/diy/ - Do It Yourself

>>1571355
goddamnit you shitnigger there were still a good couple of days left on the old bread

>>1571355
this thread's digits brought to you by the LT1355 dual low power high speed op amp

>>1571372
>12MHz
ooh nice

so I got really drunk last night, went on ebay, and inadvertently bid on an oscilloscope

consequently I now own an oscilloscope
I am also now forever banning myself from ebay when drunk

>>1571380
congrats, it's for your own good

>>1571397
I'm not up to speed 100% on various scopes, but a 4 channel, 300 MHz fully functional tektronix scope seems pretty alright for $250, especially since I was absolutely shitfaced

how did my drunk ass do

>>1571380
i mean uh...for an analog scope...it pretty good
its not a 2465B?

>>1571411
it's missing the name plate so I won't know what exact model it is until it gets here

>>1571413
the toggles (mode/source/coupling) etc pretty sure on the B they are buttons instead of toggles.

>>1571440
in that case I have a plain 2465, no A or B

>>1571404
if you don't need digital, pretty gud

Ok, I have a mains device running off its own bridge rectifier and a PWM board connected to it. Because it's running off a rectifier, the -ve voltage rail isn't 0V with respect to ground, so when I touch the potentiometers it parasitically couples to me from ground and pulls the FET gates high enough to turn on slightly. How should I avoid this? Tying them to ground would have the same effect, but tying them to the -ve rail mightn't be comfortable to touch since somewhat of a current would flow through me to ground. Is tying them to negative and putting plastic knobs on them the only option?

>>1571575
Uh, a rectifier references to ground. You've got another problem.

>>1571589
It's a full bridge, on the negative side of the waveform the negative rail is goes below the voltage of the neutral line. Pic related.

>>1571590
oh it actually goes above, same difference

>>1571590
>>1571591
shit, it is below because I mixed the negative and positive voltage rails around.

>>1571404
no but it is the right option

I have this power supply from chinks.
It converts 230AC to 5V DC
But it can only put out about 700ma and i need around one amp.
Is it safe to connect two of these in parallel?

>>1571575
OK good news, grounding the aluminium PCB backing didn't cause the flickering, but tying the pots to ground has the same effect as touching it with my hands, which confirms that coupling to ground is the issue.

>>1571627
Possibly, but I wouldn't personally, you'd probably end up with one flyback module doing all the work. You could go through and replace parts with ones with better current capacities, or put heat-sinks on the FET and diode, but I'd just buy one rated for the current you want, it will probably be cheaper than 2 of those anyhow.

>>1571628
I can't find any small power sources like that on ali which convert 230AC to 5V dc and can do more than one amp.

In that range every they sell are regular phone charger adapters with a cable, which yeah i could use, but i can't easily fit them into my 3d pritned case without taking them apart and modding them and stuff

>>1571629
How about this one?
https://www.aliexpress.com/item//32899751766.html
It's a bit bulky but it looks like the best you're going to get. Looks like it has much better noise suppression than the 0.7A one too.

>>1571630
Looks decent. But 1A is pretty borderline minimum i need and chinks like to exaggerate. Guess i could try to order one and see if it works out or not.
How do those connection holes work tho? They seem to be filled up. Normally when i buy stuff it has either pins in it or empty holes for pins, or wires to put into.

>>1571631
That will just be solder, since the other one isn't full. I'd encourage you to select your own listing though, as there are multiple duplicate listings. Keywords were:
>5v 1a ac dc module
Feel free to replace 1a with whatever also gets you a substantial amount of listings.

I want to make a really high-Q band-pass filter for demodulating from an IR LED, any thoughts as to what topology I should pick? Going to just use it for toggling a switch at a distance so I plan on encoding no data in the pulse, and instead use a specific frequency so other devices don't give the receiver false triggers.
A T-FF circuit might also be something to look into, not sure if a logic IC would take up less room than discrete transistors, or if there's some way to make one with left over op-amps and/or comparators that relies on the propagation delay.

Is there any sensor which works with arduino and can measure ac voltages between 120 - 500?

>>1571702
rectify the voltage and feed it through a voltage divider into the ADC. use an isolating amplifier to isolate if necessary

>>1571704
alternative: small transformer feeding into rectifier and then lowpass, into adc.

>>1571705
Transformer sounds good but small 500V trafos are not so common.

>>1571671
Crystal ladder filters can achieve very high Q's. 20,000+ is not uncommon. They can be a bit of a chore to design and matching the crystals can be a rather tedious process which obviously requires you have the hardware to properly characterize their motional parameters, so a spectrum analyzer+tracking generator or a VNA.

I purchased a few of these latching relays, and i can't figure out how to operate them.

I connected the + and - pins to 5V DC and then connected the input pin to +5V DC and the relay clicked. But when i turn the power off, the relay just clicked back to the previous state again like a regular relay, instead of staying in that state.

https://www.aliexpress.com/item/pieceofshit/32964268416.html

What am i doing wrong here?

>>1571770
Nothing that how it works.

>>1571772
i have no idea what you just said

>>1571772
>>1571773
oh never mind, just got it
no that is not how it works, it is a bistable latching relay, it has to stay in the last state forever

>>1571770
stupid chinks probably sent me wrong product FUCKING AGAIN

>>1571775
Relays coils cannot be powered by your ineptitude alone.

There's nothing self latching about that circuit. It's just a shitty driver. But if it delivered what it promises its the in line that sets/resets the relay. Removing the power kills the coil.

>>1571779
I probably got my whittu piggu as scammed.
I checked out bistable relays from other sellers and they all have chips and shit on them like this one
Luckily they were so cheap it doesn't hurt too much and i can still use them if i need a regular relay.

>>1571782
God damn it. More than the money, it pisses me off that now i have to wait another month for the new relays to arrive.

>>1571782
Or you can add a transistor or 2 and make it self latching.
Or use one of the poles.

So many options....

>>1571784
>Or you can add a transistor or 2 and make it self latching.
You misunderstand a bistable relay has to stay in its current state even with the power off.
This means that if you turn the relay "on" and remove the power it stays "on"

>>1571786
No relay keeps its state without power to the coil. Unless you have 2 coils. Then it's a true bistable relay. Bistable cannot be self latching. Self latching implies a reset on power down.

There's nothing bistable or self latching about that part.

>>1571792
I am using the same naming the chink vendors are using, maybe it is just called bistable then.

Basically, like i said above, the relay has to have 3 contacts, one of them being common, and then you apply a single power pulse to make it flip from one state to another and it stays in it until you apply a power pulse again. Unless the relay is in the process of switching between states, then it consumes no power at all.

>>1571794
Then stop fapping at shitty Arduino chink crap.

This is a real bistable.
https://www.morssmitt.com/product-categories/200008/latching-bistable-heavy-duty-power-relays

>>1571795
Chink shit is super cheap, works great most of the time and i like their arduino ready variants of things because you can just plug them in and control them with logic level outputs.

I order these suckers. I checked very thoroughly this time and they are definitely legit bistable relays.

Unless they send me wrong damn product yet again. I wouldn't be surprised whatsoever.

>>1571792
>No relay keeps its state without power to the coil.
Nope, that's what keep/latching/bistable relays do,
they only need a pulse.
Some work with two coils, some use one but need an inverted pulse for the other position.

>>1571770
You got chink'd.
Google the ID on the relay:

https://lcsc.com/product-detail/Relays_SRD-12VDC-SL-C_C30431.html
>Description
>Relays General Purpose Non Latching 12VDC SPDT 10A (250VAC , 110VDC) Max Plugin RoHS

You can open a dispute and get your dollar back if you want.

>>1571799
>>1571802
That is a plain old relay.

>>1571801
True the ones with magnets have one. All mine are 2 coils here. In any case bistable cannot be established with electronics.

>>1571802
What amazes me is that i always check ratings before ordering a product and it has a bunch of 5* rating with people being very happy with their purchase. Like what the shit? Is it not strange to them that their "bistable" relay requires constant power to stay in one of the two states?

>>1571804
>That is a plain old relay.
It can't be, the chinks say it in the description, but i am beginning to think the ping pongs who are making those descriptions don't know the meanings of the terms they use and if they says bistable they probably mean that you only need a short logic pulse to switch between states, but the relay actually needs constant power to for one of those states.
You know what, fuck this shit. I will just take a stepper motor, glue a wire to it, and then rotate it with arduino to touch either wire 1 or wire 2, boom fucking ghetto bistable relay.

>>1571809
Why can't u read spec sheets instead of store product descriptions? Even in Chinese it's plain to see. Besides they don't speak engrish...

>>1571810
stop spec shaming me

I bought an led strip with 60 leds. I cut a strip with 20 leds. How do I find the current and power through it? The surface mount resistors say 301

>>1571671
the quadrature sampling detector has impressively high Q and is relatively easy to build. could probably even emulate in software. just remember that you set detector rise time, and bandwidth is inversely proportional

>>1571842
you cut one section (3 LEDs), apply input voltage, and measure the voltage across one of the resistors. E = I * R
then multiply by how many 3-LED sections you want to use
>strip with 20 LEDs
>20 ÷ 7
kek

>>1571842
ok 3 white diodes per package, one resistor and one diode * 3 packages per parallel branch assumed
so 3.3v * 3 = 9.9v, 12v-9.9=2.1v i=v/r=2.1/300=7mA
so .007 * 12 * 3 (3 branches remember?) = 0.252W per 3 leds.
was i close?

>>1571842
Measure the current?
Do you not have a multimeter?

>>1571860
>ok 3 white diodes per package
That's a thing? I thought they were RGB LEDs.

>>1571809
>stepper motor, glue a wire to it
Very good. Something like that has been introduced 70 years ago, much simpler though. One button (or many, all in parallel), push-ON, push-OFF. Single coil (no permanent magnet), no energy required to keep the last state of the relay, only for the transition. https://www.google.com/search?q=impulse+switch

>>1571629
You can't find a USB power supply? I'm a burger and pretty much every one I've seen can do 1A, it's above that causes problems.

>>1571863
multi-die white LEDs are a very common thing, especially above the 1-3W classes

>>1571916
>without taking them apart and modding them and stuff
wrong turn on your way to /g/?

>>1571863
Unless they explicitly say RGB, it's only white led's in a single package. I have a strip from years ago that's laid out that way.

>>1571917
Are you an idiot? You crack open the case and everything is there. Even the chinkiest USBs I got were good at 1A > 4.9V, and are 110-230V.

>maker space has no hot air stations or desoldering guns
>no SMT parts either
>but has an actual pick-and-place machine
what in hell

>>1571924
Maybe some company donated their obsolete one?

