/diy/ - Do It Yourself

this thread's digits brought to you in part by the TUSB8041 SuperSpeed USB 3.0 hub. because we all need more SuperSpeed USB ports amirite

>>1358041
How do I resize Multisim`s UI? This is comically small. Why every engineering software has obtuse and clunky UI`s straight out of the 90s? Only CAD manages to do some half decent shit

>>1358047
nevermind senpai, found the ( ( ( enlarge ) ) ) button

>>1358047
Man I hate it when I dl some software that expects a normal PC screen, but since I've got a 13.3 inch 2560 x 1600 display all the GUIs are tiny. But it's better than them deciding they can't work at that pixel density and only showing me a blocky 1280 x 800, which also sometimes happens.

>>1358041
>OP pic
https://www.youtube.com/watch?v=LkUrHRWCqRg

>>1358146
Why say that?

I'm trying to find parts and cannot for the life of me find them for a reasonable budget.

I need UNpolarized (no polar electrolytics) caps rated to at least 40V. somewhere in the 1400-1800uF range.

And Inductors at the 1600uH rated to 5A.


I have an alternate design I could use, but it are at the same rating, but more parts needed 910uF 560uF 390uF,and 910uH 510uH 390uH.

>>1358175
>1400-1800uF
Not going to be easy, what do you need them for? Chances are there's a better way if going about things. Inductors over 1.6mH isn't too difficult however, nether is that current requirement. But I'd wind them yourself since buying big inductors can be a bit pricy.

>>1358175
What are you using it for? What sort of frequency and current?

Got a diagram?

>>1358175
You're not going to find NP caps at that capacitance that are effective at their rated values. Just stick two identical electrolytics back to back and you have a NP cap with half the capacitance.

>>1358221
>two identical electrolytics back to back
Does this work? Don't you just get half the potential in the middle of the two? Pic related

>>1358223
Two equal capacitors in series act like two equal resistors in parallel, the impedance is halfed.

>>1358221
> stick two identical electrolytics back to back and you have a NP cap
What? No.

>>1358226
the impedance is doubled. the capacitance is halved.

>>1358175
that could be buy-a-roll-of-aluminium-foil-and-saran-wrap territory
you might consider placing identical caps back to back with reverse-shunt diodes across each one, if you can find suitable diodes and you don't care too much about crossover distortion

>>1358223
>>1358231
It works, but electrolytics are very lossy and using them in resonant circuits isn't usually a good idea.
www.cde.com/resources/catalogs/AEappGUIDE.pdf

>>1358175
You want a 100Hz tank? You sure that's the best option?
Also, why so comically small inductor? Henries would be more common at that frequency. That said, a multi-henry coil rated for 5A would be big.

>>1358212
>>1358210
I found out bi polar electrolytic are a thing. I ran it by a college and said they will work.

I had to buy a lot of spares (+/-20% rating REEEE)

>>1358245
no, seriously, what do you need a 100Hz tank for?

>>1358238
Well actually since only one capacitor is being used at any one time, it the effective capacitance would equal the capacitance of each capacitor. Must have horrible linearity though.

>>1358245
Use a bigger inductor and a smaller capacitor you maniac. There's a reason nobody uses such large caps in AC circuits.

>>1358142
I have a 1920x1080 screen

>>1358265
congratulation

Assuming full sunlight, approximately how long would it take a 12v 1.5W solar panel to charge a 4400 mAh battery bank when I have it wired up with a buck converter?

>>1358364
Assuming perfect efficiency and a 3.7v nominal battery bank, 4.4Ah*3.7V=16.28Wh nominal, / 1.5W = 10.85 hours.

With sunlight not being constant through the day, bullshit in the panel rating, not using the ideal voltage, losses in converter + battery bank something like 20 hours would be more reasonable.

>>1358041
i want my indicators to utilize both the indicator bulb and brake light so people can actually see when i'm making a turn. i've sketched up the part of the circuit that handles the delay but i need the part that keeps the light turned on for around a second. ideas?

>>1358245
>>1358246
>>1358253
I'm the guy from last thread who is designing a filter off of a current source at 60Hz that has harmonic spikes messing with the circuit.

I am 99% sure I cant use an active filter at this frequency and current rating. And I have not found an example of people making active filters off of current sources.

I am looking for any high value inductor rated at 5A. No luck. The bipolar electrolytic wont work on main AC ( just found this out).


Are there any pre-made filters for this application. We are coming up on the project deadline and it isn't necessary that I do the circuit design for this part of the circuit.

>>1358457
>I am looking for any high value inductor rated at 5A.
browse the graphs on http://www.fair-rite.com/product/toroids-5978014001/ for example and avail yourself of a few calculators. 40 turns (approx. 4 m) of 10AWG spaced 5mm/turn on that core ~= 1 H (measure to be sure). that core/wire will handle 5 A at 100°C without dissipating significant power. now add a 2.5uF film capacitor and there's your 100Hz LC filter
it'll cost you $200 plus a few cents for the capacitor and a dollar for the wire, but that's just an example. there are many other cores in the 50-100 burger range that could also provide high inductances for which reasonable complementary capacitors e.g. motor run caps could be found
>it isn't necessary that I do the circuit design for this part of the circuit.
did you contact the application engineers at an inductor or ferrite company to get their suggestions? that's literally what they do all day

I have a pair of chokes, about 2-3cm in diameter. They get pretty toasty, how would you cool them if blowing air was not an option?

I was thinking of sticking them on a heat sink, but the lack of good contact is not enough, and the insulation from the wires might be rubbed off and cause shorts.
I have water cooling running around to other devices, i could use that.

>>1358502
forgot to add: toroid chokes

>>1358364
'bout tree fiddy

>>1358502
>pretty toasty
cores are usually in fine shape up to 200°C. if you're really concerned, you should get larger ones

>>1358421
Thanks buddy. Luckily my project won't suck much juice so that'll prob work.

Is anyone willing to help a software-weenie with some hardware?
I've got an old laptop (butterfly thinkpad) that I want to rebuild. The keyboard is the critical part and I'm unsure of what I need to connect it to an arduino to capture key inputs.
It's an FFC but I'm overwhelmed by component sites, exactly what item do I need?
I'm unsure of how to measure it for the required specs, assume total retardation.

>>1358524
All 3 cables. two for the keyboard, one for the trackpoint

>>1358526
standard keyboard should only need 20 pins for matrix, 4 or 5 pins for ps2/usb side.
i would expect thin cable for tracker, either or both fat cables for keyboard. as for how its wired you might need to strip it down, keyboard is typically two sheets, one with horizontal and one with vertical for a matrix with a spacer in between, when you type it pushes together.
so from that probably one fat connector is for the top layer and one is for the bottom i would assume, how it works is like any other matrix you need to apply voltage to each connector on one side sequentially and for each pin scan each pin on the other side to see if any of them are connected to detect a keypress.

in theory a keyboard is what, 104 keys so you can do it with sqrt(104) ~11 pins but you start to run into trouble like that because you cant press multiple buttons simultaneously which is kind of fucking important for a computer. just take it apart and have a look or google a picture, its pretty simple idea really.

>>1358539
Sorry I know roughly how a keyboard works, what I was looking for was what hardware I need to connect the FFC ends to an arduino so I can record the output and write a driver for it. I'm aiming to have that keyboard controlling a small form-factor computer ie a Pi or similar.

>>1358526
>22 data lines
>21 data lines
are these parallel or serial busses?

guys is there a guide on how to build a ferrite based inverter?, im making a proyect with batteries and i want a slimmer inverter, is there other way to slim down a inverter?, i know the biggest part is the transformer

>>1358572
I haven't a clue, it's the keyboard cables for this http://www.thinkwiki.org/wiki/Category:701C

1. What makes a good solder and what makes a bad solder, and is there an easy way to tell?
2. Is tinning an iron necessary or useful, and if so, how often? I've watched some tutorials and some present it as some sort of common sense axiom and others (usually still being aimed at complete noobs) just don't mention it at all

>>1358582
Shit, how much did that cost? I want one of these so bad

>>1358609
It was about 80USD for two I think. I got lucky though

>>1358544
Those cables plug in to what is known as a Zero Insertion Force (ZIF) connector. The important metrics to look for are how many connecting pins it has, how many rows of connections it has, and what the pitch (spacing) is between each pin. It's called zero insertion force because you lay the cable down on part of the connector, then clamp the other side of the connector to it. Try searching digikey, mouser, and arrow for ZIF connectors

>>1358606
Good solder connections are small, smooth, and shiney. Bad connections are big, dull blobs.

At the temperature that a soldering iron tip gets to, you are definitely going to oxidize metal. The question is what metal are you going to ruin? Would you rather ruin the metal that the tip is made of, or would you rather ruin a tiny worthless blob of solder that is protecting the tip? Tin your tip often, and it will last a long time. It also transfers heat better with a little molten solder on it.

