[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vr / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / asp / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / qst / sci / soc / sp / tg / toy / trv / tv / vp / wsg / wsr / x] [Settings] [Home]
Settings Home
/diy/ - Do It Yourself

4chan Pass users can bypass this verification. [Learn More] [Login]
  • Please read the Rules and FAQ before posting.
  • There are 58 posters in this thread.

05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
06/20/16New 4chan Banner Contest with a chance to win a 4chan Pass! See the contest page for details.
[Hide] [Show All]

Meta on /qa/ only.
All meta discussion of boards is to be redirected to /qa/.

File: 1499337917874.gif (14 KB, 611x376)
14 KB
bump limit reached on old thread >>1273089

https://www.wiki.printf.pl/index.php?title=Pasta (fresh)

>I'm new to electronics, where to get started?
There are several good books and YouTube channels that are commonly recommended for beginners and those wanting to learn more, many with advanced techniques. The best way to get involved in electronics is just to make stuff. Don't be afraid to get your hands dirty.

>What books are there?
Getting Started in Electronics Forrest Mims III
Make: Electronics Charles Platt
How to Diagnose & Fix Everything Electronic Michael Jay Greier

All New Electronics Self-Teaching Guide: Kybett, Boysen
Practical Electronics for Inventors: Paul Scherz and Simon Monk

The Art of Electronics by Paul Horowitz and Winfield Hill

>What YouTube channels are there?

>What websites feature electronics projects or ideas?

>Where do I get components and lab equipment from?
Search the web for "hobbyist electronics sources" to find plenty.
In the US and elsewhere, mouser.com, digikey.com, arrow.com, newark.com are full-line distributors that entertain small orders.

>What circuit sim software do you use?
This mostly comes down to personal preference. These are the most common ones though:
NI Multisim
iCircuit for Macs
CircuitJS (quick, dirty, interactive)

>What software should I use to layout boards?
Circuit Wizard

>My circuit doesn't work. Halp?
Check wiring, soldering, part pinouts, and board artwork if applicable, then post schematic. Supply ALL relevant info and component values.when asking a question.
File: 1497360655993.png (94 KB, 734x770)
94 KB
This thread's digits brought to you in part by the IRF7831 N-channel power MOSFET.
>dat Rds(on) doe
Should I go for an 888D or a 951? I'm leaning towards the 888D because I'm not sure if I'll feel the benefit of the 951, especially not for more than double the price of the 888D.
you should go for a chink knockoff of one of them for half the price
I've a knockoff 936 and it's done the job for a while but I need better. Thermal capacity etc is so poor and everything just feels nasty.
What brand is your knockoff, out of curiosity? My Yihua feels pretty solid.
just get an 862d+ clone soldering station w/ hot air for like $45
It's a Yihua, I guess some are just better than others and I managed to get a crusty one. Takes an age to get to temperature.
My 862d+ takes 2 seconds roughly
I just retired my old, decrepit, little blue Xytronics 168 in favor of an Aoyue 888A from SRA and I am extremely pleased with my purchase. The 60W handle comes up to temp from cold in about 15-20 seconds. The hot air works wonderfully and has an auto-cooldown cycle. It's sold with a dozen tips for $73 with free shipping in the continental US.
Is it really not worth it to salvage ceramic components like capacitors, and SMD stuff such as resistors or LEDs? I just saw some videos and soldering SMDs doesn't seem that hard, provided I can get myself the hot air station and a microscope.
I'm asking this because I could probably salvage something besides cables from a few older video game console controllers but I'm not quite sure if it's worth the effort.
Consider how you would organize them after you have pulled them. I wouldn't think it worth it except for ICs or inductors or maybe larger transistors and MOSFETs with some unambiguous marking other than just "TK1" or whatever.
SMD ceramic caps are especially worthless. They aren't usually marked and they almost never indicate their working voltage limit. SMD electrolytics don't take too kindly to all that reheating and cooling and will have diminished life.
So most of the good stuff will mostly be THT and [T]QFPs/QFNs, along with maybe SMD LEDs and the other components you mentioned?
One of my current projects uses THT so SMDs are kinda useless to me for now but I was just wondering about it.
File: IMG_20171114_001109.jpg (1.05 MB, 3120x3120)
1.05 MB
1.05 MB JPG
I order a lot of components through AliExpress, and a good 90% of the time they come in sicstock.com packaging. I tried going to the website to check it out, and I keep getting a connection refused page. Anyone else having that problem? Is it because I'm an amerifag?
Basically correct. If you've got the microscope you could look at the smaller devices like SOT23 and SC70 and try to look them up, and possibly but unlikely you'll find a PIC10F20x, a micro-ADC, or something similarly hi-spec that way. It's a safe bet that SOT23-(5|6) devices will be voltage regulators of some flavor, maybe 74*1G logic.
BGAs could also be worth the effort, but you'd need the appropriate stuff to reball them.
File: IMG_1198.jpg (1.27 MB, 3264x2448)
1.27 MB
1.27 MB JPG
I bought this transformer for a few dollars and it's pretty ancient looking. It's 250V in, and has two pairs of 0-6-9V outputs, which I plan on putting in series and rectifying to get a ~25VDC power supply. Any idea how much power I can draw through a transformer of this size without overheating it? 50W? I might buy a thermistor to be safe, but knowing what similar-sized transformers can do would be a help.