>>1571917
>multi-die white LEDs are a very common thing, especially above the 1-3W classes
Only ever saw those configured as an array and with phosphor coating.

>>1571925
Probably something like that. They also have a kickass benchtop DMM that I had to use since the mA fuse in their shitty handheld meter had blown, and boy that guy had like 8 signifiant figures.

>>1571927
As I see it, there's no reason to use RGBs over white LEDs because the three colors can't be individually addressed, but I can't see any phosphor on those chips so it's hard to say.

>>1571930
Actually, I think they usually have four wires. My old white strip has it's ends marked with RGB+, and another strip with color is marked the same way.

>>1571671

you can use a tone decoder chip like an NE567 to detect a certain frequency. dunno if they still make 'em or if you gotta ebay them from China.

>>1571921
>cracking open cases
smuganimegirl61123.jpg

>>1571930
very community. such maintain. wow
>reason
now that I think of it, 300 ohm * 20mA = 6V drop leaving 2.0V for each LED, which sounds like red to me

>>1571944
I think assuming 20mA might not be watertight, but that would make sense.

>>1571944
They're gonna be 10mA. It's closer to 3V*3 leaving 3/300 = 10mA.

>NE555
>supply current: 10mA
this is awful

Anyone see anything wrong with this circuit diagram to make a T flip-flop? Using no passives but a pull-down resistor on the input for my "button" (a wire that I tap from Vcc to the resistor leg). I find if the circuit starts up by chance such that Q = 0 I can toggle it with the button to Q = 1, but I can't toggle it back again no matter how many times I tap (not a bouncing issue), and if it starts up such that Q = 1 it is still impossible to toggle. I'm using an SN74HC00N. I can also try it with a NOR IC if that would help.

>>1572014
and the cmos ones are overpriced too. get a microcontroller.

>>1572014
Nobody uses 555 timers besides amateurs making babbies first oscillator and a some cheap simple Chinese products though even the Chinks are using them less and less

>>1572089
Hey they're free from my makerspace, not to mention quiescent current is the main driving force behind the parts I'm picking.

>>1572100
if battery life is important then you're just going to make yourself suffer by avoiding micros. i get that uc toolchains are annoying to figure out but just get a <$1 8dip pic and write 30 lines of asm and be done with it.

>>1572104
Not him but if he only needs an oscillator or one shot or whatever it seems like overkill to use a microcontroller to do that one thing.

If he can integrate some other circuit functions into the MCU then yeah it's more efficient and a good option but assuming it's for that one purpose it seems kinda stupid.

>>1572108
it's stupid when you don't need to use the sleep function. i actually avoid microcontrollers when possible but the quiescent flexibility is crucial for battery applications. also learning at least one uc toolchain is worthwhile for its own sake.

>>1572104
If battery life is important than surely the maximum current draw of 20µA for a CMOS IC is better than whatever a MCU will give, no? I'm making shit on protoboard, adding some MCU isn't terribly likely lest I use one of my ATTiny85s, which have a q current of 300µA. At least suggest that I use a JK/T flipflop IC, which is basically guaranteed to have a lower q current than a MCU of equivalent price. I don't have one available though.

>>1572109
>>1572111
Oh and it's not actually battery life that's important, it just has to be low current as I've got some sort of linearly regulated power supply that has a Thevenin equivalent of a 4.5V source with 3kΩ in series, and I can't easily replace this, but it may be worth doing so.

>>1572109
I tried to learn PICs but I found MPLAB X too complicated. It has a lot of features but that makes it really beginner hostile. Also the whole C/asm thing is really confusing. Seems like they wanna distance themselves from assembler and they wanna move to C like Arduino. But there's documentation for both all over the place and it's really confusing.

>>1572114
It's definitely worth replacing that. Voltage sources should have an internal resistance as low as possible, ideally zero. 3k is really high for a voltage source. I'd say it'd make a decent current source but it's really too low impedance for that. That's really not good at doing anything.

>>1572111
4000 series cmos is better if you care to use it, footprint aside. but the attiny has an actual quiescent of 100nA if you put it to sleep and wake on an input change. you can simplify and use something very dumb like a pic10f200 to minimize initial configuration which is one of the more annoying things of ucs.

>>1572116
mplab x is annoying but at least it's not an inhumane nightmare like fpga software. i just use the command line tools for asm.

>>1572121
The wake current matters more than the standby current in this situation. I might add a capacitive dropper + zener power supply, but since half of my circuit is already operating without an issue off the existing power supply I'd like to try to get that to work.

>>1572073
Oh I simulated this in minecraft and it gets this issue also, I'll try a version with NORs instead.

>>1572145
>Oh I simulated this in minecraft
Fucking, zoomers

>>1572159
fuck you logisim isn't working for me, plus I know the mechanics of the game well

I need help with intuition, I'll use capacitors for my example

I imagine electricity using the water analogy. The electricity flows through the circuit. It flows through one input, to one output, into the next input slightly changed, etc.
Eventually the desired output is reached.

This thought process relies on the signal flowing directly through every component, which modifies it in some way.

what confuses the fuck out of me is that in order to get a capacitor (diode, inductor, etc) to have an effect, you don't put it in series with the circuit. You put it in parallel with the circuit and ground.

Why? If the electricity is essentially flowing adjacent to the component rather than through it, how does it have an effect?

>>1572169

caps are often placed across a DC voltage to stabilize it, but in the more common use, they ARE used in series with something, a resistor or some other current-limiting element. the 2 together create an RC time constant because of the capacitor's charging and discharging characteristics, and this is infinitely useful. another useful characteristic is the fact that caps act as frequency-variable resistors (Xc = 1/(2 Pi f C)) so you can do such things as filter stuff, and separate out an AC signal riding on a DC level.

>>1572183
To be honest in that configuration the capacitor is still going to ground. Capacitors in series will be in the case of a high-pass configuration.

>>1572169
>You put it in parallel with the circuit and ground.
The water is going through in pulses because the pump is going up and down to pump the water but only pumps on the up stroke.
The capacitor is a barrel the water gets pumped into and acts as a reservoir letting the water out in a steady stream instead of the pulses from the pump.

>>1572096
kek, this, especially with the $0.03 microcontroller being a thing

>>1572108
>efficient
if his bottom line is current consumption, it's not overkill at all. there is no Rchg+Rdis+Qdis path between the rails in a digital timer

>>1572145
1. it's not allowed to assert both R and S. only one at a time
2. a pure-logic TFF without a master-slave configuration is naga happen
3. may I suggest you use falstad instead

>>1572195
I'm giving it a shot with some low-pass filters in there for delay purposes, I think I'm almost there. I'll take you up on falstad though.

>>1572201
then it's no longer pure logic. but it could work
>back in my day the only good logic sim available to the hobbyist was Rocky's Boots

>>1572202
It didn't work with just delays on either set of outputs, I'll look into putting them on input pins or multiple positions. Though at this rate I'll buy a dedicated TFF/JKFF IC. Unless there's some hysteresis-based comparator/op-amp one I can make, since I'll likely have one or the other on the board already.

>>1572216
kinda hard to find a dedicated TFF, and a JKFF would probably be overkill. I suggest a dedicated DFF

>>1572222
I'm running a quick and dirty project on perfboard, as much as I like the look of those single-gate ICs I think it will be more bother than it's worth. Instead I'll go for a 4027 JK IC since the passives needed for a 4001 or 4011-based solution would be going over"board".
Now I'm going to figure out the large amounts of passives needed to filter an IR signal effectively.

I have 2 sensors and I want to transmit their data from one arduino to another.
Since I have only little space I would prefer using an arduino nano but they have only one serial transmit and serial receive pin each.
What algorithm can I use to overcome this?
I was thinking of having the Master send the data of each sensor alternatingly. Would I be required to use a bit adressing system to make it more reliable?
For transmission I'm using a bluetooth module and I'm building the logic using Simulink, not the arduino IDE, fyi.

>>1572359
there are multiple protocols for this shit, just pick one of them instead of reinventing the wheel

>>1572359
Consider Master/Slave protocol like Modbus.

>>1571355
I've made a lot of pretty complex PCBs (even sold a bunch of products from designs I've made), but now I'm trying to standardize all my design practices so everything isn'g just a mishmash of different trace widths and drill sizes.

Is there a cheat sheet for PCB design standards?
I could follow IPC-2222 pretty easily, but those are min/max recommended specs.

I'm more looking for _common_ design selections (not high frequency or RF stuff):
via drill sizes
trace widths
annular ring sizes
space between board edge and ground pour/traces
even typical grid spacing for layout (mine's always coin-flip between mils and mm)

>>1571809
>It can't be, the chinks say it in the description
holy shit dude, you are a moron
>blaming others for your mistake
>not looking up datasheets on the part numbers of the relays CLEARLY DISPLAYED in the product pictures
>getting this mad

don't cut yourself on that edge

>>1572379
>>1572404
Oh I failed to realize that those sensors are essentially slaves aswell, thanks haha

Are there any rules of thumb to choosing the amount of turns (with constant ratio) for a transformer? Are more turns always better?
Is there any reason to ever choose something like 1:2 over 10:20?

>>1572497
Could be hundreds of turns.
There is such a thing as too many turns because of the wire gauge necessary to pack in more turns, which affects power limit, heat considerations, etc.
Can't just do one turn on the primary, you need flux linkage and you need more turns to do that. If you just wrapped one turn of 14 gauge wire on a transformer primary it would just act like a short circuit. If single-turn transformers worked you'd see them everywhere, but you don't.

Hello /diy/, can you help a nigga out? My dad brought this thing home from work a few years ago, after it got scrapped. It may have something to do with charging electric forklifts.

>>1572523
Looks like I forgot to actually ask my question: what is that thing?

>>1572431
nah, that's between you and your manufacturer
I'd pick a bunch of cheap board houses and take the max of their standard technology capabilities, which probably ends up around 6/6mil trace/space, 6mil annular ring thickness, 12mil drill (so 12/24 vias), 20mil copper-to-edge...
I usually use a 10mil grid fwtw, 5mil if I'm using a board house that can handle it. there are worse general guidelines than to make the grid size half of trace+space

>>1572523
>I have junk, what do?

>>1572497
Too few turns increases core losses. Too many turns increases resistive losses. Also, copper costs money and there's a limit to how many turns you can fit into a core.

>>1572497
More turns increase coupling coefficient which increases efficiency... up to a point. Eventually resistive losses will mitigate those gains. So you want as many turns as keep ohmic losses within reason and of course you're limited by how many turns you can physically fit on whatever size core you have. I believe coupling coefficient is a function of core permeability so if you want to use less turns use high permeability ferrites and stuff.

