/diy/ - Do It Yourself

>>1413287
this thread's digits brought to you by the MIC3287, just another of the many, many boost-mode white LED drivers to be found in every shape and size anymore

Hi /ohm/ I would like to make or obtain a DC-DC power supply to go from a couple of lipo batteries to 0-10 KV. I would ideally like to be able to control output voltage from 0-10 KV from a microcontroller. Ideally, output current would be limited by how much the battery can provide. That's to say, I would really like to get 5-10 watts out. Yes this is probably unsafe, no watts out will not be more than watts in. This is to be for powering dielectric elastomers to hopefully make a robot.

>>1413343
Making a current-limited supply is easy in that you can use direct current feedback with a shunt and comparator/opamp, but making a power-limited supply would require something to multiply the measured voltage by the current and give feedback based on that. Since I don't believe it's an easy thing to do without being digital, any fluctuating load you hook it up to could surpass this power rating before the microcontroller has time to correct it. Use a big enough capacitor and this isn't an issue, but it's something to keep in mind. You're probably looking for a flyback converter with a feedback winding and current-shunt between the load and ground, but look into what neon-sign transformers use. Constant-power supplies don't really exist as far as I know, so you won't get an off-the-shelf model.

Why is constant-power important/why do you want it as opposed to constant-current?

I am confused.

So I took a standard chink LED, which is specified to be 2V 20ma and I connected it to my power source which is 3.2V.
I ran a simple calculation

3.2/(20/1000)=160 ohms
I used 150ohm resistor which gives me 21ma

So after measuring the power source and the resistor with my multimeter, which confirmed the values i mentioned above, i measured the LED current at it was WAY off.. on one led i measured 60ma, on another one 80ma... they should burn out instantly and yet I don't even think they were at full brightness.

What in the name of fuck?
Why are the laws of physics being such a dick to me?

>>1413356
Post pics. How are you measuring the current?

Sure you're getting 60/70, not 6/7mA? 3.2 - 2V = 1.2V across resistor, / 160ohm = 7.5mA, so if you're getting 6-7mA, that's well within expected range.

>>1413356
that's not how you calculate the resistors

(3.2-2) v /20 mA = 0.06 k = 60 ohms

Are you sure you measured right? 3.2 v should never reach 60 mA with 150 ohm in series, no matter what happen in the LED.

>>1413360
>>1413363
welp looks like i totally fucked it up
I measured with the meter set to 200ma so i thought you had to shift a decimal place on what it shows you
also i thought you have to use the voltage going into the led to calculate the resistor value.. oh fuck, that makes much more sense...
i am so fucking tired of being a brainlet ;_;

>>1413355
>Why is constant-power important/why do you want it as opposed to constant-current?
He never said it was. You're just misinterpreting what he wants.

>>1413366
kek, I didn't believe in Ohm's Law as a kid because I didn't know about non-ideal circuit elements like real batteries with non-zero ESR
anyway, brains grow through exercise, which is just what you're doing. Pic related shows the general case which will help you frequently in circuit analysis. there's a dual law wrt current which is also helpful in understanding parallel loads and such, you can look it up

>>1413343
>a couple of lipos
that's a tough one. you're not likely to get away without some custom magnetics
maybe you could rummage around some of the online surplus stores like All Electronics to see if they have a coil already wound or assembly already made that you could adapt to your purpose with some sort of control elements

Can i connect wires from a solar panel directly to a car battery to charge it?

>>1413378
Internet says you need a charge controller, but, nah, connect it directly, what could go wrong?

>>1413383
>charge controller
Isn't that just in case there is a solar flare and the solar panel jumps from 10V to 1000V for a second?
If the DC fluctiates below 13V of the battery then it should be fine i would think

hi /diy/

Can I use my iPhone's adapter for my USB mini drill?
The adapter has a 5v output, the drill's ching chong manual says dc 5.5v. Can I also plug it in my laptop?

sorry I'm retarded thank you

>>1413378
you'd better add a diode

>>1413388
try it, what could go wrong

>>1413388
that 2A on the printed sheet is likely 2 amps
if you have 2 amp usb charger and 2 amp ports on your laptop you're good to go

>>1413388
That iPhone adapter is 1A, so it wont power the drill correctly.
A PC USB port will most likely give even less power.
You could try a portable charger or a less gay phone charger.

>>1413385
A solar cell has an ESR, so the voltage will drop if it would otherwise be putting more current into the battery than it can provide, which for a lead-acid battery will be the case. If you get 14-16V worth of cells and hook it up directly it will charge the battery, but you'll have to disconnect it if the charge current is too high (not going to happen unless you have a fairly big panel) and when it reaches terminal voltage. You can hardly charge a battery too slowly, so if it's a <20W panel you should be fine, else I'd be pretty careful or just go ahead and buy a charge controller. If you do hook it up directly, put an ammeter in series with the circuit and put a voltmeter across the battery.

The charge controller itself has a DC-DC converter to give the battery the right voltage to match the ideal current curve regardless of cell voltage, and it will shut off automatically once it reaches ~14V.

>>1413388
That charger won't source 2A, it will be 1A at the most at that size. The result will be it runs at a lower speed, unless an overcurrent mechanism trips and cuts the thing off completely. But 2A USB chargers are pretty abundant, and any decent USB-port on your computer will probably provide that, I know my laptop's ports certainly can.

>>1413385
>Isn't that just in case there is a solar flare and the solar panel jumps from 10V to 1000V for a second?

No, it's there so the battery actually charges properly. Skipping over the technical details, either the panel will not completely charge the battery, or it will overcharge it.

>>1413396
>>1413401
>>1413402
I checked the iphone adapter again, "Output: 5v === 1A".
Tried my laptop's USB port and it worked
I'll look for a proper adapter, thanks anons

>>1413406
>Tried my laptop's USB port and it worked
Did you try drilling something or just spin it on the air? It wont draw its full current if it is spinning on the air.

>>1413404
>will not completely charge the battery
How could that happen?

Fuck there is too much stuff.
What are the mosfet parameter names for

Min - max voltage and current to open the mosfet
Max voltage and current it can pass through
And max heat it can dissipate?

https://lib.chipdip.ru/222/DOC000222413.pdf

>>1413421
Vgs(th) "Gate-source threshold" voltage or similar is the minimum to start turning on. Gate current will be negligible - the unique point of the MOSFET is that it has an insulated gate, so it's charge-controlled rather than current-controlled. Vgss is the maximum voltage you can apply before the insulation fails.
Vdss(br) "Drain-source breakdown" or similar gives you the maximum voltage you can apply across the D-S on the MOSFET before it starts conducting regardless of the gate voltage.
Pd is power dissipated. Max current is Id.

It sounds like you might want to read up a bit more on MOSFETs - the Art of Electronics chapter is pretty good. If you're just looking to switch a big load on and off (at low-ish frequencies) without too much complexity, SSRs or even just normal relays are a good alternative. What's the application?

>>1413424
I need to use it as a relay to switch around 15V and around 20A DC
I wanted to use relay but got bullied here, so i found out about transistors and got bullied some more so i purchased a mosfet instead (which is what i got told to do) and if it won't work again i will fucking shoot myself
Not to mention that SSRs are incredibly expensive

What's the difference between solder with silver in it and solder without? does it just change the melting point?

>>1413427
>I wanted to use relay but got bullied here, so i found out about transistors and got bullied some more

links or screen cap?

>>1413431
It's discriminatory against werewolfs and should be avoided

>>1413432
>>1408236
ctrl+f "fiery death"

>>1413434

You got help and some snarky comments. You are anonymous here, so it really isn't bullying, in my opinion.

Have you tried asking on a subreddit for electronics? They discourage mean people.

>>1413434
>got bullied
You were acting like a bitch, friend:
>>1411962
>>1411399
>>1411396

>>1413439
Yeah I'm not him, I was just linking that to show that he already knows he's going to burn his house down in the middle of the night and he's just looking for someone to blame when the inevitable happens because he tried to cheap out.
https://www.youtube.com/watch?v=hI51uKAbQDc&feature=youtu.be&t=47s

>>1413427
If you're using the FQP50N06 you linked you'll probably be fine then. If you make sure your switch voltage is >10V even taking the on resistance as worst case you'll be dissipating ~8.8W. Don't forget to heat-sink & thermal grease it, otherwise it's gonna toast itself pretty quickly.
Yeah, relays are convenient since they completely isolate the two circuits, but they consume a fair amount of power themselves. Cheap Chinese SSRs do the job, but they'd be over-specced for DC applications.

>>1413431
It's the most common type for lead-free, "SAC" solder alloys, slightly lower melting point than other lead-free solders. I think there's some edge case-stuff depending on what the contact material is too, Tektronix used to put reels of it inside some of their equipment.

>>1413355
Also, forgot to mention I'm driving a capacitive load. Constant power is not that important

>>1413445
I have copper heatsink from old gpu will it hold on the mosfet if i use cpu thermal paste as glue?

>>1413467

You should have some sort of pressure. Either a small bolt or a c-clamp or anything that provides firm pressure. Just the grease alone is not a good idea.

>>1413467
>>1413472
Thermal adhesive should do though, provided the heatsink isn't massive.

>>1413487
>Thermal adhesive

ah. I missed that part. I'm not familiar with the adhesive type.

>>1413489
It really depends on the weight of the heatsink and the surface area of the mounting surface. It's typically a 2 component epoxy glue, so if he has something like pic related he should be fine, but if the heatsink is one of those more modern massive beasts with cooler shrouds and stuff it most likely won't hold(and if he's using something overkill like that he should have mounting clamps anyway to balance the weight). Of course if the mosfet is using TO220 or similar package the easiest way is to just drill a hole on the heatsink and bolt it on.

>>1413445
Will putting 2 mossfets in parallel halve the heat output for each one? Since i have a whole bag of them from china so i can put like 10 of them in series and use a 9V battery to open them since they have almost no current draw, the battery should last forever

>>1413529
Okay follow up retarded question, can I connect a wire from + on the battery run it through the load and then have the negative connected to - in the power outlet? (so that nothing is connected to the - on battery)
Because obviously i cannot ground say the 15A load on the ground of the 9V battery,.

>>1413542

pls provide a diagram, or be more careful and specific with your description.

>>1413544
Well it's really simple
I have
Mosfet
Battery
12V source
12V 15A load

This is how i want to connect it

Battery + goes into mossfet gate to open it.
source + goes into load +
load - goes into mosfet drain
mosfet source goes into the 12V source -
battery - goes into the 12V source - too..?

this is my current plan, the gate voltage is 10V but 9V should work i will jsut get a tad more resistance between drain and source

>>1413552
I really want to know what the purpose of your deathtrap is. My money is on low budget grow-op.

>>1413552

add something like a 10k resistor from G to S so that it turns off and that circuit might work if the load is resistive. If the load is inductive you should put a diode across it so the current can decay without damaging the fet.

>>1413555
Not the death-trap builder, but by
>you should put a diode across it
I take it you mean across the two poles of the load, correct?

>>1413554
>>1413555
the load is a basically a bunch of resistors whose only job is to produce heat

yeah in a youtube video i watched on mosfets they use a resistor between G and S too but i don't really understand why, since if i cut the battery power there is no way for it to stay open

HOAI, WOELCOM TOE THEY EHY EHY VEE BLEOG

>>1413558
>a resistor between G and S

that has a capacitance and if you simply remove the 9 volts the capacitance will hold the 9 volts, so you need a resistor to discharge it.

>>1413557

Yes. if it has inductance, when you try to switch it off, the current will develop a voltage that tries to maintain that current, meaning it will be plus at the "lower" end, so the diode anode goes at that end and the cathode goes to the other end of the inductive load.

>>1413564
How did you came up with 10k tho?

>>1413558
>the load is a basically a bunch of resistors whose only job is to produce heat
This kind of vagueness doesn't make your plan sound less sketchy. In fact it sounds even more like a krokodil lab now.