I seem to be having solder problems. I'm trying to tin my tip, and the solder just bubbles up and drops off instead of sticking to the tip. I thought it was my old tip, but I bought a brand new tip, and it's behaving the same way. I'm using this weller station and these tips.

station: https://www.amazon.com/Weller-WLC100-40-Watt-Soldering-Station/dp/B000AS28UC/ref=pd_bxgy_469_img_2?_encoding=UTF8&pd_rd_i=B000AS28UC&pd_rd_r=CFXENCMAAGDEGGH3CJYM&pd_rd_w=i7VNL&pd_rd_wg=IDx7O&psc=1&refRID=CFXENCMAAGDEGGH3CJYM


tips: https://www.amazon.com/gp/product/B0747R26S8/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

The tips I bought look the same as the one I had, so I thought they would work better, but I'm not seeing the return to form I expected.

What are some possible causes for these problems?

I'm hesitant to post this question on a cancerous hole like /sci/ or /g/ so might as well give it a shot here amongst fellow EEEs in case any of you happens to be well versed in information theory and shit.

Is there an algorithm for calculating the number of password permutations possible, given a password with an entropy of X bits?

I'm fairly shit at both maths and finding relevant answers on google/stackexchange so I have absolutely no idea whether the answer to this is trivially simple or incredibly complicated or somewhere inbetween...

>>1358656
That's what the given entropy in bits means. 128bits = 2^128 permutations

>>1358457
That is an awful circuit. Firstly, you can't pull that much power out of a current transformer, they're designed to be lossy with a terrible mutual inductance. Secondly, converting this current through a transformer to 120V then dropping that down to 12V is just asking to be losing efficiency. I'd just rectify the voltage from the CT and feed it straight into a DC-DC converter/battery charger module. You should be able to find something appropriate on ebay or ali. I'd also put an over-voltage protection circuit of sorts after the rectifier, a FET that shorts the rails if the voltage gets too high, like a MOV but better at sustaining a high current. A MOV across the input might also be a good idea, just in case. With a DC-DC converter it doesn't matter what harmonics you have, since the rectified voltage will be converted into the correct DC level regardless of amplitude (to some extent).

Wanting to use your existing battery charger could be problematic, but chances are it's got a switched-mode PSU in there that can run off a variety of DC voltages anyway (since it will just have a rectifier before it), so I'd do some testing to see if it will work off the voltages you'll be getting. Chances are the harmonics are getting such high voltages is because you're putting them through a transformer, where reactance is higher the greater the frequency, so the voltage at higher frequencies has to be higher to keep the same current. Using a CT really isn't a good idea if you can avoid it.

>>1358654
Too hot or not enough flux. If the tips came with a coating of something to stop them corroding in storage you will want to wipe this off, but don't file or sand it or something stupid like that.

>>1358662
Thanks, I figured that was a likely canditate but I've only really had practical experience of using informaton theory for basic compression so I wasn't really confident about the chances of anything I assumed being remotely true. Entropy makes intuitive sense but the maths can get a bit fuzzy to me sometimes.

>>1358662
>128bits = 2^128 permutations
I just read up on IPv4 and PBv6, the former having 4 billion permutations and having to be replaced because we were running out of addresses, but being replaced with something with 340E36 permutations. Isn't 64bit easily enough for the entire future of the internet?

>>1358580
>guys is there a guide on how to build a ferrite based inverter?, im making a proyect with batteries and i want a slimmer inverter, is there other way to slim down a inverter?, i know the biggest part is the transformer
>no one?

Guys, i have a shitton of dc motors, if i join them in series, do i generate twice of the voltage marked on the dc motor?

>>1358580
inverter is a generic term for a dc-to-ac converter. go web search what you're trying to do in particular (EL lighting, CCFL, mains from a battery, etc) and it's possible you'll find something that does what you call for, maybe even with stock components

>>1358606
I roll my tips around in the tip tinner the first time I fire them up. thereafter I make sure to keep a bit of solder on them including when shut down, which I wipe off in a ball of brass turnings just before I solder a joint
>>1358667
you need flux for that, my dude. or Pic related

>>1358683
>enough
sure, but bits are cheaper than the trillions of burgerbux it would take to deploy yet another version of IP. 128 bits is also enough to allow addresses to conform more closely to actual network topology and have enough bits left over for devices to generate their own local addresses and statelessly auto-configure, and other things we may not have even invented yet

>>1358694
take it to the recycler

>>1358667
>>1358704
I have rosin flux, the solder is also rosin core. I have tried a few different solders, they are all cored though.

I am trying lower temps, it's weird because everything was fine with my station set to 4 (out of 5) heat on my last tip before it acted up. Once it started acting up on the same settings, I assumed it was the tip and replaced it. On the replacement, I'm trying with and without flux, and on a 4 it doesn't melt at all (does smoke some) on the tip, but it turns to drops instantly when touched to a spot further from the tip. On a 3 I still get no melting on the tip but some melting (still the drops) at spots further from the tip. At 2, I get no melting anywhere.

I keep trying different temperatures, but the behavior doesn't seem to improve. It's either not melting at all, or instantly turning to molten drops and falling off.

>>1358683
2^64 is relevant to IPV6 becaus that's the number of unique IP addresses available in IPV6, i.e. for the forseeable future every device gets its own unique IP address. IPV4 hasn't had enough to go around for a while so workarounds like network address translation and shit have been necessary.

Bit length for a password, as the guy who you replied to pointed out, determines the number of different possible passwords an attacker must guess in order to brute force your password. In that context bit means bits of entropy, i.e. the summation of the probability that any one symbol may occur multiplied by log base 2 of the reciprocal of the probability of that symbol.

I think the recommendation is that 80 bits of entropy should be sufficient unless you've got the NSA on your ass or something.

Each character in your password corresponds to roughly 6 bits of entropy as long as your password is pseudorandon and you make sure to select from alphanumeric and special characters.

I think that's mostly correct anyway but I've not slept in 2 days so don't take all that as gospel

>>1358708
That is odd. Perhaps the tip states "for lead-free" and you're using 60/40 and that's somehow causing a reaction with the tip plating, but I very much doubt it. Is the rosin smoking much?

>>1358716
yeah, the rosin smokes for a while after applying it to the tip.

>>1358719
Well then the tip probably is getting up to temp. It says "Tinned with lead free alloy" on the page for the tips, are you using 60/40? I doubt it will make a difference but you can give lead-free it a shot. Do you have another iron you can test the solder with? Might be a dirty batch.

Personally I'm swapping to lead-free after noticing that it really isn't that much harder to solder with, it just requires a slightly higher temp.

>>1358704
but still tho, if i join them can i get more voltage?

>>1358723
No, I'm using lead-free, and this is my only iron.

There isn't an environmental factor that could cause this behavior, is there? Like if it was too wet or too dry? I don't know if that kind of thing exists, I'm just throwing it out there since it has been raining here lately, and I tend to leave my windows open to let the cool air in.

>>1358710
ah fuck wait ipv6 has 2^128 bits
>>1358683
ignore my reply my sleep deprived brain was answering a different question of its own

>>1358041
What is a book? Please explain TIA

>>1358727
the things you pirate for college

>>1358730


>>1358730
I cant swim

>>1358724
you have a shitload, why don't you try it before hauling them to the recycler?

>>1358725
you said you had flux paste? you might try, when cold, coating the tip in flux paste then running the iron for about a minute or two at well below the melting point of your solder (125°C if you have a thermometer in it). the flux should activate and get bubbly and start to darken and/or boil off. then turn it up to soldering temp and scratch the tip with solder wire, and see if it sticks
or use tip tinner, which will also be good for rejuvenating tips when they wear out

>>1358041
I'm interesting in converting the doors in my house to electronic locks unlockable via my phone. I know how to code and interface microcontrollers. What type of door hardware do I need?

>>1358724
Yes, more voltage in series and more current in parallel. AFAIK, DC motors produce voltage in proportion to rpm, and as you put a load on them they will put a load on whatever's rotating them, i.e. they're harder to turn the lower load resistance you have. Faraday's law and conservation of energy, I think. Upgrade that to Lenz's law if you're unsure.

>>1358748
Maybe an electric latch

>>1358663
hahaha
thanks, I know its not the perfect circuit, but it does work, and magnetic resonant saturated circuits in this fashion are very common for power engineering before solid state and relay technology. (which is why its near impossible to find info on it.)

The circuit we designed works perfectly fine, and the power dissipation is really not that bad.
The CT I am using is from 1950, and was used to power relay circuits, So I know it can do the job.
The harmonics are not the result of the circuit. We found the harmonics on the CT when only connected to an O-scope. The FFT output looked not great.
I just need some insight on how to design a filter on a variable current source.

Really, I would of preferred to go with my hysteresis comparator design, or our load switching MOSFET solution. But this is the idea that we chose based on its simplicity. It's my job to make it work.

>>1358779
Would I need one microcontroller for each door or what?

