/diy/ - Do It Yourself

this thread's digits brought to you in part by the humble Keystone 1656 hex standoff

>>1551622
>Is the intention
not at all
start with obvious stuff like the core MCU, power, clocking, and programming connector. then add any other facilities on the dev board you want to use in the final product (LEDs, displays, memories, switches, etc). then add whatever offboard sensors/actuators you want to put onboard. now you have a schematic, probably spanning multiple sheets, which you can now export into a pcb editor and start routing
in the process, you would probably refer to the MCU's and other (dc-dc converter/display/line driver/op amp/etc.) datasheets and appnotes for circuits, pinouts, and decoupling and layout guidelines, also probably to catalogs, aliexpress, and/or your own component stock inventory to select the right device/footprint to put on your board. along the way, don't be too afraid to change component packages, move board outlines, or add layers to make layout easier or cleaner (according to your budget)

>>1551666
READ RULE 4 BEPPU

>>1551666
>cigarettes without gore pics

Oven guy from the last thread.
So, I gave the circuit some more thought. I figure I want to keep the Teensy dev board to prototype other projects, and I do't exactly want to buy an entire new Teensy board JUST for the oven, nor do I want to have to plug in the Teensy every time I want to use the damn oven.
Isn't the point to take the MCU from a dev board and integrate it into the circuit?

Not sure if I'm writing in correct place, but why all TRS connectors I buy are junk, even expensive ones, but connectors salvaged even from free airplane headphones are so much better?

>>1551674
you shitter

>>1551751
answer for 1.5V relay anon

>>1551753
I did say I was an idiot
been reading all night about mosfets and BJTs
what's the difference and does it matter?

>>1551761
>what's the difference and does it matter?
In a nutshell.
>BJT
Constant voltage drop in fully open state. You control it with current.
>MOSFET
Resistive drop in fully open state. You control it with voltage.
>IGBT
Voltage drop in fully open state, You control it with voltage.

>>1551764
yeah this

>>1551765
In low-current application you want MOSFET, since losses will be smaller (let's say electric kick scooter controller), in high-current application you want IGBTs (let's say a train).
BJTs are for some reason favorable in amps*. (except D-class)

>>1551770
Why IGBTs for high current?

>>1551775
Let's say IGBT has 2V voltage drop, while it is on.
Let's say MOSFET has 0.01 Ohm, while it is on.
For currents lower, than 200A, MOSFET will be more efficient, since it will have lower voltage drop. For currents above 200A, IGBT will be more efficient, since it will have 2V drop no matter what.
Those number are random tho.

>>1551779
Oh I guess you're right. What's gate capacitance like in comparison?

>>1551770
is sharing grounds between different voltages safe?
I really don't want to fry anything

>>1551784
>is sharing grounds between different voltages safe?
Yes. This is how many systems work.
Your computer has 5V, 3.3V and 12V rails, all with common ground.
>>1551782
Idk, honestly. Historically IGBTs were slow, probably because of the gate capacitance, and this is why older trains are noisy, while newer are silent.

>>1551742
Ever try neutrik connectors? Those are pretty good. You can get them at music stores in the pro-audio department.

>>1551785
The real noisy trains are GTO thyristor inverters or choppers.

>>1551785
The real noisy trains are GTO thyristor inverters
or chopper drives.

>>1551784
It's safe though it can cause signal integrity issues in some designs. Like you don't typically want high speed digital logic or high power loads sharing the same ground as your low power analog circuitry.

>>1551830
The two ground networks would normally be connected. But you'd split the ground into two separate rails close to the supply, rather than connecting both analogue and digital components to a single ground rail.

tarduino GPS brainlet from the last thread here.
So someone gave me the link to these commands for my GPS module in the last thread.
http://www.flytron.com/pdf/MTK_commands.pdf
Stupid question but how do I issue these commands to the module.
Can I just use the software serial's write function like GPS.write("$PMTK251,4800*14<CR><LF>");

>>1551728
hm, well, you could, but I think you might run into a bootstrapping problem of how to solder QFN/BGA without a reflow oven...
you might consider redesigning for something like a $2 bluepill and just leave it in there. that should be more than enough to implement PID, a reasonable front panel, maybe even a graph of the actual temperature at however many points in the oven
(btw you might want to think about a MAX31855K or similar thermocouple-to-digital converter, which could be easier to get and will definitely be easier to apply)

>>1551902
if your software serial shit works, yes, more or less, but I suggest you use a USB-serial dongle to set it up AND TEST IT out of the project so you can rule out one possible problem
you will probably have to calculate the checksum, but this is explained in various places around the net that you can google or you can be lazy and https://nmeachecksum.eqth.net/
and of course you would write \r\n instead of <CR><LF> because C

To whom it appeals
Find the inductance
and the voltage

Can you use a R2R ladder with a shift register like 595 to make 1 8 bit DAC or 2 4 bit DAC?

>>1551979
sure, but it'll be no more accurate than your resistors, and best to use a '595 with a symmetrical (i.e. CMOS) output

>>1551728
>'t the point to take the MCU from a dev board and integrate it into the circuit?

Sometimes it's cheaper to get an entire "devboard" from Aliexpress than trying to source just the MCU from Digikey or Mouser. PCB is so cheap, so the space savings doesn't cover the cost for buying the individual components from legit resellers.

>>1551980
So 74HC595 and 0.5% resistors will give me adequate results?

>>1551986
Look at this, someone did just that thing I'm imagining

https://github.com/tardate/LittleArduinoProjects/tree/master/playground/R2RDAC

>>1551800
I will try.

>>1551986
>adequate
that's an engineering decision. try it and see
personally I wouldn't bother when chinkshit 16-bit dual DACs like the PT8211 are ~12 cents
t.did that R-2R stuff 25 years ago with a parallel port

>>1551828
Really noisy drive is in my e-bike BLDC motor controller, which exploded all 6 MOSFETs, battery fuse, and MOSFET in BMS.
After that I think I need an oscilloscope, because I don't quite understand how motors work...

So how stupid is this? I'm tying the cathode high on an LED so the microcontroller can read a '1' when the switch is open. I want the microcontroller to know when the switch is open or closed. Without that 10k tying it high the line that goes to the microcontroller would only be about 1V which is isn't low or high according to the controller. The max reverse voltage on the LED is 5, which I'm under (only around 2V) and there's no where for current to go if the forward voltage is exceeded. I breadboarded it, it works, but is there a reason I've never seen anyone doe this? Should I not be a tard and just up the supply voltage so I don't have this issue?

>>1552020
nothing seems wrong with that.
Unless you're driving that line high for some reason, that LED shouldn't see any reverse voltage at all.
I guess the "right" way to do that would be to stick an NPN transistor, or a FET, to pull the microcontroller pin low.

>>1552020

if your mcu has internal pullups on input lines you can enable it and you won't need the 10k in your circuit.

Anybody willing to give some advice to a fool when it comes to this:
>>1552015

>>1552007
74hc595 is 2 cents each and the resistors are a fraction of a cent each, so I'll go with 595 at the moment but I'll look into things like PT8211

Also, if I add a ceramic cap in line with the output from the R2R ladder I can generate a pure AC signal that can go into audio amplifiers, right?

This is so cool

What other circuit "building blocks" are good to know?

How good one can expect the results from using a lm331 to convert the output of supply to a uC to be?

>>1552024
Nope, won't be driving those pins high at all, they'll be inputs only. Good idea with the transistor switch, didn't think of that

>>1552027
just checked the datasheet, just about every pin has a weak pull up option. Also a good idea. I'll probably go with that as I'm trying to minimize the parts count on this thing

I need a cheap way to turn on a motor on button push, ideally only with a button
can i use this bad boi to switch a squirrelcage motor?
Motor is on the weakside, 0.3kW / 0.9A and the switchs blocks are rated for 2A light inductive load
can i get ~1000 switch cycles out of this before fail?

>>1552096
Your inrush will be quite a bit higher, but it'll probably be fine.
why are you trying to start and run an induction motor with a momentary switch? you're intending that the motor will run as long as you have the button held down?
If you're trying to pulse the motor like that, you'll have those inrush current peaks a lot, and it's not good for the motor either.