>>1413565

I pulled it out of my ass. You don't want a low value or it will waste power. I'd look at some designs and see what they use, but keep in mind that many designs switch at a high rate and will need to discharge the fet quickly to reduce losses as it passes through the linear region, and sometimes they have additional components to accomplish this.

10k should work for you, I think. Or 5k. Or 15k. You get the idea.

>>1413565
Pull-down resistors generally don't need to have a very specific value.

>>1413573
How long does the inductance last? I need it to switch only like once every 20 sec

>>1413582
>I need it to switch only like once every 20 sec
You're not gonna have any problems with a timeframe like that, the voltage spike will last only a fraction of a second.

>>1413591
>You're not gonna have any problems with a timeframe like that, the voltage spike will last only a fraction of a second.

Are you saying that because his load is resistive, or are you saying that an inductive load that is switched every 20 seconds does not need a protection diode? A spike of several hundred volts that only lasts a "fraction of a second" will still damage the fet.

In other words, it has nothing to do with timeframe, but has everything to do with the nature of the load and the level of current.

>>1413605
It does need a protection diode, he just doesn't have to take special precautions that he might have to if he was switching the load fast.

>>1413608

ok. he said inductance when he was referring to the gate capacitance, but then you referred to a "voltage spike" so I don't know what the fuck is going on.

it happens.

I have a teensy lc and I am trying to connect 12 (or more if it works) capacitive touch sensors and it only has pins for 11. I originally tried making a grid of lines but realized that if I touch one that it would trigger the other ones it went under/over.

I had a new idea to put them all on the same layer and then when I want to check one (touchRead) first set all the others to high or low so that they dont collect the charge and give a false read. I have a very weak understanding of electricity and unfortunately have to get this finished soon. I also thought to put a big either ground plane or 5v plane under all of the sensors and the columns of sensors would be connected under it so that it doesnt interfere.

Pic related is the top layer of the 'pcb' I am going to diy to test my idea. black are traces.
Anyone have any thoughts on this? I am a programmer and am very new to dealing with microcontrollers and electricity.

>>1413626
Normal way to do lots of buttons is to arrange in a grid, power rows in sequence and test columns, works if you only press one button at a time.
Capacitive sensor though depends on the sensor, if it's analog voltage output you probably need an analog multiplexer for whatever voltage you expect. If you had digital sensors with spi or i2c you can do it all in software anyway.
On the other hand if the number of buttons is increasing you might change the approach and using fewer modules try to triangulate where the press occurred based on ratio of voltage or something.

>>1413626
Hackaday had a good overview about I/O expansion recently: https://hackaday.com/2018/06/14/general-purpose-i-o-how-to-get-more/

Like >>1413627 said, the keypad "matrix scan" is the normal method but it depends on how you're doing the capacitive sensing. You'd only need 8 pins to do it.

>>1413627
With what Im making multiple buttons can be pressed at once. My sensors are just those traces connected to the touch pins on the teensy. In my first version I just took solid core copper wire and bent them into rectangles with beams filling their centers and connected them directly to the pins but there were issues with touching near one would set off others near it. I wish I had fancier sensors with i2c or spi but from what Ive seen they are expensive if you are buying a lot of them and unfortunately Im on a budget.

>>1413629
I am using the built in touch capable pins and the touchRead function to use them. It is nice because it only requires me to connect a single wire to each 'sensor' (in my case its the traces on that grid from before). I believe this matrix should work and youre right about it using the 8 pins. My main concern is if the traces are too close or something else where they will set off false positives with nearby ones.

I saw pic related when trying to learn about capacitive matrixes and it works by sending a signal on the columns and checking how much it has changed when sensing for it on the rows. Seems like a good idea but I dont know if the teensy lc is capable/powerful enough to generate, detect, and process them fast enough without using other boards or chips.

>>1413489
>thermal adhesive
it's good shit, mang. Pic related is a one-part silicone-type thermal adhesive that can be fairly easily removed if necessary

>>1413633
TTP22x chips are pretty cheap and come in 1, 4, 8 and 16-button versions
anyway, your sensors do look a bit too close together in that grid artwork, maybe try one where spacing width / sensor width ~= 1. if there is no other way to tune touchRead()'s sensitivity, you might try putting some sheets of some material between the sensor grid and the finger as a spacer

>>1413697
Sorry theres no scale in that image so I dont know if it is still too close when I tell you that that is going to be 8.5 x ~6 inches. when Im done with that it will be under a sheet of acrylic and vinyl that are about 5mm thick combined.

I just did a quick google for the TTP22x and mostly found sites that no longer exist. Any chance you could link me a datasheet or place to buy them?

>>1413717
Differentfag here. Is there a reason you need this to be done quickly, why you can't use a matrix of mechanical micro switches, and why you can't wire a second layer for the columns using jumpers like pic related(to give them a bit of distance from the surface layer, pic is supposed to be a cross section)? Sorry if I'm asking a bit much since you appear to be quite inexperienced, I'm just wondering what the hell you're trying to accomplish, and giving accurate specs seems to be a recurrent problem for people asking advice here.

will studying linear algebra make me better at electronics

Dumb question, but I just bought a box of old radioshack and "archer for radioshack" components. Different ICs diodes capacitors resistors rectifiers etc etc.

Im not really in the electronics hobby, I know how to solder and have put a few little circuits together thats about it.
Is this stuff still in its packaging from the 80s going to still be usable? Is there degradation from just sitting on the shelf?
Its got a lot of BNC connectors and stuff too, so probably a ham radio guy I guess.

>>1413733
I just finished a third-year electronics physics paper and we used nothing of the sort, just network functions that were simple enough to analyse without matrices. But the textbook did have a section on analysis with those methods. In general unless you're doing something that requires it you should be fine without it.

>>1413741
Electronics don't decay over a few decades for the most part, but electrolytic capacitors may well need replacing. To be sure you don't mess anything up, don't turn it on, take it apart. Transistors, diodes, resistors and such will be fine.

>>1413733

Maybe:
Thenevienes equations, logic circuits uses boolean (binary!), calculations of opamps/capicators/resistors. I had difficulty with everything, and are weak in math, but it helped to have experience, I got less of an obstacle. It's simpler than you think, but sometimes it'll be difficultier than you think!
Just be patient with yourself. :)

Learn as much as possible related to the topic you're working on to solve, instead of digging tru the book page by page.

>>1413760
This is sage advice. Ever wonder what the biggest difference between an amateur and a professional is? The professional knows what to Google. No sense in perusing through 100s of pages of mind-numbing equations and memorizing them, learn the basics and learn where you should look for more knowledge for issues that arise. After that: Just do your best! And then you may feel proud of your effort.

>>1413760
also btw it's Thevenin not Theveniene

>>1413741
>>1413745
ICs can have a fairly short shelf life depending on the type, it's useful to check by part number. Connectors obviously never go bad unless they rust to oblivion or something.

Maybe the /general/ will help more:

1) Did I waste money on this Klein MM600? Will I regret not spending the extra $30 to get the next step up with TRMS? Just gonna use it for typical weekend warrior and diagnosing problems around the house/car.

2) Bought this little 4 amp battery charger at HF on sale to throw on muh scoot since it doesn’t get run much. Will 4 amps start a car with a dying battery if I leave it overnight?

3) What should a new or good condition 12V lead acid battery read as far as voltage?

4) Is there some other reading I could get with the meter to tell if the battery is flat or it’s time for a new one?

I have no idea what I’m doing with electronics but I’m trying to learn what this meter does and what the numbers mean.

how complex would it be to make a single digit display run off 7 reed switches that display 1-6 and an R without using an IC?

>>1413760
Oh I was assuming he meant matrix/vector algebra, for nodal/mesh analysis. Now that I'm looking at it it isn't that tough at all, you just need to know are basic concepts of matrix/vector multiplication, the transpose, the determinant, plus Kirchoff's laws and the Thevenin/Norton equivalents.

>>1413769
Oh yeah, I've heard of sensor ICs and such drifting over time.

>>1413778
>without using an IC
30-150 transistors is my guess. What do the reed switches do? I'd start by making a compact truth table.

>>1413728
Im trying to get it done quick because I have been working on several failed versions over a few months and it is a gift for my friends birthday which is coming up soon. Using switches would work but I am trying to keep it touch so that the surface is a single flat material like acrylic or glass and it can be prettily embedded into something else.

The first image I posted is the top most layer before the surface. Pic related is a (crude) example of how it will go together. The layers are 'pcbs' I make by coating a sheet of projector transparency with copper then cutting and removing the negative with a Cricut. The transparency insulated fine and I can even make multi layered ones by cutting a small hole and soldering between layers.

I know this is all very crude and unprofessional but I am new to microcontrollers and working with physical electronics so I appreciate any help and advice.

>>1413784
Each reed switch once engaged would display the number 1 through 6 or the letter R, these being the gears for a car. Once the magnet that is under the gear lever goes over one of the reed switches it would engage the display to show the designated number/letter.

anyone have source on learning rf electronics? It seems like there mostly piece meal materials and not comprehensive.

>>1413787
I think your biggest problem is that columns 0 and 1 are connected to each other, while only column 2 is separate. Meanwhile rows are all separate. Therein lies your problem. Wiring it like this would mean none of them cross, but there's gotta be jumpers/another layer for that to happen. After that it's just a matter of polling each row/column pair in a recursive for loop to read which buttons are pressed.

>>1413809
Also this circuit feels a lot like this common "autism test" - it cannot be solved without crossing wires.

>>1413717
for your existing design, consider adding 100pF or so between the sense lines and ground, and adding 100kohms in series with the drive lines, to make it more stable and repeatable
>TTP22x
part numbers include TTP223 (1-channel), TTP224 (4-channel), TTP226 (8-channel with I2C), TTP229 (16-channel with I2C)
they're all over aliexpress, both chips and pre-made panels. here's a premade panel for the TTP229 https://www.aliexpress.com/item/16-channel-benis/32764684659.html
if you'd prefer a more professional/faster solution, one or more of Microchip AT42QT1110 are carried by various distributors and might serve your purpose. it also has an SPI port and some LED drivers, just in case you need them

>>1413778
7 transistors, a few dozen diodes, a resistor for each LED in your display
but y tho

>>1413798
ARRL Handbook, maybe their study guides up to Amateur Extra class? of course it's mainly going to cover modes and bands of general interest to the licensed ham, but you can probably extrapolate

>>1413791
Ah, that's not too bad an idea for a project. Don't want a neutral/out-of-gear indicator? Also making an R look not like an A means you'll need something other than a 7-segment display. A 14- or 16-segment display (pretty common) would work though, and it would only add another transistor or three to the works compared to just using a 7-segment A instead.

To do it with ICs the best way would be to use a 1-of-16 decoder, a hexadecimal-to-7-segment decoder (MC14495P1), and a few peripheral transistors to use the first reed switch to display an F but add the extra two segments to show an R. You'd also need to use discrete transistors to drive the LED display since the split segment(s) will draw twice the current as the single segment it would be expecting. Unless you want to cheese it and use a microcontroller.

Pic related is the semi-compact truth table for doing it in discrete transistor logic, but expanding out the "x"s is proving a right pain I'm not even going to attempt a karnaugh map.

>>1413812
>7 transistors
Sure maybe with a nixie tube, not with a 7+ segment.

>>1413812
>>1413814
Oh shit, that's right, diode logic OR gates. Never mind me.

>>1413814
>unless you want to cheese it and use an microcontroller
I have an Uno, but I felt it was too cheesy and wanted to do it without to learn more. Thank you.

>>1413817
Yeah, turns out a microcontroller woulda actually be more difficult than the diode route. Don't even need the transistors, that is if the reed switches are safe to put 150mA through.

So I've got this frequency divider circuit. I think it should divide to 32.768kHz*2/(2^9+2^8+2^7+2^6) = 68.277 Hz but when I simulate it I get 69.48 Hz. Am I missing something or is multisim just off?