>>1358785
If you're asking this, you don't know how to interface to microcontrollers.

https://www.google.com.au/search?q=setting+up+electronic+locks+arduino

>>1358786
What is the cheapest microcontroller right now? Something wifi enabled. I have like 8 doors

>>1358787
Just go and read some fucking articles from that google search you dense faggot, you are so far off the mark its not funny. You need one uC, and you need to learn how to control hardware via its pins. We're not here to spoonfeed you top to tail through a project, just get you over roadblocks.

>>1358788
How would one microcontroller control 8 doors on two different floors without a fuckton of internal wiring?

>>1358793

Not that guy, but, seriously, you really need to learn how to actually do this before you try and start sourcing parts.

Trust me on this one, because that's a really bad habit I still have a bit of. It's very obvious by the questions you're asking that you have no idea how the details of this are actually going to come together, which is critical in virtually any kind of electronics design.

Case in point: Almost any microcontroller will work for this. However, there's a world of difference in getting wireless connectivity on a small AVR chip (like an attiny) and one of the ESP32 boards.


As an aside, for small quantities, NEVER concern yourself with a few dollars in components, if it means sacrificing functionality or it makes things more difficult on the design end. In most cases, the total cost of the rest of the hardware will dwarf the price difference between a barely-workable chip and something costing even several times more. Never even mind the extra time spent getting it to function as intended. It's just not worth it for 1-off projects.

just getting timer based debouncing to work in pic assembly feels like an achievement

>>1358780
In that case, get yourself a big transformer and use it as your inductor with its windings in series, or otherwise go for a large audio inductor. From there an NP electrolytic will work ok for the other side of your filter. Note that this will make the whole thing pretty massive. Check mouser or digikey for the best capacitors and inductors for value/$, and see how efficient you can make it.

>>1358795
I was thinking of using a Raspberry Pi Zero W for WiFi + Python support. Each device would be running a Python TCP listener. When the Python script receives a certain message it'll open the door. Thoughts?

>>1358801
ludicrously overweight for the task, but it would work
ESP8266 and 20 lines of Lua or microPython would be plenty
in either case, now you've got a logic-level signal, you just need to turn it into power enough to pull in the latch

>>1358801
you should be able to run this off of a PC for one device. if you can get that working, port it to the RaspPi. When that works, expand to more doors.

>>1358805
uh, where's the GPIO?

>>1358805
I know it is doable, I'm just wondering if that would be the cheapest option. Raspberry Pi Zero W is $10 each I believe.

>>1358808
ESP8266 boards for $3

>>1358814
it's really something that you can just connect the antenna directly through a cap to the ic

Do you guys have any suggestions for ICs that are capable of having modern DAWs (such as Ableton) recognise it as a MIDI device?

Everything I look for results in a Teensy board, but I rather do it from scratch.

>>1358855
no, because such a chip would be of no particular value in the present musical instrument market. even if there were, it'd be a microcontroller with USB-MIDI bridge code burned into it
so just get an ATMEGAxxU2 and build the LUFA example code and enjoy life
also see usb.org and read the USB 1.1 spec (at least) and the MIDI device class spec

>>1358807
Why? He wants wifi

I'm making a test board using this DRV592

Do I still need these 1/10 uF decoupling caps connected to Vdd if I'm gonna be powering it through an external lab voltage source?

>>1358859
So basically what you're saying is that what a Teensy board is capable of, a typical Arduino can too?

>>1358863
Nvm I'm retarded and the answer is yes. Still not sure about the 10uF at the top though

>>1358863
If you want to run it at a frequency relevant to the output filters in your pic, yes.
Even if you don't, probably still yes. You might even need more capacitors, if your power supply wires are long.

>>1358868
>a frequency relevant to the output filters
You mean like 16kHz? I wouldn't say that's significant compared to the PSU wire's inductance. Though the datasheet does say it goes up to 1MHz.

>>1358868
>>1358866
noob here, why ICs need a tiny cap at the power input?

>>1358874
>You mean like 16kHz?
No. Those filters aren't nowhere near sufficient for 16kHz switching frequency. You dimension the coils like you would dimension a switching regulator's coil. Well, at least in laser's case, where you want the ripple to be reasonably low. If you're ok with massive ripple, you can use smaller coils and/or lower frequency. You can even omit the coils, if square wave drive is ok for your load and you don't care about EMC.
Also, the output transitions are much faster than your PWM frequency and your bypassing/filtering has to take that into account.

>>1358863
One more question about this one,

Say I want the /Shutdown and /HiZ inputs to be inactive all the time, should I use pullup resistors or can I just connect them to Vdd straight up

Same question regarding FAULT0/1 monitor pins, is it okay connecting them straight to ground, or should I use pulldowns?

>tfw my city has an abnormal number of hobbos and people that live of collecting cans and scrap metal
>the instant something is put out on the street t is taken and dismantled by some hobbo
>trash company accepts electronic trash
>electronic scrap collecting bin in my uni where I used to get parts no longer exists (it was a hobo free)
>too autistic to go there and say "hey I`m studying EE and I need parts for some power supplies, can I see if I find some old PCs here?"
kill me

>>1358891
Actually disregard the first question as it is also retarded, the answer is yes, use pullups

Though regarding the FAULT0/1 pins, they are the DRV592's output, would I be better off grounding them or just letting them float?

>>1358892
>too autistic
Buy a lab coat and hard hat and walk in there like you own the place.

>>1358865
If you stick a crystal on it and use the V-USB soft USB controller yes. Using a chip with built in USB such as the ones the Teensy used makes it easier.

ESP8266 eeprom question. This site says it's actually flash and that it is limited to 4096 bytes, but you can specify 4 to 4096.

http://arduino-esp8266.readthedocs.io/en/latest/libraries.html

Other sources say 512 bytes, but it isn't clear whether or not they just chose that as the amount they wanted to use out of the 4096 max. Anyone know the actual eeprom specs?

So, does the library site documentation mean that you always write the entire block when you commit? Can you change just one byte? Flash programming usually is done in blocks, which is why I think you might have to specify the entire block on every commit.

Finally, how are they writing to flash? My understanding is that the ESP8266 only allows that when you're using the bootloader to load the code. If they can do it arbitrarily, it would be cool if I could use a lot more of that huge flash space that my code never comes close to filling.

>>1358779
Better pick one that has a hidden way to physically bypass it for when the power fails or the thing wears out. Lol you'd feel pretty fancy with your RFID locked basement dungeon until you lock yourself in it with your victim

>>1358744
I tried that, the Flux smokes and bubbles, but still the solder just bubbles off.

I can buy tinner, is there no other reason or do I just need to buy stuff? I guess I'm just confused as to why I'm getting this problem even on new tips and I'm wondering if there's possibly a problem somewhere else (the iron for example).

>>1358910
Looks like 4K sectors are standard. Commit writes an entire sector at once, changing a single byte isn't possible with flash.

The stock firmware only uses the flash for bootloader use, other programming environments usually split it into partial bootloader + user code in one section with a SPIFFS filesystem in the remainder, the Arduino board package defaults to 1MB program+loader + 3MB SPIFFS on 4MB modules.

>>1359043

thanks for the reply. I'll use an external eeprom on I2C.

Lets say you had an aluminum casting that weighed about 1-5kg, how feasible would it be to put something magnetic in it, then make an electromagnet that could lift the casting. Doesn't need to lift by much, just so nothing is touching any surface.

How much we looking at cost wise and energy consumption wise?

Is this a bad idea right here? How would you route this

Talking about the SMD feet and the decoupling cap, C2 to be exact

>>1359109
this is absolutely hideous but it'll work. have a little pride though.

one other thing you can do is make a very large pth under the thermal pad so you can stick a fine tip in there and solder it from the other side.

>>1359112
also 1206 smd packages are very easy to solder, so since you're already using a qfn you should think about adding them just to pretty up your board.

>>1359090
>nothing is touching any surface
You mean stable floating, Levitron-like?

>>1359114

Yeah

>>1359112
I'm struggling with Analog ground/power and the so-called High-current (P) ground/power, as they should apparently be kept separate except for 1 point close to the power supply. Or does this only go for the ground and I can connect Vcc's separately?

The thermal pad should be connected to Analog ground

This is the circuit depicted in >>1358863

Now I have no idea how to connect C3 to the Avdd (pin 1). Fuck me I'm retarded

What the fuck does High-current even mean. Should I even have the Vcc trace this fat. Why are the feet so close together

>>1359118
You have the keyword, look for diy versions.

