cheaper too, since you're not basically buying pure tin.
>>
Im playing around with an ah3503 hall effect sensor. Its supply voltage is supposed to be 4.5-6V but Im powering it off of 3.3v and it seems to be working. Also in the diagram it does not have a resistor anywhere but I remember the sensor heating up before. I put a 1k resistor at the end of the circuit and realized it messed up the output. If I use an internal pullup do I have to have an external resistor.
>>
>>1628716
>basically every product out there
out where, alibay?
>The "joint gets weaker if you mix solders" is a meme
true
>pure tin
3% silver probably doesn't help the price
but really, it's only 100 more °F, git gud

>>1628727
/ohm/ RULE 2
>>
>>1628729
sorry, thought I attached the diagram
>>
>>1628731
The a1321 is a linear sensor just like the a3505 so the wiring should be the same. The right one makes it digital.
>>
>>1628729
>alibay
Yes. Even Rossman reflows motherboards with leaded solder paste, apparently.
>but really, it's only 100 more °F, git gud
Personally I don't find lead-free any more or less difficult to solder, with or without my T12 temperature controlled meme station. My change was solely motivated by the available accessories and their prices.
>>
How the fuck do I get into networks? I'm a ee grad but more focused on hardware and power systems, networks are deep wizzardry for me. I'm thinking about making some sort of mesh network for some sensors with uC and aliexpress radio modules but I do not have a single clue on where to start learning, I can barely configure my router. (I can code alright)
>>
File: 1552420307434.png (50 KB, 255x204)
50 KB
50 KB PNG
>>1628690
Thanks fren.

>>1628716
67/33 is very close to eutectic, so I stick to that.

I only know of one lead-free solder that's eutectic. But irony of ironies, it's 0.15% antimony.

https://www.sparkfun.com/products/10243

( If anyone else sells this stuff I don't know who. )
>>
>>1628811
>But irony of ironies, it's 0.15% antimony
Is that toxic? I know there are some interesting low-melt soldering alloys out there with indium and/or gallium in them, and also I think some specially designed to not dissolve gold contacts/pads. There's a wikipedia page that lists a bunch of different solder alloys (possibly also listed with other alloy types like brasses and such) and also states whether they're eutectic, and I bet at least half the ones I mentioned have semi-common eutectic blends. If not the ordinary 99.3/0.7% RoHS blend.
>>
File: 1548812372228.png (204 KB, 2440x1768)
204 KB
204 KB PNG
>>1628813
Antimony is a poison, yeh. I don't know how much vapor comes off solder containing it, but I do know that a glass of lemonade with 0.013% antimony will send you to the hospital.

Concentrations are much higher in some lead-free plumbing solders: typically 5%.
>>
>>1628815
Well if they use them for plumbing I imagine it's safer than lead.
>>
File: 1542248806859.png (97 KB, 1267x785)
97 KB
97 KB PNG
>>1628821
I have no faith in get-this-horrible-awful-evil-material-out-of-this-industry fads. There are plenty of examples of substitutes being worse in different ways.
>>
>>1628823
Like the production of lithium ion batteries for cars being more environmentally damaging than petrol cars' emissions over their entire lifetimes? I'd tend to agree.
>>
Anyone have that DIY (rookie) project ideas image? I believe there were hundred different ones to choose from
>>
>>1628836
you mean the 00...99 /ohm/ projects list?
>>
what happens if you connect 5v to one of these chink power sources that normally turn 220v into 5v?
(the live and neutral side is not connected to anything)
when using it to power say arduino, and i plug the arduino in via usb, the light on the power source turns on and i am scared that the transformer will fuck my computer up by power surging it or some shit
>>
File: dee.jpg (94 KB, 680x820)
94 KB
94 KB JPG
Tried to post on /adv/, got sent here. So:
I'm a EE major and I want to work on projects related to my major. I'm most interested in the field of Radio Frequency and the lack of intresting projects is killing my motivation. Not only for this major, but the others I've jumped from (Computer Eng, Mechatronics). I don't want to work on something that's at least hardly ever done, I can bring myself to do any of the common projects I've seen online.

Maybe I've set expectations way too high on what my projects should be. I am a undergrad after all. So any help on pegging down my expectations or feed my inspiration with some great project ideas
>>
File: SimplePowerSupply.gif (13 KB, 928x460)
13 KB
13 KB GIF
>>1628872
Do you also expect mains voltage on the other side?
See for yourself why the LED comes up.
>>
>>1628885
Ah, that makes sense, the Dick 1 blocks the current flow so all is gud.
It is a mystery to me through why wouldn't they put the led before the dick, so I wouldn't get a scary 220v power source led on
>>
>>1628888
Interesting auto correction you have there.
>>
>>1628858
yes
>>
>>1628836
pls make free energy device i'm tired of energy bills
>>
File: IMG_20190613_191306.jpg (407 KB, 1536x2048)
407 KB
407 KB JPG
Nice, it's working flawlessly
>>
>>1628888
>literally no knowledge of basic flyback operation
>unironically using it on breadboard shit
I'm even more suprised by your lack of knowledge on what inductor is
>>
File: ohm_projects.png (97 KB, 950x1300)
97 KB
97 KB PNG
>>1628919
voilà
>>
File: 1527965386774.gif (1.96 MB, 480x320)
1.96 MB
1.96 MB GIF
>>1628944
cheers
>>
>>1628944
Rolling on 57
>>
File: 1545396110492.jpg (61 KB, 620x425)
61 KB
61 KB JPG
>>1628811
>antimony
>>
>>1628797
Halp
>>
>>1628797
it's kind of a big department. if I were being even more cheeky than usual I would direct you to something like the US Postal Service's Domestic Mail Manual, but nah
same general idea though: assuming you have a wire over which one or more nodes can send datagrams, it's more or less just names, addresses, envelopes, messages, etc.
what sort of range were you looking for? what level of MCU? have you settled on any radio modules yet? I would suggest the nRF24L01+ modules for maximum ease of use and minimum protocol bullshit
what you will have then is a radio that can send and receive packets of up to 32 bytes at 500kbps on any of 125 channels, unicast or broadcast, with a user-assigned node identifier up to 5 bytes long
then you can define command codes according to the operations that interest you (identify, configure periodic report, set actuator, etc.)
>>1629000
chekd