>>1572545
Thanks for your useless reply.

>>1572523
looks like a big PSU controller, with those two boards with the toroids on them being the PSUs themselves. Not sure what other duties the main board does though, nor do I know what those "made in taiwan" boards are doing. I'd look up the datasheets of the ICs with the most pins on each board.

>>1572523
your photos are complete shit, can't read a thing

Is there any way to do interactive graphical circuit simulation like Falstad using LTSpice?

>>1572633
I don't believe spice can do that in general, best bet is to use a time-dependant pulse voltage source to switch FETs/switches on or off.

>>1572634
current and voltage sources can also be altered in a time-dependent fashion (particularly see page 4, piecewise linear)
https://web.stanford.edu/class/ee133/handouts/general/spice_ref.pdf

how hard would it be to use an analog sensor (in this case an IR sensor to measure the RPM of a shaft) and convert it to an I2C one?
i'm planning to use an arduino nano too and i would like to experiment with protocols

>>1572704
try it and get back to us

>>1572705
https://quadmeup.com/attiny85-light-sensor-i2c-slave-device/
I was thinking of doing it like this but damn this seems like alot of effort for something that should be available as a breakout board

>>1572704
An IR sensor won't output an analog value anyhow, it will output pulses at a certain frequency, no? Why bother converting this into I2C if you can basically read this with a micro as is? Do >>1572713 if you have to.

>>1572713
wouldn't even work, those cells are too slow
I believe you're looking for a photointerrupter like Pic related

>>1572741

oh, no you dont. coz with that thing you have to attach some dangerous spinning part to the rotating shaft. instead you want one of these, which just requires painting or taping part of the shaft.

I have this 1.5 inch oled display which uses and i2c connection. I just don't know the particular model so I don't know which library to pick to get this working properly on Arduino. Anybody familiar with this? what can I do.>

>>1572850
>what can I do.>

learn to focus your potato.

>>1572850

Sous vide anyone?

>>1572922
I had to look that up, people actually "cook" this way?

>>1572931
It can get very precise results, by keeping a steak at exactly XX degrees (55 or so C) for 2 or 3 hours you can get a steak that's medium-rare pink all the way through. Crisp up the surface with a blowtorch and you've got a juicier steak than any skillet will give you. They're also useful for things that you might otherwise have to use a double-boiler for, such as confectionaries and such.

>>1572724
Well the thing is, I want to use that sensor alongside another one (accelerometer or hall sensor) and I want the information to be sent from a master microcontroller to a slave microcontroller via bluetooth alternatingly.
And if I'm not mistaken I2C sensors convert the data into bits (to those bits I would allocate adresses so the slave knows which data came from sensor 1 or 2) which I can then feed into the bluetooth module and have the slave microcontroller process the data.
But I think there might be a better way, no?
I've been thinking about this a while now.
I can draw a small sketch if needed for visualization but I'm on my phone rn.

>>1572800
I googled this one and I only found the ones with 4 pins. What are those other 2 pins good for?

>>1572950
>What are those other 2 pins good for?
I believe those are shadows, friend. I think the ATtiny method is the way to go by the sounds of things.

Wow trying to make a discrete IR receiver like the TSOP4136 is nuts, I'll just buy one even if I can't customise the frequency.

>>1572508
>>1572547
>>1572577
Thanks, so is there a formula or something that can give me a rough estimate of what's a good amount of turns for given voltages, frequency, and load?

>>1571627
Put a diode in series in each one, and a low value resistor, like 0.1ohm or lower.

>>1571809
>be chinese woker
>one of the 100 admins of aliexpress page of a electronic chinesium
>job is to pass 5000 ads a day through bing translator and organize pictures
>mismatch an product on a description
>it still has the correct picture and part number
>costumer complains, company lose 2 dollars + shipping of new relay
>get fired because some dude coudn't look up a datasheet
foreign devil get out of china

>lost the desire to do any projects I had in mind
just kill me

>>1573081
just imagine that without anything connected to the secondary the primary(ies) are just coils shorting two live wires at 60/50hz (plus some core losses).
What is the impedance they'll have? Is the current enough to release magic smoke?

>>1573081
No, there is no single formula that covers it all, but there are tutorials you can study to understand the procedure and calculate things on your own. That would be your "something".

>>1573106
But is there a formula or something to overcome my laziness?

>>1573110
read a book nigga

>>1572096
>Nobody uses 555 timers besides amateurs making babbies first oscillator and a some cheap simple Chinese products though even the Chinks are using them less and less
I worked for a few years for a company that made water quality sensors and other things for the water quality industry. They needed an 'evaporation pan controller'. An Evaporation Pan looks like a stainless steel kiddie pool, and is installed right next to a reservoir. The purpose of it is to be a gauge of how much water is evaporating from the reservoir. They put a 0-12" water level sensor in it, and it needs to be refilled when it gets low.
Using a couple simple water presence sensors we already sold as upper level and lower level water sensors, I used a 555 to control a solenoid valve connected to a water source to refill the evap pan. Simple, robust, inexpensive, and pretty much bombproof. If I was designing one today I'd probably use a SOT-23-6 PIC chip instead so I could include a timeout feature (in case the upper level water sensor failed for some reason, so the pan didn't overfill), but even that could have been done wtih another nice cheap 555 for another couple bucks in parts. You don't necessarily need a microcontroller for every job.

>>1572096
>just code it bro

The ohmlets over at 3d printers don't know so i am asking here
I want to connect an ntc thermistor to my 3d printer and in the config file in the firmware where you select the thermistor you are using it sayd
1 : 100k thermistor - best choice for EPCOS 100k (4.7k pullup)

What i don't get is the 4.7k pull up parts. Thermistors don't need pull ups to work, they are just a resistor. You don't need'to pull up a resistor.
So i understand it as in you need to connect a 4.7k resistor in series with the 100k thermistor but like i said that makes no sense.

Anyone has any idea what do they mean by that shit?

>>1573111

magnetics is an insane field coz they have 10 formulas on every page. probably better to just build/measure/document until you figure out what works.

>>1573134

a variable resistor doesnt give you any useful information until you put it into a voltage divider, which then gives you a variable voltage which is easily measurable.

>>1573140
Well there is just a normal slot to plug the thermistor in that has 2 contacts, so i assume the 4.7 is built in?

>>1572883
I tried this. I don't think it uses the SSD1306 controller. According to some chink site that sells the same sized boards I should be using SSD1309, but U8g2 only has SPI available for that controller, not i2c.

>>1573141
> i assume the 4.7 is built in?

maybe. the documentation assumes you have more information than you actually have. read more.

how hard would it be to make a circuit that detects the temperature and then sends it back to my computer via radio waves?

>>1573229
ez use esp32 and a temperature probe, it has built in wifi and bluetooth

>>1573081
en.wikibooks.org/wiki/Electronics/Transformer_Design

>>1573229
yeah easy, if >>1573232 isn't your cup o' tea you can go with any old MCU and a 433MHz transmitter. Though you will need another MCU and a 433MHz receiver hooked up to your computer, unless your computer is some HAM's abomination that already has such a receiver. Other frequencies are available.

>>1573279
This
>cheap pic on sending end
>lolduino on the receiver end
You can pirate matlab on tpb and it has a package that connects arduino to simulink. You can't go any easier than dragging blocks arround.

>>1573232
>>1573279
>>1573282
thank you

>>1573282
Would that be easier than transferring data via wifi though? I've never done either.

What's a good place for buying cheap as hell electronic components in Europe?
I don't want to get CHINKED

>>1573229
Wi-Fi isn't a very simple protocol and requires an entire Internet stack on top of that. Bluetooth is a bit simpler and possibly worth an evening of reading
easiest would be an nRF24L01+. no signal processing necessary, decently long range, packet-oriented, easy to use via SPI

>>1573282
>how do I get data from point A to point B
>via a degree in signal processing
wew

>>1573309
>cheap as hell
>in yurop
intredasting

>>1573314
>>via a degree in signal processing
>wew
M8, are you dumb?
>pirate matlab with 10401394 packages
>buy 1 lolduino and some attiny or whatever
>plug arduino on computer
>open matlab
>skip scary text
>click simulink
>open arduino blocks
>google ``how arduino simulink``
>``input``->``graph``
>``run``

>>1573309
I buy all my shit from ali, never been chinked. Buy from reputable sellers and you won't have an issue.

>>1573320
it's still a shit-tier waste of CPU power, bandwidth, and everything else that an /ohm/budsman should hold holy

>>1573333
Got any recommendations?

>>1573346
For what? There are a few standard IC selling companies that you should be able to find easily by searching for whatever and sorting by most orders.

>>1573309
>Cheap as hell
>not Chinked

Pick one and only one.

>>1573346
no suggestions, because part of the "fun" of ali sourcing is trying to find the fewest vendors to fill your whole needs in order to reduce shipping cost, but a few tips:
>make sure they use their own photos, not photos ripped from some other seller or generic package cartoons (pictures of taped parts are best)
>it's a good sign if they know what they're selling at the most basic level e.g. comparator, op amp, N-channel FET and the description calls it out
>if they say "new original" then you have grounds to dispute if they're not
>there's usually an apparent floor for a particular IC type and lot size, and some sellers below that. if you're below the floor, be prepared to receive lots like Pic related

why is it so rare/uncommon for circuits to be designed on vertically stacked boards (separated by, e.g., standoffs)

>>1573309
Aliexpress has some very nice stores. I have bought lots of components there and I have never been chinked.
That being said, make sure you spend a little more and ask your cargo to be shipped via Aliexpress Standard Shipping. In the best of cases I received my things in two weeks, and five in the worst. I live in Chile so cargo to your country should take way less.
Now, I don't know how many taxes do they charge in your country. Chinese have no problem in documenting lower prices in their invoices, I never had to pay taxes for my chink shit.
I suggest you to visit >>>/g/csg for more info.

>>1573460
think about how you are going to automate that
then compare to SMD

Can someone give me a quick rundown on LED strips and CRI? Why are CRI rated LED strips so much more expensive? Example:
CRI 90+ White LED Strip - 5 meters: $28
https://www.amazon.com/MARSWELL-SMD5630-Daylight-Non-waterproof-Photographer/dp/B07BGRR7XX/

White LED strip - 5 meters - no mention of CRI: $14
https://www.amazon.com/Dimmable-MINGER-Dressing-Daylight-Workbench/dp/B07CVRVPRM/

>>1573408
>in order to reduce shipping cost
I'll be the first to admit it's nicer having everything in one package, but 99% of everything I buy has free shipping on it anyway so I'm not too picky.
>if you're below the floor, be prepared to receive lots like Pic related
Or: "sorry, chang messed up the pricing, could you please paypal us the extra $5?" so you end up paying more than you usually would, or "sorry, we have none of this in stock, we will refund you". Occasionally you'll pay less than the floor, get the right product arriving on time, and find that the seller has removed that listing immediately after it's shipped, but that's nothing to bet on.