>>1359090
that sounds stupid, what are you trying to do? maybe we can help you with a non retard version

>>1359120
pic is what it's telling you to do. your chip will pull large currents through the PVDD pins and out the PGND pins for brief transients which, due to trace resistance, will cause a voltage drop on the power supply rails. this drop can fuck up your logic circuitry. additionally the current causes the ground to rise, which can fuck you again in both analog and digital signal levels since ground is no longer close to 0V in places. this is solved by connecting the power and signal circuits to your supply through completely separate traces. you still get voltage drops from the length of power supply wiring after they're connected but the big capacitor helps with that a lot.

basically connect everything that draws lots of current to PVDD/PGND and everything that communicates or measures to AVDD/AGND. your chip is an exception since it does both, but that's why it has split supplies.

trace width is based on self-heating due to power loss. there are calculators online. be conservative with your sizing, because you can.

the feet are close together because fuck you. if you don't have a flux pen then you'd better buy one before you try soldering that with your shitty $50 weller.

how long took you to start making your own stuff instead of just following schematics and analysing circuits? it is very hard for me to think about an objective or function I want and then backtrack to a workable design. Does that come with experience?

>>1359127
i did my first pcb about a year after i picked up EE as a hobby. it was just an isolated usb to uart converter but i didn't copy anything. you'll always look at reference schematics though because not doing so is willfully opening yourself up to mistakes.

>>1358862
how's he gonna open the door without GPIO?

>>1358877
digital outputs often draw a spike of current as they swing the output from one state to the other. this current spike can pull local ground closer to Vcc, possibly resulting in false high logic levels internally or externally to the device

>>1359109
absolutelydisgusting.jpg
a. why not just use power planes
b. c'mon, man, use 1206s

>>1359126
Alright thanks I'll give it a try from scratch again later

As for traces coming out from the chip, should I try routing them like a good boy (1) or is the niggery way alright since it does seem easier to solder and would allow for larger currents. Also allows for arguably smaller footprint

>>1359141
it's prettier to do 1, so since it doesn't matter you should do that. pcbs are as much art as they are science.

>>1359137
a. I don't know but the datasheet says "Ground planes are not recommended for AGND or PGND."

Also this is just a test board to see if this thing works for the intended purpose, will be made inhouse so 2 layers only

b. I don't know if my uni has those ready, will definitely use SMD's for the final board, I thought through-holes would be better for prototyping? There's probably a drill somewhere

>>1359141
>easier to solder
slightly harder to rework, should you desire to do so later
y u no use polygons and let the router handle attachment to the pads

>>1359145
there's nothing wrong with ground planes as long as you have two separate planes joined back at the power connector. imo voltage planes on 2 layer boards end up looking awful so i just route mine as thick traces.

>>1359145
a. oh, that's a good reason, never mind >>1359146 then
I'd agree with other anon and recommend (1)
b. through-hole is easier for solderless breadboards and non-plated through-holes. if you're making the board in-house you're probably not plating the through-holes or using solder mask, so it's probably easier to use the component leads as vias than to use eyelets and SMD

noob here, I think I`ll make a metronome

>>1359195
A metronome does not BUZZ

>>1358524
What pitch is this FFC? It is variable pitch right, what do I call that so I can search for a connector? Every site I look at just has one type of pitch per item.

>>1359220
i was curious so i found a similar "two row" ffc on digi. pic is how they define its pitch, but i couldn't find any searchable fields to differentiate from a two row or one row connector. if you have calipers i'd measure the pitch, filter by that, and start digging.

>>1359235
Is "two row" how it is called? That is a big help thank you. Could you link that ffc you found on digi as a reference

>>1359241
filename

>>1358430
why not use just the diode? what's the capacitor and resistor for?
Don't you want the brake light bulb to flash the same as the indicator, be on at the same time?
With just a diode, the brake light will be on with the indicator is on but the indicator won't be on when the brake light is on.

If you really want the brake light to flash for a set time, you could use a MOSFET to connect the light to 12 volts and drive it with a monostable (one-shot) timer triggered by the indicator.

>>1358780
>The CT I am using is from 1950, and was used to power relay circuits,
More info please. Was this thing powering some control circuit powered only by the current it's controlling?

What are you trying to do with this exactly? Why do you need 120 volts AC in your system, why can't you rectify this current harmonics and all and feed it into a DC-DC converter to output 12v?
finally, if you're coming up on the project deadline you should have already started building a long time ago. Seriously, so many projects sink because people think they can build last-minute.

>>1358223
What this nig said about impedance is correct
>>1358226

Also many times you will see electrolytics back to back with their polarity being opposite. It's a cheap way to make a bipolar cap with a large value rather than buying a single non-polarized high-value capacitor.

>>1359195
>>1359201
Yes, just power a speaker/piezo element with it, you'll get something akin to a clicking noise which should be much better for keeping time to than a buzzer. Turning the driver into a tuned oscillator by adding the appropriate impedance operator or two to the piezo/speaker might also be an idea if you want it to ring at a particular frequency.

>>1359292
Yes but it only works with the diodes. Without them you get an inverse voltage across each capacitor for half the cycle. The voltage V(n002) in that spice sim is that across the second cap C2, and as you can see it goes both positive and negative. Also the impedance is doubled, not halved, because the capacitance is halved and impedance is proportional to the inverse of the capacitance. The rule of thumb being all impedances add when in series.

>>1359386
>Yes but it only works with the diodes.
If a well-known electrolytic capacitor manufacturer says it works without diodes, then it probably works without diodes.

Some OC since I didn't have space on my desk. But turns out that the bottom of the desk ontop of the solder iron kinda changed color and is very hot. Should have lowered it a bit I guess.

>>1359443
Then there's got to be some explanation as to why it works, because you're still subjecting it to +V/2 and -V/2. Perhaps they're rated to V/2 in reverse, I wouldn't know.

>>1359201
>>1359386
to be quite frank I could not find a piezo in multisim`s library. at least the buzzer really uses the sound board and makes noises in the interactive sim.f

>>1359455
The capacitors themselves kinda-sorta rectify the voltage.
See the PDF in >>1358240

>>1359459
They're akin to a capacitor with a parallel resistance, though some models have inductors somewhere as well. But I'd search for an equivalent model for either, they're probably downloadable for whatever sim you're using.

>>1359463
>The two capacitors rectify the applied voltage and act as if they had been partially bypassed by diodes
Ah, so each cap acts as somewhat of a short circuit in reverse. It says that NP caps use a second anode foil instead of a cathode, and that with all but momentary AC duties they will overheat and fail, which I suspect will also happen to a back-to-back-electrolytic capacitor.

>a non-polar capacitor equal in capacitance to half of the rated capacitance of either of the original pair
Doesn't quite make sense to me since there's no reason for the combined capacitance to be half of either of the others if one is just being bypassed.

>>1359480
yeah I know that, but I`m autistic and I wanted to hear it. I used a piezo to get signals from an acoustic guitar and do stuff with them

Is it related enough? Supposed to be a mini drill

>>1359511
/diy/ quality

>>1359512
Do I need to put some more duct tape on it?

It sounds like I'm asking for trouble, but I have an old laptop battery. They usually have 18650s right? What can I do with it?

How do I make sure a solder tip fits my iron? I bought a fine tip and it didn't fit and I don't want a repeat

>>1359109
I redrew parts of the VCC and GND traces. Do it like this. Sorry if it's a little hard to read.

>>1359615
>They usually have 18650s right?
Yes
>What can I do with it?
Make a battery

I'm interested in converting the locks in my house to electronic locks. I have 4 doors and 4 raspberry pis. Would it be a good idea to have my phone send the unlock signal to my VPS which then sends the unlock signal to the raspberry pi(s)? I originally wanted to make it local but my upstairs and downstairs have different Wifi networks

this is more of a general diy question but anyways, there is no sqtddtot upright now.


I got these dc motors from a electric shaver and a nose hair cutter thing. I was planning to so some sort of micro lathe with one of them to fuck arround with scrap. But I have no idea of how to connect an axle (in this case a piece of thicc wire to them without damaging them). Anyone has an idea or experience on re-using these motors?

>>1359646

>>1359646
what you're looking for is a "shaft coupling" with a set screw. these motors are a bad choice even just for fucking around though. quadcopter bldcs are the gold standard, and they're cheap.

>>1359647
>quadcopter bldcs
Not him but I was wondering something about BLDCs: are they basically the exact same thing as synchronous AC motors?

>>1359663
they're basically the same, and i expect that you could drive them interchangeably. as i understand it though the magnetic geometry is different to optimize them for sine or square waves.

>>1359657
>these motors are a bad choice even just for fucking around though.
Why? Also that what I was looking for, thanks. Now I need to find a way to make a coupler for those tiny ass axles. they are probably less than 1cm. I think I`ll find some sort of tubing, glue there then fill with epoxy and then use that enlarged shaft thing to couple to a bolt.

>>1359704
not enough torque to be useful as any cutting tool.

>>1359704
mcmaster has couplings down to 1/32" so they do exist. obviously best to find them from a source that doesn't mark up 200x.

Do any of you lads know why the DC link voltage that feeds inverters are generally 400V is it just a legacy thing?