>>1628882
a bit surprised you couldn't find anything actually interesting in mechatronics, with such a short, sharp feedback path for the designer. unfortunately, not everything that needs to be done is interesting, not even necessarily on the way to an interesting project. maybe you have to do some chores first like reading a few hundred-page datasheets to learn about some component or how to control some IC, or characterizing a component or building block you've not used before

>>1628944
anyone else think this list is due for a refresh?
>>
>>1628998
why isn't she holding the metal part?
fucking amateur
>>
>>1629005
>refresh?
Feel free to update.
>>
File: heh.png (41 KB, 712x312)
41 KB
41 KB PNG
Does anybody know why you would ever want to do this?
>2x part count
>2x cost vs just using bjt for higher power
>doesn't do shit for Power diss
Is the benefit that it's tracking, for multiple supply voltages?
>>
>>1628882
hey homie I'm basically you atm but I'm losing motivation bc there's the blurred line between low-hanging projects which seem easy and a waste of precious time vs. so challenging I lose interest or feel I'm in over my head. v frustrating.

But what subjects are you taking atm? How much time have you spent on projects (and when did u start ee)?

I've been building lab/test equipment bc it not only suits my needs but is only as challenging as you want to make it. If a Lm317 based psu is boring, try lm723 -> op-amp + pass bjt -> bjt diff amp w/ output stages. I hope you see the endless possibilities and opportunities to learn
>>
>>1629016
audiofags probably
>>
>>1629016
probably made a lot of sense back in the days when MOS ICs required triple supplies, could still make some sense in the event you need two rails to be a certain voltage apart, without regard to the distance from ground. e.g. building headroom for a non-RRIO op amp treating an RR signal from the lower set of rails, or a bootstrapping supply for high-side n-type pass transistors

>>1629012
some people would rather smell the antimony braps than hide the pungent aroma under burning chicken
>>
>>1629005
The idea that my school was selling to me is that a Mechatronics student learns CS, EE, ME for the application of creating autos. I don't have much intrest in creating automation, and felt the circullium was setting people up on being masters of none.
>>1629029
>Building lab equipment
That sounds like something I'll be intrested in doing. Will look into it
>>
>>1629016
>multiple supply voltages?
No, clean output.
>>1629034
>audiofags
One example; decoupling of line- and load-regulation.
>>
>>1629051
there ya go! bootstrap your workbench, working toward what does seem interesting by way of the more specific tools needed for it
>masters of none
once you're past that steep part of the learning curve and have that basic, working understanding, you can always pick up the details later if you don't feel like deriving them yourself for the exercise, and/or you can deal more intelligently with component vendors and their engineering staffs
an interdisciplinary approach did nothing wrong, ever
>>
Can anyone give me a transistor circuit for a SR latch that would actually recreate this effect? I've tried it with a few of the NOR gates implemented on falstad (TTL, ECL) and some simple other simple SR circuits, but none of them oscillate. It works with their logic gates though. Is this just an artifact of the abstraction process (aka a bunch of bullshit)?
>>
File: 1557179062179.jpg (17 KB, 474x253)
17 KB
17 KB JPG
>>1629130
real life ≠ vidya. you're seeing an artifact of your simulator's discrete-time paradigm. it's acting as a hidden delay element
with real components, you won't get any oscillation without adding phase shift and gain. the parasitic capacitances etc. in your design aren't enough, aren't predictable, and are more likely to drown out oscillation than sustain it
>>
>>1628885
>implying any commercial converters won't just have a dedicated controller instead of a zener diode
You're probably right that the secondary side will be similar though.

>>1629130
please look into more rigorous oscillator topologies
>>
File: 42c.gif (1.92 MB, 450x251)
1.92 MB
1.92 MB GIF
>Feel like I've finally learned enough about electronics to make my own analog design
>Start working on a simple VCO
>Realize that my design has a flaw when I start implementing it, turning my VCO into a horrible VCA
>That's ok, I'll just make a few changes here real quick
>Realize that my design has just turned into a shittier version of the venerable 555 timer.

WHELP, I guess I'll try again.
Either that or see if I can make some small improvement on the 555 to make it more like the VCO I was trying to make originally.
>>
>>1629144
I'm not following. You're saying the simulator is not mirroring real life but then saying that you don't get oscillations in real life... but that's what I'm getting in the simulator. And is that circuit you posted supposed to be what I asked for? I don't understand where the inputs are.

Pic is one of the circuits I tried that doesn't oscillate. Just copied and pasted the NOR gate example from falstad, don't actually know how they work, but it operates as an SR latch just fine, and when both inputs go from high to low the outputs stay low, no oscillation or unpredictability.
>>
>>1629185
I'm not actually trying to make an oscillator. I just want to know if digital logic is a lie.
>>
>>1629229
Well technically you could make an oscillator based around propagation delays, but I think it would tend towards a steady state unless you used schmitt-trigger gates (i.e. 74HC14 instead of 74HC04), and the propagation delays aren't very reliable.

>>1629212
Are you using JFETs or transconductance amplifiers? Because you should be. Relying on RC circuits is pleb circuitmongering. Capacitors + current sources/sinks is more acceptable.
>>
>>1629234
Cap + Sinking Current Mirror.
I've worked with transconductance amps before, but for this design I'd like to do the whole thing at the transistor level, as a personal challenge.

I haven't quite learned how a transconductance amp works at the transistor level yet.
>>
>>1629185
these days, sure. it's only in the past 5-10 years that controller ICs for switchers have gotten cheap enough to replace self-resonant transistor switchers

>>1629223
your logic sim is lying to you because it forces your design into a synchronous mode in order to simulate it. in essence, that turns your R-S into a J-K with a hidden clock of period 1 turn. your circuit sim is telling you the truth because oscillators like that don't happen irl
a more rigorous logic sim might be annotated with real min/typ/max delays, e.g. Verilog timing annotation
>posted
yes, there are no inputs. you can add another pair of npns across the b-e junctions (with obligatory base resistors) to hold the astable multivibrator in one or the other state

>>1629229
it's a lie, in the sense that all electronics is analog. the branch we call digital just happens to have looser standards on linearity and precision, being only specified at their extremes of input voltage, and as a rule have no ungated feedback path and run open loop (but the CD4069 inverter has been used as an oscillator and an analog building block with feedback more than once, and the CD4066 is famous)