>>1573535
Your eye receives 3 colours, red, green, and blue, and your light receptors in your eyes have peaks at these three colours. The peaks are somewhat broad and overlap a little though. Our brains process red+green+blue in the right proportions as being white, so if you just want to send light into your eyes then go ahead and use a very narrow-band light source like RGB LEDs or a CFL. But coloured objects have different bands that they absorb and reflect at in their pigments, and these can be very narrow, so if you shine a 650nm, 550nm, 450nm lights with very narrow frequency distributions at something that only absorbs and reflects light at 600nm, it will look particularly dark. Using lights like this will show poor colour rendition compared to the sun, a very broad light source. So we compare white light sources to the sun and give them a similarity percentage known as CRI, more or less. CRI 100 would mean coloured objects would look identical to how they do under sunlight, give or take a shift in the colour distribution. White LEDs work by producing blue light and feeding this into numerous phosphors that absorb this and re-emit it as longer wavelength light, so to get a very broad mix of light you need to mix a lot of phosphors.
cont.

>>1573544
Do I need expensive lights to light my anime figurines or can I get away with using the cheap ones?

>>1573544
Ordinary white LEDs usually have a noticeable sharp peak in the blue from the LED's original colour, and a curve from the green to red thanks to the phosphors. It's apparently quite difficult to get phosphors in the far red range (650nm - 700nm+ or so), hence some people still swear by incandescents and halogens. Personally I've got an RGB LED strip for lighting, and an orange raincoat hanging nearby, and it looks distinctly red with the light on; it reflects the red light from the RGB strip just fine, but none of the green. A few peculiarities are that our red light receptors are also sensitive to high frequency (~400nm blue) light, hence why 400nm "UV" LEDs look purple. Note that common digital camera sensors also have this property.

>>1573546
Cheap is probably fine, though I'm no expert on anime figure pigments. An RGB strip could look cooler, and that's what I'd personally go for, but I'm a sucker for customisable colours so take that with a grain of salt.

>>1573548
thanks for the info

finally, a "7800" IC that is actually cool

>>1573544
>could you please paypal us the extra $5?
have never had this happen, ever
>sorry, we have none of this in stock, we will refund you
have had this happen a couple of times
not listed: fake tracking number gets """lost""" in the post because seller wants to fuck off for spring festival, which I actually don't begrudge that much

>>1573091
why the resistor?

>>1573567
It limits and balances the current.

>>1573567
because the supplies are not identical and their regulation points vary by a few tens of millivolts. without them, the one with the higher point will bear *all* the load until it can't anymore, and only then will the other one throw in. with them, they share load a bit more evenly
this matters mainly because of heat and capacitor wear. probably shouldn't expect fully twice the current out of the pair

>>1571627
>>1571628
actually, multiple instances of *that* particular supply *might* effectively be connected directly in parallel, by connecting the cathodes resp. anodes of each supply's optoisolator in common. assuming they use a hysteretic feedback scheme

>>1573480
I don't think that's what he meant. Pretty sure he meant standard horizontal PCB layout with vertical board to board interconnects. Like pic related.

>>1573594
ah
interconnects are a source of lots of undesirable things (resistance, inductance, unreliability, cost) that make them second-best for most applications. cute form factor tho

>>1573603
>>1573594
I've seen something similar in USB wall-warts, though they use some plasticky flat-flex cable instead of pin headers. The boards are on either side of the flyback converter and probably add isolation, and for cramming inside a small plastic enclosure it's more efficient than a single larger board, I guess.

>>1573594
>>1573603
yeah this is exactly what I was talking about

I'm trying to make something that involves a 10x10 grid of LEDs and cramming a bunch of circuitry onto the LED board would ruin the symmetry and take up even more space

>>1573644
>not using SMTs to have plenty of room to spare

>>1573408
Thanks for the advice my man, I'm gonna try doing some shopping on there

Why isn't arduino used in the industry?

>>1573644
I'd imagine the stacked board approach would work well far that. I did a 10x10 LED spectrum analyzer on the two boards (though at right angles to each other rather than stacked). I'd just try to avoid sending high speed signals over the interconnects because of the parasitic inductance and stuff the other guy mentioned but I think given the nature of the project you're probably fine.

>>1573692
The industry can afford expensive eval boards and IDEs.
Why bother with an Arduino?

>>1573692
Because there is no need? Also
>robustness and reliability
>plcs
>relays
>computers
>dedicated equipement
I worked in a shop once where when I did not want to waste much time testing stuff, I used an arduino to prototype things. But not in a finished product ever. (Used PICs for that :) )

If a power source is 12v and a load consumes 0.5A, and i need a resistor that drops 5v in series with the load is it (12 - 5) / 0.5 = 14ohms?

>>1573735
What is your load?

>>1573738
small fan

>>1573738
the power source is 12v but i want it as if it was 7v

>>1573735
*Yes assuming the resistance of your load is at least 10x greater than your dropper resistor, preferably higher than that. This is because the resistor in series with your load creates a voltage divider and the lower your load resistance the greater the drop across the 14 ohm.

That said this is a terrible way to drop voltage because you burn 3.5W of power in your resistor.

>>1573740
Won't work with resistors.
Use a regulator.

>>1573749
That's still gonna be inefficient. He should use a buck converter.

>>1573745
>you burn 3.5W
How did you arrive at that value?

>>1573774
It should be 2.5W. 0.5A * 5V = 2.5W.

Also: a resistor is only adequate if either the load will always be drawing at least 0.5A or if the load can tolerate higher voltages. Otherwise, you need some kind of regulation (A 7V 0.5A zener in parallel with the load will work, but an actual regulator will give better regulation).

If you don't want to waste 2.5W, use a buck converter.

>>1573774
>>1573786
Typo. Either way 2.5W is still a lot of power and a very inefficient way to drop the voltage.

>>1573692
doesn't really have a great computation speed, but that can be offloaded (either very easily or very painfully), and for this reason it's not very cost efficient
doesn't have the sort of protection circuits necessary, and it isn't built to handle high vibrations or high humidity
theres limitations in the programming language itself like a limit on global variables (and it's low)
you really wouldn't want to trust it with critical data or warnings

>>1573788
>Typo.
Now the thinko. What makes you assume that at 7V the fan draws the same current as at 12V?

>>1573798
At the time I posted it I wasn't aware he was powering the fan and I was just considering it a constant current load.

Do you think a twisted pair of wires i pulled from an internet cable can handle 6amps? (not together as + and -)

>>1573826
https://en.wikipedia.org/wiki/Category_5_cable#Characteristics
>maximum current per conductor

>>1573835
500ma? Is that a joke? that is BARELY enough to even charge a phone with from usb

>>1573644
>>1573661
>and double the board area

>>1573692
>why isn't a shit-tier set of weakly written, inflexible libraries with a bloated environment used in industry
good question

>>1573798
steady state or starting?
>I don't know *flies into river*

>>1573850
ethernet wires aren't made for carrying power, they're made for cleanly carrying a signal without appreciable electromagnetic interaction and without discontinuities. PoE works by cranking up the voltage (24-48V iirc) and letting each power-using device convert the voltage to current as it pleases

>>1573853
>ethernet wires aren't made for carrying power,
but it's jus regular copper, i can see it when i strip the isolation. unless they are using some magic copperlike metal they are just regular copper wires and the actual signal noise is reduced by the act of twisting the wires in pairs and then the main isolation being shielded

>>1573856
>>characteristics
>24 AWG
>.577A

https://www.powerstream.com/Wire_Size.htm
>copper wire
>24AWG
>.577A

what's the problem bud

>>1573858
It pisses me off because internet cables are great sources of near free and infinite flexible wires for projects and now i can only use them for logic level shit at at most like power maybe one arduino

>>1573862
Where do you get cat5 for near free?

>>1573856
Yes ethernet cables do use regular copper and yes they are run as shielded twisted pair, that is what gives it better noise characteristics. I still don't understand your disbelief regarding why they can't carry more than 500mA of current? You need a certain wire diameter to handle 6A DC and even more to handle 6A AC because of the skin effect. Obviously ethernet systems didn't need to work with currents that high so they used smaller diameter wire which will make implementing the system cheaper. It all makes sense. If you need wire that can handle 6A then go buy a spool of wire that's 19 AWG or larger diameter. Or get a bunch of smaller diameter wires and use them in parallel.

>>1573869
i mean, it is cheaper than buying (even uninsulated) spools of the same diameter
the closest i could find was twice the price (considering cat5 has 8 wires)

>>1573877
it'll carry six amps (given enough voltage) but there are consequences you might not like

>>1573899
you can say that about literally anything

>>1573735
>(12 - 5) / 0.5
If the resistor needs to drop 5V (leaving 7V for the load) then the calculation will be 5V/0.5A = 10Ω, with 2.5W of power consumption. The 7V load itself will have an effective resistance of 14Ω and will burn 3.5W of power. Feel free to swap these if you expressed your statement wrong in the first place.

>>1573905
>the human body:
>it'll carry six amps (given enough voltage) but there are consequences you might not like

>>1573899
Fun fact: Electric chair execution pumps 7 amps through you.
I wonder if they use constant current source since resistance across body warries and if someone has dry skin and shitty physiology and very high resistance it will take forever to kill them with the low current.

>>1573924
ventricular fibrillation can be induced with 10mA. plenty of margin

>>1573924
Dunno that's interesting. Honestly they'd be better off lowering the current probably in the order of 100-1000mA since that range is where ventricular fibrillation is most likely to occur. 7A is more like burning you to death which is arguably less humane. For them to use a constant current source that needs to be able to push 7A through a human body of variable resistance from say 3k-300k you'd need a very high voltage supply that can output from 21kV all the way up to 2100kV which is a lot then you need semiconductors to control it which you won't find any that are rated for 2.1MV. Maybe some old thyratron can but silicon can't.

>>1573939
I wonder what electricity through the brain would do though, as opposed to going for fibrillation?

Apparently it treats depression.

I have a PC and MP3 player hooked up to a Durabrand HT-400 stereo system. How can I filter out the line interference the PC causes to the MP3 sound while the PC is on? At this point I can almost tell you what my PC is doing just from hearing the squeals and noise. The line noise only occurs while the USB wall charger is hooked up to the MP3 player. I do that so I can listen to it non-stop.