>>1552118
the motor gets normally powered by a contactor. It powers an auger and sometimes a piece of wood gets stuck in there.
The Push-button is for a cheap reverse mode
since i can fit 4 switch blocks on its the cheapest way i can come up with

>>1552133
That sounds pretty reasonable actually.
You even got that NC contact as a safety, looks nice.
what's F1 in that diagram?

>>1552148
its a thermal fuse bolted on the contactor

>>1551907
So, I wired the GPS module up to a USB cable and I'm getting this in Putty at 9600 baud which looks okay

>>1552153
>bolted on the contactor
ah so that's why you had to bypass it. that makes sense to me, if the thing overheats after jamming at least you could still reverse it.

>>1552155
I still get gibberish when I try to do it through the arduino at 9600 though.
This is the GPS part of my schematic.

>>1552020
>>1552027
You don't need the 10k either way. The resistor and LED will pull the pin up just fine.

how many of these "FQP30N06L" can I power from a single arduino uno?
I require at least 8

>>1552257
the more the slower

>>1552267
that's a big problem
we talking nanoseconds?

>Arduino
>Nixie tubes

>>1552274
what's wrong with arduinos?
you can do pretty much anything with it and it's simpler to code than other microcontrollers

>>1551656
I really want to get into electronics as a hobby. But it's so boring without an actual goal. You have any ideas what one could do for projects to get into this shit? (Pls nothing useless)

>>1552250
I do though. If I sample the voltage between the resistor and LED, I'll get 3v when the switch is open, and 2v when it's closed. The MCU sees that as high and high. If I sample the point after the LED, I'll get 1v open and gnd closed. Which the MCU sees as undefined and low. I can up the supply voltage to get it to fit in the range of the MCU or I can use internal pull ups. I like the internal pull ups since I need one less battery and no extra components

>>1552285
Kind of in the same boat as you but I have had some hands on experience due to uni.
For my bachelor's thesis I will have to make a rather big project and I have barely any clue about programming microcontrollers.
Maybe you should start with something bigger and work yourself up bit by bit?
I like this page for example, it also has alot of explanations and tutorials.
http://www.talkingelectronics.com/te_interactive_index.html

It's mostly analog electronics but it should help you gain understanding

>>1552285
hey anon. i was in your position for a long time.
you don't learn things like this by making useful things, you have to get through the seemingly pointless stuff first.
for the first year or so of programming I was making useless shit: text and 2d games, simple simulations, etc. but before the 5k+ loc projects, i had to learn by writing boring pointless programs, sorting strings, calculating stuff, implementing data structures, etc.
same thing here. just buy stuff and get started, you'll get to the mentally stimulating projects once you've learned a bit.

>>1552274
Arduinos are fine for prototyping and anytime you're not invested enough to get a custom PCB. Nixies are maximally aesthetic, but ebay nixie clocks are a waste of money. I'd only use nixies for a project like a benchtop DMM or PSU or something like that, using those neat symbol nixies with Ω and V and A symbols.

>>1552286
An ADC isn't going to have the same input impedance as a digital input. Try that and measure with a multimeter.

>>1552285
>new to hobby
>nothing useless
pretty sure my first project was blinking an LED with a PIC. Absolutely useless, but hard enough at the time, and it was satisfying to finish. You won't be able to just jump in and make something useful. Gotta walk before you can run. And if you find those little steps boring I'm not sure what to tell you. I'm still building useless shit too, like right now I'm working on a 16 bit binary to decimal converter that I'll put on a little PCB. All because I think it's kinda cool

But anyway here are some ideas
- voltage regulator PCB that fits in your breadboard
- 4 digit 7 segment display controller. Make it so it takes a number through SPI, so it can communicate with other devices. Useful for debugging when you want to test register and variable values. Pretty hard to start with though

>>1552285
I'm using the basics to cheat at video games because I'm a child

>>1552285
Not him but similar question.
How much did you suck when you got at this? I just finished a very shitty bench supply (for some reason diference amplifier did not work for sensing current, it gave weird readings, had to make some weird shit to current limit) that is ugly as fuck. [But it works, power is clean and nice] This was my first big project and I made 3 fucking vero boards before this because I either shorted something during testing, made such mess of solder that it was easier to start over, etc..

>>1552285
I'm the opposite
I have a project and I learn the absolute bear minimum just to make it function
this stuff is not fun, it's necessary

>>1552224
then use the real serial port

>>1552308
I couldn't even build a digital clock from a kit when I first started (decades ago)
like biology, the odds are you'll never know it all. there are often layers of structure that differ between specialties. just start with the fundamentals and never stop learning

>>1552315
>bear minimum

>>1552323
>never stop learning
truer words were never spoken

>>1552323
>then use the real serial port
That's what I'm planning to do but I'd like to know the thing is working first.
I'm basically making a nixie clock
>>1552274
I know, lel.

I've designed the board and got it populated.
It basically takes a 12V supply, steps it up to the 200 odd volts for the tubes and down to 5V for the logic.
I've got a RTC and the GPS on the board too.
The planning to program it on the Arduino and then move the microcontroller from the Arduino onto my board which has a 16Mhz crystal.

Clock basically has a set button that will interrupt the Arduino and get it to pull the time down from GPS and write it to the RTC.
Otherwise it just reads the time from the RTC, converts it to BCD for hours, minutes, seconds and shifts it out to the tubes.
I have everything else working, I can set the time on the RTC manually and then have the Arduino drive the tubes from it but the GPS is fucking me over

Welder anon again, it's taking shape. I'll need to print some plastic part on top of the cap bank to hold everything together and isolated.
[/blog]

just dropped £1.79 on the components I need
huge stakes, hope i got this right

>>1552274
But what about VFDs and arduino?

>>1552428
Everything is onions: arduino, stm32, esp32, pic, etc. as are nixies, VFDs, LEDs, incandescents, and electroluminescent displays.

You are literally an onions boy if you use any of the above. Only analog and RF is not onions but only analog as far back as transistors because tubes are super-onions. Also if you like analog and RF solely for the purposes of being a hipster then you're onions too.

>>1552438
>don`t do anything
>t. /do it yourself/

>>1552415
For your next photo do it with really good lighting, I want to see your soldering details.

>>1552034
You should have the tip tinned at all times. It's both a sacrificial layer and a way to improve heat transfer. At high temperatures oxygen does very bad things to the metals of the tip, so better a coating of disposable solder suffer that. The solder layer tinning the tip is also liquid at temperature, so it will deform when you touch it to the joint and improve thermal contact. You don't want too much though, or you'll be pressing a blob of solder against the joint without any flux and there'll be no room for flux and fresh solder to touch the joint.

>>1552034
You shouldn't ever need tip cleaner if you've been diligent about tinning your tips, but if your tip is fucked the cleaner is basically an acid bath that will try to eat away the outer layer to expose fresh metal which should take solder again.

>>1552415
>ardweenie rotary encoder
I'm amused that these things are so common that it's actually cheaper and easier to source the module with PCB than it is to buy just an encoder with the same specs by itself.

>>1552425
i ordered three custom boards for $1.85 incl. s/h an hour ago. i honestly feel bad for whoever has to package and mail it.

>>1552464
I just soldered the MOSFETs to the other copper rail, filed the tips of the wire a bit to increase the contact area and use M3 screws. I clipped the drain pins as I just used the MOSFET surface as drain terminal, and I will solder the gates when my circuit is done. Do I need to put a small resistor (say 22 ohm) between each of these gates and the driver's output? These resistor's function is to protect MOSFET input capacitance from shortcircuiting while it's charging?

>>1552478
It just werks(TM) and I don't feel like soldering push buttons.

>>1552491
>Do I need to put a small resistor (say 22 ohm) between each of these gates and the driver's output?
No if you're using mosfet drivers.

>>1552491
Sure you shouldn't be using IGBTs, or would they switch too slowly?

Wiring ready. Now for the part of designing the circuit.

>>1552506
The idea crossed my mind when I already had the MOSFETs, so I went with that. I would prefer IGBTs if I made this project again.