>>1629243
prepare to be demystified
http://www.ti.com/lit/an/sboa117a/sboa117a.pdf
>>
>>1629243
Mind posting your circuit diagram? I'm interested in how you're minimising leakage in your capacitor circuit if nothing else.
>>
>>1627131
>replace batteries on a ups
>*a few minutes later*
>electrical arching and burning electronics smell
>*panic and cursing omitted*
>dissection reveals burnt looking goo from su'scon capacitors

Welp, I need a new UPS. Does anyone know off the top of their heads of any cheap-ish models that use good/Japanese caps (Rubicon, Chemi-con, Nichicon, Panasonic)?
>>
File: MyVCO.png (23 KB, 642x523)
23 KB
23 KB PNG
>>1629255
Sure, here it is.
I'm still early in the re-design, hence the schmitt trigger and op-amp, rather than discrete transistors.
I'm not really doing anything to minimize the leakage, this is a really simple design.
>>
>>1629285
Well using a 10µF cap does proof against leakages for the most part, but I'm concerned about the charge being dumped whenever you short it. More concerning is the arrangement surrounding that 4th NPN, surely a significant amount of current from the cap is being lost through that B-E junction?
>>
>>1629288
Yeah, I need to calculate what the short circuit current is.
Earlier I had a current limiting resistor, I just haven't put it back in yet.

>More concerning is the arrangement surrounding that 4th NPN
The emitter follower?
Input resistance should be about ~1MΩ, since Rin = beta*Re = 100*10,000.
It doesn't draw any appreciable current in the simulator, although I do admit that it is rather useless at this point, since the ideal op-amp and schmitt trigger both draw zero current.
>>
>>1629298
Oh that's reasonable I guess.
I'm guessing that you adjust the resistance of that 1k to change frequency?
>>
>>1629302
I change the voltage going into the 1k to set the frequency, basically a dead simple voltage to current converter.
I'll need to calculate the proper resistor and capacitor values though so I'll have an actual known voltage to frequency relation.
Right now I just have 1k and 10uf because it works and I can test the control circuitry with it as such.
>>
>>1629308
>Right now I just have 1k and 10uf because it works and I can test the control circuitry with it as such
Yes it's definitely easier when you know what to expect. Certainly better than dealing with those time constants from Rc circuits in any case.
I'd be interested to see what happens when you go from a 10µ to a 100n or so; whether you start to see nonlinearities in the charge curve.
>>
>>1629312
Me too, although I feel like that might be a better test for when I actually implement it with real transistors.
Simulators are nice, but this one doesn't even model things like leakage current as far as I can tell.
>>
>>1629315
>this one doesn't even model things like leakage current as far as I can tell
How about swapping to a free spice instead?
>>
>>1629316
I need to learn how to use spice.
At college we were taught how to use PLECs.

I do have LTSpice installed on my computer, but I need to RTFM.
>>
>>1629320
Is it a mac? Because if so you'll have to do everything with hotkeys or the right-click menu because they decided that a toolbar was too difficult to code.
All you really need to know for basic operation is:
>ensure you have a ground node somewhere on your circuit
>all basic parts must have a value
>part values can be written with standard notation like 4.3u, 10meg, 5m3, etc. (meg = 10^6, M = m = 10^-3, because they aren't case sensitive)
>all transistors and such should have a model name selected (diodes don't really matter as much as transistors)
>use ".tran 0 50m" to simulate for 50ms
>use "Pulse" voltage sources in order to have voltage sweeps (i.e. from 3V to 5V)
>feel free to change the names of components from "C2" to "charge_cap" or whatever you fancy
>
>>
>>1629332
Ah, thanks.
I am using an old mac. I though the interface seemed oddly barebones.
>>
I'm a total noob, is it possible to control a 40W RGB LED panel with arduino? It comes with a remote but I can't use it for what I wanna do (automatic f.lux-like behaviour)
>>
>>1629394
Definitely possible. How many wires are going from the receiver to the LED panel? If it's only 4 then it's an easy task, requiring only PSU hardware, 3 correctly rated MOSFETs, and possibly a few passives.

Also check whether the default controller is constant voltage or constant current, because if it's the latter you may run into some issues with PWMing it and may want to current-dim it instead if the power supply allows for it.
>>
>>1629394
>RGB
I don't think you're going to be very happy with the quality of light out of that. you are still going to have blue wavelengths in the light unless it's pure yellow. consider getting a warm white+cool white panel (or build one from convenient components on ali)
>>
>>1629399
It's not that bad, I just stick to colours without any blue in them. Mainly yellows and oranges. The poor colour rendition is certainly noticeable, but not annoying. Ideally I'd have 6000K-2000K dimmable white, possibly with 500-1500K or red also, but 2000Ks and colder are rare.
>>
>>1629397
I don't have the panel yet, but in this video I can see 6 wires https://youtu.be/2Dom80vEEas?t=29

>>1629399
I know that the quality sucks, but the "warm white" is usually not less than 2700K, which is still too cold for me after 9pm

>>1629400
What's your setup like?
>>
>>1629421
>6 wires
>RGBWW
That will be either 1 common cathode and an anode for each colour channel, or 1 common anode and a cathode for each colour channel. Since it has V+ in the video, the 5 channels will be controllable with N-channel MOSFETs. Someone else may suggest good FETs, otherwise I'd just search for a common voltage range (35V?) on ali or wherever and make sure it can handle the current. Should be easy enough to drive. I'd also grab a buck converter with which to power your arduino (assuming it won't be connected to a running computer all the time). I don't think you'll need gate resistors for the FETs.
Do you have a purchase link?