What is the best method to get rid of the line noise on the wall adapter side? I plugged the wall adapter into another outlet that is on a different circuit and different breaker, but that made it 200% worse. I've never had ground loop this fucking bad and it is lower in the same outlet of course, but still terrible. I've had thought the UPS would have taken care of the problem, but seems not. Ferrite chokes aren't working.

Now, I'm up to soldering up something. I have piles of various transformers and even magnet wire, so an isolator is something I could rig up easily enough. I'd rather not start stripping wires if there's a better, less-invasive DIY solution.

>>1573954
Transformers work, but an isolated power supply might be an easier method. A ground reference to your shielding might also help, so if your USB wall charger doesn't have a ground prong you should get one with one, since USB power supplies are usually isolated anyhow. Or just put a piece of wire in the ground pin and tie it to your USB -ve output, after checking that you'll not be making a ground loop of course.

>>1573954
Commo mode choke. And shielded cables/coax.

>>1573954
>>1573959
Oh yeah capacitors block DC also.

>>1573960
Is that common in audio applications? Definitely shielded cables though.

>>1573962
I have no idea about anything audio, but if noise is a problem you should
>get shielded cables/coax or whatever you area uses
>get rid of noise sources
A commom mode input choke IS a transformer. Putting up a isolation transformer will just make the noise go through it, isolation transformers have nothing to do with noise, just galvanic insulation.

>>1573959
The wall adapter us unshielded and without a ground plug. Here's the inside. Looks like it will be easier to do things on the audio cable side instead. Ferrites on that side help a bit, I may rustle up some more and see how that goes.

>white goop everywhere

kek

>>1573954
>yet another trying to connect headphone jacks to line ins problem
try this

>>1573966
The noise could be from the PSU ground being different from the computer ground by some DC voltage (buzzing/humming can be a symptom of that), in which case isolation is the way to go. But since >>1573975 it probably is a noise issue. But I'm still not sure whether it would be common-mode or normal-mode noise.

>>1573975
that's how chinks increase creepage distance
>Y cap
did you try turning the adapter around in the outlet

>>1573984
lol Yeah, that was the first thing I did.

>>1573983
The PC is hooked to a PSU. The wall adapter can be plugged into its power (same level of noise), into another outlet on the same breaker (same amount of noise), or into an outlet on another breaker (unholy levels of noise.) When the MP3 player is plugged into the wall adapter, but not playing the noise is horrible, like touching the audio jack on the stereo with your finger loud, but as soon as I turn the player on, that loud noise disappears and all that is left is a high pitched squeal noise

>>1573979
1 mF Caps on the audio line? I'll have to see if I have any...

>>1573986
Squeal could be the switching frequency of either PSU, but since it isn't 50/60Hz the isolation methods (transformer or capacitors) shouldn't make any difference. Feel free to try though.

Well, I happened to unplug the wall adapter from the wall without unplugging it from the MP3 player...the massive noise is still there. Seems the wall adapter cable is acting as an antenna itself.

>>1573995
Yeah, I'm not going that far. Also, the noise changes as I move the mouse around.

I picked up a Psion Organiser II off of eBay, looks neat as shit to mess with. They run off of standard 9 volt batteries which made me wonder if it would be possible to use a rechargable 9V to rig one up for solar power using some of the cheap 9V panels you can find around the internet.

This might sound braindead, but could you simply just wire up some kind of interposer between the battery and the contacts to include a solar panel in the circuit, or would doing a solar recharge mod on a 9V-powered device be a little more involved than that?

>>1573999
>Seems the wall adapter cable is acting as an antenna itself
In that case shielded cables and grounding the adapter should make a difference.

>>1573984
in which case ferrites at the USB end of the cord might be worth a try

>>1574000
it's more involved than that. you also need to switch or blend between the two sources. look up "power path management" to get an idea of what's involved and some ICs that might do this

>>1574001
I'd need to buy an adapter that is already grounded it seems. I looked up the mp3 player's stuff and it seems there's a built in radio receiver. It uses the audio cable and power cable as antennae. I think I'll just listen to it unplugged then plug it in when needed.

>>1574005
Damn, figured as much. Thanks, I'll take a look at it. Wonder if any of the panels I want could feasibly power this thing anyway.

>>1573986
That's a mu not an m, it's one MICROfarad. Honestly 1uF is not ideal as at 20Hz it has an impedance of 8k and it might attenuate your audio some. Then again if you're just feeding that into some high impedance buffer then that 8k impedance shouldn't be an issue. To be safe a 100-680uF cap is a better choice.

>>1574013
100µF nonpolar caps are kinda rare though
anyway
>not just firing up winamp and playing from the mp3 player's mass storage

>>1574013
>mu not an m

Yeah, I was thinking..."mF???" lol If I can find my box of audio plugs I'll try it.

>>1574017
I have a box of that sort of thing too, somewhere.

>>1574017
It doesn't need to be non-polar. You only need to get the polarity correct for any DC bias conditions that exist. Even then if you're really worried about polarity just whack two in series (+-) (-+) and now you have a non-polar cap, granted with half the capacitance.

>>1574033
>whack two in series (+-) (-+)
true enough
and with twice the inductance, which may or may not matter much depending on input impedance on the receiver

>>1574028
yeah sry bt dat, my mouse penmanship has always been shitty

>>1574038
Any additional inductance will be very small and insignificant at audio frequencies. A few extra nH won't even begin to make a difference until you get up into VHF and beyond.

>>1574033
>just whack two in series
>for audio applications
God no, at that point you'd basically be running your audio through antiparallel diodes. The only reason to use isolation caps in this kind of circuit is when you have some DC difference between one side and the other. This DC difference is almost always greater than the audio maximum amplitude for unamplified sound, so you can just measure which side has the higher voltage and orient your caps to suit. Anyone using antiseries electrolytic capacitors in an audio circuit should be shot.

>>1574053
There is absolutely nothing wrong with antiseries electrolytics. Sounds like you're just being an audiophool.

>>1574062
Try it and then come back and tell me how good an idea it is.

>>1574072
I have already employed antiseries electrolytic capactiors in the range of 100-470uF for coupling in several audio amplifiers I've designed with no ill effect. It doesn't alter the sound in any negative way but the larger caps do extend the frequency response of the amplifier so the -3dB frequency of the high pass filter created by those coupling caps is well under 20Hz. Sounds like you're spouting a bunch of placebo garbage.

>>1574073
You design shit amplifiers then. Antiseries diodes in the signal path is goddamn foolish.

>>1574085
Capacitors are not diodes you fucking idiot. It's designed to be a dead short for AC. The polarity has nothing to do with anything.

>>1574085
Here, I even simulated it to illustrate exactly how little it matters putting electrolytics in series in the audio path just in case you're still too retarded to get it

>0% TOTAL HARMONIC DISTORTION

>>1574096
>simulators
>modelling electrolytic capacitor reverse biasing realistically

>>1574114
THERE IS NO BIASING FOWARD OR REVERSE CAPACITORS ARE NOT DIODES

>>1574114
>capacitor reverse biasing
uh?

>>1574118
"reverse bias" doesn't have to refer to a diode, it's applied to transistors and the like often also, because they have different characteristics depending on whether there's a net voltage in either direction; a bias. An electrolytic capacitor acts different in the reverse direction to the forward direction, hence it can be reverse biased. Is this so hard to grasp? In fact, an electrolytic capacitor has non-ohmic and time-dependant behaviour in the reverse direction, due to the electrochemical nature of the conducting medium.

>>1574096
Okay. I thought you were doing pic related, hence the diode talk.

>designing remote control
>using a 3-pin IR receiver
>realise the transmitter button is bouncing and so need a debounce circuit
>finally debug it and get it working
>some "bouncing" is introduced in between the transmitter and the receiver when the IR LED isn't less than 5cm from the receiver
>pic related
>have to put debouncer on receiver side now
>receiver side has less room and less available power
>transmitter side already has a spare comparator from the oscillator
>don't even need transmitter debouncer anymore
piss

>>1574134
Oh by remote I mean a single 38kHz IR pulse that turns a light on or off. It's in a room without any other IR controlled devices so I'm not bothering with encoding any data in the pulse.

>"buy american bro support your local electronics shop"
>go to support local electronics shop
>get raked over the coals and fucked 6 ways to sunday over prices
why even bother

>>1574138
meh, think of them as convenience stores when you really cannot wait for Digi-Key to overnight you a thing

>>1573939
>you'd need a very high voltage supply that can output from 21kV all the way up to 2100kV which is a lot

they put a wet sponge beneath the contacts which lowers the resistance enormously, esp if they use salt water.

>>1574134
>less than 5cm from the receiver

if you're within 5cm of a thing, you dont need a remote control. so your problem isnt real and doesnt require a solution.

KiCAD gets cuter with layers

>>1573939
>"Under dry conditions, the resistance offered by the human body may be as high as 100,000 ohms. Wet or broken skin may drop the body's resistance to 1,000 ohms," adding that "high-voltage electrical energy quickly breaks down human skin, reducing the human body's resistance to 500 ohms".

Looks like they would probably only need a few kV even with perfectly dry skin to get 7A.

>>1574180
And I don't know why he claimed 0.1-1A is most deadly, as this pic clearly indicates more current is always more dangerous..

I am getting really fucking tired of this. After getting chinked on the bistable relays where the chinks shipped me just regular relays, i bought two from a shop in my country and the description said bistable and they turned out to be fucking normal relays again. This is starting to drive me crazy. I am unable to get my fucking hands on a bistable relay i need so fucking much.

I looked at ali again and found something which this time looks legit, pic related

I checked the data sheets and everything, and it does appear this relay can remain in either of its two states without power, correct?

https://www.aliexpress.com/item/12V-Coil-Bistable-Latching-Benis-DPDT-2A-30VDC-1A-125VAC-HFD2-005-S-L2-D-Realy/32948908428.html

this is the datasheet for it https://www.datasheets360.com/part/detail/hfd2-005-s-l2-d/-5164402641065414951/, it says only latched, but another datasheet i found says bistable (but it is myssing the D in the name at the end)
https://www.tme.eu/en/details/hfd2_005-s-l2/miniature-electromagnetic-relays/hongfa-relay/

>>1574222
>L2
that does appear to be the benis you want

>>1574252
Welp i ended up ordering a whole lot of 10pcs since they don't sell them in single pieces.
If it turns to be just a regular relay yet again, i am going to attach mains wires to my nipples and end it

>>1574222
>>1574274
You do realize that instead of buyfagging you could have made as many of those as you wanted right? Because they are not difficult to construct.

>>1574085
>Capacitors
>diodes
wat

>>1574289
Who in the hell makes their own relays?
They cost a few cents and are smaller and more reliable than what you could cobble together.

