/diy/ - Do It Yourself

>have to take up to chemistry for engineers to start electrical engineer major courses
what is this shit I'm not going to be touching anything chemistry in my field related

>>1446678
you will be using acid to make PCBs numnut

>>1446678
education is a scam. bend over and take it up the class you normie. not only are you taking things you don't need, you are paying for them as well. and that doesn't even include textbooks. enjoy paying your college loans off, wagecuck.

>>1446678
Chem is fairly useful and interesting stuff. In particular, electrochemistry is pretty damn interesting, as it goes into both galvanic and voltaic cells. Avoiding galvanic reactions between different metal conductors in a humid or salty environment could be a topic covered.

>>1446665
>number of insertions vs number of pins remaining
lol

routing a 315MHz receiver board for the first time. could the proximity of the extra metal from the crystal oscillator be an issue? do i need to run a ground under the microstrip across the (datasheet specified) gap to the antenna?

>>1446878
>315 MHz
That's a lot of potential signal problems.

I wanna drive 10 3W LEDs with dimming from a PWM signal. Anyone know the easiest way to do that?

I'm thinking about getting 10 of these https://www.aliexpress.com/item/DC-DC-5-35V-To-700mA-Step-Down-Buck-Power-Supply-For-Arduino-CC-CV-Step/32828673347.html and getting a 12V 5A power supply.

I also have a AC to 33V converter I pulled out of a busted smart lightbulb. Maybe I could get away with just one of those drivers and putting them all in series?

>>1446902
https://www.electronics-tutorials..(double-you ess)/transistor/tran_7.html

+ 555 circuit

I can show you a broken circuit, it uses a 555, three resistors (four?) and a capacitor, and it can do 0-100% brightness with a 0-Vcc input (not linear though).

Or you could buffer a digital thing and drive a mosfet.

Look for logic-level, trench, something else.

>>1446902
LEDs are current mode. voltage and current are interrelated, and voltage reduces with increased temperature. you'll need a current mode driver to avoid shortening their lifetime
use a proper constant-current LED driver chip or module, like this one
https://www.aliexpress.com/item/y-u-no-like-benis/32850598517.html

>>1446878
I would assume not. if they say gap, I'd gap it

>>1446902
Do they have to be dimmed separately? Assuming not, and that they're white or blue you'll have ~30VDC across them if in series, so it might work to use that smart lightbulb converter plus an LM317 (with heat sink), provided the converter is constant voltage in the first place. In this case you'd just vary the current to dim the bulbs instead of PWM, which is arguably a better solution. But there's also a potential situation where all LEDs in series want to pull more than 31.75V at 1A, where you'll either have to go for a low-dropout current regulator of some sort or ditch the setup for something more practical than a reused PSU.

>>1446912
Or use resistors, like LED strips do. Not saying it's the best method but it certainly works. Since he should be driving them with 1A a proper CC switching converter is likely the best decision, but 30W total would mean it wouldn't be that cheap.

>>1446918
>30W total would mean it wouldn't be that cheap.
oh, shit, 3W. damn
well, there are dozens of ICs made just for driving strings of 3W white LEDs, but now we're getting out of the 'simple' realm, so a big fat resistor and a big fat MOSFET might be the right way to go in this case

If i want to switch dc poles back and forth in a coil fast enough to heat up metal shit i put in it, how do i do that?

>>1446949
did you even try to google "diy induction heater" first

>>1446951
Bully.

>>1446949
It's probably unfeasible for you, an induction heater is a fairly advanced project. Great return though, they can be used as a furnace.

Why is that stupid led burning instead of blinking?

>>1446990
presumably because no current limiting resistor in series

>>1446990
Might have something to do with the 150mA of current you're putting through it. Most LEDs won't do more than 20mA continuous and can generally handle 60-70mA when pulsed with a sufficiently short duty cycle though this varies and you should really check your LED manufacturer's datasheet to see the rated pulse current. The specification you're looking for is Peak Forward Current and it's usually specified for a 10% duty cycle at 1kHz.

>>1446994
>>1446995
Nah, connecting the led directly to 2V is fine in the simulator. I am confused by the OUTput of the 555 timer.

I feed 2 into it and i added a cap to create a simple square wave.
So on the out pin is should be getting 2V (peak of the square wave) or 0V (bottom of the wave) but if you look at the graph, the square wave ranges from 2V to 2.21V... how is that possible?

>>1446996
The falstad simulator isn't terribly good. I'm not getting results that are even close to yours. I made sure to pick a 555 timer that could work down to 2V, not all of them do and I have no idea how the components in your sim are modeled and what parameters you have control over.

>>1446998
Ok thanks, looks like it will be better for learning to actually breadboard things instead of simulating them

>>1446999
or just use spice

Hey guys,
I am currently building a portable speaker and I have this battery pack.
Now I want to replace that switch (red circle), so I can easily turn it on and off.

My second option (in case this is not as easy as it seems) is a hole, big enough to operate the switch, but I'd prefer replacing the switch.

>>1447019
So? Just do it. Switch is just two interrupted wires.

>>1447019
may as well just leave that switch on and add another one elsewhere in the power loop

>>1447024
Like this? Won't this drain too much power when "switch 2" is turned off?

>>1447023
I worry that I might damage the battery, but I assume that is not necessary?

>>1447028
The mechanical removal of the switch has nothing to do with electronics really.
As for the switch it self any switch will do, no switch you pick will damage the battery.
The only problem would be if you used say a tiny push button that can't handle the current.

So I found out that 90% of wall charger are trash. The all boast fancy current ratings at certain voltages, but the second I try to draw even half that the voltage plummets by 20% like nothing.

>>1447028
>Won't this drain too much power when "switch 2" is turned off?
no. if you interrupt the loop in one place, you interrupt the whole current flow

>>1447034
yep. chargers are not necessarily designed to be solid power supplies. as long as it meets the charging V/I profile, the manufacturer is happy. a 20% drop isn't going to hurt charging very much

>>1447035
>>1447032
Alright, thanks guys!

Is bit-banging an I2C or other interface as simple as just flipping the inputs high or low in a clocked order in your program? I mean that's what the protocols are all about in the end but actually programming it seems silly

>>1447069
>but actually programming it seems silly

your question made sense up until you threw this in.

>>1447069
>as simple as
sure. there's a fair bit more handshaking to I2C than that though, because of clock-stretching, ack bits, etc. bit-banging is pretty silly if you have a hardware interface to do all the work for you, but sometimes you don't...

>>1447070
>>1447072
I mean of course it has to be programmed as the standard says, I just have some sort of mental dissonance with real-time programs

Why do people use fans when soldering? They just suck away all those amazingly smelling vapors. I just can't get enough of them!

How much continuous current can bread board cable handle?

>>1447095
http://www.powerstream.com/Wire_Size.htm
about an amp per lead

>>1447102
says the chart is for AC only and they don't say if the thickness is with or without isolation

>>1447103
no, just the skin depth rating is meaningless with dc
insulation doesn't have anything to do with the conductor. also that's a matter of voltage, not current. different materials, different charts

>>1446678
chem and material science are useful and should be taught more in highschool

is there a way to make a common-collector amplifier with current gain that is well-defined/ie. not beta-dependent?

>>1446678
electrical materials is one of the deepest wizzardries there is in the field.

>>1447263
you can use op amps if you can supply a negative rail. this circuit is ridiculous because the current gain of a transistor is usually just a middleman that op amps can cut out, but it does what you're after.

>>1447274
forgot another 100k resistor connecting the second noninverting input to Iout

>>1447263
Negative feedback my brother.

>>1447304
Can you do negative feedback for a transistor like you do an op-amp where you literally feed your output back to the input or are you restricted to emitter degeneration?

>>1447410

you can do it for a transistor, or a series of transistors, or whatever gain stages you got, as long as you're using an inverted signal.

Does anyone know if there is a zif socket component in Kicad?

Sorry, accidentally posted in the old thread. Hi, hoping to get some ideas on how to achieve the following: I have a car accessory that normally runs illuminated at 12V. When I flip the headlight switch, the dim circuit lowers voltage to about 10.1V. However this voltage drop doesn't seem to be sufficient to trigger the accessory's dim function.
Is there any simple way to have the signal be 12V normally, and drop further once the switch is activated? I tried putting a diode inline, but that only dropped the overall brightness- I want to keep max bright voltage the same, and drop the dim voltage further.
Any help/advice would be appreciated. Thanks! pic related

>>1447517
that's odd, there are usually two separate filaments for high and low beam. what kind of car is this?

>>1447517
>dim function on same line
Then you fucked. Unless there is some option in car menu for dimming or voltage is being switched somewhere via relays (or other means).

>>1447517
Use op amp to route the power thru a resistor if the voltage drops below the point you want

hey guys,
does the soldering flux have any relation to the soldering flux plumbers use for their cuprum pipes?
is there any problem using the latter instead the former?

>>1447617
>cuprum
a.k.a. copper... i didn't remember the "english" version so I used the latin originaru one.

>>1447617
yes, don't. acid flux suitable for pipes will destroy your thin wires
can't you get rosin flux? doesn't your electronics solder have a rosin core inside it?

>>1447628
Has anyone ever used copper pipe as a conductor? it could be used in a /diy/ fashion as a replacement for a bus bar.