>>1413809
I dont see where you mean that columns 0 and 1 are connected and would appreciate it if you could point out where. I dont know of any problems with this design as I havent made it yet. I just came up with this new design and layout last night and it hasent been tested. All of the issues I had with past design was with capacitive interference which should be fixed by the rwos and columns not being on overlaping layers and having the ground/5v layer in-between the sensors and the and the jumper/teensy layer. My main ask was if anyone had any thoughts on if this is good or not for capacitive sensors as I dont not have hardly any experience with electricity and havent the slightest clue if it will work.

>>1413810
Thats the worst part of this. Pic related.

>>1413812
Thanks for the PNs. They are pretty cheap and I will keep them for future reference but wont be able to use them for this project considering the "Expected delivery: 30-50 days." I will look more to see if I can get a few of the AT42QT1110 and see if I can get a small break out board spun up for me. Thanks!

>>1413826
It's higher than it should be, so it isn't propagation delay. The output should be counting every 480 input counts, but the real number is between 471 and 472. It could be lower than either of those (and still be a sum of powers of 2) and have an extra propagation delay to push it up to the 47.6 counts, but it seems unlikely. Make a next-state table from the circuit? If you skip all the extra counts where Q6 doesn't change it's only 16 or 32 rows deep.

>>1413833
That's 471.6 counts.

>>1413822
>more difficult
in 14-bit PIC it's fairly easy. I see a subroutine of BTFSS/RETLW pairs to read pins and get the right pattern for the display, maybe 6-10 instructions for port setup, then a main loop that calls the sub and stores W to PORTC, forever, for a total of ~30 instructions

>>1413826
maybe it's simulating one or more ripple carry glitches?

>>1413840
The counter is done wrong in the simulator, with Q4 dividing by two and Q5 by four, etc, unless you compensate by scaling down the input frequency so it's possible they fucked it up some other way.

>>1413829
Are they NOT connected in a god-awful way according to your schematic? If not then so God help me and strike me down from my high and mighty throne in the Heavens.

>>1413814
>>>1413812
>>7 transistors
>Sure maybe with a nixie tube, not with a 7+ segment.

no, tranistors can still be used, coupled with resistor, determined by the mA-usage of the LED. Using a NPN or PNP tranistor is determited by CA or CC. One have to decode the inputsignal out to the segments.

For neutral, I'd try a "-" (segment G), rear could be a small r . Funny I have almost same project, a small counter/timer, here's my "draft" . ATTiny 2313, blue 0.5" LED, sockets, not done anyting yet. The LED have been rubbed white by wet sandpaper.

>>1413814
Maybe the "R" for "reverse" can be "E" for "escape", "C" for "cowardice" or "F" for "fuck it" instead?

Noob question, is a pickit4 really worth it? I looked it up but I only get mixed reviews, some say it works great others say that it is complete crap and that the Pickit2 is the best option.

>>1413861
i had trouble getting my chinese pickit2 to work, but my chinese pickit 3 works great. i see no reason for a pickit 4.

>>1413864
I'm kinda scared to buy something Microchip stopped giving support to, from what I understand the pickit3 will no longer be updated.
Also what are the problems that may arise from a generic chinese pickit3? the originals cost almost 100usd in my country but the generic ones cost about 30usd

>>1413826

as i see it, the counter counts to 960 then resets to zero. since you're taking Q9, it goes high at 256, then low again at 512, then high at 768, then resets to low at 960.

that's 2 full cycles every 960 counts, which is 68.266 Hz

>>1413856
I see the issue here. What you consider columns I consider rows. Those are connected because half of each sensor square is a row and the other is a column. Those two halves connect like that so that I dont have another of what you consider a column of jumpers to do. In that black square where they meet I will connect it on the back layer to what you labeled column 2 to connect that row.

>>1413866
mine can't seem to pull vdd above 4.6 volts, so i have to manually lower the supply voltage setting every time i use it on anything that isn't self powered or i get a "low voltage" error. besides dumb chinese things like that i haven't heard of any issues.

>>1413858
With a 14-seg display you could easily test for the R condition and feed it to segments "k" and "d", turning an F into an R. Not that that's relevant any longer.

>>1413868
Well I guess God almighty is gonna have to strike me down then because apparently we have very different ways of thinking. Sorry for wasting your time.

>>1413868
Also you still can't solve >>1413810 without crossing wires though so I guess God Almighty will let me live on his ranch and tend to His horses.

>>1413431
http://www.farnell.com/datasheets/315929.pdf

>>1413875
You didnt waste any time so no worries. Different ways of thinking clashing is how you get new ideas.

>>1413878
On second thought you wasted the time I spent falling for your autism test so maybe you should be smitten :P

I need to replace a blown 470uF 25V cap in an old TV. I don't want to order a single cap. Can I sub a parallel network of 300uF, 100uF, and two 22uF, each 50V? I know theoretically it works out but I don't know about practical ESR and voltage ratings.

And I don't have a circuit diagram to know what the cap does so I don't know if I can simplify the substitution.

>>1413896
Oh I am smitten with you, anon, maybe we shall one day find love with each other, since you seem to be falling for me. You still can't wire your circuit using just a one-layered PCB though.
>>1413899
If you're skeptical you could always simulate it, but I doubt the setup you're suggesting would prove to be a problem. Try it? ALTHOUGH "an old TV"? If you're talking about a CRT monitor I would wait a while before opening it, as they can hold quite a charge for several days.

>>1413899
Oh and to elaborate, it will kill you. Make sure any CRT monitor you work with doesn't have that kind of electric charge in it.

>>1413920
>>1413916
It is a CRT TV but I'm certain it's not any of the lethal caps. Based on other TV schematics I've seen I think it just just couples a power rail to ground. I just didn't know if the exact uF needed to be matched or if ESR is an important consideration for parallel networks.

>>1413922
parallel capacitors inherently have lower esr/esl than a single large one. your nigger rig wiring might counteract any savings you'd see there, but in a well designed system the esr of the bulk capacitance is not critical because there's smaller distributed capacitors providing local power where series resistance matters. that's no guarantee you won't have issues.

>>1413916
>You still can't wire your circuit using just a one-layered PCB though

>>1413922
>tfw you realize "deathcap" refers boths to the deadly mushrooms and to the caps that hold a lethal charge

>>1413927
look at those parallel plates

>>1413287
Who the fuck did that solder job?

>>1413926
this guy's a real pro

>>1413937
https://twitter.com/girlie_mac/status/967919459309531138

>>1413940
Jesus fuck. She needs to never, ever pick up an iron again.

>>1413916
We need a password so that one day if you find someone struggling to make an impossible one layer pcb or I find someone swapping columns and rows, then we know its meant to be.

>>1413941
She seems to be self-conscious about her soldering though. That's better than 90% of idiots.

>>1413940
I can't even see what she did to get joints that bad. They're shiny and covered in what I assume to be rosin, are they just cold? Is the iron too low power? Certainly not a problem I've ever had.

>>1413948
Nope, it's a Hakko 40W NTC, if anything that would be too hot for those joints.

I don't understand the point that the Art of Electronics is trying to make.
>most of its collector current comes from Q 3 ’s base (because only ∼0.6 mA of the 4.4 mA collector current comes from R 3 – make sure you understand why).
Okay, if we were to just have circuit (a), the base current is:
Ib = (15-Vbe)/3.3k = (15-0.6)/3.3k = 4.4 mA

In circuit (b), the BE junction of Q1 holds a constant voltage over R3:
IR3 = 0.6/1000 = 6 mA

>That is, R 3 does not have much effect on Q 3 ’s saturation. Another way to say it is that the divider would sit at about +11.6 V (rather than +14.4 V), were it not for Q 3 ’s base–emitter diode, which consequently gets most of Q 2 ’s collector current.
Why would voltage divider be at 11.6V? I don't understand what the authors are trying to say.

>>1413961
Without Q3's diode to pull the R3-R2-Q3 node to 14.4V, the node would just be a voltage divider between R2 and R3, or 11.6V.

As I see it the job of the circuit is to drive Q3's base between 15V to ~11.6V, since it doesn't need to be held too far from the 15V rail to be off. The point they're making is that R3 has little impact on the circuit when it's on and can therefore be increased without Q3's base current. The downside being this will worsen the time taken for Q3 to turn off again.

>>1413971
>Without Q3's diode to pull the R3-R2-Q3 node to 14.4V, the node would just be a voltage divider between R2 and R3, or 11.6V.
Ah, thank you.

Also, is there any preference to 2nd edition vs 3rd edition for the Art of Electronics?

>>1413961
>0.6/1000 = 6 mA
No, 0.6 mA. V/kΩ=mA and V/MΩ=µA.

>>1413899

please remember that ESR only became important lately coz of switching power supplies and the requirement that they act fast. old-timey electronics like CRTs generally dont care.
and caps last a lot longer when you're not abusing them with high-frequency ripple.

>>1413980
>Also, is there any preference to 2nd edition vs 3rd edition for the Art of Electronics?

2nd edition is from 1989. that's so unimaginably ancient, it precedes naked girls on webcams.

>>1413697
>heatsink plaster
>a chinaman literally just jizzed into a tube

>>1413776
1) For low voltage stuff pretty much anything will be fine. Assuming you'll only be taking measurements of actually sine waves not having tRMS isn't a problem - it's only for non-sinusoidal waveforms.

2) Starter motors draw >200A on startup - it's not reliant on your charger, it's reliant on your battery.

3) That depends on it's state of charge (pic). Fully charged should be about 12.7V.

4) To actually assess the condition of a lead-acid battery it needs to be tested under load, seeing how the voltage changes with varying load currents; this can't really be done at home though. Taking the fully-charged voltage, and checking for self-discharge (multiple measurements over the course of hours) gives you a rough overview though.
http://batteryuniversity.com/learn/article/difficulties_with_testing_batteries

High AC/DC current measurement: Clamp ammeter or shunt resistor + voltmeter? Discuss.

>>1414026
*Sinusoidal at 50-60Hz, I believe frequency also makes a difference to the method by which an average voltage is calculated.

>>1414031
AC: Clamp is probably fine
DC: I've heard bad things about the accuracy of DC clamp-meters.

>>1414034
Oh but an inline (not clamp-on) hall-effect current sensor should be fine.

>>1414031
Shunt if you can patch it in somewhere, clamp meter if not.

>>1414034
Hmm, isn't RMS just 1/sqrt(2) * Pk-Pk? There's no dependency on frequency in the equation, at least.

>>1414039
If the wave was just being fed into an oversampled ADC, have its peak measured, and divided by √2 you'd be right, but in that case they'd just use that data to measure the true RMS value. I think the circuitry that returns a DC average value to send to the main ADC is frequency dependant.

>>1413810

Anyone know what's the purpose of R6 and R7 in this bandgap reference?

>>1414026
Cool. I guess I’ll see how the scoot battery actually holds the charge. Was at like 12.0V when I first put the charger on and around 13.0V when I took the charger off before work. It might be time for a new battery too since that one is almost 5yrs old.

Tried out my shitty camera shutter switch, it works flawlessly. But it's still manual. I'll see about hooking it up to a microcontroller and programming a series of photographs so I can just leave it for hours. Light pollution could definitely be less, I'll see about taking some shots when I'm back out in the country next week.

>>1413776
the best test of a battery is volt droop under load, and the best test-load is the actual job it's supposed to be doing.
pull the spark plug cap off so the motor won't start, then observe the battery voltage while you crank. if it doesn't bounce back to over 12V when you stop cranking it's flat, and if it drops under 9V during cranking it's pooched.

Do they make photoresistors in both configurations?
bright low resistance - dim high resistance
dim low resistance - dim high resistance
Not sure how to search for them.

Looking for bright 100kohm+ - dim <10ohm

>>1414132
No. You need extra components to achieve that.

>>1414132
>analog sensors
disgusting
what is this? 1960s?