>>1552494
Thanks for the info.

>>1552521
How much did that copper cost you anyhow?

>>1552524
15 dollars in local currency, bought 2.4 kg of copper rails from a junk store.

>>1552469
>>1552471
I've been soldering for a while and this is doesn't I don't really get. I try to tin the tip before I turn off the iron. Every time I just get a big glob of solder somewhere on the iron, not a nice, even coating. Is that what's supposed to happen? Do I need more solder? Less?

>>1552551
If it's not coating the tip then your tip is too oxidized. There's always going to be some blobbing but that's what your sponge or brass afro are for.

psa: nichrome wire is a lot of fun to play with

>>1552438
>years down the line
>b-but why doesn't anyone wanna diy electronics anymore?

>>1551656
Is there such thing as a cheap oscilloscope? Would anyone reccomend one for someone just getting into electronics?

>>1552571
>Is there such thing as a cheap oscilloscope?
Cheapest are PC-connected USB scopes, like the hantek. Cost < $100
Never used one so can't comment, but it's certainly better than nothing.
A scope for electronics is like a microscope for microbiology, it's pretty much essential.
I use a slightly-broken tek465b from the 70's that I got for $200.

>>1552571
get a $50 usb scope, it'll suffice
a $300 scope really benefits because of the tactile controls and the fact that its a standalone unit

>>1552571
A soundcard scope + the line-in of your phone runs a few bucks or less depending on whether you build it yourself, but they're utter shit.

>>1552469
This >>1552551

I understand the heat transfer part, like if I want to solder a switch without melting the plastic body, I’ll get solder on the tip first so it heats the joint quick, but I have never been able to get it to stay coated, it balls up every single time. That was my issue. But it could be that the tip is fucked, I have been using the $3 HF iron because I wanted to learn the basic shit exactly like this before I destroy the $25 Weller that I have yet to plug in.

>>1552584
If you get some weak acid or even just flux on its own you might be able to clean the corrosion off your tip.

>>1552587
It doesn’t even matter, I’m just going to throw the thing out. I need to grab an actual station but I can’t decide on a $40 Chinese station, a $90 Hakko, or drop $70-$100 on one of the higher rated Chinese soldering + hot air stations.

I think I’m gonna end up with the soldering + hot air station but with Xmas just past and then V-Day this week and girl’s Bday in a few more weeks, I have been holding off on buying shit for myself.

I did find a good way to go about it though, I ordered a resistor kit the other day and I got some Mickey cookie cutters for her at the same time so I didn’t feel as guilty.

>>1552600
>pic

>>1552603
Probably belongs in /sqt/ but...

These shitty free blue HF lights, they have the flashlight mode with 3x little LED’s and then the area light LED strip. You goys always mention putting a current limiting resistor inline or you’ll burn up the lights, right?

The earlier ones had two resistors, one for the smaller light and one for the brighter light. In the everlong quest to make things cheaper I assume, they’re now only using one resistor on the dim side. Also check it out, now they only use one screw to hold the switch in place instead of the two.

So my question, did they remove the other resistor because they don’t give a fuck if you burn out the strip within an hour of use, or are the shittiest possible Chinesium batteries the current limiter because they can’t even put out the juice to burn it up?

>>1552600
Replace your tip, that one's oxidised to shit. If you're in the market for a new iron, I'd personally recommend the chinese T12 station.

>>1552634
The latter, though technically they're relying on the equivalent series resistance of the battery. You're like a bootleg bigclive.

>>1552634
cheap consumer LED products usually depend on battery internal resistance and LED forward voltage to avoid LED failure. also, it's a freebie, who gives a fuck

>>1552274
>wasting time
>thinking i'm cool because i'm hating on popular things

>>1552648
I’m not gonna replace the tip on that $3 POS. If it ceases to work, I’ll start using that weller right next to it in the pic.

I think I’m going to grab pic related within the next couple months. There are lots of similar stations from $60-$100 and some anon has this one and says it has been solid for him.

>>1552651
Ok, I figured it had to be the incredibly shitty batteries those things come with. But if somebody were to actually use those lights and replace the dead shit cells with some good Energizers, would it fry the LED’s real quick?

And idgaf about those lights, I stop by HF for $10’worth of zip ties and glue and grab a free one and I have bits of half a dozen I took apart and another 3 or 4 brand new ones hanging on my toolbox to be sacraficed or fuxed with. It’s funny to watch that company try and cut half a penny here and there over time and you can see where they switched suppliers on the same items.

>>1552662
Those dirty bois really did cut back on these free lights. Look at the 3 damn screws that hold the case together, the newer ones are way smaller.

The little LED strip on the new cheap one has way less individual spots as well, I’m assuming they’re dimmer? Would that make a difference between needing the resistor in one and not needing it in the other? Or they don’t give a fuck one bit because nobody pays for the things anyway.

>tfw just tried to power both LED strips at the same time off one 3xAAA pack and they won’t turn on
Hmm so I guess alkaline batteries really are trash at putting out any current. I’m gonna hook up an 18650 to the one with no resistor and see if it burns out before it melts the plastic reflector.

Do schmitt triggers with a positive lower threshold voltage exist?

>>1552679
Yes. If you're looking at the diagrams of building one from an opamp/comparator, swap the ground with the midpoint voltage you want.

So I use one of these for work. what's the deal with the class II or double insulated test? I've never seen it fail anything (even though when I plug the alliance in after it might spark or trip the power).

does it really not fail anything? the class ii button even passes thin air when nothing is plugged into the machine.

is there even any point in plugging the appliance in or doing the machine test for class ii items? seems like a waste of time since right know the result it will give me. may as well just throw a tag onto it

>>1552685
Oh it's called an asymmetrical schmitt trigger.
Your solution is alot simpler than the ones I found. They were using 3 resistors.
How exactly would it work? Does the voltage on the ground simply pull up only the lower threshold voltage?

>>1552671
>>1552662
you should buy a book instead and stop shitposting. From what you tell, you recently got a grip on ohms law.

anyone ever their own two- or three-chip linux computer onto a board? what did you use?
the OSD335x is a bit overpowered and overpriced for my application. I only need about a single 133MHz+ ARM core and 32MB, maybe 64MB of RAM. mainline kernel support would be really great but not a deal breaker

What was the name of that chink cordless soldering iron that cost around ~$60 and was shilled here back in the day?

I think it had a display for temperature and usb charging. Was it a meme? Can't find it for the life of me

>>1552714
TS-100 or TS-80 (newer)

>>1552687
aren't you supposed to probe exposed portions of the device when testing double-insulation?

>>1552716
>TS-80
Looks fucking sweet, thanks

But what's the catch, why am I only seeing recommendations for weller and hakko cordless irons when i search for it?

>>1552720
yeah you do that for all devices class 1 and 2.

it's just that the class 2 test button seems to pass anything and everything. it's like there isnt even any point in really doing it.

>>1552716
Don't fall for this fishing rod meme. T12 handles are much shorter from the tip to the gripping section, which gives you more control and precision, and T12 cartridges are cheaper and available in wider range of shapes.

>>1552721
Oh it's not really cordless, got it

>>1552732
The only really cordless soldering iron would be butane powered.

>>1551912

>>1552491
>File: 20190210_010209[1].jpg
>doesn't measure anything before drilling/screwing
activates my autism

>>1552752
>The only really cordless soldering iron would be butane powered.
You must be counting the cord to the charger as part of the iron.

The reason I even looked online was because I used to have a Weller that used two Sub-C NiCad batteries.

How do you guys even make the small mechanical parts of your projects? I have a bunch of small salvaged dc motors from cd drives and shavers, I was planning on making a BEAM robot with them, but I realised I have no clue on how to make and attach axles, wheels, gears and cogs.
>only common tools
>hand drill
>files
>saws
>apartmentfag
>no 3d printer
>no lathe

>>1552792
And if it matters I'm a ape when it comes to knowledge of materials.