I see here a 1/8" x 3/4" bar for one foot is about $55 at one store.

I see a 2" diameter copper pipe, for one foot is about $15

Maybe they gouge people that need bars, since they are less common and more industry specific, but I know the pipes are thinner. In low amperage situations, the pipe might be easier to procure, install, and maintain.

Looks like others have already covered this:
>https://evmc2.wordpress.com/2011/07/10/copper-tube-bus-bar-idea/
>https://www.jewtube.com/watch?v=jP1vVb-z0yc

>>1447682
It's a pretty common choice for induction heater work coils.

>>1447682
the bus bar designed for electrical applications is usually pure virgin copper. the copper pipe or tube designed for fluid handling probably won't be.

>>1447691
>pure virgin copper.

I thought this was bullshit, but "virgin" copper has never been recycled.

>>1447682
>Has anyone ever used copper pipe as a conductor?

>>1447718
https://www.youtube.com/watch?v=kzpc3-Pj9jg

I want to make my own simple and slow (data rate <1200 bps) serial data bus. It would be used for on-off-controlling addressable devices attached to the bus.
Do you think the circuit shown on SD12 interface specification is total overkill for this? Is there anything simpler?

>>1447750
The complexity of that thing is almost entirely due to protection shit. Omit protections according to your desired level of fragileness.
If your devices consume very little power, you might be able to drop the regulator too, and use 5V supply instead of 12V.

Personally, I've usually used RS485 for similar purposes. It needs more wires (2 for data) and unless you use slew-rate limited drivers, you need to pay attention to line terminations, stub lengths etc.

>>1447750
>It would be used for on-off-controlling addressable devices attached to the bus.
>Is there anything simpler?
Depending on what kind of devices your want to adress, for a 3-wire bus I would rather consider :
- RS-232 (strict or at TTL levels) with only TX,GND, and a VCC suiting your remote devices. If a 4th wire is possible, consider adding a RX, or a RIng wire so the remote device can acknolwedge it was diddled.
- I2C : this is the industrial workhorse for such a job, works with 2 wires (Bi-dir Data, Clock) and a common ground. Doesn't carry a 12V line, hence an extra wire may be needed depending on how your remote devices are powered, but to be fair there are so many implementations floating around (tarduino, linux, MS-DOS/whatever) it will be less of a headache than anything else.

this made me curious..
since ground is literally a wire stuck into a ground.. how much ground do you actually need to have for it to work?

>>1447851
About the size of a planet. Self-capacitance is proportional to radius. The self-capacitance of Earth is ~700 uF, which corresponds to an impedance of ~3.8 Ω at 60 Hz.

>>1447851
Ground is a circuit that returns back to the supply transformer, a stick in the ground at the transformer and another stick in the ground at the building is a pretty shitty implementation, needsto be pretty long to get resistance low enough to properly pop the fuse/circuit breaker.

>>1447935
Yes and no. Some distribution systems do use a ground as a return path, but this is between the generator and substation, or substation and transformer, not from transformer to house. In a house you'll have both split-phases coming in from the transformer along with the centre-tap; neutral. Both at the transformer and at the distribution box at your house there will be grounding rods from the neutral lines into the earth. But this ground pathway is not relied on for carrying current, it is for the emergency case when there is a neutral fault between the transformer and your distribution box.

But the ground connection is also used to sink high-frequency interference and reference an otherwise floating voltage source to a safe 0V, which is the intended purpose in the suicide-shower image. By running water past live before then neutral and then a ground wire, you are ensuring that the water coming out of the shower is referenced the same as any copper pipes that you yourself may be electrically connected to by virtue of standing in the puddle in your shower, preventing a shock. While the current will typically flow through live to neutral, thus not requiring a ground wire, if a neutral fault appears the current will instead flow through the ground, triggering your RCD/GFCI instead of flowing through you. A large enough mass of earth on this wire would prevent you from shocking yourself through self-capacitance, but best it is instead connected to the ground/neutral wire in your distribution box, where a fault would trip the breaker, or connected to the neutral or a ground rod in which case the breaker will not be tripped but you still won't be electrocuted. Having a huge self-capacitance does not specify where the voltage will be referenced to, and in the case that your post-transformer distribution system is floating with no ground reference, it could possibly have a potential closer to live than neutral, or even higher.

I have a remote control from a LG blue-ray and doesn't works. Anyone knows if these missing parts from this green lines could be the problem? I already try to clean this shit usingIsopropyl alcohol.

>>1447948
Light up your subject for better depth of field, exposure, and generic clarity.

>>1447948

it looks like the trace was eaten away by acid, but since it's located under the green solder mask, then it probably came like that out of the factory, and is probably ok. if you had a tool to do a continuity test, you could have tested it.
i'd look elsewhere for the problem.

>>1447947
The ground connection is to carry current in case of a fault, enough current to trip the circuit breaker/fuse.

>>1447949
I have a remote control that doesn't works(I thought the problem was dead bateries, but I already buy new ones but nope, that's not the problem) and I have no idea what could be the problem and I am here to get some help from you guys.

Before open up this control to see what the heck was the problem, I didn't had any kind of answer when I was pressing the buttons. After clean the printed circuit board using using Isopropyl alcohol I notice that arrow keys works now, but not the other keys, so I guess that I am on the right way.

I am the same faggot from threads ago that was quering abot how to fix this control remote.

>>1447955
>if you had a tool to do a continuity test, you could have tested i

Nope, I don't own such thing.

>it looks like the trace was eaten away by acid

Seems like you are right:
Here you can see an early photo of the printed circuit board>>1447960 before I clean it using Isopropyl alcohol. Do you think that those missing parts could be the problem? Any idea how I can fix that? thin Aluminium foil? a pencil? soldering?

>i'd look elsewhere for the problem.

Any better thread to ask this? I am a bit new on this board.

>>1447959
I did say that.

>>1447960
That stain is probably just burnt soldering flux. I'd make sure to clean the conductive pads on the other side of the rubber membrane keypad.

>>1447966
I aready did that, but seems like didn't help a lot.

>>1447960

>>1447969
>Falling on Planned obsolescence

Anon...

Could someone help me with creating a MIDI controller and the idea behind it?

>>1447960
>arrow keys works now, but not the other keys

that's a strong indicator that the pins on the big chip at the bottom need to be resoldered coz one or more legs have come loose.

the fact that squeezing the unit causes it to sometimes work confirms this theory.

disclaimer: if /ohm/ is anything like /dpt/, you get a bunch of posts from drooling retards asking the most inane shit every day. this is one of those posts.

hi /ohm/. I recently got interested in electronics after finding some old components (assorted ICs, jumper cables, breadboards, resistors, etc etc). right now I'm reading this https://www.allaboutcircuits.com/textbook/, which is probability shit if electronics content is like programming content (if it's not available as a book, it's worthless), but I'm liking it so far and it doesn't seem to be saying anything dumb (from what I can tell with my previous physics/digital circuits/google knowledge). hopefully I didn't pick too terrible of a resource to read.

I'm stuck trying to reason about the voltage between points in a few circuits. pic related.
in circuit A, the voltage between any points in the set {1, 3} and the points in {2, 4} would be 9v (ignoring the sign), correct? the current is 0, because the circit is open.
in circuit B, I reason that the voltage between points 1 and 2 is 0v, because they are electrically common, and through this reasoning, 1 and 2 are electrically common with 3 and 4 (respectively) and so the voltage between 3 and 4 must be 0v. but on the other hand, it would seem like the voltage between 3 and 4 should be 9v, because that's the electrical potential difference between the two terminals of the source. I'm very much leaning towards it being 0v (ignoring the resistance of the wires), with the current being +inf. but why is my second line of reasoning flawed?
circuit C is what lead me to think about all this, because I was reading about parallel circuits and wondered what would happen if I simply removed one of the resistances. I'm assuming that the answer to my question about circuit B will help me understand circuit C, but I included it anyway (in case anyone has anything to comment on it).

(cont)

>>1448016
>in circuit A, the voltage between any points in the set {1, 3} and the points in {2, 4} would be 9v (ignoring the sign), correct?

Correct, zero volts across an ideal wire.

>in circuit B, I reason that the voltage between points 1 and 2 is 0v, because they are electrically common, and through this reasoning, 1 and 2 are electrically common with 3 and 4 (respectively) and so the voltage between 3 and 4 must be 0v. but on the other hand, it would seem like the voltage between 3 and 4 should be 9v, because that's the electrical potential difference between the two terminals of the source. I'm very much leaning towards it being 0v (ignoring the resistance of the wires), with the current being +inf. but why is my second line of reasoning flawed?

If you assume both ideal voltage source and ideal wire, you get V3-4 = both 0 and +9, a contradiction, so the assumptions must be incorrect. In reality both the wires and the battery have resistance, Rwire << Rbattery, so V3-4 ~= 0, I = 9/Rbattery.

Similar thing applies for C, in reality R1-2 would usually be << 1 ohm, so Iresistor ~= 0, Vresistor ~= V1-2 ~= Vbat, I1-2 ~= 6/Rbattery

>>1448016
lastly, i have a question about something that'll probably never be relevant in my life, but that intrigued me. it's about something said on this page: https://www.allaboutcircuits.com/textbook/direct-current/chpt-3/common-sources-hazard/

they say this about pic related:
>If the distance between the two ground contact points (the downed cable and the system ground) is small, there will be substantial voltage dropped along short distances between the two points. Therefore, a person standing on the ground between those two points will be in danger of receiving an electric shock by intercepting a voltage between their two feet!