I'll probably hook up some array of switches so I can have a DC power supply with a ground of various amplitudes.
Okay, thanks for the help. I should really get me some sectioned jewel boxes to sort all kinds of shit I might use later, but at least I'm off to a good start.
Also, how do you salvage coils? The moment you apply some pressure to them, they get squished and it's hard if not impossible to get them back to their original state
Sadly I have no hot air station yet and I would have to be incredibly stupid to shove a PCB in the microwave as that's the only thing I have to heat stuff other than the stove.
No oven? I bet you could whip something up with nichrome/propane and an electric motor to whack the side of the PCB every couple of seconds, that might be interesting. Now I've got another shitty project idea, hehe...
Curiously, I've never once seen that, after buying hundreds of dollars in components through aliexpress. It could be that I'm buying in smaller lots. Same/different vendors?

>sectioned jewel boxes
Probably a good idea, but you may find that simply buying sample packs of SMD components would be a better investment of resources.
>how do you salvage coils
Unless you have an LRC meter, I wouldn't. They're literally just wire.

Enjoy your delicious chemicals next time you want to bake something.
>They're literally just wire.
>someone who's never had to wind a 30mH inductor
Yeah... To be quite honest I wouldn't feel any comfortable putting a PCB in the oven, then baking some potatoes or cookies in it.
Yeah, exactly what I thought, but I might someday get lucky and find some of those slightly expensive components, such as low wattage potentiometers or the like.
Thanks for all the pointers
look at the wire gauge on the output to get a hint.

also, do you get 18VAC across the 9s?
>air core
and one more thing, how many ohms from 6 to 9, and from 0 to 6?
I've bought components from different vendors, and they come like that (mostly caps and resistors). I'm starting to think that there's really only a few vendors and they all have 20-30 stores, hence the very similar prices and sometimes the very same pictures/descriptions. Either that or they all are middle-middle men, all drop shipping with the same place. I've gotten some really good deals only to find the item not available anymore a few weeks later when I want to get more, only to find a cloned listing from another vendor at the exact same price.
File: IMG_1199.jpg (1.03 MB, 3264x2448)
1.03 MB
1.03 MB JPG
The wire is 0.5mm wide or so, which corresponds to 24 gauge, or 3.5A ampacity at 60°C or so, but that assumes open air not all coiled up.

The primary is 160Ω, and isolated from the outputs as expected. The two sets of (0, 6, 9) are isolated from one another. The resistance between 0V and 6V appears to be around 0.4Ω with a rudimentary correction for lead resistance, but that is with a shitty multimeter. The resistance from 6V to 9V is around 0.1Ω, and these measurements are similar on both sides. I'd trust them to within 0.4Ω or so since the value does hop around a little.