>>1552792
>How do you guys even make the small mechanical parts of your projects?
i'm a meche and i have no idea. the golden rule of mechanical design though is to set your requirements so you can create simple, coarse, conservative components. that means not trying to make things marginally compact, precise, or fast. if you can do that then you can just invest in a shitty printer and some generic chinese bearings/acme rods/shaft couplings/flanged NEMA motors. "design for manufacturing" is key, and the salvaged parts you have are designed to interface with assemblies that use a very different kind of manufacturing than the one you have access to.

>>1552798
I feel like a caveman trying to make an axe living in a city that has little selection of building materials and parts, being a poor EE student that has only a bycicle.

At this moment I'm trying to figure out what I can use to couple to the shafts of the motors, I tried the inner contacts of a screw terminal, but they are asymetrical and wobble a lot.

Literally a poor version of clickspring.

>>1552807
shaft couplings are $1 each or less on ebay. pic related is the standard, but there are others that have rubber grommets for damping and allow for more axial misalignment. if you can't afford ebay then give up now.

>>1552695
The original ground was just setting a 0V threshold, apply a different voltage, get a different threshold. You're probably gonna need 2 resistors to get the threshold voltage, so the three resistor total solution is one less part.

>>1552762
These are really shit, no temperature control.

>>1552811
Uhmm. Nice

>>1552807
>At this moment I'm trying to figure out what I can use to couple to the shafts of the motors, I tried the inner contacts of a screw terminal, but they are asymetrical and wobble a lot.
My dude, you ought to read "Mechanisms And Mechanical Devices Sourcebook" (jewgle the pdf)

Uni gave us microbits to program. Decided to make a gaming system out of it. Currently working on pong and some dodging game

>>1552838
And another video showing the menu screen. If you happen to have a microbit, the code can be found on: https://github.com/indjev99/Microbit-Game-Suite

I'd like to learn and practice BGA soldering, but I don't really know of any cheap way to practice it. I thought about buying cheap orange Pi Zeros and spare CPUs but I've read they're not interchangeable. Is that true?

Are there any cheap circuits on sale with BGA components? I need something that I can test after replacing. Like being able to turn it on or something. Old broken circuits won't help me find out if I soldered it right.

I want to make a rpm sensor for motors with this photomicrosensor and a arduino but I cant find its datasheet and also dont know the meaning of each pin.

Can you guys help me?

OMROM EE-SX416 made in japan

Would the lm331 pic monstruosity work? I swear to god I can't get a decent measurement using the opamps I have here in the diference amplifier configuration

>>1552909

>>1552911
I literally googled the name you gave me.
One pin is supply, one ground and the other output.
https://omronfs.omron.com/en_US/ecb/products/pdf/en-ee_sx4009_p1.pdf

>>1552909
This got me thinking, I have been putting off that Arduino crap until I find a purpose. Tach looks cool, otherwise possibly pic related, a LUX meter. Y’all niggas think a nigga like me could pull those off?

http://embedded-lab.com/blog/building-a-simple-digital-light-meter-using-arduino-and-bh1750fvi-sensor/

>>1552922
As far as I can tell you don’t actually do anything except post tool pics. Maybe you’ll eventually get it after enough swearing, fucking up and getting help from other people.

I'm trying to pick up some cheap LCDs off of alibaba to connect to a raspberry pi. Would I need anything other than a female FPC connector and a breakout to connect to the IO pins? They always have a lcd driver integrated into the panel.

Pls

>>1552910
more people would help if you had a real schematic

>>1552224
Is your Arduino serial port configured for 9600, no parity bit, 8 data bits, and 1 stop bit?

>>1552942
Oh for sure it will be much trial and error and take me 6 weeks to complete, but it may work. I still never properly fixed this light but I think I did recently have a breakthrough when I started to understand what resistors do, but my Ohm’s Law math never seems to add up, or at least the math gives me different numbers than the meters, but I know I’m doing something retarded that I will eventually figure out.

>>1552942
H8ers keep h8in

>>1552961
I'm a dirty phoneposter away from home and computer, I wanted to have a idea of what to do when I got there so that's is why I'm asking for help
>>1552980
>>1552981
Is your hobby just buying stuff? I mean, ohms law is just a linear relation. You can't go any more basic than that.

>>1552997
I have 10 or 20 more things I want to grab off Amazon and then maybe I will actually build something. Costs money to make money.

i see a lot of 5v regulators on digikey with a 6.5v maximum input voltage. what's the practical use of these? a single lithium is below 5v and two in series are over 8v.

>>1553009
>i see a lot of 5v regulators on digikey with a 6.5v maximum input voltage.

a lot of regulators that essentially need a regulated input voltage? got a part number for example?

>>1553024
tc1054 for example. it's the cheapest 5v reg on digi so i saw it and the rest of its series when i sorted by price. there's a fair few others that take up the first page of results along with 30v regs with monstrous quiescent currents.

>>1553026

ah, battery regulators; maybe that's relevant, as well as the output voltages you can get from that device: VR
is the regulator output voltage setting. For example: VR = 1.8V, 2.5V, 2.7V, 2.85V, 3.0V, 3.3V, 3.6V, 4.0V, 5.0V.

It doesn't look so bad if you're powering a 3 volt or 1.8 volt circuit, but yeah, I see your point.

since maplins is gone i've discovered an actual components superstore not far away
thank god because maplins was shite

How does a blown NPN transitor act?

I'm building a step-up converter using and it doesn't fuckin work when i ported it to a perfboard, works fine on protoboard

short circuit testing with a multimeter i found that my collector and emitter appear shorted, while that's not the case when checking the transistor on the protoboard

Is this how a blown transistor acts? I don't want to dig up my shit again if possible

>>1553044
Just an update, seems the transistor is fine, i cut its emitter leg off and the part attached to the transistor doesn't beep, while the one below beeps

back to debugging it seems

>>1552917
Thanks a lot anon. Excuse me for being an idiot.

>>1553032
>an actual components superstore
just buy them online?
rs/farnell/rapid is basically free delivery if not over a small amount.

got some hacker maker african javascript shit in the mail
toot what kind of connectors are these

>>1553094
hmm...

R8 my driver circuit

>>1553124
I having trouble seeing a way of making the pnp dude consume less power. The npn above wastes 1nW while the pnp 1mW.

>>1552885
You could try swapping CPUs or RAM between Pi Zeros.

any one know of a decent 5v boost board? Iv been through a heap of cheap ebay ones but none of them seem to be able to output anywhere near 500ma. My requirements are fairly simple:
-low input voltage (.5v - 5v)
-atleast 500ma output, preferably upto 2a continuous

pic related is the worst one yet. breaks down at only 100ma output!

>>1553083
because fuck waiting
also I can see things before I buy them, which is probably safer

>>1553138
>0.5v input voltage
>5v 2A output.
you want a usb charger that draws 20A wow i wonder why you are struggling...

>>1553124
>not just using a PFET and an NFET

>>1553094
just solder it to some jumpers like a man

>>1552885
>BGA soldering
You really can't. You need specialized equipment and training, and there's a high risk of failure in which case you may have ruined the part and/or the PCB, and even if you haven't you have to recover the part safely and re-ball it, clean the board, and try again. You can only do that so many times before something gets wrecked permanently. BGAs are out of reach of hobbyists.

>>1553151
obviously not expecting 10w out from .5v in....
I test all the boost converters i have using a bench supply.
When i said low input voltage i meant between .5v and 5v

>>1553138
>>1553151
also 2.5w is only 2a @ 1.2v

>>1552960
the arduino thread might know better about them. iirc MIPI DSI compatibility does half the work for you but you might still need to construct initialization strings for the on-display controller. lame fags like myself stick to TFT or an MPU bus interface

>>1553009
presumably in case you're running a 5.5V voltage bus (because you can) and need to do very tight point-of-load regulation

>>1553094
wild-ass guess: JST SH series? measure the connector pitch if you really want to know

>>1553171
not OP but how about a slightly less ambitious goal of correctly attaching a BGA device to a board with > 95% success rate?

>>1553182
that's "energy harvesting" territory. maybe that search term will help

>>1553183
And you're planning to keep your total impedance under half an ohm how exactly?