>>1359663
Pretty much the same thing. BLDC may imply being designed for a more trapezium shaped EMF so it can be driven more smoothly off square waves vs a sine shaped EMF on a PMSM.

synchronous AC motors also includes ones using a field winding instead of permanent magnets.

>>1359716
Most inverter designs behave as buck converters towards the AC side, so you want DC link voltage to be above the peak-peak voltage of your AC line, so

>>1359721
so you want the link voltage to be above peak-peak AC line voltage. 400V DC is reasonable for a 120VAC system.

>>1359721
Ok so for a 230V rms AC line i a get peak-peak voltage of 325 V so even with voltage drops in the inverter were missing a 50+ V drop. Is the inverter bucking the output voltage to the regulation 230 V?

>>1359666
>>1359720
Cheers. Follow up: why not substitute the permanent magnets on a BLDC with a squirrel cage or wound rotor and get a mini asynchronous motor? Not enough space?

>>1359720
>Pretty much the same thing. BLDC

well not really.

an AC motor is designed to start and run powered by a fixed frequency voltage, like 60 hz.

any decent brushless dc motor expects the power to be synchronized to the position of the rotor. good ones use encoders or some means of detecting the rotor position. cheaper ones use back EMF, but they all synchronize the power waveform to the actual position of the rotor.

Which makes a huge difference. They are not "pretty much the same thing", unless you are driving your brushless dc motor in some crap-assed mode.

man I never even stopped to think I could get wire at a lowe's

the price isn't half bad either, a far sight cheaper than other places I've seen

>>1359727
>regulation
nominal line voltage is 240Vac ±10%. 264 * sqrt(2) ~= 375 which is within 10% of 400Vdc, leaving some headroom for Vds(on) and current sensing resistors. you would ideally like the voltage into any converter to be as high as practical. the prevalence of commodity off-line switchers make the components to handle 400Vdc quite practical and cheap

>>1359747
>14 awg
haha just kidding

>>1359735
Permanent magnet is generally better for power density and is simpler to control.

>>1359746
I said same thing as AC synchronous (or PMSM), which are synchronised rotor to power frequency. AC induction/asynchronous motors are different.

>>1359619
Buy one specifically made for your iron, or series of irons. There's a few main series of tips out there, which are used by multiple different brands making tips for those brands' irons interchangeable, though I'm no expert. What iron are you using?

>>1359800
An x-tronic 3020-XTS. It's a no-name brand from Amazon

>>1359807
why would you spend $60 on that

>>1359735
I'd wonder more why they don't put permanent magnets inside induction motors, you think you'd get less core losses.

>>1359746
Is this why?

>>1359797
Well induction motors are technically a type of synchronous motor, in fact they're the archetypal asynchronous motor. I doubt you'll find any sinusoidally-powered frequency-feedback AC motor out there, if that's what you were implying.

>>1359807
do they look like this

>>1359808
it was a gift

>>1359812
it only came with one tip, but yeah, the one that came with it looks something like that. I saw those tips on amazon and I'm not sure whether to get them.

>>1359811
Asynchronous means not synchronous. An induction motor needs to run slower than synchronous speed to produce torque.

PMSMs are sinusoidally powered with feedback AC motors.

If you put permanent magnets inside an induction motor rotor you get what's called a line start permanent magnet synchronous motor, it starts up as asynchronous until it gets near synchronous speed with the bars providing the torque, at which point it snaps into synchronous operation locked to the permanent magnets and the asynchronous winding/bars stop doing anything.

>>1359815
Hakko 900-M series is popular among chink soldering irons. it's probably a good buy
if you can measure, confirm that the base part where the element goes is a little over 4mm inside diameter, about 6.5mm outside diameter, and (less critical) 25mm long

>>1359822
Ah, because the position of the magnetic pole in the squirrel cage shifts as it moves making it not spin at mains frequency multiplied by whatever number of poles used. You're right, thanks for clearing that up.

How are those 858D hot air stations? I want hot air, but I don't want to spend an arm and a dick, and my Chinese solder station has been great overall, so I'm thinking about gambling it.

>>1359643
Just make sure you have some kind of backup because that system will not be absolutely reliable.

>>1359249
yeah i forgot to mention that i'd need the light brake light to turn on after the indicator like pic related. i learned that by wiring a capacitor and resistor in series would add a delay (capacitor farad X resistance in ohms equals delay in seconds) so i guess the last thing i need is that mosfet

>>1359899
nvm by looking at this gif i realized that my circuit is complete the way i drew it originally... so i guess that's one problem solved

>>1359876
i got that one today (same 'zany' brand)

seems pretty decent, i'm cutting out the mold flash around the fan intake

$25 off ebay

>>1359889
How would you make it more reliable?

>>1359889
if he's got a separate pencil station, he can always fix it :^)

>>1359909
give it a once-over. open it up, check for cold solder joints, loose connections/hardware

>>1359909
Just make sure there's one door somewhere you can get in with a key. Pis are shit for reliability because the SD cards fail. He's not saying it's absolutely not reliable, he's saying it's not absolutely reliable. It'll probably be fine, you just don't want to kick in a door if it ever fails.

>>1359909
First off, I wouldn't use wifi at all, I would hardwire it. When you go wireless you always trade some reliability, if something doesn't move around it should be hardwired.
It should not depend on your VPS, because that could go down or your internet could go down, you should have a server inside the house to control any locks, cameras, and alarms. This server should have a UPS, with extended batteries if you live in an area prone to outages.
You should have some way of getting in that doesn't involve the internet, like a traditional key lock or at least a keypad.
Finally, test your shit. Don't wait for a power outage to discover that some critical switch or router isn't on the UPS, or one of your pis doesn't reboot properly.

>>1358041
chocolate

>>1360147
get the FUCK out

>>1360265
nevermind stay, we will enrich the thread

>>1359748
Thank you this is exactly the kind of explanation I needed

DIY IR blaster
Is this going to fry my headphone jack?

>>1360308
wtf am I looking at?

>>1360318
DIY IR blaster.

Looking at getting into FPGAs, done embedded C with 32-bit micros, mainly Atmel and STM32F.
1. What family should I go with Xilinx or Altera? I understand ISE is a pain past Windows 7.
2. VHDL or Verilog
3. What development board would be good to get started?

>>1360330
i bought a board a while ago and messed around with fitting a 4000 logic circuit i'd made into it. i found verilog to be wonderful, but you'd have to put a gun to my head to get me to use quartus prime again. be a guinea pig and try xilinx.

i just got the $20 ebay special and a chinese usb blaster. both worked fine, both had chinese QUALITY manufacturing.

>>1360320
Will not work.

>>1360308
add a series resistor, you fucking nigger

>>1360344
Don't listen to this fool!!! follow your dream!!!!

Are you going to blast by outputting sound??

>>1360330
Ha lol only retards get into FPGA when python exists!!!!! Just learn u some good programming

>>1360352
>>1360350
>cremes
get out whitey, this is our thread

where do you guys publish the stuff you build?
Do you have your own project homepage or do you use one of the sites for this?

>>1360348
how many ohms?

>>1360350
there's apps that output an audio signal that matches remote signals

>>1360353
Fuck of shit-dick Whities rule electronics with an IRON FIST

>>1360330
of the big two, Altera is cheaper at the low end. clone USB Blasters can be gotten for under $5. there are more minimal dev board options for Altera CPLD/FPGA chips, which themselves are also somewhat cheaper
verilog, of course
depends on how much you expect to cram in there to start. a chinkshit Cyclone-whatever board would be good enough for most experimentation

>>1360333
they all license the same shite from Synopsys or whatever

>>1360363
won't work for IR. the receivers generally look for the IR brightness to be modulated by a particular frequency which is not an audio standard. you really should learn something about what you're doing before you do it

>>1360374
>won't work for IR
They sell them, what's different in theirs?

>>1360374
You can modulate amplitude with an audio sample. You limiting you imagination saying the IR blaster can't do AM or FM. I suggest taking a simple course in ADC and DAC systems :^)

>>1360330
1. Altera, if only for the tools, which work better than Xilinx's in my experience. I have only had to work on each for uni work though.
2. VHDL for me. It is what was taught, and happens to be in demand where I am.
3. I used the DE0-nano (and DE1-SoC, but it has two arm cores on it, way overkill for anything you could want at first) from Terasic, liked it, can't remember the Xilinx board.

Now, another option is to go with Lattice and icestorm, an open source toolchain instead of all the bloated suite mannufacturers push on you. I have not made the leap as I think it only supports verilog, but there are very cheap little boards available. If I was starting on my own, I would go with these I think, simply because of the leanness and transparency of the toolchain. Seriously, I've been meaning to learn verilog to have a quick and lightweight prototyping setup.

>>1360356
>implying
they're going to die with me

>>1360363
IR remote controls modulate the light with a 38KHz carrier signal. 38KHz is ultrasonic, it's out of the audio band which only goes up to 20KHz. I would be very surprised if you were able to output 38KHz from an audio jack.