>What's your setup like?
Pitifully weak 5V 1.8A USB LED strip draped over my oscilloscope. It's the model with the RF remote+receiver, which occasionally is triggered by something other than my remote, which is a pain. The red is also way too weak, so the yellow looks like a lime and the orange looks like a yellow, and the remote just has those presets so I can't go any more orange.
But I'm slowly working on a more powerful RGB setup with a driverless RGB COB, which may or may not be unusable thanks to the 100Hz flicker. The external driving circuitry required is also stupidly complex, so I don't recommend purchasing one. But it will have nice adjustment knobs and an on-off remote, neons on the front, and a stylish laser-cut acrylic case. I'd be working on it now if the maker space wasn't closed for a few weeks and I didn't have an exam in 9 hours.
>>
>>1629077
>bootstrap your workbench
Is this a reasonable goal? Do people often use DIY equipment instead of buying their instruments in EE? Not him, but I don't want to get my hopes up.
>>
>>1629430
Thanks for the advice. This is the panel i was looking at: https://www.ledpanelwholesale.co.uk/led-panel-60-60-rgbww

>Pitifully weak 5V 1.8A USB LED strip draped over my oscilloscope
what do you use it for, is it just for decoration? And what are you gonna do with the COB, isn't that stuff used in live performances?
>>
>>1629432
People often make their own PSUs since they can use standard ICs or even pre-built modules. ±0.1V accuracy is pretty easy to obtain with an MCU ADC. Making one with a rotary switch or two and an old multi-multi-tap transformer (e.g. from a tube tester) is also somewhat popular. Simple function generators are also somewhat easy to make with a microcontroller driving a display of some kind with a rotary encoder and a few buttons. Scopes and more advanced function generators are less likely to be plausible to make, if only because of the fine temporal resolution required.
RCL meters are a neat project as long as they're not dealing with mΩ, nH, pF, etc. There are various degrees of depth you can go when characterising an inductor (ESR, parallel capacitance, core effects) but that's really just a DSP/software problem. Your accuracy comes down to what tolerance parts you use

>>1629461
Both used for lighting my bedroom.
>65 eurobux
That's a bit pricy. I wonder if replacement LCD backlights can put out enough power to be used?
>>
>>1629464
>replacement LCD backlights
Scratch that, they don't come as panels, just as the edge-lighting LED strips. You'd need a translucent panel to shine them through, which arguably isn't a difficult task. A piece of acrylic plastic sanded on one side with foil on the other and LED strips around the perimeter would do a decent job for under half the price.
>>
>>1629464
>That's a bit pricy
I can't find anything cheaper that delivers in EU
>>
File: IMG_20190614_185901.jpg (296 KB, 1536x2048)
296 KB
296 KB JPG
>>1628929
Finally finished.
A fancy smart plug which i can easily turn on or off from anywhere in the world through my own api.
It was really easy to hook it up to my google assistant too so i can say shit like "hey google turn the fan on" and in couple of seconds it clicks (and i obviously plugged the desk fan into it)
I am very happy with the result
And its size is not that much bigger than commercial plugs.
>>
File: 1546378057703.jpg (192 KB, 1356x827)
192 KB
192 KB JPG
I'm looking at buying a red pitaya as a general electronics fuck around device. I want something cheap and compact with a lot of features because I'm an apartment bitch and I don't have any space.
Any better ideas? EEVBlog shat on the device 3 years ago and the default web interface looks fucking terrible so I'm skeptical about it.
>>
>>1629577
Good work!
Now I understand why type F plugs have this weird hole.
>>
>>1629617
>weird hole.
what hole? you mean the ground pin?
>>
>>1629472
>I have junk
20W (nominal) light engines can be gotten for like a buck each on ali, in your chioce of color or white
>>
>>1629686
Yes, (PE) pin in the socket, hole in the plug.
>>
>>1629691
>20W light engines
Not getting any concrete results from that search.
I think I'd prefer LED strips since they don't require a heat sink and have a good variety of styles and colours, but I am somewhat open to suggestion.
>>
>>1629727
there exist light strips with warm (like 3200K?) and cold LEDs, in the standard 4-pin configuration. if you wanted to you could augment with some monocolor strips in red, orange, and yellow
for engines you could try "hong li cob" and see what you find. you might be able to get away with simply thermal-gluing them to a metal sheet or plate if you're not pushing them to horticultural power levels
>>
>>1629767
You can also get strips with RGB+warm+cold white all together.

For my desk lighting I use a strip of the WS2812 clones with RGB+warm white, it's pretty nice, the white is pretty warm, somewhere around 3000K, and full white + full blue can get it to a real bright ~6500K
>>
>>1629767
>hong li cob
Woah those are cheap. What are they used for?
The same seller also sells ~130V and ~180V models, which look to be for fairly simple step-down converters running off rectified mains. Which could make a fun project, making a variable current-limited mains buck converter.
>4-pin
Don't you mean 3 pin? I found a couple of sellers selling single-color 2000K or 2500K LED strips, in particular this one:
https://www.aliexpress.com/item/strip/32858633850.html
Which I might go for, but they will require a power supply of some sort. And considering the power selection is "14W" and nothing else, regardless of how many metres, and there aren't any images of the strip face itself, it's hard to know what to go for. The description page says "14W/m" and "cut size 5cm", but also "120LEDs per metre" and "120 light beads per m" on two seperate occasions, with lengths of 1-5m. So I guess I'll want a power supply in the realm of 24V 3A and some sort of dimming circuit, since I can't seem to find any appropriate constant-current-dimming power supplies. Multiplied by two, since I'd want the 6500K and the 2000K.

>>1629863
I don't exactly need
>>
>>1629876
light fixtures, etc.
>3-pin
4. they use the same tape as the RGB strips but mount W+CW LEDs on them. one of the pins is unused
https://www.aliexpress.com/item/High-quality-24V-benis/32450008748.html
>cut size
is probably the distance between cut lines and the size of each parallel cell. at 120LEDs/m there are 6 per cell, which makes sense for a 24V strip
>constant-current
I have yet to see a strip designed for constant-current. they tend to be designed for constant-voltage for ease of use (e.g. parallel connections, trimming to fit, no need for a CC driver). just add a few kHz of PWM switch and enjoy
>>
>>1629893
>at 120LEDs/m there are 6 per cell, which makes sense for a 24V strip
Oh that's actually a perfectly reasonable number of LEDs. I figured it was far too many, but with 6 per 5cm segment that's quite plausible.
>I have yet to see a strip designed for constant-current
It wasn't that the strip had to be designed for constant current at all, but rather that it would be easier to control its brightness by varying the current than by varying the voltage, thanks to the Shockley diode equation. I'm unsure of what happens in this middle-ground between PTC and NTC characteristics, but would a strip with both PTC resistors and NTC LEDs be able to handle CC and CV operation without thermal runaway? You're already dropping 3/4 of the voltage across the LEDs but they run just fine on CV supplies.