What's the chances that there's a thermal fuse in here? Before you ask yes I already looked for a datasheet and there isn't one, but here's the box if anyone else feels the need to look for one. The code is "R133" or "TR133" (first is on the transformer and the second is on the box), so that might be a more generic identifier.
That's probably about right, or there are just a shit-ton of resellers reselling what's locally available, chopping up reels, etc. I found a vendor that split up my order of like 8 items into like 5 packages, so I may be about to experience that pleasure myself. (Joke's on them, I paid a total of $2 shipping for that order)
I generally order components in pretty small lots (like 100-200 pieces per type), mostly semiconductors or basic component assortments, and usually they just come in cheap poly bags on cut tape.
OK, so it looks like your 6Vs are taps, which limits your total power out if you wanted to use both 6 and 9. But yeah, the 9s will stack, and you'll get a little less out than 18*1.4=25.2 after you've factored in diode drops and how much ripple you want to smooth out for the load you're expecting.

you can plan to dump some heat where the transformer mounts to a case. kinda doubt a thermal fuse, but who nose.
So father has this old foken power supply to charge car batteries. Has this switch to 6-12-24V.
Says it stopped working.
K, checked diodes, one contact was disconnected. Solder it on. Turn power supply, turn switch to 12V, check output voltage says 6V dc. Turn it to 24V ,multimeter says about 12V dc output.
So I switch my multimeter to ac, swith supply to 12V and multimeter shows 14V ac.
Is something wrong with my multimeter?
Checked all diodes, ~+0.500.

Gonna disassemble that 6-12-24V switch and check it.
But, Where is AC coming from?
Probably either a blown cap, or another loose connection somewhere. Draw a schematic, measure voltages, plot on schematic (both ac and dc in this case), and something should jump out at you.
The symbol on the right side of that image doesn't really exist in electronics. I don't understand why it's listed under "basic electronics symbols"
File: tx-sr707.jpg (571 KB, 2000x1333)
571 KB
571 KB JPG
I'm excited, I just got this Onkyo TX-SR707 running brain-free. This series is really simple to analogue convert, you basically just need to bridge out the relays on the amp board and inject a signal into the shielded cables leading to the power amp, and you're done. It even has an analogue protection circuit for you to use, no processor required.

Oh yes.
Those chargers often simply don't even have an output filter cap. Hook it up to a battery and see if you get a current reading on your multimeter. There's a decent chance he's just cooked his battery because those "transformer and diode" chargers are trash.
Yep, this charger is very old, it's just simple transformer with diodes. Dates back to soviet russia and I bet it's older than me.
I looked up internets and yeah, my miltimeter shows what it shouldn't supposed to show.
Simple test: turn multimeter to dc, shows 12V on battery, then turn on ac and it shows 26.4V on battery. Internets say multimeter multiplies dc by 2.2 and shows it as ac.
Anyway, everything is kinda ok now, but I have very big voltage drop. When I measure voltage after diodes I get what it suppose to be. But when I connect on board ammeter and voltage meter back, I get huge 45-50% voltage drop.
There is even some kind of copper plate for ammeter. Probably shunt.
I'm gonna revive this old motherfucker.
File: xxx.png (37 KB, 844x431)
37 KB
I am still having problems with powering my motor with +-3V.
I've got an incoming signal of +-3V (from an Op-Amp) that does not have enough current to power the motor, so I rebuild this H-Bridge, but I didn't know how to connect the four inputs of the bridge, so I tried something out and ended up with two inverted inputs and two normal inputs.
But the motor only turns in one direction, depending on how I hook it up.
I already measured the outcoming voltages (arround 4-5 volts, positive and negative) and have no idea, why the motor does not turn.
File: ssss.jpg (26 KB, 571x264)
26 KB
So I left only ammeter connected and everything is fine.
But once I connect 1 an 2 contacts on pic related I get huge voltage drop.
What is wrong with that device?
File: voltage1.png (56 KB, 603x384)
56 KB
Missed the new bread.
Ive got a question. Will pic related detect zero in 230AC home voltage? if yes how will the voltage betewen A and B look? Is it going to be spike to certain value when zero crossing and then back to zero or is it going to be @ certain value untill zero crossing and sharp drop to 0V followed by come back to certain value?
Anyone in this thread can help?
Reposting from last thread.