Brainlet here, Ok so I have some equipment that expects battery voltage from 4.25ish at 100% and it shuts off at 3.5v.
I'm trying to replace the battery with a larger capacity one but6 it seems to work differently.
I just discharge tested the cell and it seems to be full at about the same 4.2ish volts, but at 3.5v, its only spent about 1700mah. Going all the way to 2.96v squeezes out the full 3200mah rating.
So the equipment is basically ignoring almost half the capacity of the cell.
How can I make a controller that raises the battery voltage as it discharges so it basically works within that spec?

>>1553212
>>1553138
ask this guy you want a booster! boost circuit, voltage pump, 'joule theif' to suck out the low voltage and pump out lovely slightly less low voltage.
how it works is it draws out the low voltage but more current. because physics and laws of thermodynamics and free energy or something..

>>1553195
>not OP but how about a slightly less ambitious goal of correctly attaching a BGA device to a board with > 95% success rate?
You still need the correct equipment, which is not cheap. You need the training to use the equipment. You need to heat the PCB with the equipment hot enough, but not too hot or you damage the PCB. Even if you get it all correct there's no guarantee that all the solder balls melt and make proper contact with the pads and adhere, and the only way to tell when your board doesn't work is with an even more expensive x-ray machine, and even that machine isn't 100% on determining if the part was installed correctly. Because of the nature of the BGA package and the PCBs you install them on (which are all always multi-layer) you cannot 'beep out' your failed PCB to determine if all the balls are contacting, or even really if there is a short between adjacent balls. As previously stated if you fail removing the part again may destroy it or the PCB or both, then you have to re-ball it (which you may not be able to do adequately in your hobby setup), and then attempt to install it again, which may destroy the part, the PCB, or both, and if you damage the part from too much heat you'll never know if your PCB failed again because the part failed or because, again, it wasn't installed correctly. This is why having so many parts available as BGA only has kind of ruined electronics as a hobby at least for high-end projects; you have to get an assembly house to do it for you, that's not cheap as a one-off, and there's still a chance your board won't work right.<br><br>

In development they use conductive elastomer sockets and retention hardware to get around soldering BGAs to test PCBs, but those are all 100% custom parts, very expensive ($4 digits), not accessible to hobbyists because of the expense.

So you see using BGAs in hobby projects is not really practical or even really possible.

>>1553212
If you have to go all the way down to 3.0V to get the full rating you probably got scammed on the ratings. 3.5 is a bit high, but the cells will last a while if only discharged to 3.5, so I'd run with it.

>using a joule thief to step up voltages to drive N-channel mosfets
Really got my almonds this one

>Drive speakers with RCA
>3.5 hanging loose
>Produces audible static unless plugged in.
How do I fix it without just plugging it into something? If I do plug it in it messes with my drivers.

>>1553451
Put another 3.5 jack somewhere in your car that's connected to ground.

>>1553459
Not car. Desktop.
The RCA is from a focusrite soundcard to the speakers.
Did you take drivers like that?
What I meant was my alsa audio drivers, they're a bit fucky wucky like that.
Also I don't want to just connect it to something else.

>>1553462
Oh shit I thought I was on /o/ for some reason

>>1553132

I've read that doesn't work because they're not just CPUs but SoC. I have no idea about it but for what I've heard SoCs aren't swappable.

>>1553171
I got a cheap desoldering station with heat gun and preheating plates. It's shit but I've seen russians in youtube succesfully do it on consoles. I know no matter how good I try with such tools I will get a working half assed thing at best, that's why I'm looking for cheap BGA boards to train, just for fun and to get some basic experience.

Thank you both for the replies.

Probably a stupid question, but how hard is it to make pic related? It's supposed to be a 6v DC gearmotor that's powered by a (2 cell?) LiPo or Li-Ion battery, which is charged via USB. I don't know a lot about electronics and might hire a freelancer to make the pcb. How long would it take someone who knows what he's doing to make something like this?

>>1553576

I have no idea what I'm looking at but there are all kinds of charging modules for sale for lithium batteries. If you use two in series (or two in series with more pairs in parallel) you need to find a 2S module.

Pic related. You just have to google "li ion charger module xS" where x is how many cells you want in series.

>>1553576
Test with modular design, nowadays you can test a prototype just by tying chinkshit modules together and then trying to reduce them into a PCB if you know what each one of those do.
>I don't know a lot about electronics
I would hire a freelancer if you want to launch it as a commercial product and you want it for yesterday though. There may be some current considerations you learn by experience or by trial and error.

Do most headphones have an amp in them?
I assume they do because all the pairs I have had include a volume adjuster, but could it just be a potentiometer in series with the speaker?
I don't know enough about how that works

>>1553633
An amp needs a power source. From an audio plug there is nowhere you can grab that power, so they must use only passive components.
If they are USB powered that's a completely different story.

>>1553559
SoCs are generally swappable between same board models, there's not much specific to an individual board unless there's encryption/other security matching a ROM or something else on the board.

Which chinkshit hot air station should I get for soldering some D2PAK's and other smd parts in general?

>>1553633
>I assume they do because all the pairs I have had include a volume adjuster, but could it just be a potentiometer in series with the speaker?
It's just a potentiometer

>some dude with a werid name makes the falstad circuit sim
>super usable, allows to see power and current flows to gain intrinsic knowledge
>barely has UI
>richer than god companies can't make a decent FPGA toolset or a clean,fast,decent circuit sim
why?

>>1553655
Pretty happy with my 858D.

>>1553675
There's good money in making your tools kinda hard to use so your users don't want to learn again to switch to the competitor's tools.

>>1553559
>that filename

>>1553655
Where do you guys buy these Yihuas from?

>>1553655

>>1551785
>>1551784
specifically you referenced pc tho and it's a good point to mention that (at least I think) most pc's or at least the USB port is earth referenced, which can fuck certain things up. I've fried a couple expensive as in-amps like this

>>1552428
i love vfds but I've never got around to using one

How the fuck do I fix ringing and still have fast rise times? (In mosfet bridges)

>>1553124
I'm no expert but i remembered this current boosted charge pump schematic and maybe it will give you idea

>>1553684
bought mine from a local china shit importer on trademe.

>>1553720
reduce your inductance, or active clamp.

>>1553767
Was it an 8786? Can't find any other Yihua rework stations on Trademe at the moment. There's also a Saike 852, which looks a bit better than the 8786.

>>1553774
The 858D, already had a 936 iron bought from Hobby King, have since replaced it with a generic T12 station.

>>1553675
The falstad sim is okay if you're just beginning in electronics but once you get into designing and characterizing more complex circuits it falls way behind proper simulators like Pspice or hell even Multisim or LTspice. You speak a bit too highly of it.

>>1553778
plus it's a visual nightmare

do I need an N channel or a P channel?
I want to switch very low loads as fast as possible

>>1553862
n-channels are better for the same price because of physics and market demand. how fast is "as fast as possible"? because the short answer is always nch but the long answer includes annoying things like gate drivers and kelvin sources.

Is this the correct output for a shift generator? It's a bunch of D flip flops connected together. Data is read when Phi_1 is high and transfers data when Phi_2 goes high.

>>1553863
i guess the technical limit that I would want is 16.6ms

>>1553865
you can just drive an n-channel off a microcontroller gpio then. beware of threshold voltages when picking mosfets. some mosfets need ~12v to get the lowest on resistance whereas others are happy being driven off logic voltages. if you don't have logic level mosfets then you can just add an npn/pnp buffer to your gpio.

>>1553867
i need 7 or 8 of them
can a single arduino do that?

>>1553868
add a 150 ohm resistor between the gpio and the mosfet gates (tied together), and a 100k between the gates and sources.

driving a mosfet gate is driving a capacitor, so technically whenever you change state you're shorting your io pin. the pins don't generally care but at higher capacitances (10-100nF here) i would imagine the instantaneous heating in the io might damage it. the resistor will limit you to under 40mA which is the absolute max gpio current rating. the resistor increases the turn-on time but it's still only going to take ~10us to turn the gates on which is less than a percent of your switching frequency. fets dissipate power when turning on, on the order of 1/4*V*I*t, which you can dig into if you want, but you should be fine.

the 100k is to hold the fets off while your arduino initializes.