>>1360392
A lot of phones can output 192kHz audio now. that's enough for a course IR blaster. I would recommend you open your mind to many possibilities :)

>>1360393
I'm open to it, but just would be surprised there's no low-pass filter.

>>1360330
m8 don`t do that to yourself. FPGAs are suffering and pain in electronic form. Anyways, whatever you do stay away from quartus or any board that requires it. I still have ptsd from that

I found this motor in a dumpster. How do i know what each of these wires does and which ones need to be hooked up to a power source?

>>1360404
>look up model
>look up manual/datasheet
it`s probably there

Specs

>>1360408
>>1360404
>>1360404
>tfw only find construction refuse in bins and streets
:(

>>1358995
sorry to bump, but curious if anyone could offer more advice on this issue. I'm hesitant to buy tinner before I really have to.

>>1360408
>>1360404
Baldor is a well-known maker of motors. You should be able to get a wiring diagram for that motor off of their website.

what's the difference between a voltage regulator and a resistor

>>1360330
I prefer Actel (microsemi) FPGAs
Xilinx tools blow

>>1360546
everything

>>1360546
both of them drop voltage when you pass current through them. literally the same thing pretty much.

where do you all get your wire from? Amazon is expensive, and I don't feel like waiting for aliexpress

>>1360363
>how many ohms?
Doesn't matter, whatever you have in your junk box. See it as a soldering exercise.

>>1360546
A resistor has resistance to the flow of electric current. Learn Ohm's Law.
A voltage regulator is an active device that uses feedback to produce a constant output voltage as load and input voltage vary.

>>1360552
I have never bought wire. I get it all from the trash or recycling.

>elecrow's cheapest shipping is now $12
so much for $5/10 pcbs

Recomend me a good soldering iron. I have 30e

What would you recommend for soft starting a single phase resistance-start induction motor? I have only 600 watts to start this 1/4 HP motor.
I had hoped I could just put a big inductor or resistance in series and have it slowly spin up, but no matter what I do it needs 10+ amps to start spinning at all. I guess this is because the rotor needs a strong yank to get it to follow the rotating field, if the current is too low the rotor just hums as the field passes it by.

>>1360642
A hand crank, or if you have the money a smaller motor geared properly to give it a kick.
Or if you want a real project a few car batteries and an inverter, a changeover contactor to switch onto mains and start charging the batteries again for next start up, perhaps a vdf can do some of this for you....
Of course it doesn't have to be a battery, diesel generator might be an idea.

Of course if you explain what the limit is, there are breakers rated for motor curves, wiring can take small overcurrents, calculate the heating time.

>>1360658
The limit is it's supplied by an inverter.
This is for an experiment in getting the most out of an inverter. Anything with a motor like a fridge or furnace draws a much higher current when starting than when running, many of these things can be run on an inverter but not started by it.
My idea here is to have a large flywheel on this motor, spin it up slowly, then run it as a self-excited generator and use the energy stored in the flywheel to supply the inrush current for larger loads.
These japanese engineers did something similar, so I know the principle works: https://www.jstage.jst.go.jp/article/ieejias/127/8/127_8_844/_article
Using a DC motor to start it spinning probably would work, I just wanted to do it purely electrically if there was some way.

>>1360673
You can handle an auxiliary inverter, you cannot handle a flywheel.

>>1360710
An inverter is out of reach for loads that draw 25+ amps at inrush.

I'm trying to wire up a Dsub 15 connector for VGA. Do I need to wire the ground pins?

>>1360844
Yes.

>>1360633
TS-100
WIth a 12-21v battery pack you can make it portable.

>>1360924
>TS-100
>overheating meme iron

What's a good book that just gives me a bunch of small projects I can do for fun.

I'm an electrician so I know how electricity works, and I've done PLC programming so I know how logic gates work, but i've never done projects just soldering electronic pieces together.

Basically I just want to do a bunch of fun small projects to familiarize myself with the different components and possibilities.

>>1360978
Typical chocolate poster, maybe look at the OP first next time.

>>1360978
>Project/idea websites?
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/
From the OP.

>>1360981
>>1360988
thanks & apologies

>>1360400
Doing a proper ~20kHz low pass filter for a 44.1/48kHz DAC takes up space and power, running the DAC faster then doing a simple passive analogue low pass higher up if they can still pass EMC is easier.

>>1360844
You need at least one of them. Preferably use all the ground pins, with the red green blue grounds connected to the shield for the micro coax on the red green blue.

>>1360642
Variable frequency drive, get rid of the resistance start. What kind of mechanical load is on it? What causes the 600 watt limit?

>>1360633
Hakko FX-951/T12 clone. Or 936 clone for cheaper.

I need to encase a project that is using mains power. If I am using a plastic case with metal screws that are completely covered internally by the screw posts when assembled, do I need to ground the screws? The original converter that I am modifying has this setup and does not ground the screws, but I want to be sure that this is considered safe before using it in my own case. I was not able to find the relevant codes.

>>1361152
>do I need to ground the screws?
No, much like a double-insulated grinder.

>>1361158
Thank you.

For mains-powered devices in metal cases, code seems to require that doors and removable panels have grounding wires. Is this true for panels that bolt onto a grounded central frame, if it meets the two-thread rule for grounding screws?

>>1360978

>>1360377
pop one open and find out rather than mentally masturbating

>>1360380
>muh imagination
>>>/g/
I think it's more likely, in the smartphone case, that they either send the baseband data over the headphone jack and modulate inside the blaster, or they speak their own protocol between the headphone jack and a controller which sends the IR commands at precisely the carrier freqs and rates the controlled devices expect
in either case, you're stupid

>>1361161
So long as the metallic connection back to the ground wire through the frame is good, you're fine. If it's a panel that might be opened in service having a separate wire would be a good idea.

>>1361165
There are lots of android DIY blasters just using antiparallel LEDs, it works fine. For Android there's the ZaZa app that sends modulated IR over audio. video example https://youtu.be/3-yvT13t8zU?t=4m39s

>>1361161
>Is this true for panels that bolt onto a grounded central frame
If they're grounded in two places you can end up with ground loops causing problems, but that shouldn't happen unless you're passing current through to ground.

>>1360380
Hey, an AM or FM modulated IR audio sender might make a nice little project, I'll look into that thanks. Never thought about doing FM with PWM for some reason, I guess because the square waves would be messy if you were transmitting it through RF, which still might happen if you don't shield the optical transmitter properly I guess. Though optical AM doesn't really need a carrier frequency at all so that would be a piece of cake provided your IR noise is inconsequential and you can keep things linear, which I guess means current sources and sinks.

>>1361216
We did PWM IR audio transmission as a university project, was pretty cool.

>>1361127
are those clones any good ? I found a bunch of 936 Baku/Hnawai/... clones all going for ~30e
I guess anytihng is better than the crap I own now - 3$ chinease soldering shit

>>1361252
I use a 936 clone and am pretty happy with it, only ting I sometimes want different is more heating capacity, which the T12 based ones are better at with the heater in tip.

>>1360392
>38KHz carrier signal
19 kHz drive is enough with two diodes. Why do light bulbs flicker at 120 Hz?

>>1360929
works on my iron.

>>1361164
rolling because its not like i have a mountain of other projects already going, not to mention my test tomorrow

>>1361272
Because the bulbs emit at both the peak and the trough of the AC voltage sine-wave. As such for every one wavelength they turn on and off twice.

>>1361375
Same thing happens with a pair of antiparallel LEDs like on the DIY IR emitters.

>>1361396
Yup, next generation of IR phone plugs will be able to learn codes using the mic ring.

How do I wire up a motor, dpdt switch, regulator, and psu? I have power going from psu to regulator, then with regulator I have the motor wires going to the power of the dpdt switch? Then from dpdt motor wires to the motor?

I am confused on the placement of regulator then dpdt or dpdt then regulator.

>>1361473
LEFT-STOP-RIGHT or only ON/OFF?
Why the regulator?

>>1361476
DPDT is to use motor forward/reverse.
Regulator is to control the RPM.

>>1361478
dpdt is on-on

>>1361478
Then switch between regulator and motor.
Motor can reverse but never stop.

>>1361473
if your motor isn't especially small then it's most likely a bad idea for you to wire it through a regulator. can you give more details, or ideally a picture?

Any pointers for getting started in PIC development.
I've designed a board with a PIC microcontroller in the hopes of learning to develop them.
I find the whole layout of the IDE a bit overwhelming.
What's with all the header file, linker files, etc?
Can I not just write one source file and compile it?
After a few hours of messing about with templates and sample code, I managed to get a project compiled (basically just shift three bytes out one bit at a time to three daisy chained shift registers).
When I loaded it to the PIC, it didn't so shit.

I pulled the pic out of the socket and wirelinked an arduino into the socket on the board to the the exact same thing and it worked fine.