I also wonder about the COBs with no apparent current limiting/PTC features. Are they meant to have some sort of CC power supply? What about the automotive ones? The battery voltage fluctuates far too much to imagine them working with LEDs.
>>
>>1629876
>variable current-limited mains buck converter.
could be exciting, take that as you will
an isolated converter would be a bit of trouble to put together (this is about 7 projects down my get-to list) but a buck would be ezpz once you've found the right converter IC (the HV9910 is a good choice, bring your own high-voltage MOSFET to reach current levels that shouldn't even be possible)

>>1629895
>had to be designed
doesn't have to be, but (for efficiency) it does help
>PTC resistors
sounds dicey, not least due to thermal matching, PTC dissipation and how bou dat inrush current on cold start
>no current limiting
COBs do generally depend on a CC driver
>automotive
they probably just run them over spec like low-tier household LED lamps. B+ has maybe 2.5V of uncertainty, depending on how shot your battery is
that said, there are two-terminal current regulators in SOD packages, from ON Semi and others, that can simply be placed in series with an LED string connected to a voltage source with a little bit of headroom (too lazy to look up right now)
>>
>>1629981
>the right converter IC
I was considering going for a shitty comparator-based bang-bang circuit, but they're not too good when you want them to be adjustable as they kinda need tuning for the frequency, duty-cycle, output voltage, etc. Are there not flyback ICs with provisions for a variable output? Or is it the transformer with feedback windings that you're battling with?

>>PTC resistors
I was just meaning normal carbon-film resistors. They have a slight positive temperature coefficient that prevents the LEDs from going into thermal runaway. Though perhaps this would happen anyway with a constant resistance.

>two-terminal current regulators
Apparently you can just use a JFET as a constant-current diode. Not that I think they did that, it's probably closer to an LM317 with the sense resistance chosen to give the desired current.
>>
>>1629876
see also https://www.aliexpress.com/item/SZYOUMY-24-Key-IR-Remote-Dimmer-DC12-24V-Dual-White-CCT-Color-Temperature-Controller-With-FOR/32879494805.html which you can augment with transistors of your choice so you can PWM an entire block of flats
>>
>>1629984
I don't like quantised color choices, I'd prefer a custom remote with knobs on it. Custom because they don't exist. 256 different PWM values would likely suffice, though I may want to split them up logarithmically. Ideally I'd add inertia to the rotary encoders to get fine and coarse resolution depending on the speed at which I turn them, but I could also use the push-in-centre encoders with the button=not in telling it to count more quickly.
>>
File: 1552444146147.png (453 KB, 2124x1375)
453 KB
453 KB PNG
>>1629983
you could probably make a hysteretic approach work, if you're willing to accept a fairly high minimum output current
some flyback ICs use a bang-bang approach, using a transistor optocoupler and a TL431 or similar to pass a voltage-too-high signal across the isolation barrier. if driving a high current, you might need some extra circuitry on the secondary side to amplify the current sense resistor, in order to keep its dissipation low (monolithic drivers use reference voltages on the order of 200mV), or you use a feedback winding
sadly, switching power supply transformers are, for the most part, not off-the-shelf parts
>slight positive TC
look up the curves and see if they match
>Not that I think they did that
looks like not much more than a self-biased transistor. Pic related

>>1629990
>knobs
>porn music starts
capacitive touch sliders?
>porn music intensifies
if size doesn't matter
>porn music has escalated
you could go all out and get some high-end rotary encoders like the chinky $12 ones (cheaper than $46 compact multi-turn panel pots). put some modern weighted metal knobs on there and there's your inertia
>split them up logarithmically
2.0 gamma is computationally easy: just square the requested intensity. it's not precise but it's better than not. if you have a bit more than 8 bits on your PWM, you're set
might be fun to learn your way around the special features of the STM32 timers, like the quadrature decoders, DMA for making remote control protocol bits, etc.
>>
>>1630009
Op-amps on the secondary to amplify and schmitt trigger the signal (with low hysteresis), feeding an opto that feeds the oscillator enable? With step-down converters it might be easier to have a significant amount of your peripheral circuitry on the secondary, at least considering the required low-power linear dropper needed to feed the active components.

>you could go all out and get some high-end rotary encoders
>weighted knobs
Size would be the biggest constraint. Ideally I'd go for some thin multipole magnetic discs in laser-cut acrylic with a bushing or two and a hall sensor nearby for a 3mm-thick encoder, but that's likely a dream. I saw some of the single-dimmer remotes (either warm/cool or light/dark) have something like a capacitive dial, but I wouldn't trust its resolution.
As far as weighted knobs, I'd prefer the software inertia to such a clunky hardware implementation. The transmitting micro could fairly simply measure the direction and frequency of the turns, and square the frequency, before sending over data in the form of [bool=encoder, bool=direction, int=number-of-steps] every 10ms or so. Probably via RF, since the PWMing lights could interfere with the IR operation.
A standard arduino has only 8 bits of resolution in its analogwrite function, but I'd likely be able to custom bit-bang that up to at least 10 bits at a reasonable frequency. With some sort of non-volatile memory to store brightnesses between powerdowns.
>>
File: 20190528_185209.png (3.11 MB, 4392x3124)
3.11 MB
3.11 MB PNG
Inductor are more complex than I thought
>>
>>1630018
that PWM waveform had better not be the sine output of your function generator
>>
>>1630019
Hysteresis current control
>>
>>1630022
Well regardless it's those high-frequency components that will be going straight through that low-freq ferrite.
>>
>>1630040
Yes, it's all about core material
>>
>>1629983
>Are there not flyback ICs with provisions for a variable output?
There are, Innoswitch3 comes to mind (other manufacturers should have equivalent as well)

For cheapass version I had some success with simple flyback converters with optocoupler control where you design transformer for maximum output power and via optocouler on-time bring voltage down.

Fastest way is to use dc/dc converters after flyback
>>
>>1629608
>125MS/s
Does it have a single-channel mode where it can do more than that? Because that's pretty low otherwise.
>>
>>1630125
>Fastest way is to use dc/dc converters after flyback
Yeah I'm getting that feeling.
Is it even possible to use a feedback winding to provide current feedback?
>>
File: led.jpg (115 KB, 656x472)
115 KB
115 KB JPG
Hey /ohm/, got these LED strings from a friend who wanted to throw them out because the battery compartment broke, so I intended to just wire them to a new one.