>One practical precaution a person could take if they see a power line falling towards the ground is to only contact the ground at one point, either by running away (when you run, only one foot contacts the ground at any given time), or if there’s nowhere to run, by standing on one foot.

what I don't get is why you'd be shocked standing on both feet. you would be grounded, so would the voltage between your body and any point on the ground be 0v? why would the current suddenly change its path to go through your legs? I understand that a human body is probably less resistive than the ground, so current would prefer the legs, but how can they be "pushed" through your legs if there's nothing pushing them (thinking of volts as the pressure of water in the water flow/electricity analogy)? this, however, conflicts with what I "know" about current going to infinity when the voltage is 0. also, why would standing on one leg make you safe? why wouldn't the electricity simply go through the base of your foot?

>>1448021
>Correct, zero volts across an ideal wire.
you mean nine, right?

as for the others: basically, just don't try to reason about ideal short circuits because it doesn't make sense. glad to know I wasn't going crazy.

>>1448022
The ground is a pretty bad conductor, much worse than the downed wire that's still pumping current into it, so you'll have a big voltage gradient along the surface of the ground.

>>1448023
Zero volts along each wire, nine volts between them.

>>1448024
thanks

>>1448016
It is pointless to consider a situation with a short-circuit like B and C while holding onto your "perfect wire" and "perfect voltage source" concepts, since the two are mutually exclusive. Either the voltage source will limit its own current, the resistance of the wire will become significant, or more likely, both.

I'd look up the terminology used for nodal and mesh analyses, since that's somewhat applicable here. Doubt you'd care to learn the actual analysis mechanisms themselves though, a bit of a nightmare lest you're proficient with matrices and vectors and the like.

Otherwise, thinking in terms of Kirchoff's voltage law is pretty handy, with voltage being dropped by each component.

>>1448014
>the pins on the big chip at the bottom need to be resoldered coz one or more legs have come loose.

I dunno bruh, are you sure?, the pins looks fine.

Also:
If I need do it, how I can do it? I can't see any pin on the other side of the printed circuit board where is located the big black chip.

>>1448055
you solder SMT on the component side. clean any solder/slag off the iron tip, put a very small amount of fresh solder on the tip of the iron to start, and touch the tip of the iron to the end of each lead just until it melts, working your way down each row. give the IC 30 seconds to cool before doing the other side just to be safe because I don't know your equipment

>>1448060
>I don't know your equipment

I only own a pencil tip soldering iron.

I am fucked?

Also:
What does mean SMT?

>>1448065
Surface mount technology. It means parts that are mounted on the board instead of with pins that go through it. The term SMD (surface mount device) is arguably more common.

>>1448065
it can be done. take care not to heat the IC any more than necessary and avoid bridging two pads together. try sliding the tip on the board, along the axis of each lead, then pull back as soon as it melts. do not hold the tip on the pad/lead for more than 3 seconds
see 9000.5 hours in MS Paint related

How would i go about detecting clapping?
I want to turn on shit when someone claps.

So i was thinking, microphones generate some voltage and current when they hear sounds right?
So how about connecting a microphone to an op amp and setting the reference voltage to be same as one produced by a clap sound. Would that work?

>>1448073
What does mean those numbers?

>>1448084

never heard of The Clapper? i got a couple from the thrift store for a buck a piece.

>>1448090
They don't sell those here and ordering them online costs money and takes a while. I want to build that shit right now with just a regular microphone

>>1448086
numbered list of directions. slide in, then slide out

>>1448092
no, it wouldn't
look up clapper circuits and see why

>>1448092

right, but it's gonna be a difficult job distinguishing a clap from some other sound. gonna require a lot of experimentation. perhaps you can find a schematic somewhere online, or copy someone else's work who's done a similar project. a long shot, but better than playing with filters and timers for a week.

a quick google search shows it's been done before.

>>1448095
So if i want to trigger it on any loud sound, not just claps, then it becomes super easy basically right?
Literally just the mic and the op amp

>>1448096

yes, but probably annoying to have lamps and shit going on and off while you're dogging your boyfriend.

>>1448096
why don't you just try it and see

>>1448098
Yeah.. shit... one option left then, is there some simple way (with like caps, resistors, and other common components) to detect the proximity of human skin through 2cm of wood?

>>1448101
found it!
this will be even better than getting clap

>>1448101
I think claps are typically far louder than any other sound you'll hear inside a house, I don't believe clapper circuits have any sort of filtering to that end.

For your capacitance solution you'll need a high resistance R and fairly high frequency AC. You can either use the person as a capacitive route to ground, thus decreasing the voltage after the resistor, or use two plates that for which the person's proximity increases the total capacitance, hence letting more current through and detecting a voltage drop this way, or simply sense the person's proximity via AC mains coupling. There are also optical and ultrasonic methods.

The fuck is this pitch?

I have an odd idea for a project. I can mostly get by with a small SMT atmospheric pressure sensor, but is there such a thing as a tiny affordable air flow measuring thing? We're talking a max of maybe 6 cubic inches.

>>1448134
Two pressure sensors in a known pipe make a flow sensor, provided laminar flow.

>>1448124
intended to be mounted as a component on a motherboard

I am struggling to understand how a low/high filters work. Basically you put a cap in series with a microphone, to filter out the power noise for the microphone. But how does that work? From my understanding power flows through the cap until it chargers, and then it stops the current completely. So then shouldn't it block all power from the microphone completely? How does it get discharged?

>>1448139
You seem to completely lack basic knowledge or your low/high/band pass filters with mic are not the example you wanted to make.

Capacitor in series with microphone allows for AC (audio) signal to pass while blocking DC voltage (power supply, bias etc)

>>1448124
2mm?

>>1448139
Look up what impedance is.

After that it should become clear to you that passive filters are simply frequency dependent voltage dividers.

>>1448096
>then it becomes super easy basically right?
easy for me, apparently difficult for you - that's why you're here

>>1448139
You're right about how caps behave, but only in the DC case. When there's an AC component, the capacitor charges and discharges in proportion to the AC signal, so there's still current.

I would recommend Physics for Scientists and Engineers by Serway and Jewett. Read the RC circuits section.

>>1448139
>>1448252
And if you don't already know how to steal textbooks: http://gen.lib.rus.ec/

How do you finalize your circuits?
You start on a breakboard, but when you want to make the final form what do you do? I assume not everyone has access to PCB making machinery.

I bought a TIG welder pedal, like this one:
https://www.ebay.com/itm/TIG-Foot-Control-Pedal-7pin-connector-For-ATPW524-ATPW522-TIG200P-WSME-CT-Welder/232859854931

It works, but it has a weird intended/unintended behavior. Pic related is how it's wired up, the right pot is the foot pedal control with a 2k pot, and the adjustment knob is the left one with two wires, 10k.

What happens is that the adjustment knob(R2) alters the minimum amps on the welder, and the foot pedal(R1) works relative to it. For example, at 10%R2, I get a minimum of about 30A on the welder, and a maximum of 32A. This is completely useless. At 60% R2 I get 60A min, 80A max. At 100% R2 I get 120A min, 250A max. This is without touching the welder itself, welder being set to minimum current.

What I'd want to happen is using R2 to limit the current range, but not changing the minimum current at all. So if I set R2 to 10%, I get 10A-20A, if I set it to 50%, I get 10A-100A, and so on.
Can this be done relatively easily with passive hardware only? I thought about doing it with an Arduino and either PWM or a digital pot, which would also have the advantage of custom range mapping, but if I can keep it simple I'd rather go with passives.

(In reference to address pin numbering on parallel memory chips):
Would (for example) A0 be considered the LSB, or would A12 (or whatever the highest pin numbering happens to be)?

>>1448287
You can etch PCBs without any expensive tools, but for simple things I just use the basic prototyping boards, can get them cheap from chinks.

>>1448291
What is your favorite prototyping board store on the Chinese citadel?

>>1448287
You don't manufacture the boards, you send your files to a fab house and have them do it. For money.

>>1448288
No idea. What is this circuit connected to? Please label it better too.

>>1448289
Prolly A0 is LSB but I'd check the datasheet.

>>1448293
>For money.
How much will they pay me to make my designs?

>>1448294
Depends on size, how complex the board is, etc. For hobby stuff you can get one less than $50 easy.

>>1448295
>$50
Sharpies and acid here i come.

>>1448287
Kinda depends on the nature and complexity of your circuit. Most people will just design their own boards and then pay a company to make it. It's generally pretty cheap unless you want a large board and or lots of layers. Some places also make you buy a minimum number of boards, so you can't just get one. That said there are more companies that make custom PCBs for the hobbyist market nowadays so restrictions on minimum order sizes aren't as universal as they used to be.

You can etch your own PCB using chemical solutions that aren't particularly expensive. I've never done it so I don't know what it's like but I'd probably say you shouldn't too it for a particularly complex circuit.

Manhattan style construction is great for RF stuff and often beats a properly constructed PCB.

There are prototyping boards that are basically single sided PCBs that you can solder components to. Some are even laid out like breadboards with the buses on the side and the connected strips on each side in the middle. Sometimes people use these for their final designs. They're fine for low speed applications.