>>1574299
I think what he's implying is you can make a relay bistable by using a regular relay and some additional external components. Not that you physically construct your own. It was badly worded.

>>1574289
How do you even have time to complete any projects when you are in copper mines all the time mining copper for your wires?

>>1574301
>I think what he's implying is you can make a relay bistable by using a regular relay and some additional external components
Without relying on constant power? How?

>>1574306
You can take a wooden match a wedge it between coil and the contact to force the relay to remain closed even with power off

>>1574308
And how do you open it again without manual intervention?

>>1574309
You wedge it there only ever so slightly, so that when the coil energizes, the match falls out

>>1574317
And how do you close it again without manual intervention?

>>1574317
Then you need to stick it in again to get it stay closed again.

Congratulations you made a switch except more expensive and awful.

i have a pioneer receiver from 1978 thats main power relay keeps shutting off randomly. sometimes it happens every few mintues and sometimes it doesn't happen for a couple weeks. sometimes it turns back on by itself and sometimes turning the power off and back on resets it.

seems to only happen when im outputting to the speakers, if they are switched off and i use headphones it hasn't switched off.

i have't really worked on much electronics and definitely not something this old. would going over the whole board with a voltmeter be a good idea or am i better off just getting it repaired by someone else?

>>1574323
If you aren't ready for that project you could hurt yourself or fuck up the amp even worse trying to fix it.

>>1574326
well would it be easy to as least try to narrow the problem down?

>>1574329
You could do that yes. Make sure voltages are what they're supposed to be at various points on the board

If I use two 555's, can I make a device that will delay for a certain amount of time, put out a high logic level for a certain amount of time, and then turn off? I believe so, but when the 555 is used as a "delayer", what does triggering it again do? Does it reset the "count"?

>>1574334
Yes what you're asking is possible. Read the data sheet, a 555 is way easier to understand once you know how the internals work.

>>1574318
you just keep moving the goalposts, this is ridiculous.

>>1574323
>main power relay k
Just replace the power source with a cheap chinkshit

>>1574338
Can I make a single-button flip flop as well? What is its hysteresis like? I have a bunch on my component drawer but I want to find out about its intrincacies before assembling everything

>>1574340
so if it's a short causing it will stop happening?

I have a handheld printer that can be powered by 6 aaa batteries or by connecting a 9V power adapter, but if i try to connect the adapter with no batteries inside, it doesn't work, how come?

>>1574334
use attiny instead

>>1574352
use ne556 instead

>>1574342
1. yea probably
2. 1/3 and 1/3 vcc
3a. use falstad circuit simulator
3b. there aren't many intricacies, the 555 is mainly a voltage divider, two comparators, and an SR flip-flop. there are dozens of useful things you can do with that.

>>1574361
If I want to make a simple model elevator, for example

I'll need a flip flop that will set to "on" when the top button is pressed, and to "off" when the bottom is pressed. The state of that flip/flop will determine the direction the motor will spin.

Either of those buttons will trigger another flip/flop that will drive the "enable" pin of the h-bridge, and when the elevator touches either of two end stops, it will reset the state of the first flip/flop.

i.e: the enable flip/flop will receive the input of an OR gate with 4 inputs, and the direction flip/flop will receive input from two buttons.

Basically I need a SR flip flop and a JK flip flop with both J and K set to high, right?

>>1574334
> If I use two 555's, can I make a device that will delay for a certain amount of time, put out a high logic level for a certain amount of time, and then turn off?
Yes. Two monostables connected by a capacitor. So the falling edge of the output of the first generates a brief logic-low pulse.

> but when the 555 is used as a "delayer", what does triggering it again do? Does it reset the "count"?
No. A 555 goes into the "charging" state (OUT high, DIS' open-circuit) on the falling edge of TRIG, and into the discharging state (OUT low, DIS' low) when THRESH exceeds CONT (or (2/3)*Vcc if CONT isn't connected). You can pull the RESET' pin low to reset it to the discharging state, but then you'd need to pull TRIG low again to start another pulse.

>>1574361
2. 1/3 and 2/3 vcc, rather

>>1574368
>Yes. Two monostables connected by a capacitor. So the falling edge of the output of the first generates a brief logic-low pulse.
This works pretty well

Now I'll need some trickery to use it as a flip/flop for a single input

>>1574384
>Now I'll need some trickery to use it as a flip/flop for a single input

I tried this in Falstad and it works a bit, but it's easier to simply use a D-type flip flop. The "flip flop" using the 555 ends up acting like an oscillator.

How do I design a PCB so that I have SMT components on the bottom and they need to stay in place during wave soldering?

>>1574455
Aren't they glued in place before soldering? Some ICs are placed in the diagonal because somehow it helps.

>>1574346
It's more likely some component overheating and failing, or failing to meet a power/stability expectation because it's value has drifted.

>>1574308
you can put a wooden match on a servo arm attached to an arduino

>>1574358
use padauk PMS150 instead

>>1574367
that's a pretty shitty model elevator. anyway draw it out and trace it out, maybe draw a state diagram. much becomes clearer when penis put to paper

>>1574402
this anon gets it

>>1574455
you put glue on the board before placing. KiCAD supports an Adhes layer but the default footprints don't have one

>gotta spend 10 quid on components to finalise my design
>can't do it
whyyyyyy

>>1574171
No I'm saying I need to add debouncing for it to work when over 5cm from the receiver, it only works fine when within that distance. I plan on using it up to 3 or 4 metres away, so my design needs revision of some kind.

I'm also not sure how hard I can drive an IR LED for maximum brightness. Is it acceptable to put more than 20mA through it at a 38kHz 50% duty cycle? Will I get better performance if I increase the momentary current but decrease the duty cycle?

>>1574558
>Will I get better performance if I increase the momentary current but decrease the duty cycle?

yes, try 10% duty cycle at 5x rated current.

>>1574558
>Will I get better performance if I increase the momentary current but decrease the duty cycle?

yes, try 10% duty cycle at 5x rated current. but only if you need it. at 3-4 meters, you shouldn't need it.

>>1574575
Wait, I just checked its datasheet and apparently it's rated for 50mA continuous anyhow, which I guess makes sense for a 1V diode. It also says 200mA current is possible with 10% DC at 1kHz, which is pretty neat. Assuming this random LED in my parts bag is the same as the one on this datasheet, but I doubt there's much difference between them.

Also I found this:
https://www.labsphere.com/site/assets/files/2570/the-radiometry-of-light-emitting-diodes-leds.pdf
Which is really interesting looking, might use it to calculate the radiant power output of this LED.

>>1574584
Ok looks like it's in the ballpark of 6mW of optical power. That doesn't look too bad compared to visible light LEDs either.

Dunno if this is the right place to ask.

So I have this device which counts how many units of electrical energy is used in the shop. Somehow, in the course of one month it has gone from reading 8k kWh to 112k kWh. The previous 4 months has provided plausible readings.

Any explanations as to why this is?

>>1574626
sparky shit is probably better somewhere else in /diy/ but you've either got a ballast, motor starter, or similar reactive device taken a shit, or someone wired their weed grow into your circuit

>>1574626
Amp clamp the wires on every breaker, see if you can narrow down what's draining the most power.

>>1574632
>>1574636
I don't think you get the question. There is an indicated energy usage of 100 MWh over 45 days. That comes out to a constant load of 92kw (125 horsepower), or 408 amps.

I'm asking if there is anything which would cause the meter to crap out like that.

>>1574648
If that's out of the capability of your wiring then you're probably right. The current shunt/transformer could have shorted internally, open it up and see. It's hardly any use to you intact.

>>1574626
ok there are 730 hours in a month
you have used 104000kWh
so you are using 142kW per hour
thats a dual pole? 32A breaker, at 230v thats 7.36kw.
so you are apparently pulling 20 times what should trip the breaker.
i hope for your sake that the meter is simply fucked.

>>1574632
>either got a ballast, motor starter, or similar reactive device taken a shit
there is a possibility that its a power factor issue?
absolutely depends on the device how it measures power usage, there are 3 different 'types' of power depending on inductive/capacitive loading a device needs to measure the voltage as well as current.

or it could just be fucked.

you want a diagnosis of exactly what is wrong inside it? no idea get it cracked open and let us have a wee look at the gizzards.

>>1574538
Are you the same poster that complains all the time about setbacks and being too depressed and constantly frustrated by microcontrollers?

>>1574656
>kW per hour

>>1574656
It's obviously useless and unreliable at this point. I'll take it down tomorrow and have a look inside, if the thread is still alive.

>>1574656
>>1574663
Redoing the equation:
>730 h / month
>E/t = 104000 kWh / month
>P = 104000 / 730 = 142 kW = 142000 W
>V = 230 VAC(RMS)
>I = P/V = 142000 / 230 = 620 A
Not sure where that 7.36kW comes from bro.
>there is a possibility that its a power factor issue?
It would be tough to get that bad of a power factor, and even if you could the breaker trips based on current, not current multiplied by power factor, so it's basically guaranteed that the measuring device is kaput.

>>1574669
>if the thread is still alive
It will be, this is a slow board. After the thread reaches bump limit we usually get another 3 days out of it, if not more.

anyone here ever fucked around with the laser diode from a CD/DVD drive?
I just want a high power laser but can't find them anywhere

>>1574670
Oh that was the power that the breaker can handle, don't mind me.

>>1574671
Ali should be your guy for anything from 1mW to 1W, how much power do you want?

>>1574673
however much I can get honestly

>>1574674
Then let me change the question. How much are you willing to spend, and how bulky is it allowed to be?

>>1574682
about $50-$100, but I'm real flexible. It's just for shits and giggles. It needs to be handheld though. I remember there was a company that sold giant lasers whose casings were the size of flashlights but I haven't heard about them in a long time

>>1574684
The 1W+ lasers are used in laser cutter CNCs, and they're about the size of an apple or larger. Have a look on aliexpress and ebay and see what you can find. Taking their power ratings with a grain of salt, of course.

>>1574658
no, I lack confidence

>>1574350
The adapter probably isn't going deep enough. It needs to go far in to switch the power from batteries to plug. Also the polarity could be reversed on your adapter, compared to the thing.

>>1574684
Just buy one of the illegal laser pointers that can blind people, they're probably still around.

>Just spent the better part of 2 months taking online Udemy courses on RF design theory after work
>Run into the same problem as I had in school: bored to tears because I'm not building/making anything and I absorb nothing
>In turn, learn pretty much nothing on the subject
What would even be a good project that will get me some good exposure/appreciation for RF circuit design?
I wanna make antennas on pcbs and all that junk.
I though about making some bluetooth controllers, but I dunno.