But looking at them, this is just exposed copper wire running electricity through them. I have very little understanding in actual electronics - how do people not get electrocuted touching them when they're plugged in, closing a circuit with their finger by touching both wires? Is there a resistor of some kind near the batteries to lower the voltage significantly? Am I endangering myself by just connecting them to a live AA battery?
>>
>>1630146
Also I realize I sound dumb and ignorant, because I am, which is why I'm asking.
>>
>>1630136
partially
As load increases on secondary, feedback winding will also generate higher voltage
After than I have no idea if it's possible to get accurate estimations on actual current going through (especially with variable output voltage)
>>
>>1630149
How about a shitty parallel ADC feeding an array of optocouplers, which feed an R2R ladder "DAC"? Ideally your ADC would be passive like the R2R, but I'm unsure how you'd get that to work without some sort of array of comparators and reference voltages.
Doing it serial would also be an option I guess. Just have an MCU on both sides of the circuit, use one for the ADC to measure the current, and the other to receive the data and modulate the frequency and duty-cycle as it sees fit. That wouldn't be a terribly difficult circuit either, though a not terribly efficient one as far as power consumption and BOM cost goes. But I would like to be able to tweak the CC and CV characteristics and modulation methods purely via software. It would be dead easy to interface with a digital remote also.
>>
File: tegaki.png (11 KB, 400x400)
11 KB
11 KB PNG
>>1630156
Actually, just having the main controller MCU on the secondary side that directly pulses the main FET with the optocoupler could work, provided you get your bootstrap MCU power supply right and optocouplers could be made to fire fast enough. Like pic related but probably with a totem-pole between the opto+pulldown and the main FET.
>>
>>1630156
>How about a shitty parallel ADC feeding an array
Needlessly complicated

>>1630163 example would be preferred way
But generally I would still prefer DC/DC on secondary side, especially with CV/CC, currents sensing and other requirements.

If you insist on having control over mains switching - I suggest reading ST AN4930 for reference
>>
>>1630178
I just figure control over mains switching would be more energy efficient and less noisy, and would also be a neat project for project's sake. The sort of thing I might modify into a variable benchtop power supply with the help of a linear regulator of some kind for ripple rejection.
Though perhaps feedback via an MCU would be too slow for a case with a highly dynamic load.

You're certainly right that stacking two converters would be ideal as far as cost and simplicity goes.
>>
>>1630146
I will give you a fast answer.
First, the copper is very likely to be coated with resin or plastic. So it wont conduct or short circuit when tangled.

Second, if it's a small light battery powered, you have no chance of being electrocuted. Your skin protects you against electricity in the following way :
Shortly put, DC burns, AC contracts your muscles against your will.

At low voltages (<24-50V) Voltage is too weak to go though your skin.
At low current (depending on voltage, <10mA for AC, 2A for DC 3-4W I guess?), even if the electricity flows through you, you dissipate the heat, and "everything is fine" (see capacitive touch).

If there is enough voltage, it will go through your skin and will "command" your muscle against your will". Funny to move your legs to exercise, lethal if the current decide to control your heart.
If there is enough current, an arc can form and burn you flesh (don't look picture of electrical burn by the way. Unless you don't want to sleep that night).
If both, you are screwed even before touching the wire.

An electric car is battery powered, and is fucking dangerous (battery wise) because of high voltage and current, even if DC.
Your house live is dangerous because it's high voltage AC.
Your car battery (to start your thermic engine) is mostly not dangerous because it's low voltage. Don't touch it with wet hand tho, it have lot of current and can burn (as in destroy) your hand.
Your cell phone battery can still burn you. But it have a protection circuit witch will prevent it.
An AA battery is tottaly safe (electricity wise. I'm not responsible if you shove it elsewhere).

A funcky urban says (legend or not, I don't know) is that you can die if you touch a 9V battery with your tong.
It's a really particular case of your tong being wet and full of nervous terminals and close to the brain, and a voltage that is sufficient to pierce your skin in these conditions. that seem plausible to me, but I have nothing to back it up.
>>
File: LEDpcb.jpg (350 KB, 1024x768)
350 KB
350 KB JPG
Im new, and I want to surface mount LEDs for flashlights onto MCPCBs.
Ive looked up surface mount soldering, but from what im reading bare LEDS are really really frail and finicky with heat. They have domes on top of the LED that can get fucked up with heat too.

Pic related is someone elses picture of the MCPCB I want to swap an LED on, its from a particular flashlight.

What would be a safe way to do it, that isnt prohibitively expensive.
>>
>>1630210
Do it by hand if you know how to solder :
Apply flux to PCB and LED.
Put solder on one AND ONLY ONE PCB's pad.
While gently massaging your solder with your iron to keep it fluid but not to hot, place the LED with tweezers.
If you can't have it placed correctly in less than two second, remove it, wait a bit and repeat.
As soon as the LED is placed, you want to remove your iron while maintaining your LED in place.
Inspect your one leg soldered LED and maybe try it.
If everything is fine and well aligned, just solder the remaining legs, one at a time leaving pause in-between.

Do it with a hot air gun if you are crazy :
Apply flux on both LED and PCB
Apply solder paste on the PCB
If needed, apply glue to the LED
Put in place the LED onto the PCB
GENTLY blow your hot air gun drawing circles atop of your PCB. (remember to fix your PCB with klapton tape or any other method of it not to fly.
You can use tweezers if the airflow is too heavy to maintain the LED.

Do it with a plancha if you already have on and know how to do it :
Apply flux on both LED and PCB
Apply solder paste on the PCB
If needed, apply glue to the LED
Put in place the LED onto the PCB
Turn on the plancha and wait for it to get hot.
Turn it off when the solder starts to meld, with thermal inertia, it will be ok.