>>1414073
Which camera? You can install magic lantern on canons and get an intervalometer through software.

>>1414137
Which components? Are they low power?

I bought a fairly decent (I thought) UPS, Cyberpower 1500 VA AVR, and in battery mode it shows about 20% loading when powering my computer and a few other small loads. Pic related is the waveform; I don't know why the voltage is so low, maybe I had a voltage divider on it when I took the shot several months ago.

So I got a new computer that draws less power than my antique, and the power supply is an EVGA 450W which is way more than I need, since I use the onboard GPU and none of the fans ever speed up hardly at all. Typically the sensors say about 30 watts and 10 deg C cpu temperature.

The sucky part is that this new power supply makes an awful clicking sound, probably about 60 hz, when on battery mode. Any thoughts, other than that I should have gone for the sinewave UPS; should I be concerned if it kicks in when I'm away, and is on for hours?

>>1414145

I'll probably get one of these if you guys don't have any bad news about it. User ratings on Amazon are pretty decent.

https://www.amazon.com/CyberPower-CP1000PFCLCD-Sinewave-Outlets-Mini-Tower/dp/B00429N192

>>1414143
Can't be arsed to think, as it's almost certain that such circuit wouldn't be the best way to achieve whatever you actually want to achieve.
So, design your circuit around the fact that LDRs and other photosensitive components generally conduct better or produce more power when illuminated.

>>1414045
Considering the idea of that reference is to run Q1 and Q2 at different currents: set up the different currents. Also, produce current-related voltages for that servo amp.

Halp
Can someone explain to me what tHR (Read Data Hold) means in this diagram?
The hold times for outputs (tHW, tHA, tHRW) make sense to me - when the processor starts the next cycle, there will be a small delay during which the old values will stay on the bus. But what does hold time mean for the processor's inputs? Following a read cycle, after the falling edge of Φ2, do peripherals need to keep the value read on the data bus for all of tHR (in addition to tDSU (Read Data Setup) before the falling Φ2)? Or is tHR a upper limit, meaning peripherals need to stop outputting to the data bus within that amount of time? Or what?

Diagram is from http://archive.6502.org/datasheets/mos_65ce02_mpu.pdf (page 9)
Modern version (what I'm actually using) is http://archive.6502.org/datasheets/wdc_w65c02s_may_17_2013.pdf, but I found its timing diagram (page 26) to be utterly incomprehensible

>>1414154
It means that your incoming data must remain valid at least tHR after Φ2 falls. Processor latches the incoming data at that edge. Because of that, it's defined as the minimum time, instead of the maximum.

>>1414157
Cool, thanks

>>1414145
clicking sound is the power factor correction circuit in the psu having the worst day of its life. the actual power-maky bit of the psu doesn't give a fuck, a dirty squarewave rectifies just fine.
protip: 'sinewave' is a marketing lie, it's outputting a stepped squarewave. you can pay more money for finer steps, which will turn your click into various pitches of squeak and squeal.

I'm having a bit of trouble with one of the power supply circuits in a screen printer.
It's just a transformer stepping 115V down to 42V which is then rectified, filtered through a choke and smoothed with capacitors.
When I disconnect the outputs from the rectifier, the pins measure the voltage correctly but when I connect the outputs, the voltage completely drops off.
I assumed it was one of the servo packs pulling too much current but I supplied the packs from a bench PSU and they pull around 6A max. The rectifier is rated for 25A and the transformer is able to handle 9A output.

I also noticed that if I connect only a capacitor across the outputs of the retifier, The voltage completely drops too.

The caps seem good as I can charge both of them up to 48V using a power supply limited to 10mA.
I've also switched out both the transformer and rectifier as I had spares.

So basically it seems that the voltage drops out completely even under the smallest of loads.

I just wanted to make sure I've exhausted all possibilities in this part of the circuit.

The 115VAC being fed into the transformer is coming from a solid state relay so I'm going to check that next.

>>1413552
I am starting my project, wish me luck

>>1414169
>the actual power-maky bit of the psu doesn't give a fuck, a dirty squarewave rectifies just fine.

ah, thanks. maybe I'll keep the clicker for a while, as long as the computer seems fine with the clicking when I test battery power.

>>1414199
What is pretty cool is that on youtube they said you need a resistor to keep the CPU on, but mine stays on even without the fake load, so I don't even probably have to put it in

>>1414175
Sounds like your SSR is off and all you get is the leakage current of its snubber network.

>>1414206
*PSU

>>1414206
>they said you need a resistor to keep the CPU on,

from what I've read newer PSUs don't all need the load to stay on.

>>1414218
>newer
this one is probably older than you are and I are

Is it really safe to connect grounds from a 24V power source with massive current, and a tiny logic board that has 3V3 and only tiny current tolerance?
The manual says so but I am VERY scared.

>>1414225
>The manual says so but I am VERY scared.

You always connect grounds when more than one circuit shares common signal or power voltages, and there should be one ground level for everything, and no ground loops.

>>1414229
But then whats stopping dozens of amperes going through the load, from sprinting through my logic board and lighting the paths up like a Christmas tree?

>>1414232

good point. don't connect the grounds, just connect everything else. I mis-read your earlier question. You NEVER connect grounds.

Glad you caught my error in time.

>>1414225
Circuits don't have "current tolerance" on supply, they'll only draw as much current as is needed. It's necessary to share grounds to pass signals between them, excluding optocouplers and such. OTOH if you connect a 3.3V device to a 24V supply rail it's fucked.

Do you mean you're switching a 24V load?

>>1414075
Cool. Noted and saved.

I shoulda taken a reading yesterday before I put the charger on. Once I turned on the charger, it was right around 12.0V and this morning the charger was closer to 13.0V. At the very least it’s reading 0.3V higher than it was before.

It’s def a mix of old battery and not enough use. I’ll replace the thing eventually.

>>1414225
>Is it really safe to connect grounds from a 24V power source with massive current,

you've received contradictory information, so let me set you straight: this is perfectly safe, and in fact, is almost certainly required for it to work.

it's only the other side of the supply, the +24V, that you have to keep away from your sensitive bits.

Is anyone familiar with image sensors? I'm slightly confused by Sony's specification and can't really find information on this.
Pic related is a spec from one of Sony's image sensors. Why is Sensitivity in this rated at a higher millivolt value than the Saturation signal, which is supposed to represent the total well capacity of a particular pixel?
Below are details of how these are measured. Is there something I misinterpret?

What are the names of these components?

1. These look somewhere a cross between a transistor and IC chip.

2. Looks somwhat like a crystal. Except it's just silver foil.

3. Tiny little SMC. Three prongs on the bottom, one on top.

4. Looks like a transistor. Except it's got a freaky alien face. 'N3' inscribed on front.

>>1414344
Pretty sure 1center is a rectifier, you should be able to see a ~ over two of the pins, and +/- over the other two. 1right looks like a resistance network, but i'm not sure.
Again, you gotta google the codes written on the chips.

>>1414344
Google the part numbers. K9650M is a 2ch audio IF filter. 4-lead part may be a bridge rect. AN7310 is a LCD backlight controller.
2, no idea.
3 looks like a power component, likely a regulator or reference.
4 is an IR receiver.

Parts out of an old clock radio, tv maybe.

I can't seem to think out the answer to this situation myself. But it should be obvious, idk.
>have the need for 120v DC
>want to simply put 10 12v batteries in series
Is it possible to charge all 10 batteries from one 12v source without costly isolation?
I can just use one diode per battery, right?

>>1414139
>analog sensors
they're all analog at some point, anon

>>1414343
if you could image under those conditions and there were no other limitations, you would get that output, and you can scale according to your actual conditions (aperture, intensity) to calculate the expected output
my guess is they don't measure it under the stated conditions, or their saturation signal minimum is extremely conservative

>>1414354
no, where's the other terminal on the battery go?
you'd best just charge the whole stack

>>1414359
It's a 4000w load in a vehicle, I was planning on just getting a bigass alternator, but I'm not sure how to step it up to 120 in a cost effective way

Where can I find which voltage/current is safe on the body?
I want to voluntarily shock myself, but I get no info what voltage I need at my wrist to be uncomfortable without any danger. How would I build that? For high Voltage I'd need AC to use a trafo right? I'd like not to use those toy shockers or stun guns if possible. I don't want to get a dangerous shock or have involuntary muscle spasms, just to be uncomfortable. The goal is to train myself like you would your dog with a shock collar.

>>1414364
Depends, but for safety, you probably want about 400 volts.
What you really want is capacitors, they will be good for this because they Seperate you from the dangerous power generation method, you only get a very brief shock.
Just pick up some large 400v capacitors, and have them charge and then dump to your arm when you want to be shocked, should do fine.

>>1414368
So basically: Power source(AC) -> trafo for high voltage (400V) -> full bridge rectifier or similar for DC -> load a capacitor. And use transistors to switch from charging to discharging the capacitor? Any source on how this doesnt kill me? I mean I know I'm asking anons, but I really don't want to kill myself.

Sounds about right, you'll be fine.
May want to have a second mains cut just to be safe.

>>1414375
Thanks man

>>1414364
some TENS units go up to about 80 volts. as always, for safety, avoid even potentially passing current across the heart
frequency is actually important. generally, the higher the frequency, the less likely there is to be muscle involvement. lower frequencies (hundreds of Hz) produce more of a "thud", higher frequencies (several kHz) produce more of a "sting". also, current density matters. use the smallest contact surface area to concentrate and intensify the effect
I recommend you look into TENS units and the designs thereof. most of them use an audio output stage transformer to step up a current pulse. for your experimental pleasure, http://www.chemelec.com/Projects/Tens/Tens.htm , also see the book "Juice!" by "Uncle Abdul"

Is it normal for flux to boil while soldering? I'm using some somewhat cheapo 63-37 with 1.8% flux core, and soldering at 350°C. If I happen to heat some solder for a couple seconds too much, it starts literally boiling. Even before that, some droplets of flux will fly out. Both my board and work surface are pockmarked with dried flux droplets - which doesn't bother me as long as it's not harmful.

Am I doing something terribly wrong?

I'm going to convert a radio into an internet radio / bluetooth speaker.
It has a TDA7266 amp that I'm planning on reusing, along with buttons and a 16x2 LCD (16-pin). It's got a 9V supply and an LM1117T 3.3V regulator and an LM2937ET 5V regulator.
It would be nice to be able to run it off lithium batteries, but I'm guessing the current setup is probably too inefficient.. Should I consider just buying a small d-class amp and running 5V?

I'm planning on using these:
https://www.ebay.com/itm/ESP-WROOM-32/202115988537
https://www.aliexpress.com/item/PCM5102A-DAC/32877909874.html
And I guess I should order a breadboard and jumper wires as well.

I'm looking at running this, but I would like to be able to switch between internet radio and bluetooth speaker functionality:
https://github.com/MrBuddyCasino/ESP32_MP3_Decoder
The author has this to say about the issue:
>"The problem is that Bluetooth consumes a lot of RAM, and the web radio needs a lot of it, especially for AAC. It is currently not possible to dynamically allocate and de-allocate the Bluetooth memory, which makes switching between the two modes during runtime impossible. You might get around this if you use a module with external PSRAM."
Does this mean that you could make it work but just need to reboot the thing?

>>1414403
Or could I just get an ESP32-WROVER and not worry about that issue at all?

>>1414403
>9V supply
definitely get the power situation sorted first. it's more efficient to step down than to step up. if you keep the amp that's there, you would probably use a 3S Li+ pack (and enjoy the attendant charging complications), and you might consider replacing those regulators with something more efficient if there's much current drawn through them. check datasheets, yadda yadda
it would be easier to charge a 3P pack, probably, maybe
>switch
it sounds to me as if you would need to reboot to a different image, unless you can have a digital input or NVRAM flag that you check at startup to determine which mode to go into before turning up the BT. if you can't do this, go for the WROVER

>>1414344
1 - SIP IC's (look for datasheets yourself)
2 - Is a polystyrene capacitor
3 - Just lookup the number printed on it, acquire datasheet
4 - IR sensor, with built in IC. Used for the receiving end of a remote

>>1414142
Nikon D3100. That's an interesting idea, I've never thought about custom firmware for a camera.