>>1553278
The cells are rated to go down to 2.5min without damage. I cant run with it at only 1700mah, I dont care about the life of the cells, I just need the capacity in the compact form factor.

>>1553876
thanks, I'm cataloguing everything from the generously helpful anons here, so if anything goes wrong it's my own fault
another thing, and this is completely superfluous, i want to add LEDs in series because this thing I'm designing is a visual spectacle as well as a functional robot. Will that eat up too much voltage to trigger the FETs?
but as I say, it's not essential

>>1553921
>Will that eat up too much voltage to trigger the FETs?
the easiest way to check is to see what gate voltage the fet's datasheet uses to specify its on resistance. as long as you have that after the diode drop you're fine.

the fet gates don't draw any dc current so of course you'll need to make your gate-source pulldown resistor smaller to draw current through the leds, but that will lower your gate voltage further by forming a divider with the gate current limiting resistor. do your math.

also you're obviously switching a lot of power so when you do the math for fet conduction losses you should multiply the fet resistance by ~1.6 to account for resistance increasing at temperature. the datasheet will have the specifics in a chart.

>>1553855
up the voltage

>>1553864
looks about right since your phi clocks are divided by 5. you might be running into setup/hold time violations though

>>1553921
as long as your input supply is enough you should be fine. you might also consider one of the thousands of boost or buck-boost converters specialized for LED driving e.g. MP3202 with enable pins that you can directly drive with MCU I/O pins

>>1553864
I meant shift register, not shift generator

>>1553925
Do you know how I could shift it over even more? I know I can increase the length of the transistors but then it violates the size constraints. This is what I get if I make the length of the nmos 96 um and the width 4.8 um. But it's too big

>>1553931
hmm, then you must need some dead time between phi1 and phi2, a couple of inverters on one of them might do it

>>1553559
>BGA reballing
It's never the balls. Shit's always abused and the IC's internally delaminated. What fixes it isn't the reballing itself but the thermal expansion during the process forcing the layers in the IC back together. Same reason baking sometimes works, but it never lasts.

>>1551677
>>1551666
was rule 4 put in place specifically for this namefag
lmfao

>>1553778
yeah pretty much upboat

>>1552838
uuh wtf is that ruler? im guessing it's inches but is the voltage actual as labelled and gnd'd on the right. if yes that's kinda cool if not 3/10 completely dissapointed

>>1554040
>inches
shamefur

>>1551674
papa bless

>>1552415
Noice, have you seen this one:
http://www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/
https://www.youtube.com/watch?v=4UgBG0oFAwU

>>1551979
Have done this for a cheap video DAC. Use an R2R resistor pack instead of discrete components for better linearity.

I'm looking at "The Circuits and Filters Handbook" of Wai-Kai Chen pedigree.

I see there's two editions for sale on Amazon and Ebay.

The first edition is substantially cheaper, about $40.
The 3rd edition (can't find the second) is $345.

Normally I would be 100% content with just buying the older version, but the thing is the two editions are 15 years apart. That's an eternity in electronics.

Have any of you guys read the books? What would you all recommend I do?

>>1554100
inb4 download the PDF
I am, I just really like having physical copies of these EE books
it makes me look smart when people come over

I never used terminal blocks, how are the terminals linked? pic for reference

How the fuck do I amass a massive collection of basic shit so I can build things without having to wait a week for the post?

Like even wires, for testing stuff, I just build it out of stripped down cat5, but where do I just bulk order a ton of various sized wiring for cheap?

What else should I buy just a thousand of and keep in a bin?

>>1554178
It just happens. And who the fuck need all those wires man? Just keep building stuff and buying like 10 or twenty of the cheap components you need. Also there MUST be a local place that sells basic components.

>>1554174
like the left, each pair is independent

>>1554103
If you want to just look smart, get the cheap edition. None of your guests will ever care that it's the 1st and not 3rd edition.

>>1554178
>How the fuck do I amass a massive collection of basic shit so I can build things without having to wait a week for the post?
ouch, a week? I get my stuff next day from digikey.
I get in bulk: E6 resistors and capacitors and basic bjts like 2n3906, 2n2222, 2n3904. Diodes like 1n4007, and 1n4148 for signals.
Potentiometers. General purpose opamps. Crystals like 32.768khz if you're doing timekeeping, other frequencies if doing communications.
LEDs in a bunch of colors. Pushbuttons. Switches.
If you find a decent logic level mosfet for switching loads, get a bunch. TIP120/121/122 parts are garbage IMO, skip those.
Build a few projects and you'll see what parts show up often. Some parts you only ever use one of.

>I just build it out of stripped down cat5, but where do I just bulk order a ton of various sized wiring for cheap?
ebay/changgood $1 for bag full of various-length wires with crimped hookup pins. I tried this but 10% of them were junk (flaky or no continuinty) so I don't trust that shit.
Solid, tinned wire cut to length is what I use for breadboarding.

>>1553633
>Do most headphones have an amp in them?
If they're USB or wireless headphones then yes, otherwise no
>I assume they do because all the pairs I have had include a volume adjuster, but could it just be a potentiometer in series with the speaker?
That's not where in the circuit you put a volume control, you put it before the amplifier.
>I don't know enough about how that works
Clearly. Do some studying.

Hello, I'm making a thing that needs optocouplers, it's my first time using them. I'm switching it at 1khz to test things but it'll go up to 100khz.
Even without connection to the rest of the circuit the output displays pic related. How do I fix this?
I'm breadboarding it before soldering, maybe it's stray capacitances from the board?

>>1554234
I am retarded. But if I lower the load, it can't pull to ground. Help aa

>>1554234
>>1554236
Post a link to the datasheet for the optoisolator or at least the manufacturers' part number.

>>1554250
I wrote on the first pic, sorry if it was small. PC817
https://www.farnell.com/datasheets/73758.pdf

>>1554183
Thanks!

>>1554234
>>1554236
>>1554254
What is the phototransistor connected to? Need to know it's impedance and capacitance.

You might need a high-impedance buffer between the opto and your input. I'd also reduce the value of the pullup resistor so there's more current available.

Also no such thing as a true squarewave, everything has a rise/fall time, you just have to work around it as best you can.

If nothing else works you might have to find a different opto or isolate your signal in some other way.

Am I right in trying to keep every trace in a PCB sepparated to other traces with a portion of the ground plane?

>>1554234
That's exactly how they're supposed to work, senpai

>how to get a perfectly square wave

I don't know, maybe with a schmidt trigger or with an OPAMP?

>>1554008

By reballing I meant soldering new BGAs.

I hate BGAs as much as you guys but I just want to play with some balls man, blow on those hot lead balls (for cheap!).

>>1554275
>>1554259
>>1554250
I'm retarded x2. I thought I had tried a 1k resistor, but it was a 100 ohm one. Who the fuck had the idea of making blue resistors????? A 1k ohm resistor allowed it to pull down to ground with enough speed. Thanks people.

a negative resistor
just in case..

>>1552323
Okay, I owe this nigger a pint.
I wired the GPS module into the serial port and it worked fine, I got it printing the output onto an LCD.
The module is running at 115200 by default though.

Now I've run into another issue.
I've a button for setting the time, it pulls pin 2 low on the Arduino and has a pull up to 5V.

I've attached an interrupt using that pin to happen on falling and run the function in pic related.
Basically it's meant to write zeros to all 6 tubes, pull the time from the GPS (I know this works as I tested it on the LCD), write this time to the RTC (don't care about the date) and then return to the main loop after 5 seconds.

When I push the button however, all the tubes change to zeros but stay there.
I assume it's stuck in the interrupt.
Any idea what I'm doing wrong?

>>1552323
Okay, I owe this nigger a pint.
I wired the GPS module into the serial port and it worked fine, I got it printing the output onto an LCD.
The module is running at 115200 by default though.

Now I've run into another issue.
I've a button for setting the time, it pulls pin 2 low on the Arduino and has a pull up to 5V.