Please give me some decent guides for getting started.
I can't bare to be an arduino babby all my life

>>1361518
you don't have to use mplab, i find it easier to use mpasm, pk3cmd, and SDCC.

your biggest hurdle is in initial configuration. don't forget OSCCON, ANSELx, and any comparator peripherals on your relevant pins which have to be disabled individually. here's some code you can reference (particularly the init section):
https://hastebin.com/loxojuwunu.vbs
it's in asm but you can still see the registers it's modifying and some comments for clarification.

with asm you only need one file :3c

>>1358041
I have a (stupid?) question.
What do i input in this for it to charge the cells?

>>1361547
electricity.

>>1361551
>electricity.
Ok, thats one piece of the puzzle.
Now i just have to figure out the voltage. :D

I'm pretty new to this.

I have an arduino Mega and I'm trying to control a 12V load with a 5V PWM signal.

Is there any reason this wouldn't work? Is there a better way of doing it?

I need to make a circuit that produces one pulse for every 8 clock cycle. It has to be done in verilog. How would i go about doing this? Also, any good verilog tutorials?

>>1361561
heating elements are generally just a resistor. so you're feeding your resistive heating element through a resistor, which is also a heating element. it makes no sense.

also the base collector resistor is bad for your application. it's generally used in analog circuits to keep the bjt from entering cutoff, but here it'll just steal some current from your heating element.

>>1361563
if you're using quartus use an lpm_counter module with carry-out as the pulse output. this tut is good:
http://www.asic-world.com/verilog/verilog_one_day.html

>>1361554
>the voltage
>the
no, you use a proper Li+ charging circuit. do some basic web searching on lithium battery charging and use what you've learned to set up a device like
https://www.aliexpress.com/item/rechargeable-nipple-clamps/32670043457.html
which, even still, requires you to disconnect it when charging is complete

>>1361563
my Verilog's a little rusty, but try something like
>reg c[2:0];
>reg q;
>always @(posedge clk) begin
> q <= &c;
> c <= c + 1;
>end

>>1361565
Im circuit guy.

Thanks, I didn't realise it was that simple.

Will having the heating element in series with the transistor allow 12V across it, or will it be limited to 5V because that's the transistor saturation?

>>1361568
the transistor will drop between 0.3 and 2 volts depending on the model. lower is better so you dissipate less power in it (heater current * saturation voltage). in your original circuit you're still dropping voltage in the pullup resistor though. more than you would be with the npn.


alternatively you can use an n-channel mosfet, which usually will drop less voltage and thus dissipate less power.

>>1361570
A couple volts is no problem. I just was under the impression that the Collector-Emitter voltage was limited to the same voltage as the Base-Emitter.

10V across the heating element is fine, 5V is not.

>>1361571
the emitter cannot reach a voltage higher than the base voltage minus Vbe. that doesn't matter here though since the emitter is tied to ground. your npn's Vce voltage will decrease the more current you put through the base, until it saturates and reaches a minimum value that's called out in the datasheet as Vce_sat

>>1361574
Right right, that makes perfect sense.

Thank you buddy. Much appreciated.

>>1361565
Thanks for the tutorial

>>1361567
I think that's right. How did you figure it out? Particularly the c <= c + 1

>>1361567
Dont these boards have charging circuits on them? The diagram says i can theres a power input, surely i dont need an additional charging circuit... right?

>>1361585
it's a protection board and what looks like a cell balancer. lithium cells have a slightly complex charge process that you need to follow, and that board won't do it. it's just to stop you from shorting them out.

>>1361581
being a register, you need to assign it a next-state. since we're counting pulses, we want a counter, so we make one from an addition and a non-blocking assignment. (the q non-blocking assignment is in the same always block just for convenience)
if we wanted a divide-by-5 counter, we would instead
>c <= (&c) ? (8 - 5) : (c + 1);
so that the counter would count 3,4,5,6,(7),3,4,5,6,(7)
but we would probably also want to incorporate some reset logic in this case to guarantee that no more than 5 clocks pass before an output pulse
>always @(rst) begin
> if (rst)
> c <= (8 - 5);
>end

>>1359746
>AC motor is designed to start and run powered by a fixed frequency voltage, like 60 hz

You mean Synchronous motors.

Asynchronous motors come in all flavors, and all require a feedback loop to be run correctly.

>>1361547
>>1361554

Input the output voltage from a charging controller.

that protection board does not have a Charging IC that I can see (yuk yuk yuk).

The only thing that module does is "banlance" the cells using resistive load to draw down cells standing at higher voltage relative to the rest. still needs a proper voltage feed to charge. Li-ion cells don't like being charged in the same way as most other chemistries. it's really more about current regulation while charging then about voltage regulation, but that's not important for you, just go buy a proper charging module. you can get one that'll charge a 25V series-stack, and the protection board should handle balancing while you charge.

>>1361272
>>1361375
that's really clever. thanks for the explanation.

>>1361567
>rechargeable-nipple-clamps
wait, what?

>>1361635

What's a minimum frequency I can expect to get good results from for a class-D amplifier? Is it ~40kHz because of mr. Nyquist, or higher such that it's easier to filter out? I'm also wondering why the LC filter on the end doesn't ring at its resonant frequency when it gets fed a square wave.

>>1361644
Ah, doing some testing on spice I'm getting an idea of what to expect. I'm wondering if sending a class-D PWM signal would be a better way of transmitting an IR audio signal than FM. I suppose the received quality as a function of average oscillator frequency won't change between the two, just how it's driven. Demodulating the square wave FM signal might be somewhat challenging also.

I'm drawing a schematic for my soldering station that's broken. Can anyone help me decipher the electrolitic cap?
And is the part marked "N4" a diode?
And lastly, it looks like the ground for the LM78 chip is connected to one of the AC inputs, and the Vin is connected to a capacitor with the same AC input on the other side. I don't see any rectifier. This seems incorrect to me.

>>1361687
There are already ways of transmitting audio over IR. Find out the existing methods before you try to re-invent the wheel.

>>1360403
>he's not a masochist
>>1360547
I hope you're joking microsemi synthesis/map is terrible

>>1361563
module pulse_gen(clk, reset_n, pulse_o);
input clk, reset_n;
output pulse_o;

reg [3:0] counter;

always@(posedge clk, negedge reset_n)
begin
if(~reset_n)
counter <= 4'h0;
else if(counter[3])
counter <= 4'h0;
else
counter <= counter + 1'b1;
end

assign pulse_o = counter[3];

endmodule

>>1361690
Update:
N4 is the diode that rectifies the AC voltage and I had my VAC+ and gnd mixed up.

>>1361695
isn't that a divide-by-9 counter?

>>1361699
yeah you right. I just saw the power of 2 and couldn't resist. reset it to 1

Just found this:
http://www.themirt.com/index.php
Anyone tried it or something similar?

>>1361723
no

>>1361723
Not me but someone I know made one of the hand-held versions of this from the back of some hacking zine.

for as rats-nest kugled as it was, it worked like a charm on certain lights. still does, as far as I know, I don't live near the guy anymore.

Some cheap ass cities don't have IR preemption systems installed and emergency vehicles will just roll through lights cautiously, so it's sort of location dependent. there's also newer Preemption systems that are microwave frequency (2.4 Ghz) and operate using wireless protocols so they can encrypt, but I don't think those are common yet.

>>1361644
>What's a minimum frequency I can expect to get good results from for a class-D amplifier

I'm confused. do you mean Max frequency?

> I'm also wondering why the LC filter on the end doesn't ring at its resonant frequency when it gets fed a square wave.

Because the PWM should be very, very far past the cutoff of the filter. this isn't even to avoid oscillation, it's to ensure high fidelity of the signal.

>>1361922
I suppose he means minimum carrier frequency

>>1361924
ah, yeah, he probably does.

in that case
>>1361644

2 orders of magnitude above max signal frequency was the rule I was always given. This was for scientific apparatus, not for audio, so I have no idea what you'll want there, but Mhz Class D's are very common, so I doubt you'll have trouble finding something.

>>1361927
>2 orders of magnitude above
Right, that sounds good, will keep in mind. What kind of apparatus was this?

>>1361933

It was set up to amplify signals measuring phonon dispersion in some work on multiferroics. Too much to get into here and I honestly don't remember it all that well.

the frequencies were insanely high and we were tripping all over ourselves trying to get something that would sample fast enough. I'm still not 100% sure how we ended up pulling it off.

>>1361690
No idea what the 555 is for buy I'd guess it's a 33 uF 35v cap

>>1361690
What's that reed contact for?

>>1362003
Some fancy stations detect if the iron is in the holder and shut down after some period of inactivity.

Hey, I'm trying to design an analog feedback system for active noise cancellation and I'm looking for a good primer resource on control systems. I know a decent amount of signal proc stuff but not control systems in particular. Thanks!

>>1362072
But the iron is neither magnetic nor close to the board with the reed switch.