In any cases, solder more than needed.
The newer to solder, the more you want to solder (or try to solder).
Like, buy a bag of 5 PCB and 5 LED if you ned 1.
Good luck, and learn well.
>>
File: make-cplatt.png (234 KB, 680x1300)
234 KB
234 KB PNG
>>1630148
>dumb and ignorant..
>>1630193
>you can die..
From Make-Electronics (free excerpt)
Motto: "... I want you to burn things out and mess things up, because this is how you learn the limits of components and materials."
>>
>>1630242
As soon as you burn components, I'm more than okay with your motto.
But hurting itself for life is not a good idea.
Shock yourself as much as you want, it's a good way to learn. But shock you on areas that are non lethal and with currents and voltages that wont leave permanent damages on yourself.
I shock myself with a 320V fully charged capacitor 100µF (flash lamp one) and an other time a 4000V 100nF across the heart, and I can tell you, it's very unpleasant, and I have to sit for 5 minutes to feel "okay, I can resume my task now".
If I were to be unlucky this very same day, I would just have died.
So, shock yourself on hands, or legs (circuits do happens to fall from the bench sometime).
But don't shock you near head or across heart.

By the way, if you shock yourself, remember to drink water afterwards. The current ionize your cells, and you have to fill them up.
>>
Searching for milifarad caps on ali is so annoids, all listings come out a microfards since mf and mf
>>
This must be a scam right? no way they sell 40f cap for $1.5
https://www.aliexpress.com/item/2PCS-super-scam-benis-capacitor-farad-capacitor-2-7V-2F-3-3F-4F-4-7F-5F-6F-7F/33014262131.html
>>
>>1630156
optocouplers are analog (current transfer ratio is an important figure of merit) and monotonic, just not necessarily very linear. only one is required

>>1630210
bottom heat (hot plate, air, free choice) AND soldering iron

>>1630178
>Needlessly complicated
it's that one guy's trademark
>>
File: 1560605650129.jpg (202 KB, 574x800)
202 KB
202 KB JPG
>>1630328
weirdly, LCSC doesn't appear to offer supercapacitors and ultracapacitors. no benchmark to be had there. I still don't see that price as completely out-of-sorts (for a 2.7V device)
order one. be the dave for us and test actual capacity. then open it up and show us the gigantic MLCC inside
>>
File: thing.jpg (47 KB, 195x427)
47 KB
47 KB JPG
>>1630193
Firstly. this was a thorough, exhaustive, base-level explanation in lamen's terms that I did not expect to get. Whether it's instant paste from some digital text is irelevant, you have my thanks.

Secondly, I don't think the copper is coated at all. Sure, I won't plug it into a wall socket and close the terminals with my tongue, I just want to connect it to a couple batteries; but what about other things that touch it and might be moist - curtains, bedsheets, animal fur? If they conducted the current from the (possibly) exposed wires, would that be enough to start a fire or otherwise be dangerous?

Are simple batteries DC?

There's some component in the battery box between the switch and the wires (pic related) - what could it be, and is it needed? (I assumed it's a resistor)

Despite the recent /pol/ invastion, /diy/ is still the best board by far.
>>
>>1630447

you're the worst scaredy-cat i've ever read. low-voltage DC circuits (especially ones with little batteries) are completely safe to play with. you dont have to worry about shocks until you're in the 50+ volt range, and even then it's just a tickle. you can even short small batteries (AAA to D size) and they'll just get a bit hot.

>I don't think the copper is coated at all.

thats crazy talk. i guarantee it is. else everything would be shorted together and nothing would work. it probably uses a very thin enamel coating.

> (I assumed it's a resistor)

looks like a 5.1 ohm resistor to limit current.
>>
>>1630506
>you're the worst scaredy-cat i've ever read
Well, which do you prefer - someone too scared to connect a simple battery to an exposed wire, or someone hasty and arrogant they don't give a shit and kill themselves? regardless, I'll give it some current and see how it works. Thanks again.
>>
File: PoorCad.png (8 KB, 923x466)
8 KB
8 KB PNG
What would be the best way to do this while not taking up a lot of board space? There also isn't a lot of current involved, mainly for connecting to industrial sensors and the like.
>>
>>1630447
Thx m8.
No, it's not a pasta, so happy to help.

As for wires, >>1630506 is right, if it runs on batteries, not only it's completely safe, but it is obviously coated (just try to put new batteries and when everything lights up, put a paper clip between the wires for a short time. If light goes out, it was bare if not, LEDs will still light up.
You only have two wires here, so the only possible way is that LED are mounted in parallel (meaning each (+) connected together and each (-) connected together). Let's simplify and say that it will act as a single 'big' LED.

The law you are looking for here is Ohm's law : U=RI
or Voltage = Resistance * Current
It only works for DC (Direct Current, meaning "stable" current, as in batteries and most supply < 12V)
Lucky for you, I think the module works on 3 AAA batteries.
So DC current @ 3*1.5V = 4.5V
[spoiler]FYI batteries don't delivers 1.5V, but instead from like 1.7V when full to 1V when flat depending on chemistry. But here, we don't care, let's keep it simple.[/spoiler]
So, U=4.5 V or is it?
Nope.
The LED will need some voltage to light up. It will depend on the kind of LED.[spoiler] Red are the oldest and lest demanding of visible LED and can need as low as 1.7V. Blue (or purple) the most demanding at 4.5V. [/spoiler]Here, you have white ones, so we can take 3.6V as a wild guess.
So U = 4.5-3.6 = 0.9V

Okay, great, you spot the resistor, and a nice Anon tells you the value : 5.1 ohm[spoiler]
Green : 5, brown : 1, gold(when on the third ring) : *10^-1 gold(when on the fourth ring) : +-10%
so, the resistor is 51*10^-1 or 5.1 ohm +-10% JFYI[/spoiler]

U = R * I
so U / R = I
So, here 0.9 (Volts) / 5.1 (ohm) = 0.176 (Amperes) = 176mA (milli Amperes)

Great. Now, count the LED, if you have 25 in your strip, each one will have 1/25 of the calculated current. So 7mA.
As most of the time these kind of LED can take up to 10-20mA, you are good.

If you want to improve your skills, do the math for a 5V supply (a USB port).
>>
>>1630525
A step-up converter of course.
https://www.ebay.com/itm/MT3608-DC-DC-Board-2A-Step-Up-Converter-2V-24V-To-5V-9V-12V-28V-Boost-Module/123540180282
Beware, it will very likely induce noise on your power rail.

>>1630539
Aaand I shat myself up, no spoiler on /diy...
>>
>>1630544
I think you're misunderstanding the issue.
I'm not trying to power anything just read the state of say a switch but it needs to be flexible enough to work with multiple voltages.
>>
>>1630547
Nigga, you talk about "5V, 12V and 24V I/O" and "industrial sensors".
Thing like presence sensor or other delivering a 24V signal (analog or digital) needs a 24V supply.