Finally there's point to point wiring and wire wrap which can be a pain to wire and like the above is not suited to high speed analog or digital signals.

>>1448293
>>1448288
The DC comes from the welder, 20V. The 1k resistor is a stand-in for the welder's signal input, it's not actually 1k but it shouldn't matter, the welder reads the return voltage and changes the current accordingly. And R1/2k is the pot in the middle, R2/10k is pot on the left.

>>1448296
I'd say it's worth. Have a professional-looking final product you can take pride in and show off to employers.

>>1448292
I just buy from whoever has those green boards for the lowest price. I bought the cheaper orange-brown boards once, but those are complete garbage. Aliexpress tends to be the cheapest.

>>1448292
not that poster but PCBway will send you free/cheap prototyping boards on request with your custom PCB order, and the custom boards are decently priced, like $5/5 for quantity (5) 5cmx5cm, double-sided, 6mil trace/space rule, plus shipping. in burgerland, even with DHL it often works out cheaper than the local PCB pool OSHpark
see also pcbshopper dot com price comparison engine

>>1448298
Ok. Now put that on the schematic so we don't alll have to do it ourselves.

>>1448296
If you make efficient use of your PCB real estate then you will most likely spend less than $50.

>>1448304
>>1448288

I just posted this in qtddtot but figured I'd post here too.
I'm trying to build pic related but I don't fully understand plate-transformer-output impedence matching. I have 60ohm headphones and since I don't use speakers I wanted to change the output (from 8 or 4 ohms) to something useable. I know I need a different output transformer but I don't know how to look for what I need. Should I just build it first to make sure it works and then focus on transforming it to a headphone amp?

http://www.dogstar.dantimax.dk/tubestuf/miniblk2.htm

That's a link to the build if you want to look at the whole thing.

>>1448308
Is the bottom left node floating? Am I understanding correctly that the 1k resistor isn't actually present, and its two ports are the inputs to the welder? Are they the only two inputs, i.e. the welder is getting all its info from those two ports?

>>1448313
I imagine it has some resistance against the ground, I haven't measured it. The welder only receives one signal cable, the output of R1, and the welder provides the 20V DC and the ground. Whatever the voltage (or resistance) it measures is what makes it set the welding current. It's the same operation as any other case of a pot hooked up to control something indirectly.
Pic related is what it looks like in practice, there's no physical 1k resistor between ground and welder input within the foot pedal.

>>1448313
*Also the bottom node on R2 is floating, not connected to anything.

>>1448309
With transformers impedance is the square of the turns ratio Zp = [(Np/Ns)^2]*Zs.

Primary load impedance as configured in your diagram appears to be about 4.9k. I'll assume that's an optimum figure and we want to keep that. That means for an output impedance of 60 ohm your primary turns, Np, will equal sqrt(81.67)*Ns. So you'll roughly want a 9:1 transformer assuming you are using the same, or very similar core.

Alternatively, keep the transformer you have and just use an op-amp, emitter follower, whatever to buffer the output and provide a nice low output impedance so your headphones won't load it down. These days they don't use impedance matching for audio anymore, it's only used for RF. You pretty much always want low impedance outputs feeding into high impedance inputs. I think the technical term is "impedance bridging?"

>>1448315
If the welder input is sensing voltage it might be possible to get it arbitrarily close to what you want. You'd have to rearrange the circuit and maybe swap out the pots for different values.

>>1448320
I'm glad that you responded but I think I'll need that in retard. What do you mean by primary load impedence? What do you mean by "assuming you are using the same, or very similar core"?
My understanding of opamps is much less then my understanding of tubes, (which is only mediocre) what would do to provide a low output impedence?
I was trying to understand eevblogs video on opamps so should I watch that and then I'll understand?

-A very grateful anon

>>1448322
I thought some arrangement like this might work, but it might need, as you said, different pots, as well as some extra resistors. Maybe I'll just go with the arduino since at least I comprehend how that works.

>>1448296
i've had a few cheap 4x4" boards made for $5-10 each from elecrow. you can get quality gold plated boards for $5/sqin from oshpark.

>>1448296
not worth the trouble mostly
>>1448402
there's also pcbs.io which is cheaper than oshpark but still in US.

its good to not outsource everyone's job to the far east. shit made there always ends up subpar and half broken if you buy enough

>>1448296
>$8 of etchant
Pencil and dremel here i come.

>>1448416
>$0.50 pencil, $0.02 electricity
Board, meet fingernail

Someone gave me this:

40 NOT gate chips
40 OR gate chips
40 NAND gate chips
40 AND gate chips

What can i do with them?

>>1448533
Not much that isn't a gimmick. Cut the pins off and see how many you can fit in your ass.

>>1448073
>>1448093
Nevermind. I already fix it:

Do you remember that the trace was eaten away by acid on a few parts on the printed circuit board? Well, those missing parts was the problem all the time: I only fix that putting thin strips aluminium foil on the missing parts. Now my remote control works fine. Whatever, thanks by the help.

Anyone know where I can find male-male pin strips for these SIP sockets?

I'm trying to turn a square wave into a sine wave, I know I can use an array of 3 simple low pass filters(Capacitor + Resistor) but I don't know what capacitors/resistors to use

I'm trying to convert a 35-1000 hz signal outputted by an Arduino Uno

Anyone?

>>1448662
A filter has a certain characteristic "corner" frequency, given by 2πf = 1/(R*C). To filter a square into a sine, typically we want close to 1.0 gain at the fundamental frequency (f), and as low as possible for the next harmonic (3f for a square). Problem is, this means the filter values will need to change depending on your output frequency. Other methods which avoid this include: ~100kHz PWM/class-d style DAC with 1kHz filter, some sort of variable ring oscillator, PLLs probably, or go about your plan without the use of a sine in the first place. Adafruit probably make DAC shields if you're really desperate.

>t. brainstormed the same thoughts but decided to just buy an EL inverter instead

>>1448662
You've got a square wave. It's comprised of sine waves at the fundamental frequency f + sines at 3f, 5f, 7f, etc. So filter the higher order sine waves so you're left with the one you care about at f.

https://en.wikipedia.org/wiki/Low-pass_filter#RC_filter

Pick some capacitor value you have lying around like .1u and choose resistors based on that.

>>1448668

>>1448667

Alrighty

>>1448669
What do you need it for? Also stop the reddit spacing.

>>1448662
search for Digital to analog converter (DAC)

>>1448662
For better results put it through an integrator first. That'll give you a triangle out which you can than low pass filter with the cutoff just above the fundamental.

Integrator is basically identical to the op-amp inverting amplifier except you replace the resistor in the feedback path with a capacitor.

>>1448670
Don't even use Reddit

I just space things like this
Specially my code

I like grouping thoughts

Or sentences

It's nice
Y'now?


Anyhow, I'm just trying to generate a sine wave whose frequency I can alter with a potentiometer so I can then distort it

I have an issue with some microcontroller code I'm writing. I've got for loops that are executing the code inside them even when the condition fails. Example at https://codeshare.io/5MyMWR

I put that if statement in to catch this problem, and it seems to solves it. I have no idea why this is happening. Could it be some weird hardware or compiler quirk?

>>1448675
Then why even use an arduino? A 555 timer making you a square (or that swoopy triangle) wave is more than enough. Now you could just go for a resonant LC oscillator with a tuning capacitor instead, look for an oscillator topology that suits you.

Can anyone recommend a good, cheap board Fab house that can do 2oz or higher pours at a reasonable price? I don't like oshpark because last I checked you could only do that thickness on their .8mm boards or whatever (the wafer thin kinds) and they do enig which isn't appropriate for my design. I would like to know especially if you have had a good first-hand experience with the manufacturer.

Googling for board houses just sends me into an endless sea of Chinese places with super long lead times.

>>1448619
all over the place, search for "round pin header" on aliexpress

>>1448662
over six octaves? lol no

>>1448675
>chanspreading is nice
no, no it's not

>>1448676
that bit of code looks alright, turn on warnings in your compiler. if outer_index and m are not both signed or both unsigned, bugs can happen

>>1448684
did you check pcbshopper.com?

Man I just spent like 2 hours trying to get this displayed on my OLED screen, but no luck. Some whacky bitmap represented by a 16x64 array of 2-digit hex values, can't seem to find any success with converting software.

>>1448695
surely it shouldn't be too hard to use your favorite scripting language's appropriate image file parser, read out pixels from the buffer, convert to black/white (if required), and reformat as your display expects? I don't have any code handy that would help you because I mostly do this for bitmap fonts, but the principle ought to be the same

>>1448670
>reddit spacing
Do you know where this meme comes from?
In the past ribbit & the chans worked together for the greater good, which was a threat to the establishment (see:OWS).
Since then, the internet has been flooded with D&C shills to prevent any unity. They use race, religion, sexuality, political leanings, and anything else they can to amplify differences.
The "reddit users are bad" narrative is a divisive tactic to make an enemy out of your fellow man to redirect away from actual enemies.
Please don't help them.