Best soldering station for around 50 burgerbux?

>>1414645
the ts100 supposedly. I've never tried it though.

>>1413940
>https://twitter.com/girlie_mac/status/967919459309531138
would bang

>>1414682
It isn't technically a soldering station, and I've heard it has issues with the handle getting too hot. You're paying a fair bit for the "reprogrammable" feature that you could emulate with a $2 arduino and a normal soldering station pencil, and it doesn't come with its own PSU. Add the PSU and it's probably closer to $70, and by W/$ it's pretty low on the scale when you add the PSU cost.

So I kinda made a thing that measures blood flow. I think if you are really autistic you can get blood 02 concentration but I`m a brainlet. Until my pic programmer arrives I`ll have to use a uno to interface with the PC. Next tep is to learn how2solder, process a signal (anyone knows what software you use for that? I`m using spss but that`s for statistics, and maple is living hell), Ignore the pseudo spanish

>>1414707
>>1414707
Here is the circuit used if anyone cares. The plan is to use the PIC to process the data and send it to the pc or save it in a SD card or EEPROM. I still need to make a thing for you to stick your finger in so vibrations don`t mess things up. Also gain on each stage is 100 or so, giving a total gain of 50db(I think?).
Could not find my phone charger and using those circuits planners takes more time than paint. Also you could use a photodiode or something you scrunge up from a remote I thing.

>>1414709
fuck there is a 6.8k resistor to ground after the 1uF capacitors.

>>1414707
>>1414709
I made a pulse oximeter recently and the signal to noise ratio in this data is very appealing to me. We used a 1MΩ resistor to pull the phototransistor high, so your 1kΩ is blowing my mind. I can't imagine the initial signal amplitude was very high.

We processed our data by sampling it with some in-lab software/hardware combination that had an ADC and outputted a csv-ish file. We took a minute of data and threw it into python where we obtained the FFT of it, so it wasn't a live method. I think if you know what you're doing you could just use the microcontroller and use a time-weighted average on the time between pulses (instead of a running FFT as that will have too low a frequency resolution) and feed that directly via serial to your computer (if not a display on the MCU).

To do oxygenation I think you need both a red and infrared light and switch between them and compare their amplitudes, which is a right pain and I never got the data I needed to calculate oxygenation by doing that. I needed to both have measured a baseline between red and IR, and measure both lights' amplitude with no finger present, the latter of which never crossed my mind until the report was due in a few hours.

>>1414707
>SpO2 concentration
requires a red LED, possibly with a critical wavelength, and a rather less critical IR LED around 950nm (iirc)
read the app notes from some of the semiconductor companies, regarding their integrated sensors. TI has a better one but I don't have the link handy. https://www.nxp.com/docs/en/application-note/AN4327.pdf
>signal processing
start with some auto gain/offset control (peak detector with decay -> weighted-sum LPF), then you can find the events more easily

>>1414712
Here's the data we got and its FFT. Personally I've no clue what kind of software you could use to process a live stream of data and graph it.

>>1414712
I did kinda the inverse of that because I`m too lazy to program the arduino right now. I got raw data from the serial port, converted it to .csv, and did a moving average on it, That was it I think. No fourier (yet). The original plan was just to use an osciloscope to see the signal, but even used ones are expensive here so I used the Uno. I need to add a button to start recording, record for some time, spit out the data into the pc and do stuff with it until I decide what kind of thing I want the MCU to do.
>>1414715
yeah I plan on doing that but I need to make the finger holder thing or just tape the thing to my neck and see what gives, because tiny vibrations change the peak greatly. And the plan is to integrate the signal to get the mean blood pressure (which is what those medical things show below EKGs lines) The mean pressure is kinda similar already but well.

>>1414717
Thanks for the info man, and out voltages turned out quite similar. ~100 in my graph is 0.5V. Here is the raw signal. The right image is there because I was explaining it to someone.

>>1414718
Well it's not like we didn't both have amplifier stages, any similarity in output voltage is coincidental. So how do you get blood pressure from this anyhow?

>>1413287
Whats are some cheep tools I should get?

I have a shitty soldering iron and a multimeter

>>1414720
in think you can find info in medicine books about bloodflow on capillaries.
I`m going to talk to my teacher about this tomorrow, see if he has some info.
In my head this whole signal business is pic related.
numoot won`t let me post links, see image name

>>1414389
Thanks for the source man

YES! IT'S ALIVE the mosfet test was successful. But for some reason, without the pull down resistor the gate never ever closes. That should be impossible, the capacitance should leak out like in a second

>>1414831
I don't get it, the current isn't dropping at all even tho the gate is not connected, how the shit is the mosfet keeping itself open without losing gate charge

>>1414831
>That should be impossible, the capacitance should leak out like in a second
Why?

>>1414836
because the other anon said it would and nobody on 4chan lies

>>1414832
>>1414831
I thought the whole point of pull down resistors was to avoid exactly this

Babby's first 6502 CPU instruction
Gonna make sure writes also work, then try hooking up some actual RAM (right now the Arduino is the only thing there)

>>1414832
are you sure it isn`t a triac?

>>1414448
>definitely get the power situation sorted first
Yeah. I think I'm just gonna keep it hardwired for now. Seems like you can't get a cheap balance charger either.
I found some "bare" wrovers for $3/piece so I guess I'm just gonna go with that to dodge the other issue. Unfortunately the cheapest breakout board I can find is $2+$3 shipping, so I might just niggerrig it like pic related..

>>1413445
Tektronix are such bros.

I want to get the TS100 soldering iron but I don't really have much of a budget.

https://www.ebay.com/itm/MINI-TS100-Digital-OLED-Programmable-Interface-DC5525-Soldering-Iron-Station/232803167012

This one's about half the price of the official one on Amazon, but I'm not aware of any other chinese clones on the market. Is this a clone or just a scam? And if it's just for casual stuff like mice and mechanical keyboards, do I really need to spend money on temperature control or can I just get a Hakko FX901 and call it a day?

>>1414941
Fucking why? It's a stupid normie meme.

Is it common for "yellow" LEDs to be pretty orange? I want something bright yellow. I know this is a stupid question but I don't have many different samples and I don't have any physical stores nearby I could go look at, so I have no real idea whether I just got myself some shitty leds, or whether they're normally orange but proper yellow can be found with some digging, or what.

Noob here
Can someone explain this very simple 555 circuit to me?
http://www.555-timer-circuits.com/bike-turning-signal.html
I don't really see where there staggered activation of the LEDs is coming from, surely once 3 goes high it's going to turn on all the transistors and ground all the LEDs at the same time

>>1414960
Wait, nvm. I think I got it.

>>1414945
Yellow always seems to be the most random of all the specified colors. You should be able to filter by the wavelength that you'd like on the supplier's page, though.

>>1414965
Yeah after posting that I tried exactly this, but I can't find anything. What I'm looking for would probably be 575-580nm, but everyone's selling 590+ or 560-. Maybe I'm looking in the wrong places, what's a good supplier that can ship for cheap/free in the UK?

>>1414401
I also sometimes get random droplets of flux 'popping' off the tip of the iron when I'm soldering - would be interested to know if this is normal too.

When exactly am I supposed to apply extra flux? I know that I need to tin the iron, and tin the component before soldering them, so when does the flux come in? I thought it's purpose was to stop solder sticking to stuff, am I supposed to apply it around the stuff I'm soldering to contain the solder?

>>1414968
I'm pretty sure that wavelength range is green.
RS Components aren't cheap, but they do free next day delivery and have a good search interface.

>>1414980
I'm going off pic related. My current shitty leds that I'm complaining about are 588, and this seems to match up with the graph. 575 is really bright yellow, 580 is kinda normal yellow - very yellow yellow if you will - but either would work, maybe 578 if I had to pick a single perfect frequency.

I checked RSComponents actually aaaand I was going to say they don't have any, but I looked again and found that they do - just only in SMD packages. I forgot to specify that I want 1.8 or 3mm ones. It's for a keyboard, so while SMD will technically fit, I'll need to solder it to the hole pads on the PCB which seems iffy.

>>1414990
You might have to struggle through with 0805 packages or something then - through-hole LEDs are available in a much more limited range of colours. If you can't find them on RS, Farnell, Mouser, or Digikey then you're out of luck.

>>1414911
>1.27mm perfboard
neat

>>1414401
350°C is a good temperature for lead-free solders, not leaded. try 320° to start, go lower if you can and still get in and out quickly
>>1414977
flux is to clean oxides off of surfaces so that solder CAN stick to stuff. for average THT stuff, what's in the wire is enough
add flux on old boards, old components, or especially large or small work. for SMD work, flux pens are cash

>>1414980
Is there some vast Japanese subculture based on tiny schoolgirls climbing around electrical components?

>>1414831
>>1414832
MOSFETs are voltage controlled, thus they often suffer from a problem called a "floating input" which means that unless there's a pull-down resistor to ground the gate pin it might stay in the "on" potential even after the voltage is cut off since there's nothing to ground the gate pin through.
>>1414839
Exactly.

>>1414997
Yeah I guess I'll do that then. The annoying thing is I already soldered most of the switches in, expecting to use THT leds that go in after the switches, so now the plate being right above the PCB makes it supremely awkward to solder smd stuff and probably impossible without good smd pads. Meaning I'll have to desolder all the switches I JUST added to disassemble the thing and stick the surface leds on.

Nothing to it though, gotta deal with my lack of foresight or compromise on colour. Thanks for the advice anon, I'm off to hunt for non-chink leds.

>>1414997
By the way, what package would you recommend to substitute for 1.8mm (because yeah I had a look and fucking nobody does 1.8mm in any sort of reasonable frequency)? Or would any reasonable package mostly work?

>>1415041
0603 (inch) might fit well in place of a 1.8mm LED but you should measure the pads where you're trying to put them to be sure

>>1415024
>Is there some vast Japanese subculture based on tiny schoolgirls
yes
you didn't seriously think all those high-tech devices were assembled by robots, did you?

>>1414832
you can think of MOSFET gates as three capacitors, with the gate on one side, and the source and drain (and substrate) on the other

>>1415043
> 0603 (inch) might fit well in place of a 1.8mm LED but you should measure the pads where you're trying to put them to be sure
It's just a standard led pad, do they vary that much? Anyway it seems to have 2mm diameter holes, and then something like 1mm inside to inside between holes, and at a guess (because I don't have an instrument precise enough) 0.5mm actual separation between the pads.

By the way, not entirely related, but do 2*3*4 rectangular leds diffuse more light to the sides than a standard round 3mm? I'm reading conflicting information.

>>1415043
>>1415079
To add to that, would a 0402 work? I don't have any idea what the package numbers mean, but so far it's the closest thing I've found in my desired wavelength.

>>1415101
>package numbers
place a decimal point to the left of the first digit and to the left of the third digit. read as rectangular size (usually inches but sometimes cm)
let me show you what you'd be working with. cm scale numbered at top. from left, 0402 resistor, 0603 LED, 0805 capacitor
>0402
maybe, but it'll be a very delicate job, and you're going to need proper equipment for this. tweezers, magnifier, small wire solder, some liquid or paste flux, and a steady hand, and probably some very fine wire, if your base plate is opaque and you actually do need to extend the LEDs to the keycaps rather than backlighting
>rectangular
you may need to experiment, but my instinct says that the hemispherical convex end radiates more widely than a flat end would. you may be able to modify that behavior with metallic tape

>>1415128
Anecdotally, flats or concaves are better for a wide angle than convex LEDs. I had a friend making a head-tracking kit for flight-sims with two IR LEDs and a playstation stereo camera and he needed wide-angle LEDs, for which he had to sand down the hemispherical end. Of course any LED with phosphor over its die will have a wide angle anyhow.