I've attached an interrupt using that pin to happen on falling and run the function in pic related.
Basically it's meant to write zeros to all 6 tubes, pull the time from the GPS (I know this works as I tested it on the LCD), write this time to the RTC (don't care about the date) and then return to the main loop after 5 seconds.

When I push the button however, all the tubes change to zeros but stay there.
I assume it's stuck in the interrupt.
Any idea what I'm doing wrong?

>>1554317
Are you writing the ISR correctly? Is it saving all the registers then restoring them before exiting? Is the ISR taking too long and some other interrupt occurs? Writing ISRs can be tricky as I recall.

>>1554275
FYI there's no such thing as a 'perfect squarewave', all waveforms have rise and fall times that are finite and measurable, even if they look perfect on a 'scope. You zoom in enough on a fast enough 'scope and you see and can measure the rise and fall times.

>>1554275
>>1554338
Yep. This is because you need infinite bandwidth. Look at the Fourier series.

>>1554317
>>1554336
Never mind.
Apparently doing shit in the actual interrupt is a no no.
Should be setting a flag and immediately returning to the main loop

I posted this previously for other reasons, last question before ordering the parts

What gauge wire is needed for this project?
would having a gauge too large impend its function?

I don't have any wire laying around and i don't just wanna get some random gauge or speakerwire at Homedepo finding out its not right.

>>1554366
Yes, if it's longer than one operation it's too long.

>>1554388
btw its a injection molder temp controller

>>1554388
330W * 4 = 1320W
I = P/V = 1320 / 110 = 12A
check a chart, 16AWG is 16A, no smaller.
you can use a smaller gauge where necessary but no problem using something bigger if the terminals can accommodate it, be as well using the same for everything except the thermocouple will probably have some requirements.
any reason you are switching the (presumably) low side?

Is there any relay with a wide coil voltage (3.3 to 12 VDC) with 250V 10A contacts or will I have to bite the bullet and incorporate a linear regulator?

>>1554394

ACTUALLY that pic is wrong the heating element is 380Watt

The heating elements im looking at go up to 500C which is more than enough using 4 to only reach about 260C max for me.

don't understand why the Original designer insisted on 230V. very confusing.

16AWG should still work right?

>>1554403
SSR module maybe.

>>1554366
>>1554389
Well, as I remember it, the first thing you need to do is disable interrupts, then save all registers, then you can do some things, then restore all registers, enable interrupts, and exit. The problem is however that if something else that's time-critical generates an interrupt in the meantime, the system crashes (or at least gets unstable). On a PC there's an RTC interrupt on a timer so you have to get in and out of the ISR fast; on a microcontroller it's totally dependent on how the code is written. So, yeah, best to set a flag bit somewhere and let the main program deal with the actual processing.

>>1554403
Get a 3.3V one and limit the current? Magnetic stuff really only care about current, the voltage rating is due to ohmic losses and insulation. For 12V the only problem would be burning out the coil.

>>1554284
Hello, same noob here. Can someone help me with the half bridge? How do I stop the bjts going into saturation? It makes switching slower. A baker clamp makes no sense for a bridge

>>1554426
why BJTs?

>>1554428
Do you know any cheap, common tht Jfets or Mosfets? I only know th 2n7000 but it has a max current of 200mA, I don't think it can drive big chonky mosfets.

>>1554430
if you need to drive chonky mosfets buy a mosfet driver https://www.digikey.com/product-detail/en/microchip-technology/MCP1406-E-P/MCP1406-E-P-ND/1228637 for example

>>1554433
and if you want to drive a half bridge, get a half bridge driver.

>>1554275
>right
you're not wrong, but maybe autistic. 99.9% of the time you won't need to do this. also sometimes when dealing with high voltage you'll want them separated by no copper in order to prevent arcing

>>1554412
>don't understand why the Original designer insisted on 230V. very confusing.
because it's a nigger-rig made for third-world niggers to nigger recycled plastic into other plastic? just guessing here

>>1554420
interrupts of the same and lower priority (where applicable) will almost always be disabled upon entry to an ISR
if using C, declarations are required, see http://www.nongnu.org/avr-libc/user-manual/group__avr__interrupts.html

>>1554403
Why do you need such a wide voltage range on that coil? there might be a better way to do what you're trying to do.

>>1554403
use a single-BJT current source

>>1554416
>SSR module

Thanks, that's a good option.

>>1554443
>interrupts of the same and lower priority (where applicable) will almost always be disabled upon entry to an ISR
>if using C, declarations are required, see http://www.nongnu.org/avr-libc/user-manual/group__avr__interrupts.html
Hey, it's only been like 25 years or so since I fucked around with low-level code like an ISR and I'm working off memory. I don't even mess with any of that anymore.

What would a completely noob to electronics need to know in order to make a device that can supply car components with a pulse width modulated signal, with a choice in voltage output between 5V and 14V?

A device like this would be really handy for diagnosing certain faults or testing components, like controlling a throttle body actuator or aircon compressor valves.

>>1554560
Care much about a really stable output frequency or wide duty cycle range? If not you could do it with an adjustable regulator + a 555 oscillator.

>>1554585
Doesn't have to be a particularly great signal, most auto electronics work on pretty rough signals going by what I've scoped. I think a frequency range between 1-1000 hz would be what I need, and of course between 0 to 100% duty cycle.

>>1552521
Isn't it bad practice to have the tabs electrically connected like that my dude?

>>1554588
I'd probably do it with a microcontroller (generic arduino board is fine) + a couple MOSFETs and linear regulator to get your variable voltage.

>>1554589
Nah, it's fine, even better than using the leg where possible. The SMD versions of the same package don't even have a full middle pin most of the time.

>>1554430
>>1554433
>>1554436
Not him but
>/diy/ how I make and learn stuff
>buy it
Really?

>>1554600
driver chip does a better job for cheaper than making a discrete one.

>>1554600
sometimes, in some applications, didactic value and use value are at odds with one another

first /buy it yourself/
/do it yourself/ can wait

>>1554601
>cheaper
>2 ten cents transistors vs a few dollar

>>1554617
what 2 ten cent transistors can switch 6 amps in 10ns

>>1554620
You can switch almost 2 amps with cheap bjts at much more than the frequency the dude asked.

>>1551656
hey, what is the optimal wire gauge for digital signals?

should I use 16 awg solid, 20 awg solid, or 36 awg solid?

thicker wire is better for higher current, but wouldn't thicker wire have increased capacitance or something, for the digital signals?

for example, say I get a dip avr and a dip spi sram to manually log some text into, and i want to run some cable between them and solder it, any cable would work i presume but if I'm going to choose, whats the right one?

>>1554626
Wire gauge only matters for current and heating. Also, what frequency? Anything digital needs to
>be short
>no square turns
>no big loops
>close to inputs/outputs

>>1554629
<20mhz, and 3v

>no square turns
so if you place the parts you want the cable to be like a relaxed curve shape between the two points, cut to length?

>>1554631
With wires it won't be much of a problem, because they don't easily kink. But 90 degree turns in a pcb is a big no because of reflections. Just get the smallest wire you can easily handle and solder that can hold the current you need.

>>1554620
SMD MOSFETs like the AO3400/AO3401, kek

>>1554641
doesn't really matter until tens of MHz or tens of cm

>>1554626
36AWG for the signals, but use a thiccer ground return, or one same-size ground return for each signal line
how long of a cable? if it's less than a foot you probably have nothing to worry about

do IC's need some kind of socket or do they plug directly into breadboards?

What do you vibrators for?
Anyway just snip the connector off and strip the cable, then solder it, done.

Why does Atmel AT24C04 show up as two different devices (0x50 and 0x51) on the I2C bus?

>>1554714
bank selection, kek

>>1554671
DIP ics fit without a socket, and components with 0.1" pin spacing

>>1554671
if DIP they generally plug right in but you might need to help them go in straight. if you don't have a tool you might need to use your thumb and forefinger on the pin sides to guide them down evenly

I do not vibrators.

Can someone explain the differences between an AC variable frequency drive and the electronic seed controllers that you typically find on most RC brushless motors? Also why don't RC planes use VFDs, considering that they seem a lot more efficient?