>>1362112
Well, something obviously is magnetic and close enough. And like the other anon said, this is typically used to detect whether you're using the iron or not.

>>1362092
That's not exactly primer stuff if you want to do it well, but in lack of better suggestions, you could try Modern control systems, by Dorf.

Feeling like a dumbfuck.

This is from LT1374 datasheet. Does it mean I have to keep it at above 2,8-something volts or can I leave it unconnected?

I'll post block diagram(which I think suggests that at 0V it operates normally) in a minute - for most non-specific application diagrams the pin isn't connected anywhere.

>>1362159

>>1362139
>detect whether you're using the iron or not
The sleep/wake function is typically controlled by a motion switch in the handle, where else. The switch on the board would require that the holder is part of the base. Such stations exist.

I spent $300 on a vape and dropped it within a few days of use; the idea of fixing the devices gives me hope that I can somehow redeem my stupidity.

What are the odds that I could replace the broken LED screen? I have the tools to disassemble the unit and can borrow a buddy's desoldering iron. The device heats up fine and can be used at it's maximum temperature. From what I can tell, the LED is the only component affected by the drop, the three buttons allow me to use the device blindly, I'll probably hang on to it.

Can the LED be replaced, and if so, could I source a replacement that fits in the original case?

>>1362159
>Does it mean I have to keep it at above 2,8-something volts

yes

>can I leave it unconnected

Also yes. the regulator should provide a tiny bit of internal current to keep the pin high behind those comparator inputs if it isn't connected to anything.

>>1362226
you would have to find a matching chinesium screen. if you can find a part number somewhere on that mess, go nuts. otherwise, probably not.

on the bottom pic, is that damage curiosity scratching or pixie-release

>>1362239
Yeah that's what I thought. Since I need 5 voltage(stepped down and inverted from 24V) levels and 5 regulators I guess I can just connect them all to 5V and it should turn on normally,

>>1362242
If the screen cannot be fixed, is it possible to connect the heating element and temperature sensor to a microcontroller? In pic related there are two wires to the element and two to the temp sensor.

I could run the device and take mulitimeter readings and then use that as reference. Maintaining a consistent temperature over time would be tough but doable.

>>1362244
yeah, probably. there's probably a reference schematic on the data sheet with that if I had to guess.

>>1362257
everything is already connected to a micro-controller on that board. if you're talking about that level of repair, and you really have the skills for it, then you could probably install a different LED screen.

really, what you should do is look at getting a DNA box mod board or something similar and just replacing the whole board unit. the $300 you spent isn't going into that board, I can tell you that much. it isn't going into the socket, either, by the look of things.

if you have a fancy fedora-box looking for electronics, there are plenty of options out there.

>>1362269
The beige socket looking things in the above photo contains the temperature sensor and heating element; the device doesn't use 510 vape juice tanks.

That herb vape is small and great at it's job, my hope is that the ceramic heating element is why this device is so costly. If that element is worth something then this project is interesting enough for me.

The 7x12mm screen isn't too small to be desoldered, no?

>>1362266
They generally leave it without connecting it anywhere. I guess I can simply buy one in TO220 packaging and simply check it myself

>>1362297
I went a head and grabbed the data sheet.
Just leave it unconnected. Pic related. default state is on.

>>1362284
what's the brand of the vape?

>>1362333
nice trips; Arizer Air II

can someone explain or point to me on resources that explain how to fit a non linear element like diode into circuit analysis? Circuits like the voltage doubler are pure black magic for me.

i'm using the base and collector of an npn as a diode because i need more reverse voltage than the standard bc-e configuration. datasheets don't seem to publish the base-collector diode forward voltage. is it safe to assume it's the same as the base-emitter?

>>1362366
out of curiosity i simulated it and the numbers aren't the same but are very close.
2n2222a @ 10mA: 756mVb-e vs 766mVb-c

>>1362343
if you want to see the effect more clearly, turn the input voltage up to 10V or more. the 0.6V band gap/voltage drop of the diode is hiding most of the action of the doubler when driven with such a low voltage. silicon nonlinearity can be neglected in this sort of circuit, you just simplify to Vf = 0.6V or so

Help me lads. What does the state table look like. I have an a swer but it doesnt match anyone elses

>>1362411

>>1362411
my answer. is this correct?

>>1362416
it looks right to me
>has the same functionality but fixes any problem
kek

>>1362407
yes, but how the fuck does it work

>>1362633
>When the input voltage Vi reaches its negative peak −Vp, current flows through diode D1 to charge capacitor C2 to a voltage of Vp.
>When Vi reverses polarity and reaches its positive peak +Vp, it adds to the capacitor's voltage to produce a voltage of 2Vp on C2s righthand plate. Since D1 is reverse-biased, current flows from C2 through diode D2, charging capacitor C1 to a voltage of 2Vp.

>>1362634
Why I`m getting zero current reading here? I`m retarded please help, sedra is a useless microelectronics book aaaaaa

>>1362637
because you have no load

>>1362637
SPICE requires a dc path to ground for every node in the circuit. add a resistor across C2

>>1362641
>>1362639
>>1362637
I`m a brainlet, but still no current

>>1362642
dunno maybe multisim isn't as smart as it thinks it is for analog
try putting a resistor in series with D2 (1 ohm maybe) and measuring voltage across it, or maybe you've got ideal diodes in there instead of real ones

>>1362645
uhmmmmmmmm I was indeed using ideal diodes.

>>1362646
>>1362645
changing them for a 1n4001 did nothing, but on interactive simulation it gives current and power readings.

>>1362649
good start.
wait, amps?! zoom in on that current like 1000x and I bet you'll see something

>>1362652
Brb killing myself now. I woke up by accident 4 a.m and decided to learn this to help a qt 3.14 this morning for her test later in the day, I haven`t studied this yet and I`m drunk with sleep. I think that when I graduate they`ll take away my diploma. ''ur too retards to have this anon, there was a mistake"

>>1362653
is there a way to change the scale diferently for each probe like on a osciloscope? That interactive sim scope is shit and that whole deal only works for basic shit

>>1362654
if it's using SPICE under the hood, maybe you could just ask it to plot I(D2) * 1000 or something
here's hoping you get a handy out of this, at least

>>1362656
thanks anon!

I want to design a custom pcb for a project I'm making, but I never really used any software for pcb's.
Which one should I pick: KiCad or Eagle?

both

>>1362747
you are telling me to learn both to make a single pcb?

vegeta zawsze mi sie kojarzył z niemcami, nie?

did all you guys go to college to learn this shit or are there self-thaught here?

>>1363015
oops, wrong thread

>>1363000
use pen and paper

>>1362653
>ur too retards to have this anon, there was a mistake
Don't worry I know plenty of engineers who are retards, you will be in good company.

>>1362725
altium

>tfw no pic instruction to rotate W by K

>>1362725
KiCAD doesn't put stupid limits on your board. otherwise they're approximately equivalent, modulo Autodesk being shitty as usual

>>1363016
I had a mentor in my mid-teens who taught me some nice cowboy tricks and dumped his component collection on me when he got out of the hobby, but otherwise I'm self-taught. still learning, 30 years later.

>>1363060
>test bit 0
>skip if z
>rotate
>test bit 1
>skip if z
>rotate
>skip if z
>rotate
>test bit 2
>skip if z
>swap nybbles
yep
>expecting a barrel shifter on an 8-bit micro
smuganimegirl.jpg

NEW BREAD
>>1363100
>>1363100
>>1363100
NEW BREAD

>>1362343
>Circuits like the voltage doubler are pure black magic for me.

It helps to think about it this way:
Capacitors store charge, don't think of it as storing voltage, but charge. You can store charge on the cap, and then connect one side to a higher voltage, you still have the same charge so you still have the same potential but now you have a new reference point.

There's a reason these are also called "charge pumps." Charge is proportional to voltage and capacitance, but that also means that voltage is proportional to charge and capacitance.

If you fill a jar with 10 cookies and set it on top of another jar with 10 cookies you now have 20 cookies. The cookies are the charge, the voltage comes from the cookies and the number of them you can fit into each jar.

>>1362725
KiCAD

It's free, there's lots of resources available for learning and you will never ever reach time where you will outgrow it.

>>1363720
I forgot to mention that you can do the cookie thing many, many times. Meaning these circuits can go beyond doubling the voltage. Look up "voltage multiplier" it's just a voltage doubler with more diodes and capacitors.

In fact you cna make negative voltages by reversing the cap so that the charge presents a negative voltage to the circuit, it's not uncommon to make charge pump negative voltage regulators because you don't have the right part laying around.

There are limits to voltage multiplication though and again that has to do with the quantity of charge and the capacitance. You can fill up hundreds of jars with cookies but when you have thousands of people ready to eat those cookies you'll never have enough cookies to keep the right voltage you want. If your circuit eats all the charges then you won't be able to build up any voltage. this means there is a limit to how much current your circuit can draw.