>>1415128
Ah nice, thanks for the explanation. Man making this themed keyboard is a bitch, why is plain bright yellow such a niche colour?
>>1415151
That's kind of what I gathered, except I wasn't sure. Thanks for confirming it.

>>1415248
If you're yellow LED anon note that you can get yellow-by-phosphor LEDs, instead of the common yellow-by-bandgap.

>>1415028

Is this loss?

>>1415248
heads up as someone that's done 0201s. i don't think it's remotely possible to solder those without a hot air gun and i don't think 0402s would be any better. you just don't understand how fucking tiny they are until they're in your hand. i think i ate a few because they got sucked into the needle i was using to place them.

>>1414707
i don't think you can ever get BP from a pulse oximeter signal. you can get HR and hemoglobin saturation, that's it. iirc, you can get the BP information from ultrasound, but it seems like it'd be very sensitive to where you placed the sensor.

as far as SaO2 sensor placement is concerned, fingertips are a common spot. sometimes i will put one on a patient's earlobe or forehead if the fingers are problematic. and i forget who makes our monitors, but there is some obvious signal processing going on. a newly placed sensor will put the monitor into a signal seeking mode, then the signal strength will be indicated by 1 to 3 stars. so something about it is obviously not straightforward.

the best software to tinker with signal data? matlab is pretty good, but that might be an issue for you. there are clones. if you like stats, then you should know there is a free package called R that is pretty popular. and then there are matlab clones as well.

>>1415264
Using a reference table I think you'd be able to calculate blood pressure if you knew how thick the tissue was, and if you use a reflection-based pulse oximetry perhaps the tissue thickness doesn't vary much.

But yeah the only pulse oximeters I've ever seen clip onto your finger.

>>1415251
Is that the "warm white" ones? I've actually got some on order too (I just bought a bunch of different ones and then I'll compare them and either pick the best or give up), they looked kinda promising on the photos. Interestingly the "recommended"

>>1415262
I realised that I had some (white) 0402s stashed away from a chink kit (that came with almost no specs or info so I didn't know the package names until I recognised it from the other anon's picture), and you're absolutely right, I'm definitely not touching those for this kind of stuff.

>>1415264
>>1415270
see image name
>>1414795

how do i find out if my wire is thicc enough to handle some curvy currents?

>>1415275
>bitdotyslash2tC6TNI.png
there is an L after the dot btw

>>1415275
Are you a genie?

>>1415277
Look at the wikipedia page for AWG, and look at the ampaoity section, it should give you a ballpark. But I always just overspec, which is easy to do at low-voltage. Melted a pair of alligator clips at high voltage though.

>>1415284
>voltage
*current

>>1415284
no I`m not autistic

>>1415277
You should be able to find resistance p/u length of the wire to get the resistance over the required length. From that, decide how much voltage drop you can tolerate, and Vdrop/Rwire = I.

The higher your wire resistance the more heating you'll get i.e. you can get away with smaller cores but they'll self-heat a lot.

>>1415286
No I meant voltage. Though it doesn't really make sense. I fried those alligator clips when pushing 10-20A through them into a microwave transformer, but since I've never done anything with lead-acids or lithium-ions I guess for me low-voltage = low-current. Let's just say I meant current.

>>1415262
0402 are actually doable, IF you have a way to keep them on the pads, and IF you have the proper pads and enough distance to keep them off the iron, and IF it absolutely positively has to work

>>1415277
you measure the thiccness and see if she'll hold up under the desired speed
http://www.powerstream.com/Wire_Size.htm

>>1415290
>No I meant voltage.
The you meant wrong. I can easily have hair thick wire in a 1000V circuit and not even get it warm.

>>1415293
>t. can't read
It is truth that I've never had a problem with insufficient wire gauge at low voltages, and that I have had such a problem with high voltages. Now the voltage isn't the root cause of why I had such a problem, and I addressed that in my post, and that it would make more sense for me to have meant current. Subsequently, your post is meaningless.

Also holy fuck we're already at 250 posts.

>>1414718
just found this in my tabs for you. behold, MSP430 pulse oximetry http://www.ti.com/lit/an/slaa274b/slaa274b.pdf

>>1415304
>I've never had a problem with insufficient wire gauge at low voltages
well, if you'd like to, I suggest you try automotive
>250 posts
yeah, we restart about weekly

>>1415307
>we restart about weekly
But we restarted 3 days ago. I guess it takes that long for us to 404, but we were only at 400 posts or so last time.

I'm looking to make a capacitor bank that I can plug into other future circuits like a coilgun or some shit like that. I have all the details for what capcitors I am going to use, but when it comes to cheating and discharging, I don't know what source I would want to start with, whether a bunch of 9v or a car battery, what is a good choice in source for a project like this? Also how should I calculate resistor value for the resistor placed before the capacitors in parallel?

>1 wire communication probe
>it has 3 wires
the fuck

>>1415442
http://www.tpub.com/neets/book2/3d.htm

>>1415519
>supply
>data
>ground
if you want 1 wire use parasitic power and earth/chassis return

>>1415519
Just switch off your brain and parrot the propaganda.

>>1414344
2 is probably a very accurate cap. I pulled a similar one out of a radio some time back and it measured spot on 1.000 something-F

>>1415290
>when pushing 10-20A through them into a microwave transformer,
If you push twenty amps through them with an auto battery (12vdc) they'll melt just the same way
the current fucked them up - not the volts

I have 4 toggle switches each controlling a different led. What can I do to only have one of the leds on at a time? (turning one on turns the other off)
First I'll have to replace the switches with momentary buttons, right? I also looked into transistors, mosfets, multiplexers... pretty sure there is a simple solution out there but can't find it.

>>1415552
It's a polystyrene capacitor and typically ones looking like that have 2...10% tolerance. Polystyrene capacitors with 1% or better tolerance exist, but they usually look quite different.

Btw, I just measured one marked 1nF 2.5% and meter said 1.00nF.

>>1415584
Probably something like this, logic gates are your friends if it's just driving something low power like LEDs. Basically, pushing a button turns the LED connected to its SR latch on while turning LEDs connected to the other SR latches off. The diodes at the outputs are to make sure outputs don't get shorted with each other in case you push several buttons at once.

>>1415604
nice work but at some point a $5 nano starts to make sense.

>>1415608
a what?

>>1415584
>>1415608

but if you dont wanna spend 1000 frustrating hours learning how to program computer chips, you might look at an interlocking switch from Ali Express.

So I have a small arduino nano project. It's supposed to be a wearable around the wrist and therefore I need to power the arduino.
There's two problems:
First: I never powered any arduino project externally before, so pic related is the basic idea. I'd need a chip and a battery to power it and be able to charge that battery. Any recommendations for that? What do I need to look out for? I thought about a small Li-ion cell, but that's where problem 2 comes into play:
I'd wear this while sleeping. Li-ion cells start burning when punctured and I'd rather not have a mine strapped to my arm and wake up to burns. Should I still go with Li-ion? Should I use another battery type? Should I stick with it? any otther security measures I should take?
Thanks for any help.

>>1415617
A arduino nano will cost you 2 bucks and is a fully integrated microprocessor. You can simulate a whole lot of switches etc. You get analog and digital outputs, pmw waves, serial communication etc. It makes sense from a economical standpoint to switch unless you explicitly want to learn about it.

>>1415584
the simplest solution is use a microcontroller
the simple circuit solution is to use DFFs and an OR gate to clock them

>>1415608
yep
but 4 DFFs and an OR gate isn't too bad, if you really really aren't interested in writing code. either one beats wiring up a bunch of shitty NAND gates for hipster cred instead of using stuff you can get in packages (yes, I'm talking to you >>1415604 )

>>1415676
this is true. lolduino the environment is irredeemably lame but AVR chip programming isn't too bad once you get into it

>>1415676
it is useful to learn to work with individual gates if you want to deal with programmable logic

>>1415608
also
>>1415604
being level-triggered, doesn't appear to handle the case of multiple buttons pressed at once very well, unlike the edge-triggered solution

>>1415584
>momentary buttons
Yes, one for each LED and another one if you need to reset the circuit (all LEDs off). Pushing R activates Q9 (unused) and after release you're back to the power-up default state Q0 = H.
>simple solution
If you want a 5V version I'd suggest to use a HC4028 and a single shared current limiting resistor to GND.

>>1415686
ah Arduino is okay, it gets the job done. obv AVR allows for more flexibility.
I'm not saying learning about this is bad, I very much encourage it, but it's not economical for this application.

I ordered some of the $3 WROVERs for my radio project thing but I have no idea what I'm doing.

I was planning to niggerrig it with these things:
https://www.aliexpress.com/item/6-X-8CM-spacing-1-27-PCB/32774106087.html
https://www.aliexpress.com/item/1-27mm-single-row-female/32564662566.htm
https://www.aliexpress.com/item/10pcs-1-27mm-40-Pin-Male-Single-Row/32785126666.html

On second thought it would be nice to have something I could use on a standard breadboard for testing. I suppose I could chop up one of the PCBs and solder on some 2.54mm headers and wire it out to those..?
Other than that, from what I can see, I just need a 3.3V supply and a cap and resistor for the enable pin. Not sure what physical size caps and resistors to get nor what 3.3V regulator I should get either.
I literally have no usable components for this on hand, and I'm hoping to not have to make another order in 1-2 months time once when my shit arrives and I realize I'm missing something stupid.

Current shopping list:
ESP32-WROVER (bought)
PCM5102A DAC
1.27mm protoboards
1.27mm female + male headers
Standard solderless breadboard
Assortment of dupont wires
5V to 3.3V regulator
caps
resistors

>>1415725
>1.27mm
perfect to solder some common ribbon cable onto, if you can't find the particular adapter board. chop up an old IDE cable or the like, or even a floppy cable if you don't care about the SD card port. the IDC connectors are compatible with *male* Dupont leads for ease of prototyping, plus there are adapters designed for rasPi to convert 40-pin 2x20 2.54mm pitch pin header to dual-inline rows for breadboarding, and includes a donor cable for the above https://www.aliexpress.com/item/40-pin-raspi-gangbang/32358660457.html
>dupont wires for 1.27mm
if you find these let me know
if you have no onboard antenna on the module, you might want to grab one of those goofy little cheap heatshrink-wrapped 2.4GHz antennas because the internal FM aerial won't work too well

>>1415725
>physical size caps and resistors to get
depends, any THT would probably work if the leads are thin enough (1/8W resistors would almost certainly fit, and I've seen thin-leaded 1/4W resistors out of china lately too). cap lead diameter can and does vary, if too big you might need to "surface mount" them by cutting leads down to a cm or so and bending the ends into feet (when doing this, hold the lead with thin-nosed pliers by the component side of the bend, not the free side of the bend)
>3.3v regulator
be lazy, get this https://www.aliexpress.com/store/product/benis-DDD/1678083_32837803938.html
>hoping to not have to make another order in 1-2 months time once when my shit arrives and I realize I'm missing something stupid
kek, that's always how it is, but you're doing a decent job of visualizing everything

>>1415757
>perfect to solder some common ribbon cable onto
How would you go about that? I'm just kinda visualizing the thing dangling from a set of IDE cables right now.. Thoughts on the chopped up 1.27mm protoboard and 1.27mm male headers + 2.54mm headers?
>if you can't find the particular adapter board
There is one, but it's a double row thing, so not ideal for a breadboard.. Comes out at $10 for 3 as well.
>dupont wires for 1.27mm
>if you find these let me know
Well, nevermend the female 1.27mm headers then..
>if you have no onboard antenna on the module
Ah, it does. Got the wrong picture.

>>1415776
>depends, any THT would probably work
Right. Was thinking SMT, but THT probably does make more sense for versatility.
>kek, that's always how it is
Well, I almost forgot about the usb to uart adapter already.. And I should have ordered all this stuff 2 months ago, when I first thought about this project. But I guess that's a classic mistake as well.