>>1448676
Could be some obtuse type issue, I think you could also save some cycles by counting down from m instead of incrementing up to m, evaluation only requires checking an ALU flag

>>1448698
Turning an image file into a bitmap is pretty simple stuff, especially for a .gif, but I've no idea what format this 0x00 hex shit is. I'm more looking for information as to this. The display is 128x64 (or at least that's what the code thinks it is) but the hex array is 16x64. This would be fine if each value in the array corresponded to 8 pixels, but this doesn't appear to be the case at all.

>>1448724
>This would be fine if each value in the array corresponded to 8 pixels
Typically it is so, but those 8 pixels can go MSb or LSb first and some displays have two controllers so that you have to address the other side of the display differently than the other side.

>>1448724
16 bytes per row = 128 pixels per row, you will get a binary file simply by converting each hex value to binary, the controller probably wants the data on a byte by byte basis instead of a constant bit stream. The displays I've used require initialising instructions on whether the data is horizontally or vertically addressed, i.e whether each byte corresponds to 8 vertical or horizontal pixels

>>1448725
The software I'm using works with an example bitmap and displays a crisp image, not that this helps much.

>>1448727
Pretty sure it's horizontal, but here's an example of a file full of 0x01s with a few stripes across the whole thing of 0xFFs.

>>1448711
>In the past ribbit & the chans worked together for the greater good, which was a threat to the establishment (see:OWS).
In the past, ribbit wasn't infested with millions upon million normalfags, either.

>>1448727
>>1448741
Ah, I might have it. Looks like it's a combination of columns of multiple vertical rows.

>>1448711
"muh diversity" is the real D&C tactic that prevents people from coordinating with each other without extensive explicit communication
"when in Rome, do as the Romans do" used to be common sense
so, don't rebbit space
easy
as
pie

>>1448724
if you know the controller type, you can look this up. I think those are based on a Solomon Systech driver chip, and I seem to remember that you could switch the memory addressing modes around in some interesting ways

>>1448746
Well it's getting late, I'll attempt again tomorrow, if I've the time.

I made this capacitive detector with a small piece of aluminium foil and it works great.
Basically two pins on the arduino are connected with a 1M resistor and from the resistor goes a 1 meter wire attached to an aluminum foil, which is glued on the inside of a closet door.
So when a human places a hand near the foil, he acts as a cap and leeches away some of the voltage, causing the second pin to receive the voltage at a small delay. This delay is measured by the arduino.
But the problem is, after i leave it running for a while, it starts throwing out higher values that normal (like 10x higher) even though there are no humans or anything else nearby and i have to manually restart the arduino.

https://playground.arduino.cc/Main/CapacitiveSensor?from=Main.CapSense

Any idea what is going on here? Is my apartment haunted? Is some asshole ghost jacking on top of the arduino generating electromagnetic fields which are triggering my shit?

>>1448906
Reseting Tarduino fixes the problem? Maybe interrupts are getting in your way. Also, check the connections.

>>1448906
commercial touch sensors auto-calibrate periodically when at rest. you should try adding such a function in case heat or whatever is changing the RC time constant of the resistor+electrode
also how are you exciting the electrodes?

>>1448909
The arduino sets a pin to 5V and waits for the other pin to reach 5V and measures how long it takes

>>1448910
ok, good
just for giggles, did you try pulling to GND and timing that instead? Vil might vary less than Vih over temperature especially if they designed it asymmetrically in order to be TTL-compatible
still, the auto-cal is probably a very good idea if you're not using precision voltage references to determine state change

>>1448916
I am not sure if i can do that, i am using a library which does the actual pin work and timing. It also has auto calibrating built in.

If it was wiring issue then restarting the arduino with that little button shouldn't fix it, and the code i am using is super simple so it's not some overflow problem or something, i am just very confused.

fuck I hate mathematics

>>1448906
>playground.arduino.cc/
You didn't read the article, did you.

Has anyone here bought omrons/huano switches in bulk from aliexpress?
I'm wondering... I saw some that were rated 30M clicks but those were all taken down. Is 10M/3M the most I can get from AExp?

>>1448749
Ok, so each block of 8 rows of hex corresponds to a row of 128x4 pixels. It's x4 not x8 since I've got a 32px high screen, not a 64px like the software is expecting. I'll tweak that out when I get around to it. All it really means is that only every second hex value means anything, the ones in between are rounded down. Each hex value represents a 1x4 column of pixels. Now all I need to do is figure out the conversion from each group of 4 pixels to each hex value and I can write a python script to turn my .gif PhotoImage into an OLED shitpost.

>>1448992
Oh and for some fucking reason the very last hex value corresponds to the first row of 4 pixels, everything is shifted over by one. I'll just modify my python string afterwards I guess.

>>1448992
So here's a sequential ramp from 00 to FF in two rows, there's all this shitty repetition caused by the the 32-pixel high display instead of 64-pixel high.

>>1449029

there should be a digital /ohm/ for stuff like this, which has nothing to do with electronics (usually) and everything to do with reading a spec. If you don't have a spec (and they somehow don't exist on 2018 internet), and are blindly trying to figure this thing out, that's cool.

pic-unrelated

If I try to bring my MCU debugger and a small breadboard circuit in my carry-on, will TSA give me shit?

>>1449036
Probably right ay. There is the ESP8266 shill thread, but I'd much prefer to stay away from that.

>>1449036
there was /hard/ for a while

>>1449124
Oh yeah, when did that die?

>>1448689
>all over the place, search for "round pin header" on aliexpress
What about for mill-max low profile female headers specifically?

>>1448954
i did and i did everything it told me to except adding the optional capacitators because i don't have any

>>1448135
How does that work?

>>1449268
Any pipe will have a "resistance" to flow, so by putting something to measure the pressure, analogous to voltage, you can see the potential drop across the known pipe and figure out with the pipe's "resistance" how much current you have. Same concept as a current shunt, though I don't think it's necessarily linear, or "ohmic". Did that in a physics experiment where we found the critical velocity of water and calculated Reynolds' Number. For us our pressure meters were just vertical pipes open to the air, such that the water pressure pushed against gravity and was measured in "mm H2O".

>https://en.wikipedia.org/wiki/Darcy–Weisbach_equation
>https://en.wikipedia.org/wiki/Hagen–Poiseuille_equation#Relation_to_Darcy–Weisbach
Both of these look handy, but you'll need to pick a pipe cross-section to ensure laminar (non-turbulent) flow, the lack of which will mean you'll have a sharp nonlinearity and some nasty corrections to make to your pressure data.
>https://en.wikipedia.org/wiki/Reynolds_number
We used the Poiseuille equation in the physics lab.

>>1449229
In my view the main message of that article is that the whole setup is not reliable; way too many tweaks for handwaving reasons.

The cumulative effect you described may point to a code bug. A possible test could be to let it run without any sensor foils connected, just the hi-Ω resistor between the two pins. Monitor the output to see what happens.

>>1449331
I already installed it into the closet so I want to try adding the caps before ripping it out for debug. I went out and purchased a bunch of 100pF caps.
With the first one i connected together the recieve pin and a ground pin on the arduino, like the article says, easy enough.
But i am confused with where to add the second cap. It says "parallel to the body body capacitance" but what the hell does that mean?
Pic related

>>1449408
One of the things i suspect might be causing problems would be the auto calibration done in the library.
I assume it works in a way where it reads the values it gets and then pulls the average out all the values i read per an amount of time, and sets them as a new ground value. So if there is some rogue capacitance there for a minute or two somewhere, it totally fucks up the autocalibration and then i get the high values all the time.
If the cap doesn't work, next thing i try will be disabling the autocalibration.

>>1449187
machine pin headers

>>1449126
dunno but there was actually some interesting stuff going on in there occasionally. OP must have just gotten bored. sort of like I have after OPing /ohm/ for most of the past year

>>1449408
at the R, sense pad, pin junction. see Cpin? put a ground symbol on the other side of that and put it there

>>1449408
So i solder one cap leg to the tin foil and the other leg to the arduino ground? Do i understad that correctly?

>>1449420
>>1449421

>>1449422
placement isn't critical. attaching it on the board like any other component will be fine. just connect it between the input pin and ground

Where could one get USB-C breakout boards made to fit in a cable housing?

How do you build an analog sinc filter? No DSP rubbish. RLC only. op-amps are fine too.

>>1449759
Literally impossible. You can get near to it with very high order cascades of (buffered) filters though. Elliptic filters maybe. Course if you're filtering at much over 1MHz or so you'll need special op-amps to handle the frequencies, if not RF amplifiers of some variety. What for?

>>1449421
So far it's working, i have one cap between recieving pin and ground and i has been 12 hours and so far no high reading errors
i guess the cap manages to filter out any capacitative interference generated by the jumper wires and cables laying around the board

>>1449791
The C simply reduces the sensitivity. For longer wires to the foil I would use either a shielded line or a twisted pair connection with one wire grounded at the sensing device and open at the foil. One pair from a LAN cable has about 50pF per meter.

>>1449759
A sinc filter is non-causal (the output depends upon future values of the input). A perfect sinc filter has infinite delay, as the output depends upon input values infinitely far into the future.

>>1449759
the standard approximation is just a high order low pass

>>1449815
The wire is only about one meter long.