>>1554731
AC VFDs generate sinusoidal waveforms, varying both voltage and frequency. BLDC drivers typically generate trapezoidal current waveforms (i.e. apply full voltage at the transition resulting in dI/dt=V/L, then PWM to maintain constant current).

dI/dt is always V/L

>>1554771
Why would you use a BLDC ESC over a VFD?

>>1554771
The other main difference is that ESCs are closed-loop (they sense the rotor position via either Hall-effect sensors or back-EMF), while VFDs are open-loop (i.e. you just set a specific voltage and frequency and expect the motor to keep up with it).

>>1554818
"Brushless DC" motors are designed to be driven from a ESC, AC motors are designed to be driven from 3-phase AC. Ultimately, both types are synchronous AC motors (usually 3-phase, although some BLDC motors are 2/4-phase aka quadrature).

>>1554478
>>1554474
>>1554425
>>1554416

Since I'm wary of mixing mains and low voltage (I don't want to inadvertently send 230V to the USB input of my amplifier) I've found pic related that's a better way to trigger my subwoofer when I turn on the amplifier. Thanks for the help though,

>>1554851
My man, you can do it with the relay and a bipolar transistor/fet. Just limit the current passing through the coil, and use the relay with the lowest voltage you can find.

Posting here because it's not mains. How does this type of doorbell work? Someone took it down to paint that wall a couple weeks back and it hasn't worked since. I'm guessing they nudged something out of position. The only thing I can see that could really be messed up is the contacts. Are they both supposed to be touching the contact on the ringer arm when the thing is unpowered? Both not touching? I don't really know where to go from here

>>1554873
the contacts should be touching when at rest
the current through the electromagnets pulls the arm away from it's rest position.
The contacts open as the hammer is moved toward the bell.
The hammer strikes the bell anyway due to inertia.
You have to find the 'sweet spot' so the arm moves to open the contact at the proper time and returns to hit the contact again when resting or between cycles of current.

oh wow they updated ltspice for mac

>>1554879
Thanks. That works but it's arcing like crazy. Is that a problem? I never looked at it in use before so I don't know if that's normal.

How do I learn pcb/circuit design from scratch?

>>1555053
>it's arcing like crazy.
as long as it's a small blue or blue with some yellow arc it's fine
> sparklers coming off of it are not good
you might try adjusting it for best ring with least arc
I noticed in your pic, one arm is bent and one is straight.
You could try 'carefully' adjusting one way or the other for best results.

>>1555060
read a beginner's book to get the fundamental math, and dick around in a simulator. if you're actually into it you'll be driven to research topics you read about further on google. see op for books/sims. actually designing a pcb will be useless to you for a while.

>>1555060
When you do get to the point that your circuit works and you understand it well enough to want to get it etched, find a YouTube tutorial for KiCAD, shit's great. But I imagine you'll be at least a few weeks of that point.

Hey guys I'm zero knowledge about electronics but can do basic soldering, my Seasonic Focus+ PSU died last night. When I tried to open it up I found one resistor that is burned (red,red,gold,gold in pink body) will this be fixed by just replacing the resistor?

>>1555174
one would assume so

>>1555174
also a burned baby cockroach underneath the PSU board around those circled area in pic related. When I clean up with alcohol it looks perfectly find.

>>1555180
>>1555174
Unless that is a fuseable resistor made for that, something else also went to the shitter. Usuatly some PTC fuses or similar stuff. Check continuity from the input to that resistor with your multimeter (diode symbol, it beeps if there is a continuous path)

>diode symbol, it beeps

how stingy are optocouplers?
do they require a certain saturation of light to trigger or will any amount do?

>>1555229
what do you mean "optocouplers"? you'll have to be much more specific

>>1555230
ASSR-1228-002E

>>1555180
now you've done it, you're going to need to install a new baby cockroach

>>1555232
>solid state relay
there are optocouplers that have a very approximate transfer function (current out vs. current in) but an SSR is designed for a crisper action
>The relay turns on (contact closes) with a minimum input current of 3mA through the input LED. The relay turns off (contact opens) with an input voltage of 0.8V or less.
the threshold probably varies by process/voltage/temperature. do what the datasheet says and you'll receive the performance the datasheet specifies

>>1555239
only asking because I'm testing the theory with 5mm leds, and I'm opting for optos because the mechanic relays I'm using currently are too slow
they either break connection during the 0.05 second pulses or don't make it at all
according to the spec sheet these SSRs do better than 5ms in the worst cases, but looking at the LED array I've set out they don't achieve peak luminosity in that window

I guess if they promise it I have to trust them

when making a voltage divider on a breadboard, is the left topology viable?

Or does the wire physically need to be between the two resistors such as on the right?

>>1555242
>luminosity
the LEDs inside optocouplers are usually infrared, therefore lower Vf at current. they wouldn't spec it if they weren't willing to back it up. returns in volume are expensive
t.former factory rat

>>1555251
it's fine. the resistance of the breadboard strip is somewhere around the tens of milliohms. in typical circuits won't even notice it unless you're pulling enough current through that row to melt it

>>1555262
can photoresistors see infrared?
that's new to me, what spectra can't they see?

gamma

>>1555274
they're phototransistors, my dude. photoresistors are twice as slow as shit. anyway most phototransistors are most sensitive to near IR and don't see blue very well at all, Pic related is a very typical graph

>>1555262
What I meant by my picture is that circuit diagrams invariably show the output of a voltage divider being between two resistors (by definition). So, does it necessarily need to be physically between the two resistors on a breadboard, or as long as the two resistors are connected to the same junction on the breadboard it'll operate correctly?

>>1555299
in your first case, the few milliohms of breadboard is not going to show up in your output, down to like 5 digits. in the second case, the few milliohms in series with each of your intended resistors is also not going to show up in your output, down to like 5 digits
you can consider the breadboard row to be equivalent to the junction dot on a schematic for all practical intents and purposes

>>1554641
>90 degree turns in a pcb is a big no because of reflections
That's a meme. It's a no because of acid traps, and holy shit, where did all of the other myths even come from?

>>1551656
I just found out you can use a "student" (just list any name/school you want) license to get full Eagle functionality for three years, just as I did before when I needed to work with Inventor. It's fine as long as it's non-commercial use. Shit, I even used the same account that I remembered the credentials to. No fuss, just a working Eagle without the "free" board size constraints. Seems pretty based of Autodesk.

>>1555334
It's actually a smart business move on their part. Get the incoming workforce comfortable and experienced with your product, and eventually the companies that hire them will use it too

If I want to power a sensor with a 9V battery should I use a normal voltage divider with resistors or rather a zener regulator circuit?
Or even a linear voltage regulator?
Is there a rule of thumb when wanting to power low current devices?

>>1555414
the zener regulator (or even a properly sized divider) will work. the problem is that they will waste a lot of power. you can use a discrete npn regulator to improve on this (pic) but a linear voltage regulator ic (intended for batteries) will be far more efficient. you didn't ask but switching regulators will be less efficient here unless they're a niche type that you can't buy as a cheap module.

>>1555334
>>1555413
Or just use KiCAD for free forever. Eagle is obsolete.

>>1555462
Thanks anon

>>1555488
"npn regulator". there are a lot of varieties. it's just a zener regulator with an emitter follower npn after it. it'll draw 60-100 times less idle current for the same output capability. if your sensor needs a precise voltage you should still use a regulator ic though.

new >>1555499
>>15554>>1555499
>>15554>>1555499
>>1555499
>>1555499
99
99
>>15554>>1555499
>>1555499
>>155>>1555499
>>1555499
>>1555499
5499
99>>1555499

>>1555499

Can you destroy a mosfet if the mosfets drain is attached to a normally open switch and theres a constant pwm signal hitting the gate while the the switch is not turned on? The source is of course grounded.

>>1555484
this

>>1555414
either way you'll be dissipating power in the regulator, which cuts directly into your battery life by a factor of Vin/Vout. short of adding a buck converter, of which there are few efficient low-current options, there's not a whole lot that can be done about that

>>1555653
nope, Vgs limits are independent of the drain voltage

>>1555500
too early
thread will be haunted by duinists