>buy serial port interface IC
>don't buy the oscillator I need to run the thing
Whoops.

I know it's possible to build your own oscillator out of capacitors and shit. Is it plausible that I could rig one up that's good enough to run this serial port thing for now? Would need to be 800-1600Hz minimum. And what do I type into google to find out the easiest way to build such a thing? Ideally one that's less complicated than pic related, if possible

>>1415820
this circuit sucks less. google "transistor oscillator calculator" and use the table in the first link that 4chan thinks is spam.

>>1415803
>I'm just kinda visualizing the thing dangling from a set of IDE cables right now
sorta, pretty much. slit the cable 2.5cm down into 20-conductor groups (or 10 or even 5, your call). leave alternating groups long, cut the others down. solder to castellated holes. repeat. probably a better idea to use the perfboard with headers, though it's possible 2.54mm pin headers won't fit in it
I recommend a jig to hold the males in place against the castellations, either some stacked perfboards or solder the socket headers into place, unless you've got another way to keep them stable and aligned
>There is one, but it's a double row thing, so not ideal for a breadboard.. Comes out at $10 for 3 as well
if you're just stringing modules together, short Dupont leads should be sufficient for anything point to point (for power/signal distribution points I've used strips of 2.54mm pin headers just floating in space with a bare wire soldered to the board side)
otoh if those are assembly-required adapter boards, you could populate the two rows on different sides of the board (with your own pin header strips) for breadboard friendliness
>SMT
with the forewarning that I've never worked with 1.27mm perfboard, 0603s could fit diagonally, but it could be tight. 0805s might fit horizontally/vertically so as to skip a pad, but might not. 0402s might be good for neighboring pads, but are tough to work with. THT is, arguably, the safe choice, if you're not certain how this is going to go together. how's your soldering iron? how's your SMT skill? got tools? got bodge wire?

>>1415831
Thanks, I'll check it out

>>1415838
Alright. I'm gonna give the perfboards a shot. Think I would need a picture to really understand the cable method. And I guess I'll stick with through hole.
Pic related is the adapter I mentioned.

>>1415849
oh there's the antenna, noice
for future experimentation, if you have that break out board, you could also buy a set of 40 male-to-female Dupont leads (which will usually be constructed with some length of the ribbon intact) and split them into 18-lead groups, attach your own crimp-on IDCs with a vise or channel-lock pliers, then plug the IDCs into the usual dual-row headers and have free ends of whichever gender you like, for attaching to modules and/or plugging into breadboards
SMT is only reasonable/"fun" if you have boards made, imo. they can be bodged onto perfboard but it's something of an adventure

>>1415855
oh god no

>>1415855
>oh there's the antenna
It's a bit odd because there's two versions. One with built-in antenna and one that looks exactly the same except it has the connector for the external antenna soldered on.
Picture looks a bit too fun tbqh.

>>1415855
I don't have the breakout board, and honestly it seems like too much trouble, at least for what I had in mind.

I prepared the heatprobe for the rpi, now to pray for it to work. Also te burning plastic is making my head hurt.

>>1415964
Yay it's working. Now i have to put shrinkwrap on and write a bit of phyton to use the data

>>1415275
>2tC6TNI
>>1415278

thanks, i will print this out and give it a read. fairly new stuff, too, 2014

>>1415270
>But yeah the only pulse oximeters I've ever seen clip onto your finger.

OTC stuff, yeah, it's a finger clip. the stuff we use in the hospital is flat and can be slapped on the forehead, or bent in half to fit a finger or earlobe. the probes come with a few inches of pigtail that fit a keyed DB-9 connector.

>>1415967
So... This is probably a bad idea. But then how do I attach it without drilling into it so it holds well enough to not fall off in temps up to 120C?

>>1415971
Thermal adhesive? CPUs get above 120C afaik.

>>1415972
they don't even sell that stuff in this shithole
all i have is superglue

>>1415673
USB battery bank, or one of the cheapo protection/buck/boost modules intended for one.

>>1415971
Okay, relay is working too. Next stop: i2c clock display

>>1416046
aw... shit... i think i got the wrong display
https://www.aliexpress.com/item/1pcs-TM1637-4-Bits-Digital-LED-Display-Module-For-arduino-7-Segment-0-36Inch-Clock-RED/1969258031.html?spm=a2g0s.9042311.0.0.62b04c4dfk1Dsf
i don't think it's i2c and probably won't work with PI.. oh well it was like 60 cents and luckily my project will work without it, but i will have to use ssh to read the value i wanted to display, which will be annoying

>>1416063
Scratch that! It doesn't obey i2c but me people smarter than I bitbanged it into obedience by hacking the signal with regular gpio pins. I love the rpi community, so resourceful. Imagine how shitty my bomb would be without a clear Hollywood display.

>>1416069
>Imagine how shitty my bomb would be without a clear Hollywood display.
Don't forget the beeping every second. No Hollywood bomb would be perfect without the beeps.

>>1416069
Ha...

>>1416085
Oh my god. That keyboard is fucking DISGUSTING from up close like that.
I don't even want to touch it anymore.

>>1416069

So do you have a trial date or did they drop the charges?

>>1416088
Silly pajeet.
Everyone knows that only whites can make bomb jokes.

>>1413573
lol, clueless retard

>>1416085
I am progressing well, the display now reads a temperature from the probe and i made a python script to turn on the relay if the temp drops below set value. Now i need to code in some safety guards to turn off the relay permanently if it is on but the temp isn't rising

>>1415973
>using a sensor with a huge thermal mass to control a heated bed
>in your shithole you can get heated beds but not high-temperature adhesive
1. dumb idea, you should be using a thermocouple
2. what is this, a Teasmade?

>>1416213
thermocuples are analog, i need a digital probe. the bed even supports a thermistor, so if i wanted to go analog i would use that. but that is gay
the digital probe gives me an actual number without any hassle all i had to do was plug it in
and since this was the only digital probe they had it is what i got. The only downside is that the reaction to temperature changes isn't as fast as say tiny thermistor, but i don't really care about that since i don't need the precision tolerance to be that high, i just need the bed to cruise around a warp temp to prevent warping

>>1416221
see also the MAX31855. all you have to do is plug the thermocouple in one side and clock four bytes out the other side
well, as long as you've got the time for it to stabilize, and you don't mind running the sensor at the upper end of its operating range for however long periods of time, with the likelihood of excursions over the limit if you use simple hysteresis control, you do you

Hello /diy/
I want to make a rpi car computer for my 03 civic. I plan to embed it into the radio deck if that's the right term, with the official 7" touchscreen. I'm guessing I need an amp in between the pi and my car's radio connectors. Would I have to get a 4 channel amp, or something like that, for my cars 4 speakers (and sub)?
I'm realizing how daunting this project is, like how do I shut down the rpi safely when I turn the ignition off? There's almost no good touch interface that works well for the rpi, but I can make my own pretty easily.
Anyone have experience with something like this? Any good resources to use?

pic related is what i have so far lol

>>1416299
>Would I have to get a 4 channel amp
if you don't already have one and you expect to drive the speakers, yes
>how do I shut down the rpi safely when I turn the ignition off
power it with the constant 12 volt lead, switched by two relays. use the accessory 12 volt to drive the coil of a first DPST relay to 1) turn the power supply on and 2) trigger a GPIO watched by triggerhappy or similar daemon. connect set 1 of first relay contacts in parallel with the contacts of a second relay (SPST is fine) controlled by the pi to keep the power on for as long as the pi requires. turn this second relay on first in startup and off last in shutdown

>>1416254
>MAX31855
it's expensive as fuck, i am not throwing out that kind of cash to achieve extra precision i don't need, when i already have a working solution

>>1416315
$2 from china.

what is the CX slot in a multimeter for?

>>1413287
I'm using an arduino and a sensor that outputs data via ethernet. I tried using one of the example codes provided by the arduino ide for the ethernet shield but i'm not getting anything on the serial monitor.

I'm so bloody new to networking and the more I try to read into it the more I hate it. Nonetheless, I have to figure this out because the sensor i'm using is so precise I can't get it anywhere else to have it put out the information in a straight analog signal. So there isn't an alternative for me to use something else.

If anyone has any guidance that I may use I would be extremely appreciative.

>>1416350
Its for measuring how greedy the shekel goblin is being

>>1416313
Brilliant, thank you. I'm sure I'll have more questions in the future.

>>1416354
Anyone? No.

Nobody knows anything about interfacing through ethernet?
Figures.

>>1416361
So if someone doesn't have an answer within 15 minutes you write it off? Boo fucking hoo, go commit sudoku.

>>1416369
I'm still here. Can you help or what?

>>1416371
no we don't know anything about anything. Go away.

>>1416372
Figured as much.

>>1416315
1. $2 from china
2. range, not precision

>>1416354
don't use lolduino for networking if you have no idea what you're doing. use a big girl computer
also don't accidentally your pants on the way out

>>1416371
I am very sorry that I can not help you with your sensor that you have not provided any specs for. Maybe we can still be friends?

>>1416377
I don't want to use a computer, it's too impractical for the application. I need to use a mcu. I'm using arduino currently for proof of concept. I will later migrate to a more advanced micro processor.

>>1416379
I'm not sure it is an issue with the sensor - but here the model number: GXMMS.0208EA2

>>1416384
ah, so this is your first ever rodeo, okay
pro tip: proofs of concept work better when hardware is over-resourced than under-resourced. it is easier to tighten up a working design than to start with a machine that won't give you adequate feedback and has only a half-assed implementation of internets protocol
I suggest you start with hardware more like a pi instead, which you can easily use wireshark to see what is coming out of the encoder, which you can later shrink down to something with a non-joke internet interface like an STM32F4; OR use one of the other RS-485-based buses that might be easier for an 8-bit micro to communicate

>>1416404
>first ever rodeo
right...

You have told me nothing. Ethernet protocol isn't the issue. I can build hardware from scratch to be reading those hard wire signals directly if i wanted to, but I'm way to crunched for time and it's not worth the effort, when there is already existing hardware to do it for me.

It's not connecting to the internet.
The sensor outputs its data via ETHERNET, which i'm sure the arduino can handle receiving and reading packets from it. I will be polling the data and applying an output of some voltage in reference to the sensor's position. This is something all well within the capabilities of an arduino - i'm sure. I'm not familiar enough with pi. Also, having a pi which will require placing an OS on it and then to later control an IO is an idiotic route, which i'm sure you understand why.

>>1416414
People use pis for that sort of thing all the time, not that that makes it a good idea. Regardless it has its own RJ45 port and I imagine people have already put a great deal of work into using them to transmit data through a network for IOT devices. Such code should be available.

I'd start with some more elementary arduino codes, something more easily trouble-shootable. Once you have the know-how to construct your own programs instead of just using example programs you should have a much easier time trouble-shooting it. Post your code if you really want, it's better than not enough information.

What have you got on the other end to receive the data anyhow?

>>1416422
Alright I've clearly messed up on explaining what the setup is and what I'm doing.

I have boundless amounts of experience in low level coding (my job is in embedded systems). I have zero knowledge in using the various ethernet protocols.

I have a sensor, it outputs data via an Ethernet cable (EthernetIP). The sensor has it's own ip address and mac address.
I connected the sensor in an arduino ethernet shield which is mounted to the arduino uno.

I ran the example code (DHCP address printer) - can't get anymore rudimentary than that. Nothing shows up on the serial monitor. I placed some debug println in the code to see where it gets caught. It stops here at the code: if (Ethernet.begin(mac) == 0)
more specifically, Ethernet.begin(mac).

The full code can be seen on the example from the ide.

Am I missing anything? So recap.
It goes from sensor (via ethernet cable) to arduino sheild (which is mounted on an arduino un). Done. That's it. No internet, no hub, no switch. Just one cable.