Can you edit the schematic to include the twisted wire? Because if i attach the wire to the foild and then ground it at the arduino, it will create a basically voltage divider where one wire goes to the sensing pin and one to the ground and since the sensing pin has 1K resistor on it, then pretty much all voltage will get lost to the ground and the sensing pin will never raise to 5V

>>1449815
like this. you attach the non-foil wire to the ground at the receive end, but you leave it unconnected at the foil end. the cable capacitance adds to C
but if it works now, you don't need it

>>1449869
Oh i get it now. The cap helped a lot but i had one misfire so far so it's not perfect. I will try the wire thing as i have lots of network cables around here albeit unshielded

>>1449905
I get my twisted pairs from cheap unshielded network cables (UTP) too, either rigid (AWG24, pic) or flexible (AWG26) which has litz conductors.

What is V_in_high and V_in_low of your receive port? For a typical digital port it isn't 5V and 0V but somewhere in between, like >3V=H and <2V=L. In this case 1V would be the limit for environmental interference. A smaller R on the send port could improve the situation because it is the only load of the capacitive antenna receiving the always haunted environment. Noise immunity is the principal problem of such a simple setup. Did you ground your GND?

>>1449905
if noise is a problem, you need the sensors as close to the MCU as possible, within a few cm in any case

>>1449905
better yet try just doubling that cap. this is one of the variables that a design based on a commercial touch sensor will need tuned for anyway

>>1450042
>if noise is a problem
you abandon the clumsy charge-discharge method and use a reliable 60¢ burst mode charge transfer proximity sensor chip.

I have an arduino powered by an usb cable and reading the voltage on a port with its ADC. The reading oscillates wildly, without connecting any voltage source to it.

If I take the wire out of the port, it reads 0V as expected. As soon as I connect a wire, without connecting it anywhere else, it starts oscillating between 0 and 1V.
If I ground the wire, 0V as expected.
If I touch the wire, it starts oscillating so hard the sine wave becomes rectangles between 0 and 5V.
If I touch the ground on the arduino, it starts reading about 0V, with a very small oscillation of a few mV. If I then touch the wire with the other hand, it flatlines to 0V.

What could be causing this? The frequency is too high for the standard mainline interference, it's at about 50-100 kHZ, I can't measure it easily for now. How do I prevent it other than standing next to the arduino and grounding it with my body? I can post the code if it's relevant.

>>1450128
or a 6¢ chip, but otherwise excellent plan of attack

>>1450139
>50-100kHz
sounds about right for switching power supplies near you. how does it feel to be reduced to an antenna? :^)
>wat do
load down the input, of course. a 10k-100k resistor to ground would be a good start and would have little effect on most intended inputs. if it's too much load for your signal, add a buffer or go larger on the resistor

>>1450176
Alright thank you very much. I sadly don't have any actual resistors right now so that kinda sucks, I guess I will try a few materials to see if I can get something in between 100k to 1m for the time being.

What exactly do you mean with an antenna, am I receiving all these signals? I would have thought the arduino is picking them up over the wire that connects to the port.

>>1450186
>am I receiving all these signals?
Of course, night and day.

>>1446665
Anyone good with uCs? I was just assigned a week long custom project

We're using an atmega 1284, but I wanted to use a 20x4 LCD display, specifically this one. It for sure works with the 2560, and I'm going through the library and I'm not entirely sure how feasible it would be to get it to work with the 1284. The 2560 has more pins, so maybe I'd have to use USART with two 1284 uCs. Gonna look at the library later.

Is this too much of a hassle if I have less than a week? Also, if you guys have any good ideas for build-upons/peripherals for a simple sidescrolling game or top down dungeon crawler, I'd love to hear them. I'm still brainstorming.

>>1450247
New concept: FM radio that uses your body as the antenna.

>>1446665
What frequency should I tune in to for my school's security personnel? I think they're still on analog because it's not listed on radioreference as a trunked system.

>>1450395
Look up your country's radio spectrum chart and see what bands they could be using. CB and ham bands are possibilities, as are commercial bands. Also ask the ham radio general.

>>1449287
Huh. That's an awesome thing to keep in mind for future stuff, but it'll probably be a bit beyond this project. Tiny fish tank air pumps will be pumping into maybe 6 cubic inches, and the pressure in the area will vary.

>>1450461
Sorry, vary between maybe 0-10 PSI.

>>1450461
All you need is a piece of fixed pipe and a method by which to subtract one pressure from another, such as a microcontroller which I assume you're already using, should get you good results without impeding flow very much, though it could be a bit bulky.

Also https://www.aliexpress.com/item/yeet/32819992509.html
they certainly exist and pretty cheaply too. Just look at the datasheets for them and go for one that gives you sufficient accuracy for your flow rates and doesn't impede flow significantly.

>>1450128
So you connect the aluminium foil to this IC and if it detects a hand nearby it outputs some voltage? So it's the same setup as before, except it somehow magically filters out the ghost capacitance?
How will it be able to tell which of the signals are ghost signals tho? I don't want to buy it just to find out it doesn't solve the main problem.

>>1450042
i can't do that, the wire has to be about a meter long since it's hidden inside furniture

>>1450045
added another 100pf cap in parallel to see what it will do

>>1450031
I am actually not sure. The arduino ports output 5V and to be considered high on the digital pin it's somewhere between 3V and 5V i think

>>1450031
yeah i will vivisect an old internet cable today and try it with that, it's also great because those wires are really thin

I also notices that when touching the aduino it switches the on and off, so the about 10 breakboard wires i have snaking around are picking up lots of capacitance, i am hoping that once my prototyping boards arrive and i completely remove 90% of the wires and replace the UNO with mini, it will remove a large bulk of the ghost cum from my detections

>>1450308
You could use one of those small OLEDs, the 128x64 ones. They started building them larger, so it would probably be just as big, except it would look much nicer and you could add some graphics.

>>1450308
HD44780-clone displays tend to be very similar
>2560 vs 1284
the 1284 has one I2C port. you could use it, but be advised I2C ports are kinda slow
otherwise you'll need minimum 7 pins to handle a HD44780-type display: R/W, RS, E, and D4-D7 ("4-bit mode")
you could do worse than to try it
>uCs
using lolduino libraries is not using "uCs". if you were using uCs you would be fighting with the hardware, not the software

>>1450469
that should help
also consider some digital debouncing on the output side of the library
if reliability is a concern, the right thing to do would be to use one of the touch sensor ICs mentioned

>>1450484
take your (You) and gtfo with your keyword-driven shitposting

>>1450490
>take your (You) and gtfo with your keyword-driven shitposting
Show us on the doll where your daddy hurt you
Jesus christ what's your problem? I was just trying to help him.

>>1450516
OLEDs have true blacks, which is racist

>>1450516
and you were failing
>oh, you have a week?
>buy this thing
>spend at least three days sitting on your thumbs waiting for it instead of working on it
>a lolduino user will still need libraries for it
>you will still need a synchronous serial interface
you solve exactly none of this guy's needs. you're not even trying to. you're trying to appear cool and fresh with your knowledge of toys with no substance behind it. go back >>>/g/

So here's the thing guys, I'm trying to figure out what project I'll be taking as my end of degree expo for my associates degree.

Me and my team mate are pretty dead set on doing something similar to the Raspberry Turk project:

> Robot arm that can play against a human being.


Difference being that we're planning on doing our own dedicated PLC (Which we've already sorta figured out) and we'd also like to make our own system to detect not only where the pieces are but also which pieces are where.

Right now we're thinking of using NFC but we're afraid of it having too much range on the readers and not being able to see where the pieces are because of it.

Anybody got thoughts on this?

why doesn't ltspice have models for such basic components as pushbutton switches and potentiometers

>>1450541
NFC or other RFID is probably the best bet, but you could connect all squares together with 5 wires (32 bit bus = 32 pieces) and then additional power and ground array so that you can power up a single square, and if there's a piece in the square it lights up the logic wires, this however creates an accuracy requirement in piece placement on square which is stupid imo

>>1450543
>SImulation Program with Integrated Circuit Emphasis
it's a simulator, not a vidya

>>1450541
how about a camera and OpenCV?
if you had an RSSI reading from the NFC readers you could easily find the position where the response from some piece is strongest

>>1450525
Fuck off nigger
He said he doesn't have enough pins, most oleds only use i2c so that's a big advantage.
Also the libraries usually work for all atmegas, so it literally just werks.
If he doesn't want to write it in C++, which is understandable, there's a few rewrites of the more common library from adafruit in C, hell I rewrote it for myself, and it works great.

>>1450571
>He said he doesn't have enough pins
he's wrong because he's got an I2C backpack for it and the 1284 does have an I2C port. if he's that hard pressed for pins he's fucked anyway and doesn't need to be looking for
>build-upons/peripherals

>>1450571
>Fuck off nigger
is that really necessary

>>1450545
Yeah, we pretty much went down the same rabbit hole.

>>1450555
>how about a camera and OpenCV?

Because that project would most likely be refused by the department. Not only would it be too easy to do but there's already an opensource project named Raspberry Turk that's essentially that.

>if you had an RSSI reading from the NFC readers you could easily find the position where the response from some piece is strongest

That could work actually, Is RSSI easy to implement with NFC?

Yay! Chinese Santa came to visit today. I feel bad about not leaving any sake and rice crackers out for him ;_;

>>1450637
No more multimeter in series like some dirty peasant

Is it right to think of amperage as being PULLED by a load and voltage as PUSHED by a power source?