If you wanted to read any entry from any kind of voltage, you should have say so directly.

If you just want to read a switch, why in hell do you want to use it with 24V?

If it's some kind of presence sensor, most of the time you can choose an open collector style, so don't care about the voltage.

If really you need to swap sensors using different voltage, use a phototransistor and a diode in the sensor side.
This way, no matter your sensor, you have something isolated.
>>
>>1630193
> A funcky urban says (legend or not, I don't know) is that you can die if you touch a 9V battery with your tong.
If that was true, I'd have been dead by the age of five. 9V is enough that you can actually feel the current when applied to the tongue (but not to dry skin; you need around 50V for that).

>>1630447
> Secondly, I don't think the copper is coated at all.
It's coated. Copper wire used as an electrical conductor always has some form of insulation. Copper "cable" (single-core or stranded) is surrounded by a tube of plastic, "bare" copper wire (used for inductors, transformers, solenoids, etc where you need many turns) is enamelled.
>>
>>1630193
>he doesnt check to see if a 9v battery is still good with his tongue
>>
>>1630626
do you kiss your no gf with that mouth?
>>
File: NC-559 vs THE STUFF.jpg (428 KB, 1000x573)
428 KB
428 KB JPG
>>1628457
>>1628468
>>1628499
Hey anons, I just got my Amtech goo today. It's not the secret sauce my pal shared with me, but it IS pretty amazing stuff. I'm very happy with it and just wanted to thank you for telling me what "the good flux" is.

After him not telling me, and what seemed like deliberate obfuscation of flux brands by youtube rework techs, I feel like I've been let in on a wizard secret.
>>
>>1630670
open board
show vajoints
>>
Im taking my electrical journeyman test for Miami Dade County, anyone one have any tips?
>>
>>1630679
electronics is about combining electron devices that generate and process signals and power according to universal physical laws
sparkology is about knowing applicable code for your jurisdiction. start with RULE 0 in the OP
>>
File: Gieger.jpg (1.01 MB, 1385x1000)
1.01 MB
1.01 MB JPG
>>1630673
>>1630677
I'm just using it to de-solder a bunch of components right now. Don't have a board to work on atm; just a bunch of shit to repair.

This is what my joints look like when using rosin poo.
>>
Since a capacitor acts as basically a short at the start should i worry about my wires being melted before it chargers?
I have 0.1F cap across my power source and the wires are rated for about .7A tops, do i need to put some resistor there or somewhere?
I know you don't have to for small caps like 1uf since they saturate way too fast, but big caps like mine will draw current for a while before they are fully fed
>>
File: doit.png (16 KB, 247x242)
16 KB
16 KB PNG
>>1630737
>I have 0.1F cap across my power source
and did anything smoke?
>>
>>1630737
oof...
https://en.wikipedia.org/wiki/Decoupling_capacitor
>>
>>1630512
I'll take the hasty arrogant asshole since they won't live long enough to go on /diy/ and post stupid questions.
>>
File: capacitordiagram.png (11 KB, 448x306)
11 KB
11 KB PNG
>>1630737
Does this look something that shorts anything?
(dielectric doesn't conduct electricity)
>>
>>1630757
>>
>>1630757

t.never spent one hour in electronics class
>>
Does the marking LS on a SMD diode in any way, shape or form indicate the type of diode?
>>
>>1630829
Of course. Why do you think diodes are marked at all?
>>
File: 20190616_143808.jpg (1.98 MB, 3024x3483)
1.98 MB
1.98 MB JPG
Bought this old portable refrigerator and it worked before I repainted it.
No clue whats wrong with it, converter makes a buzzing sound but no cold and when I screw down the red lid it shorts to ground.
>>
File: 20190616_172101.jpg (3.73 MB, 4032x3024)
3.73 MB
3.73 MB JPG
>>1630843
Blue wire with plastic cover shorts when connected to ground but connecting where it should go, temp control gauge, it does nothing. Same buzzing sound from converter connected or not.
>>
>>1630844
>>1630843
where does your "blue wire with plastic cover" comes from?
Do you have multimeter to measure voltage/resistance with?
>>
File: 20190616_171507.jpg (1.6 MB, 1878x4032)
1.6 MB
1.6 MB JPG
>>1630982
Down to bottom of converter with rest of the wires. Took some damn poor photos, thought they were better.
I do have meter but I don't know enough to understand the results. What I know its getting power alright.
Should mention is dual system 12 and 230v. Currently using 12v with another converter because 230v cable is missing.
>>
>>1630993
I'll make a electric circuit sketch later to better understand the pics.
>>
>>1630539
While I appreciate the sentiment behind all this non-spoilered information, this is well beyond my knowledge, understanding or necessity; any chance of me remembering any of that 5 minutes from now was out the window as soon as you started putting in 8th grade level equations, kek. I even failed to follow what it was you were trying to demonstrate - the leftover current after all the LED's are lit? I'm a sysadmin and these ideas and concepts are so foreign to my world I only know of them vaguely from old kids' books I read.

I still don't understand though - if I connect them directly to 3 AAA batteries without the resistor, could they stand the current or would it fry them? (or is it the high voltage that's dangerous to electronics?)
>>
>>1630998
>or is it the high voltage that's dangerous to electronics?

When in doubt, fall back on the hydraulic analogy. Voltage is water pressure. Current is flow rate. Resistance is the restriction of flow imposed on the water by limitations of the components.

Wire = pipe
Resistor = a constriction valve along the length of the pipe. Some can be adjusted, most are fixed.
Load = a widget that uses hydro power to do work

Your LED is a load. For sake of argument think of it as a little fish tank pump that uses household water pressure to push other fluids around

The danger is the combination of voltage/pressure and current/flow. The POWER aka the amperage consumed by a circuit is what has the potential to fry components. Your little LED/pump can be overpressured and break, unless you limit the flow and pressure of water going into it. That's what the resistor is for. If you just connect it up to your main inlet pipe it's going to crack open.

You need a resistor that will adapt the pressure and flow into your LED. You choose the resistor based on the source pressure (voltage) and the power rating of your LED in milliamps and volts.

How do you figure that out? Ohm's Law and the Power Law. See anon's post you skipped over.



Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.