>>1574690
>Taking their power ratings with a salt shaker or two
fixed

>>1574750
perhaps start by making some test equipment? there's a $200 VNA project out there somewhere you might consider building
with that, you could set up a pcb photo-etching rig, make antennas on pcbs, and test the hell out of them all day long
or, ESP8266 chips are dirt cheap. skip bluetooth, go straight to wi-fi and build sensor things or super cheap mesh network nodes

>>1574753
>>Taking their power ratings with a salt shaker or two
You remind me of pic related

>>1574754
kek

>>1574755
lmao

>>1574753
I can look into making a VNA, but I honestly don't want to spend a lot of time making equipment and debugging said equipment when I'm testing shit. I've tried shit like that before and it's just a headache I don't want to deal with.
I used to photo-etch pcbs when I was a lot younger, so I mean I would inevitably have to set something up (unless I get pcbs manufactured through China for dirt cheap and skip that process)
Is there a reason I should skip bluetooth? Is it a poor exposure to dabbling in RF circuit design? I have a bunch of NRF24L01 modules laying around that I could use for a BLE project (perhaps a low energy wireless game controller?)
I actually do have a set of wifi mesh transciever modules as well, I could make something with that but I dunno what.
I know it's used in this IOT/Smart home shit (I used to work for a startup that was selling cheap chinese smartlights, hence the wifi trancievers I have), not exactly keen on making my own smartlights now. Don't really like the whole aspect of connecting devices to the internet that aren't computers anyway.

>>1574761
Perhaps something closer to radio circuit design would be more up your alley? Start with an easy lower frequency circuit and move your way up the difficulty/frequency curve from there. Personally I'd quite like to see an economical low-frequency SDR, since the cheap dongles are only 25MHz minimum and I'd like to listen to SW. It would be quite an in-depth project, but I believe there's a fair bit of documentation on the subject. Else you could just experiment with demodulation and data decoding methods for RF, since chances are there will be some of that ready to be received in your local area's ham bands.

t. not an expert

>>1574761
Just buy a VNA. Gonna be better than anything you could possibly make on your own. I scored a decent Agilent VNA for about $400. Admittedly that was a bargain but they happen every now and then. If you can't wait for a deal then expect to sink thousands.

>>1574761
afaik, nRF ≠ BLE
if you know the Bluetooth protocol and its various classes well enough, sure, go for it. the nRF8001 (might be NRND by now) + your choice of MCU might be an easy way to get into that. your call, whether you want to write protocol stacks or make radios
>connecting devices to the internet
but you don't, of course. you connect them to your home network and don't let them talk outside of it. that sort of shit gives startups the heebie jeebies
>they don't NEED US anymore REEEEEEEEE

>>1574769
if one really wants to buyfag (and on second thought maybe that's a good idea), also http://www.pocketvna.com/ , $500 for a PC-based VNA from 500kHz to 4GHz

>>1574772
That doesn't look half-bad for the price.

>>1574746
>polarity
yeah turns out thats it. every single adapter i ever had was + on the inside and - on the inside, but the retards from Brother have their shit the other way around so you are forced to buy their adapters

>>1574791
That's done in some obscure (possibly marine?) applications and actually helps mitigate galvanic corrosion, or something along those lines. But in this case I think they're just being proprietary for money's sake.

>>1574814
Jokes on them, i can just snip the cable and flip the wires.

What happens if you are connecting some wires in your computer and accidentally fuck up and connect the 12V yellow wire on the PSU to the black ground wire on the PSU?

>>1574843
Then your PSU shuts off.
ATX PSUs are protected against overcurrent.

>>1574845
So it's not possible to damage it that way? Or do do you mean it will blow a fuse?

>>1574848
I believe it should only hit the resettable fuse - so it should work again after a bit of time to cool down.

>>1574848
If it's an acceptable quality PSU and you're using shorting it with a reasonably thicc wire it'll just turn off. With a thin wire you might not get enough current to trip so that wire can get hot hot hot.

Am i reading this correctly?
brown wire is + power, blue wire is - power and black wire is +input power output if it detects something.
I am not sure what that crossed out rectangle is. Is it a resistor that limits the max current from the logical output wires so i don't have to add a resistor myself?

>>1574885
by crossed out rectangle i mean the one between the lines with the chink letters in it, on my probe it's just a line across that rectangle with no letters

>>1574885
customarily, industrial sensors use a 4-20mA standard, where 4 is typically the low end and 20 is the high end (and 0 is the sensor's broken, raise an alarm)
assuming that is the case here, if you tie black to ground with a 220 ohm resistor, black should vary between .9V and 4.5V depending on whether there's an object in range or not. should be quite safe to test

>>1574889
I checked it and when no metal is detected the voltage between black and ground is 5V, if metal is detected the voltage is 0.6V
the probe wants says 6 - 24V DC but it seems to work just fine on 5V
Also, if no metal detected the resistance between the two above is 0L and with metal detected it is 16 mega ohms

Which means that logic wire is normally pulled high and when metal is detected the probe pulls the logic wire to ground through a 16Mohm resistor right?

>>1574896
interesting, with 12v it reads 21 Mohms instead

>>1574885
Standard connection for industrial sensors: BN (+) and BU (-) is the DC power supply and BK is an open collector output that can sink up to 300mA to drive a load like a relay.

>>1574973
>sink
Do you mean supply?
As in i can connect a device between BK and ground and this device will then use BK as + and ground as - and can operate on up to 12V when the probe is NOT detecting anything

How the fuck do I get a job just doing pure pcb/circuit design stuff?
Finished my bachelors in electrical engineering, did a project in my garage (a sorta working 2 axis writing robot), and yet I can't find a job.
I had to take up a mechanical design job because it's been over 2 years since I graduated and no one would hire me for circuit stuff.
>inb4 you have a good job already stop whining
I went to school so I can get a job doing what interests me. Designing sheet metal enclosures bores me to tears.

>>1574986
There's 'source' (current flows out of the terminal) and there's 'sink' where the current flows into the terminal. In this case the load (resistor, lamp, relay..) is connected between (+) and the output terminal. The simplified diagram on the sensor shows exactly that.

Do relay coils need a resistor in series?

>>1575055
Oh i understand now. It basically shorts the BK to ground when detecting something, and it connects it to input + when it detects nothing.
So i should connect BK to a pin that is set as high input and then, when the probe detects something it will pull than pin low

>>1575082
No, the winding has its own resistance.

Anyone here have particular preferences as far as comparators and op-amps? DIP this time. Thinking of picking up some:
>LM339 quad comparators
>LM393 dual comparators
>LM324 quad op-amps
>LM358 dual op-amps
And these seem to be the most common DIP ones on ali at the very least. Is it worth doing a price comparison on some different models? Any other general purpose ICs I'd benefit from having? Probably going to beef up my MOSFET supplies too.

Also I found this:
https://www.aliexpress.com/item/test-clip/32355336386.html
Which I think is designed to program 8-pin micros, but it would also be somewhat useful in conjunction with a logic analyser. Be nice if it had 14 or 16 pins though.

>>1572169
water anology is retarded gay shit

>>1575112
I have a bunch of lm339, works alright for my low frequency stuff. But you should find one that doesn't need pull-up resistors (one with push-pull output instead of open collector)
>>1572169
You need to understand their intrinsic behavior. Then you'll see that they have more in common with springs than water. But yeah, analogies only go so far.

>>1575130
>one that doesn't need pull-up resistors
You're probably right about that one, and I think the 393 is like that also.

any recommended guides on doing prototyping with SMT products? Through hole is getting cluttered and expensive

>>1575145
For prototyping I'd still use THT because you can use a breadboard (or several) for most of it and just reuse all those parts. If you're using SMTs then you basically have to either make your circuit as a bunch of sections that you plug together with a breadboard or wires, or you make the whole thing on a singe board, both requiring you to make custom PCBs.

>>1575145
>Through hole is getting cluttered and expensive
truly sad

>>1574755
fuggen saved

>>1575112
Get JFET input op-amps like the TL072/74 or LF353 instead of the LM324/LM358. The difference is performance is worth the small bump in price.

>>1575176
I agree with you but there are some circuits you really just can't breadboard because of the really high parasitic capacitance and inductance of the breadboard. For that if you don't wanna diy your own PCB your options are deadbug or Manhattan construction basically.

>>1574897
don't use ohms to characterize voltage- or current-mode components

>>1575088
>high input
you mean pullup, but yeah

>>1575112
if you do most of your work at 5V or below, get some low-voltage op amps e.g. MCP60x, MCP600x, LMV321, etc. you will probably need some SOIC-8 and/or SOT-23 adapter boards and some pin headers to make them socket-breadboardable, ali has all o the above for pennies each

>>1575232
>instead of
no, get both. JFET op amps don't work well when inputs are near V-, LM321-style op-amps don't work well when inputs are near V+. also TL0xx have a supply voltage requirement in excess of 5V while the LM321s do not

>>1575232
I've seen JFET amplifiers around, but what's their advantage? Extra high input impedance?

And I think I've seen prototyping-friendly Manhattan construction kits, they come with a bunch of adhesive pieces of copperclad that you stick down onto your main ground board, and presumably you can restick them a few times or replace their adhesive. The soldering is still a pain though, perhaps female-to-female pin sockets could be used as a temporary replacement for prototyping purposes, assuming they exist.

>>1575239
Oh that's something to keep in mind, I forgot about rail-to-rail op-amps. You're right that I mostly work with V+-V- = 5V, but I don't usually need to use the space around the rails as I'm typically working with signals around a 2.5V rail.

>>1575232
>>1575240
And that LF353 looks like it has a pretty high standby current, is this a standard feature of JFET op-amps?

>>1575240
>Extra high input impedance?
basically, yes, also low bias current which is nice for very hi-Z input sources e.g. photodiodes, pH probes, and other sensors. a bit overshadowed by CMOS op amps lately. the MCP601 is a good lo-volt jellybean upgrade from LM321-types and can be gotten in PDIP for those so inclined. ditto the MCP6001 etc., plus RRIO
the only catch with RR-input op amps is that they often use two input stages depending on common mode voltage and don't transition between them perfectly transparently. Microchip's advice:
>The transition between the two input stages occurs when V CM = V DD – 1.1V. For the best distortion and gain linearity, with non-inverting gains, avoid this region of operation.
Manhattan construction kit: if you are willing to solder and don't need excellent mechanical stability, simply flip your single-sided copper-clad blank board over, and use copper tape and a pen knife to shape your interconnects on the non-copper side. wrap around from the ground plane as needed. Cu tape is sold for pennies per meter, cheap enough to be considered disposable, as are most SMT passives

>>1575274
How about use the upper side of the copper-clad board with an insulating layer of paper or masking tape between so you can easily solder to it if needed?