>>1450601
haven't tried it but I don't see why not

>>1450786
No; they're both pushed by the source. A load doesn't "pull" current, it "sinks" it.

Also:
> amperage
It's called "current".

what has better employment prospects for my CE specialization
>Operating Systems, Compiler, and Processor design
or
>Embedded Systems

I don't see why I can't do both but if I have to choose, which one?

>>1450798
surely there's more work in embedded. almost nobody wants a new OS, compiler or ISA
except maybe google

>>1450691
Are those things as accurate as the ol' series DMM?

>>1450786
Kinda, but that makes them seem opposites.
A load may be said to pull Amps, and a supply may be said to push Amps.

Volt is an electrical PRESSURE unit of measurement for a power source.
Amp is the unit for flow, or CURRENT

Using the water/air analogy:
Volts is the water pressure, the potential, behind the spigot.
Volts are like Lbs/in^2.
Amps is the measure of how fast you can get water from it. Amps are like gallons per minute.

Double the Volts, and the Amps will double. They are directly proportional.

The 3rd part of the equation is resistance. Resistance is the measure of pipe smallness. A small pipe has large resistance.
Double the cross sectional area of a pipe (half the resistance) and the Amps will double. Current and Resistance are inversly proportional.

The relationship between the 3 things is explained with Ohms Law.
Volts * 1/Resistance = Current
or
V = I * R (I stands for current)

>1 Volt is the pressure needed to create a flow of 1 Amp through a resistance of 1 Ohm

>>1450805
Reminder to stop using the hydraulic analogy/
https://en.wikipedia.org/wiki/Hydraulic_analogy#Limits_to_the_analogy

Just work with electronics for long enough to build up an intuitive understanding without having to resort to comparing it to something else.

>>1450808
>intuitive understanding
Doesn't help explain anything to beginners. Anon asked for a basic understanding, not a degree.
Improve answer or go fuck yourself

This must be you too
>>1450791
Current can be sunk or sourced. You're confusing a specific application with general rules.

>>1450808
>inductance, fields
Water carries with it momentum, which is analogous to inductance. Inductance can also be made with a flywheel paddle inside the water flow, which can be coupled to another paddle as an analog to inductive coupling like a transformer. Laminar flow and friction between water is similar to the force of a moving electron on its unmoving neighbours. Information can be sent down a water pipe faster than the speed of the water particles itself, analogous to the drift velocity and signal speed of electrical signals.

>leaking pipes
This is avoided if your system has no open areas for pressure to escape. No faucets, only closed hydraulic turbines and pressure pumps.

>resistance against pipe surface vs all of metal
The actual resistance mechanism is irrelevant for establishing an understanding of current and voltage; you don't need to understand resistivity to understand Ohm's law.

>quantum mechanics
hahaha we're trying to teach someone how current and voltage are related, possibly with some ac circuits and reactances in there, not describe a P-N junction or demonstrate superconductivity.

>>1450870
>Water carries with it momentum, which is analogous to inductance

At some point you have to do what the other guy said, and accept electronics and electricity on its own terms. And inductance is in no way like momentum. The energy stored in the magnetic field might be, but just let it go.

>>1450805
>A load may be said to pull Amps

This is so fucking wrong, wrong, wrong. If a load could pull current it would not need a voltage source, but only a source of electrons, like the average copper wire.

>>1450791
>Also:
>> amperage
>It's called "current".

>>1450874
>a welding company is the authority for electronics terminology

>>1450876
>some shithead on a mongolian image board posting anonymously is the authority

>>1450878
I'm an electrical engineer. If I unironically called called current "amperage" at my job I would get laughed at. What are your credentials?

>>1450791
>>1450874
>>1450876
>>1450878

>>1450882

>>1450872
Inductance is described by V = L*dI/dt, and you'll find an identical relationship within momentum. To prove it, a spring oscillator also is just like an LC oscillator, except the spring is the capacitance and the momentum of the mass is the inductor, and since both result in a sinusoidal oscillation (and we already know that a spring has a 90° phase change between speed and displacement) the "momentum inductance" must also have a phase change of 90°. A hydraulic ram pump is analogous to a boost converter.

>>1450891
>Inductance is described by V = L*dI/dt, and you'll find an identical relationship within momentum.

The amount of momentum an object has depends on its velocity and mass. So, when you push on the object and transfer energy to it to make it move faster, it has more momentum.

Give the the electrical equivalent of changing the inductance of an inductor. I'll let you have the last word on this because I know I'm right and you know you are right, and that's where this will end for me.

>>1450896
For a capacitor Q = C*V and U = 0.5*C*V^2, we can write it instead as Q = dU/dV. But there is no analogue for this with inductors, i.e. X = dU/dI = L*I so I was referring to it as inductance. Momentum is analogous to L*I (perhaps with a density of water scale-factor in there) so mass is analogous to inductance, because speed is analogous to current.

>>1450882
Why would you want to be specific about the units of measure when you're talking about a concept?
It's fucking stupid.

>>1450900
The important thing being that Q, p, and L*I are all conserved quantities, which is important for inelastic happenings.

>>1446698
Based

I have a +/-6V supply, a micro that needs 3.3V, an amplifier that can take 3.3-5V, and need the output of the amplifier to be around 0V. Is it stupid to run the micro off a 3.3V regulator, and the amplifier between the 3.3V regulator and a -1.7V regulator?

>>1450984
Not really, but why not just use a DC blocking cap (high pass filter) to keep the output around 0V? Also a crude circuit diagram would help, because chances are you'll need DC blocking caps and extra power rails if you're using the micro's ADC anyhow.

I've been thinking about image intensifiers a lot lately. I know the "classic" ones work using a microchannel plate in order to multiply photons, but I've heard mentions about digital image intensification recently. It makes me wonder if there's a way to make the cheapo Raspberry Pi NoIR type cameras function more like an image intensifer by digitally amplifying the signal it captures.

Do you think we'll see a board like the Flir Lepton, but for near infrared any time soon? Alternatively do you think there is a way to modify the NoIR camera to be more sensitive/amplify the existing signals it captures?

>>1451147
While I don't have much of an understanding how camera elements function and where the digitisation happens, but I imagine it happens not someplace that you could cut off and amplify with any ease if at all.

You could try making an optical amplifier by pumping a medium to population inversion, one with a transition frequency below that of the light you're trying to receive but above the minimum for the sensor array, and hoping the thing works. It might have to be inside the aperture. Electrically pumped CO2 might work, albeit dangerously. If this sounds interesting it's pretty much an IR laser but without the resonant cavity. Trying to get a semiconductor population inversion to do this for you may also be interesting.

You can definitely buy standalone camera elements from the usual suspects, and I've heard that some cameras have a removable IR filter that you could replace with a visible light filter and use as an IR camera, though it will be very near IR. Else you might be able to find standalone IR camera elements, though it's somewhat unlikely. But since all camera elements are a block of silicon anyway, I imagine they just put the ADC and multiplexing inside it.

Added the twisted wire although not sure if it will have any significant shielding effect

>>1450984
personally, I'd change the amplifier to one that can hang with ±6V
you have a bunch of stuff but you aren't saying how they are connected to one another, which matters, especially in the case of any signals between the micro and the amp or any external signals

>>1446665
Is there anything wrong in this schematic other than the 0.22pF cap?

>>1451216
Forgot pic

>>1451218
Pic related
and more importantly, why

I need to change one of thoe blue small trimpots on a chink board for a normal sized potenciometer, but fuck.. the solder spots are so tiny.. how the fuck am i supposed to solder shit there without it all becoming one big solder blob?

>>1450464
The size is the biggest problem for my application, I needed tiny. It's one of those things where the backup option is 95% as good, but for 5% of the cost.

>>1451240
git gud
and don't use a crayon

>>1451221
RF Transceiver. Walkie Talkie, basically, but I'm pretty sure I fucked up somewhere.

I have an effect pedal which partially stopped working. It seems like the logic and analog parts work but it can't record loops. It's cheap chinese shit.

I wanna "reflow" the solder on the board before i give it up. I have a camping stove. Can i just hold it above the flame? Are there more intelligent solutions? I don't have a heat gun. I also have a soldering iron and have about 10 years experience with soldering throughole stuff and some bigger SMT parts but this board uses too small gaps between the legs, i would instantly bridge 3-4 legs with just the tip of the iron.

>>1451258
Probably not on a flame directly due to oxidation and buildup, but you could try putting it in a sardine tin and heating that. YMMV.

>>1451258
try a pan of hot sand

>>1451258
you are simply not equipped for this operation. sorry, dude. take it to your local cell phone repair chink and have him do it

>>1451258
put it in the oven to reflow that is how it's normally done

>>1451275
>>1451285
>>1451287
>>1451294

I've googled some, i decided to not put this chink shit into my kitchen oven so it's over. I might poke around with my soldering iron.

>>1451300
Pussy.

I need help for Arduino + DMX512 stuff.
Basically, my arduino slave don't work with there is something else plugged after it, I thing it's a wiring problem.

> master > my shit
it works
> master > a DMX slave > my shit
it works
> master > my shit > a DMX slave
none works

here is my wiring:

XLR IN
1 2 3
| | |
| | |--- A
| | |
| |---------B MAX485
| | |
|--------------GND
| | |
1 2 3
XLR OUT


>>1451330
TIL [code] don't works on /diy/

>>1451330
Stop being a lazy cunt and draw a proper schematic

>>1451218
>>1451257
>but I'm pretty sure I fucked up somewhere
That's an understatement. For starters, what is the point of Q2?