I do not wish to use a pi. I will not run simple sensors through an entire os.

>>1416414
>not connecting to the internet
any network using IP protocols is AN internet, which is ≠ THE Internet
>will not run simple sensors through an entire OS
IP is simple enough amirite
so why aren't you using one of the other variants of that sensor that would be easier for a shit-tier micro to talk to?
>I ran the example code (DHCP address printer)
I bet that's a DHCP client, not a server. plug it into your LAN and see if it does something. I bet it does, unless your admins have it locked down
if there is no way to set the sensor's IP address statically, you will need to run a DHCP server on your micro to establish the network, in addition to whatever other application code to connect to the sensor and deal with the data. maybe the mega328 is enough for this, maybe not
>to [sic] crunched for time
the reason I keep telling you to hook this sensor to the LAN is so that you can probe the damn thing via telnet a lot easier than you can build a new sketch, plus you've got a known good networking stack and network. you're trying to make everything work at once. you should know better

I pulled this little DC fan out of a scrapped laptop. I want to get it running on it's own but I'm new to electronics. The specs are 5V 0.5A. There are three cables. I know about positive and negative. But what's the third wire do? And how do I discern which wire is which?

>>1416445
>But what's the third wire do? And how do I discern which wire is which?

>>1416439
Why can't I use this code provided by adafruit? I put my sensors mac address and it should read it's IP. What am i missing?
/*
DHCP-based IP printer

This sketch uses the DHCP extensions to the Ethernet library
to get an IP address via DHCP and print the address obtained.
using an Arduino Wiznet Ethernet shield.
*/

#include <SPI.h>
#include <Ethernet.h>

// Enter a MAC address for your controller below.
// Newer Ethernet shields have a MAC address printed on a sticker on the shield
byte mac[] = {
0x00, 0xAA, 0xBB, 0xCC, 0xDE, 0x02
};

// Initialize the Ethernet client library
// with the IP address and port of the server
// that you want to connect to (port 80 is default for HTTP):
EthernetClient client;

void setup() {
// Open serial communications and wait for port to open:
Serial.begin(9600);
// this check is only needed on the Leonardo:
while (!Serial) {
; // wait for serial port to connect. Needed for native USB port only
}

// start the Ethernet connection:
if (Ethernet.begin(mac) == 0) {
Serial.println("Failed to configure Ethernet using DHCP");
for (;;)
;
}
// print your local IP address:
printIPAddress();
}

void loop() {

switch (Ethernet.maintain())
{
case 1:
//renewed fail
Serial.println("Error: renewed fail");
break;

case 2:
//renewed success
Serial.println("Renewed success");

printIPAddress();
break;

case 3:
//rebind fail
Serial.println("Error: rebind fail");
break;

case 4:
//rebind success
Serial.println("Rebind success");

printIPAddress();
break;

default:
//nothing happened
break;

}
}

void printIPAddress()
{
Serial.print("My IP address: ");
for (byte thisByte = 0; thisByte < 4; thisByte++) {
// print the value of each byte of the IP address:
Serial.print(Ethernet.localIP()[thisByte], DEC);
Serial.print(".");
}
Serial.println();
}

>>1416439
Also,
This sensor was obtained a while ago without any say from me.
I'm crunched for time in proving a concept, not checking to see if the sensor works.

>>1416450
yep, that's client code. DHCP servers are what assign IP addresses, and on that point-to-point connection, you apparently don't have one
so, supply one, either in the form of code on the lolduino (which might be interesting to make work alongside some application), or plugging both into the LAN side of a router (for testing purposes)

>>1416453
got it

>>1416457
I couldn't follow what you were saying. Can I make the lolduino (which I'm not sure why people crack at it. It's plenty fast for what it is. And has all drivers and peripherals handled. Pretty okay if you ask me.) a host and just have it periodically poll from the sensor?

>>1416457
Or is it perhaps, I need to make the arduino a host and the sensor will be the client?

>measure resistance between arduino's GND and 5V pins
>843 ohms
This is bad, right? I fried something by connecting my arduino up the wrong way, didn't I?

>>1416457
I fucking get it now!

I'm trying to communicate through a EthernetIP protocol. Now, how can I go about that?

>>1416466
that's a proprietary spec atop generic IP, I've never had to work with it, but it looks like it's based on some other proprietary specs, see odva.org
but, there's an open-source (permissively licensed) stack for it https://github.com/EIPStackGroup/OpENer which you might be able to port or just borrow from, but just by way of warning, it's starting to look kinda big for a mega328. good luck, sir

>>1416461
>not sure why people crack at it
Because the Arduino bootloader is bulky and inefficient. A proper AVR programmer will make the result faster and take up less storage. Arduino is microcontroller easymode. Serial-to-USB ICs included on Arduino boards are also meme-tier.

>>1416464
>measure resistance between arduino's GND and 5V pins
There's not going to be a resistor between the pins mate, there's no reason for you to get a coherent result from that. Testing resistance puts a small voltage across the pins and measures the current that flows. Depending on the direction you could even be across a closed-circuit diode or an open-circuit transistor.

>>1413378
Each cell is 0.5 so you're going to need a 24-28 cell panel.

How to learn verilog? Any good books? Software?

>>1416489
this is how i learned it. it's beautifully simple, even if the ides are all worse than hitler.

http://www.asic-world.com/verilog/verilog_one_day.html

So the cpacutator has 220uq on it but if i stick it into the cx input on my multimeter it says 130, what the shit?

>>1416567
try it the other way round

>>1416570
same reading

>>1416571
your meter might not work very weil with polarized capacitors then
or, the cap is bad
or, your meter is bad

>>1416573
tried another cap, exactly the same reading
I guess i am just going to believe the chinks knew what they were doing (stealing)

I need a 5V T-Flipflop for a frequency divider that can take 20MHz input. Any suggestions? Should I just go for some 74ALS?

Gonna do a heatbed test run, I really hope this yellow cunt can take 10 amps

>>1416583
why not VHC?

>>1416587
>Supports 5-V Vcc
Nice, thanks.

fucking resistor colors
just fucking write it one them you dipsticks

>>1416594
Black Boys Rape Our Young Girls But Virgins Go Without, Get Some Now

>>1416594
don't you wish you'd gone SMT?

>>1416594
It's just the rainbow, but with brown and black in it because it's 2018.

>>1416597
>>1416596
Look at this shit. The colors are equally spaced and i cant even tell which some of them are

>>1416599
for me another annoying thing (on top of having to remember all the colors and what they mean), is that it's difficult to tell which side is which. Like where is the tolerance and where is the first digit...
Fuck color codes

>>1416605
exactly
it would be MUCH easier to print the text on it
plus it's racist against color blind people

Yay it's working and nothing is melting

>>1416616
*yet

>>1416619
stop bullying me

>>1416450
>I put my sensors mac address
put your shield's mac address instead you derp herder.

>>1416605
eventually, you learn the colors, just like any other code
there aren't that many tolerances in practical use, though, so your tolerance band is almost certainly going to be brown (1%) or gold (5%) in daily practice

>>1416616
smells good?

>>1416616
Okay, now I am going to hook that shit up to a mosfet, and use all the machinery I connected to the RPI yesterday, which I control through a python script, to turn on and off the power for the bed.
Fucking hell though, this thing is turning into one big know of wires and devices considering it justs monitors the temp and switches power. It will probably be smart to get like a plank and nail the components on it

I hooked it all together and it's working great, the rpi maintains the desired temperature by turning the psu on and off but temp probe is pretty slow, slower than i expected, i will have to add thermal paste under it to improve heat transfer, but it is maybe only problem during the first heat cycle when it goes from 20 to 80 which is a fast and a big jump

>>1416634
Yeah. when the bed i cold it overshoots the target temp by 15 - 20 degrees before the probe catches up, but once there, since the bed cools off very slowly, the probe has no problems keeping up and since it then need to raise the temp only by about 2C to maintian the desired temp, it catches up pretty fast and allow me to stay in a -/+ 2C precision range which is awesome
I am really happy i didn't have to resort to analog sensors like some caveman, or to expensive analog to digital extensions for the PI

>>1414175
>>1414207
Well turns out it was the relay.
All's good now

>>1416622
>eventually, you learn the colors, just like any other code
I've worked with resistors for many years.
When I'm looking through a pile of mixed values for a certain one my mind says the numbers instead of the colors.
It's like learning a second language.
You're fluent when you 'think' in the second language.
As long as you're translating in your head, you don't really know it.

Is there a difference between a flip-flop with Preset and Clear and a one with Set and Reset?

what are some cool sensors/accessories to put on my carpi?
so far I've got:
>RTC
>Accelerometer
>Backup camera/blindspot cams
>ultrasonic rangefinders around the car

>>1416772
software defined radio

I am currently working with an Arduino and a GSM-module and I want to be able to send data with http to a server, which then a winforms program or a website is going to retrieve information from. The problem is that I have no idea how to make a server/database or where to start. Any tips on where I can learn about making and managing a simple server?

>>1416813
so the arduino is going to receive SMS and then upload them to the server?

>>1416813
Check out David Banas on Youtube, he has a lot of good tutorials e.g.: this one on MySQL:
https://www.youtube.com/watch?v=yPu6qV5byu4

Travery Media also has some like this PHP/PDO crash course:
https://www.youtube.com/watch?v=kEW6f7Pilc4

And this random dude has a decent one on making a basic login system with PHP if you need to make it private:
https://www.youtube.com/watch?v=Pz5CbLqdGwM

A lot of providers will include MySQL pre-installed, all you need to do is go in and add a user account. Namecheap are pretty decent and have a lot of documentation.

>>1416639
>muh trendy 21st century progress larp
come back in a few months when you overcook the sensor

>>1416813
you're gonna need to tell us a lot more about your plan, and while we might be able to help you with getting http connections out through your lolduino, frankly, >>>/g/ might be a better place to ask about web development stuff

>>1416888
>>1417002
Im collecting data from some flow and preasure sensors with an Arduino. My vision is to somehow get this information on a database or a server with the gsm-modules gprs somehow. Im using a GMS-module with a sim card because it's not going to be located near any wifi or ethernet connections.

>>1416896
Thanks!

>>1417013
Oh right so you're using the 4G to upload the data to the server? Does the data need to be publicly accessible or do you want it to be private? Does it matter if you lose some of the readings (e.g.: if you host it on your own PC and need to shutdown/reboot)? Do you know any programming languages so far?

>>1417002
>/g/
>getting help
pick at most one one

My TV is getting no signal from its dish, and I checked under the house and on the roof and all the cable is intact. I measured 66kΩ between the shield and core, is this normal? I'll be disassembling the receiver soon.

>>1417187
Well the receiver itself doesn't have any fasteners in it, so it's probably ultrasonically welded shut or something similar to keep the IP rating. Looks like I'll need a new one.

>>1417187
>I measured 66kΩ between the shield and core, is this normal?
Don't antennas usually have an impedance somewhere in the range of 50-100 ohms?

>>1417141
in my experience /g/ can be helpful when they're not too busy memeing about Haskell

>>1417200
1. active LNA on the antenna element, resistance measurement is meaningless
2. resistance ≠ impedance even for a plain antenna
3. 94% sure >1417196 is a summerpost

>>1417200
That's AC impedance, from capacitance and inductance in the wire/antenna. I measured DC resistance, which should be expected to be larger. Looks like these things have a pretty standard downconverter in them that turns the ~10-12GHz into ~1-2GHz.

>>1417215
>94% sure >1417196 is a summerpost
Legitimately not, just realising after googling what's inside one of those LNBFs how useless a resistance measurement is. I originally assumed it just contained an RF network of some variety along with the receiver, in which case there could certainly be a resistance from ground to signal as part of some sort of filter.

One thing I noticed is that the receiver is able to be rotated from 0 to 30° (in that image it's at 25°) independently from the angle of the dish itself, which I thought wouldn't make a difference but maybe it's a polarisation thing.