>>1575281
sure, that works, that's also useful for jumpers and multi-layer designs. I also hear there are copper tapes with an insulating adhesive that might also be used, if you can get them and if you can keep edges/corners off the substrate. I'd probably recommend polyimide tape instead for ease of removal and for heat resistance
there is a price to be paid: 10-15x or more parasitic capacitance increase for that low-L low-R path to ground, which could matter at UHF or above

>>1575313
I was thinking masking tape would be easier to carve away a small spot of. As for the extra capacitance, wouldn't it be a good thing at SW and below?

>>1575315
Kapton is manageable, cuts easilyy with the point of a pen knife, also doesn't leave glue behind. don't like your circuit? take a putty knife and shave it off... if you built on the non-copper side of the board anyway
capacitance, like all things, is useful or not depending on where it is. parasitic capacitance on a negative feedback node would very likely be counterproductive

>>1575322
Ah you're right about that.

if my scope is rated 300 mhz, is it a dumb idea to get 100mhz probes?

>>1575451
No. But it won't perform well near and above 100Mhz

r8 my to be capacitive welder

>>1575472
very capacitive

>>1575472
not enough capacitance
no operating switch
no welding probes
shitty schematic

I am fucking TIRED of tripping over my solder iron cable. Are there any battery powered solder irons like there are power drills?

>>1575684
dosen't your country have google?

https://www.amazon.com/WELLER-BP865MP-Battery-Soldering-Iron/dp/B00HYPC7EG

Rate my idea for an educational electronics toy.

You have a base board with 4 inputs and 4 outputs. It's nothing but 4 transistors and the necessary resistors and RJ-type jacks. It's powered by a 6 AA batteries on their own pcb with a on/off switch.

The outputs can be motors, LED's and buzzers. The inputs are just signals from sensors (LDR, touch, buttons) , and in between the sensor output and the base-board inputs several modifier modules can be added. There are basic logic-gate modules (2-input and, 2 or 4 input or and a not for inverting purposes, but not xor, xnor, nand since only autists care about them) and flip-flop modules (d-type and sr-type). There could be special modules such as a h-bridge just for motors and a sequencer based on a decade counter but I don't want to expand too much into that.

Everything has a drill pattern based on a half-inch module so that it can be bolted down to a base board with nylon bolts.

I know there are other things like this as well but they're all very local solutions, and none of them is popular in my country. My target audience are schools, and reportedly a lot of money can be made with this, judging by what companies show in education fairs.

>>1574885
Crossed rectangle is the load

No, the output on that thing will just close the path between the bottom of the "load" and GND, it's an open collector sensor

>>1575718
Looks neat. How would the D-FF work? T-FF would be easier (with some hysteresis) since you can simplify it to a single CLK input with T always held high. I guess you could have a CLK button atop the D-FF.

Hi guys, I need a power supply for my portable device, it needs stable 3.3V rated for ~120mA for dirt cheap (<$1). Preferably in SMD so I can put in on the PCB so I can save space. The device is very bitchy about the PWM form dc/dcs chinks make. Haven't decided on the battery type yet, pretty much depends on the converter itself, but li-po works best, so it has to accept 3.7-3.0V on input.

tl;dr cheap DC/DC 3.3V 120mA

Any suggestions?

>>1575939
Also, 2-cell li-po input (7.4-6v) is also okay.

>>1575939
Converting to 3.3V from 3.7-3.0V will be less efficient than a standard buck converter (also isn't it usually 4.2-3.0V for a lithium ion?) so the 2S method is probably best. While a buck converter should work, you may need another stage of filtering in the form of an LDO to drop the final volt or so if your device is picky about ripple. I'd consider simply using a 3.3V linear regulator, since they're dead cheap and you're only pulling 120mA. You'd be losing more than half your power though.

>>1575878
If you do it like this then it toggles when the clock changes from low to high, the t-type would change state when transitioning from high to low. I wonder if I'd need something like an opamp to help with bouncing or if adding a 100n capacitor would be enough

>>1575962
I imagine you can get single-gate schmitt trigger (inverters) in tiny little SMT packages. Though I think using a SPDT button without pullups but one capacitor (pic related) mitigates any bouncing errors. SPDT buttons should be more common. Still, stopping the light sensors and such from triggering multiple times quickly would also be a good use for a debouncer-style circuit.

>>1575959
I believe 4.2 is the charging voltage, once charged, a li-po holds 3.7. I have no intentions of charging the batteries while using the device.

Could you please also elaborate on filtering. All I've been doing is adding 1uF capacitors on both ends of the converter. Do you mean to add an inductor?

>>1576015
>I believe 4.2 is the charging voltage, once charged, a li-po holds 3.7
A lipo holds at ~4.2 until it is being used.
It gradually drops to ~3v where it's considered discharged.
3.7 is just the 'median' voltage.

>>1576015
A fully-charged lipo will hold 4.2V. By filtering I mean getting rid of ripple. While a passive filter like a capacitor will get rid of a fair bit of ripple, if you have a lot to get rid of the capacitor size may become unreasonably large. A linear regulator will have a lot of ripple rejection for its size, at the cost of wasting some power within it. If you go with a cheap buck or buck/boost converter this may be necessary.

>>1576019
Yeah, for once I should have researched the topic before pulling it out my ass.

>>1576028
Wait. Are you suggesting to add a linear converter down to 3.3 after buck converter?

I should have also mentioned that power limitations are strict, as I need it to last as much as possible until battery swap.

>>1576031
Either a buck converter to 4-5V or so then a 3.3 LDO from there, or a 3.3V regulator right from 8.4V to 3.3V. If you have the space to, the former is obviously a better idea. But if the ripple from your buck converter is acceptable, you won't need the regulator.

can someone explain to me what circuit loading is, nothing I'm reading makes much sense

>want to start working as an apprentice electrician
>order NFPA 70 and some other stuff
>have basic understanding of electronics, but don't remember the math yet
>want to actually start working
>it's a literal fucking shit shoot on finding an apprenticeship
>only one that I've been able to find is IEC And pay is 11 dollars an hour
I'm out of DFW, and am completely lost, none of the unions have fuck all information anywhere, and the only local one hasn't updated their website in 6 years. What do I even do.

>>1575939
SEPICs aren't that inefficient

>>1575718
>educational
>nothing is explained about how they work
2/10

>>1576052
sparky tfw no job whining goes elsewhere

>>1576052
join the ieee

How do you guys remember everything? Do you always just keep references close, or do you just actually do it so much that you don't end up forgetting? I always seem to have huge gabs between each time I do this stuff and forget literally everything.

>>1576063
I can remember every pokemon, their type matchups, their special abilities, and most of their move sets.

It's trivial to remember electronics stuff when you're surrounded by it all the time

>>1576066
I guess it's just me then, I am unable to remember anything, even game related after about a week away from it I forget everything. Even in my last electrical job I would forget certain wiring a few days after I did it. It's made it a nightmare working in technical shit.

>>1576063
>Do you always just keep references close, or do you just actually do it so much that you don't end up forgetting
yes
I like to say electronics design is half physical science, half library science. that includes knowing how to index and read references effectively (I have about 3GB of datasheets saved and sorted by manufacturer, and someday I'll turn Solr loose on them) and also annotating and documenting one's work, which is very helpful to airheads like me
but it does help to be around it all the time, no doubt

>>1576053
>nothing is explained about how they work
The purpose is to develop abstract thought like how Lego develops fine motor skills and spatial reasoning without teaching you about the tolerances and draft anglles of the bricks themselves. It's like littlebits but less delicate

I just spent 2 hours trying to make a 555 oscillator do some duty cycle when I push a button and I'm going back to using a comparator oscillator. Fuck 555s, fuck their quiescent current, fuck their awful pinout, fuck the dependancy of frequency and duty cycle to the same components, and fuck their input impedance. I've got to the point where I either use that diode circuit to make the duty cycle less than 50% so I can pull the reset pin low to leave the LED off instead of using a PFET on the output instead, or continue using the PFET with an inverter on my button to try and pull one of the other pins low to leave the output pin high such that the PFET keeps the LED off. Both of those will use more passives than the comparator method, and are far harder to tune. Most of the time was spent trying to calculate what resistor values would give both the closest duty cycle and the closest frequency to my specifications, since they're not variable independently of one another.

>>1576169

use the old astable-followed-by-a-bistable trick and quit your whining, faggot.

>>1576169
>actually doing the calculations by hand
You're doing it wrong.

>>1576144
It would be important to use some sort of connector such that Vcc and Vee are always connected properly, and that the signal line(s) would always be safe for the devices themselves. Perhaps using a MCX or RCA connector would be good for power, so the poles can't be reversed.

>>1576204
I made a fucking spreadsheet that gets me the numbers, but knowing "exactly 3.279kΩ" doesn't exactly hope when all I have is standard value resistors in the parts bin when I need a frequency precise enough to trigger an IR receiver. In the end I decided on 2 220pF capacitors in parallel, with a 68kΩ resistor as R1 and 10kΩ trimpot for adjustment as R2, but the pF capacitance must have been too low for the input impedance, which set my duty cycle significantly lower, postponing my efforts further. It was then that I stumbled into the button issue.

>>1576201
And use a larger IC and more passives than a comparator oscillator? Not to mention higher quiescent current, higher capacitance requirement, worse pinout, etc. Sorry, but at this point there is zero reason to ever use a 555 timer for me.

>>1576211
>there is zero reason to ever use a 555 timer for me.

except for this: the 555 will be available 55 years from now, and you'll just have to pop in a replacement if it dies, and not have to rework the circuit coz it'll follow the same formulas as it always has.

>>1576211
>there is zero reason to ever use a 555 timer for me.

except for this: the 555 will be available 55 years from now (with better specs), and you'll just have to pop in a replacement if it dies, and not have to rework the circuit coz it'll follow the same timing formulas as it always has.

with a comparator, just replacing it with the exact same part, but with a diff date code, will throw the circuit completely off.

>>1576211
>using a 555 timer in a high precision application

lol. full retard. Don't bother using any relaxation based oscillator for high precision applications. Component tolerances are too shit and frequency will drift with temperature. Crystal oscillators are probably the easiest option and if no crystal exists that matches your IR receiver frequency then synthesize it with either a PLL or a prescaler. If a regular crystal isn't good enough for your IR LED (I actually suspect it's overkill and the actual reason you can't get your shit working is because you're a retard but I digress) then look at TCXOs and OCXOs.