Also, learn to draw schematics, for RF transceiver this is confusing as hell

Hi, I need to build a coil gun for a diy turret that I'm making. Unfortunately I'm having a problem finding a simple schematic since I'm not too great with electronics. Could anyone point me in the right direction?

Also, where can I find a cheap battery power supply for the coil gun since most battery packs that I've seen are >$35. The coil gun doesn't need to be too powerful since the projectile needs to have an energy of less than 2.5 joules for it to be legal in my country. Thanks.

>>1451194
I don't think it will be for a while anyways. There are systems like the X27 https://www.x20.org/color-night-vision/ that are currently marketed towards military. If they get adopted hopefully by the end of the decade Gen 2 and 3 tubes will become cheap and abundant enough that more hobbyists can mess around with them.

>>1451439
http://www.electroboom.com/?p=101

>>1451439
>legal
if it's legal it's not work making

does reading the voltage of a rechargeable battery quickly drain it? I have a lead acid battery and a NiCd battery and on both when I read the voltage it doesnt stay put, it just starts to drop quickly. Help my retarded brain understand.

>>1451575
if you're using an analog meter, surface charge could be drawn off especially quickly
that, or you've set up the leads/selector switch to measure current

>>1451579
>if you're using an analog meter, surface charge could be drawn off especially quickly

analog meters like digital meters are high impedance on voltage ranges. and what the fuck is "surface charge" on a battery?

Can an electrolytic cap be used in a manner where it is intentionally reverse-biased, so that it acts as a short if current is applied one direction, and acting as a capacitor in the other if it's kept within its voltage range? There's plenty of videos of people blowing up low-voltage caps with several tens of volts DC, but exactly how much abuse can they handle?

>>1451218

well, the antenna connects to Vcc, which is crazy. also, the input to the LM386 also comes from Vcc. this is even crazier.
R2 and R4 should be connected together, and to the collector of Q2, coz as is it's (you guessed it) crazy.
i'm sure there's more. Q4, for example, doesnt make much sense.

>>1451300
>I might poke around with my soldering iron.

this is a good idea, but you need to cover all that shit in flux. flux is magic and keeps the bridging from happening 100% of the time, nine times out of ten.

>>1451601

no. and pls stop abusing the poor helpless caps, they've caused you no harm, you nazi motherfucker.

Im using a MAX31850 K-type thermocouple converter to read temperatures in a high power furnace. Problem is, when the furnace is on, the magnetic field fucks up the readings and it jumps all over the place.
But when i connect my multimeter to the thermocouple the multimeter reads the right temperature.
How do i fix this?

how can I measure how much current a solar cell is producing? In the middle of the day, a shitty harbor freight cell produces 20V, which seems like a lot to me but it MUST be something shit like 100mA or something, right? How can I test this?

>>1451596
>analog meters like digital meters are high impedance on voltage ranges
some are - some aren't

If he's using a dedicated battery test meter these present a load on the battery during the test.

>>1451625
>measure how much current a solar cell is producing
use a current meter

>>1451661
in conjunction with a load though, might want to buy some beefy power resistors, or just use some water. Measure voltage and current at the same time to find peak power output.

>>1451703
>in conjunction with a load though
no load will be no current

>>1451732
exactly

>>1451624
>But when i connect my multimeter to the thermocouple the multimeter reads the right temperature.

so leave your multi-meter there all the time.
or try it with an equivalent resistive load across it. and add a small cap in parallel, coz it cant hurt.

>>1451624
just add a cap acros the TC at the cold end
10nF should help

>>1451763
>so leave your multi-meter there all the time.
>equivalent restive load across
You are fucking retarded

>>1451806
Already have one.

Can any one speak to the effect of changing R9 in the pic related filter? How sensitive is this? They used a somewhat uncommon value (24k) rather than the more common 22k, and I assume this is pretty critical? What effect would slapping a 22k in there have?

>>1452296
>dB/Oct
lol

22k should be fine, it will only change the filter frequency by 8% at most, maybe 4% at the least. Calculate the transfer function if you're worried.

>>1452300
>>dB/Oct
>lol
why is this funny?

As a followup question: am I correct in interpreting this as basically a two-pole LPF, one pole formed by R28/C7 and the second by R9/C6?

Please no bully. I am only a lowly technician.

>>1452305
Because dB/decade makes more sense because dBV is a decimal multiplicative unit to begin with. You'll get filter slopes of -20, -40, -60dB/decade.

I don't think you can generalise it as just two non-interacting filters because they do interact and R10 has an affect of some sort. Plot it in spice and see how the frequency dependance changes with different Rs and Cs if you want a more visual understanding than the transfer function.

SO i have arduino mini, and i need to attach the pin rail to it, since it comes with just empty holes.
Is there some way to do it without soldering? because the holes are so close to each other i will just end up with one big short circuit

>>1452341
then use a damn flux when soldering and sensible amount of solder

>>1452342
A little wick also helps (can just braid thin copper wire strands and add a little flux). If the through-holes are plated you can just add a tiny amount of solder to begin with and just hold the heat on for a little and the holes will just fill with solder without big beads. Wick will help with this. No way other than soldering, though conceivably you could use tiny copper pop-rivets.

>>1452307
decade = 3.32 octaves ezpz

>>1452307
if one's signal is denominated in octaves (ahem, audio), I submit that it makes more sense to work with octaves. for much the same reason, burgerbros often find metres are more unwieldy than feet

>>1446698
Enjoy having to do back breaking labor till you’re 70.

>>1452390
>burgerbros often find metres are more unwieldy than feet
actually 'metres' relate closely with yards for approximations
it's the mm, cm, that cause me to have to do an internal math conversion
also: liter is close enough to quart for general visualizations
>I don't want to talk about Fahrenheit and Celsius

>>1452409
>0°C: cool
>100°C: dead

meanwhile in the sensible system:
>0°F: cold
>100°F: hot

>>1452436
>human feels > water reals

>>1452439
yes. in my experience the physical phenomena that metric units are associated with aren't very relevant to engineering (eg i can work in °F and mils or mm and °C equally well) so celsius conveniences a tiny population of scientists over the 7 billion retards that just want to know how hot it is. nobody even uses degrees to boil water, just knobs with arbitrary low-high scales. not that any of this matters.

>>1452442
>so celsius conveniences a tiny population of retards over the 7 billion people that just want to know how hot it is.

>trying to use an arduino to datalog in simulink
>for some reason simulation data logger does not record anything in "external mode"
>no one has the same issue
>pirate version
Welp. It's time to learn that shitheap that is Pyserial.

>>1452522
Yes, that gray area is seven billion retards that use a scale of -20-40 for so cold you risk dying easily vs so hot you risk dying easily.

Same with food. 1 oz, small amount, 16 oz, a lot. 1g, nothing, 1kg, way too much.
>go to the butcher
>ask for 1 lb vs 500g
>go to grocery store
>get 1 gallon or a 3.75L bag

>get a cup
>8, 16, 20, 24 oz or 1, 2, 3 cup
>get 100ml, 250ml, 500ml, 750ml

>>1451624
Shield it, ground it. Copper braid has lots of fine wire. use your fingers to stretch it out a bit, wrap it around the thermocouple, and keep a tail to solder the ground wire.

>>1451601
Put a schottky diode in parallel.

>>1452522
>how hot
The 4% love their legacy units. When it's 30°C here and a gringo asks me how hot, I convert it on the fly in my head and tell him it's 86. We learn that in school, the gringos don't.

>>1452585
What a desperate nonsense.

>>1452436
>-20 °C: Cold
>40 °C: Hot
>b-but muh everyday use

Do Americans can't into negative numbers or something?

Okay /diy/
SPI related question.

I have bunch of things on my SPI bus and master device is talking to all of them one at a time.
The issue is, one of the peripheral (PN532) is not able to send signal through MISO line.

I can see the signal coming on the MISO line but the bus is not fully lowered to zero.

To describe it better, MISO line is idle HIGH, Peripheral tries to drive it low, it is too weak to drive it low to ground so I see a signal hanging to 1.5Volts from 3.3 Volts instead of going all the way down to zero.

What could be causing it?

>>1452818
One of your other devices might not be properly letting the MISO line float, or if you're terminating with just a pull up resistor may be too low value.

radio triac dimmer!

i plan to control it with an arduino and ebay 315MHz tx module hooked up to my spare pc. should be able to switch 250W in a sealed enclosure. i ran icsp traces all on top of my rf ground but i'm not really worried about the resultant ground loops or capacitive coupling given the nature of the signals on those lines during operation. i haven't run a drc or done a sanity check on it yet so i'm sure something's wrong with the board.

>>1451330
do you have the proper 120 ohm on the out of your last device? is the ground through the DMX lead the only shared ground?
>>1451332
hiroshimoot plz

>>1452837
>terminating an SPI bus
y tho, just turn on the weak pullup in the master and forget about it

>>1452910
gud. you might consider some RFI protection on the logic power supply. there are SMD power chokes but a small inductor or ferrite bead could be adequate

THIS WAY PLEASE
>>1452945
>>1452945
>>1452945