I was thinking something more along the line of a CD drive spindle or a laptop hard disk, but they spin at very high PRMs and I'm not sure the speed of the cassette deck would generate a current.
Oh geez, I knew they slipped but I didn't think a small dynamo would generate enough resistance.
I could add a connector to charge the cassettes and have 2 or 3 and have it charge over a cigarrette lighter connector or something. But that sorta beats the purpouse of the project
File: DSC_0283~01.jpg (609 KB, 2160x1900)
609 KB
609 KB JPG
>tfw don't have a pickit3
File: AC FAN.png (19 KB, 699x460)
19 KB

I'm trying to make two 12V PC fans run as long as possible to run a bucket and ice portable AC. I have 4 6v batteries two sets in series and then the two sets in parallel. I'm just trying to double check my work, and I'm willing to accept any suggestions to simplify things.
Short 5V pulse every 10 ms (at 50Hz mains frequency)
File: fans.png (3 KB, 444x204)
3 KB
looks good
>Where is AC coming from?
You're measuring the ripple in the unfiltered output.
The switch selects a different tap on the secondary of the transformer.
Transformer/Diode/Meter battery chargers don't filter the output.
what kind of motor do you use? is it intended to run in both directions?
I am trying out kicad for a new project, I am looking for a specific part that I need is there any website where you can search for parts and get a library with these parts for kicad. I am looking for a HCF4094BEY
It's a XDrive 2025-22.
In my prior tests it did run in both directions, with basically the same signal. After attaching the wheel to the motor I needed more current, so I built the H bridge.
I assume I made an error while reproducing the H bridge?
So i have tried to use a lower resistance MOSFET (TSM4435) in here >>1274462 but it still gets super hot. Around 80C even with fat pcb traces connected to the pins. I dont think this is normal.

The load is pulling 7A, Vgs is 11,2V the internal resistance should be some 21mOhms, 7^2*0.02 = 0.98W, it shouldn't get this hot.

What seems to be the problem?
You need to learn to add new components at some point in any case. Better sooner than later.
So, basically, you are telling me to make one myself
One watt to an SO-8 is quite a lot. You'll need decent heatsinking (PCB with lots of copper) and it will still run hot.
Theres two 20mm long, 5mm wide copper trace on drain and source, it has enough cooling, more than through the package.
File: pinout.jpg (78 KB, 787x509)
78 KB
got my hands on a UCC28C40 and want to experiment a bit with it. first time dealing with this kind of IC

i would like to use it in voltage control mode for now. the datasheet suggests pic related for that, saying it would give me a sawtooth on CS pin by dividing down the sawtooth on RT/CT pin and removing DC bias with the cap

i don't really trust this suggestion though. how is this supposed to work? the sawtooth on RT/CT pin looks fine, the one on CS doesn't look like a sawtooth at all. but to be honest i wouldn't expect a sawtooth from this circuit

am i misunderstanding something here?
File: 1uuKA.png (122 KB, 581x477)
122 KB
122 KB PNG
That's not very much. The usual test patterns have one square inch of copper, or over 3 times more than you used. Pic related.
And even with that much copper it would still run hot.
Right, ill expand it, but im not too confident this heating is normal.

With Junction to Foot (Drain) Thermal Resistance: 22 and Junction to Ambient Thermal Resistance (PCB mounted): 50 it shouldnt raise its temperature by 60C
>i would like to use it in voltage control mode for now. the datasheet suggests pic related for that, saying it would give me a sawtooth on CS pin

perhaps that can be done; I'm not familiar with this chip but am interested now that you've shown it to me. pic related may or may not help, but it seems to say it is an INPUT pin in voltage control mode.
>but im not too confident this heating is normal.

1 watt is about the limit for so-8. they are small, and can only move so much energy out no matter how large the traces are.
ok I'll try that
internal voltmeter may be shorted
Got pics of the circuit you built?

There are some issues with the circuit as drawn there, you can't leave the + input of the opamp floating, connect it to ground. +3V out of the logic isn't enough to turn on your high side MOSFETs, you'll need more like +6.
If you want to get rid of the pulse, stick a capacitor between A and B.
>A voltage ramp can be applied to this pin to run the device with a voltage mode control configuration
this is what i am trying to achieve, but without using an external ramp. the idea is to use the already existing sawtooth on pin RT/CT
Sorry about the floating input, the program just assumes, that it's connected to ground, so I didn't bother.
And the circuit is for a project at my university, so I don't think that I'm allowed to post it online (although I doubt that anyone would check /ohm/).
I gotta test some more tomorrow.
Do you mean by logic the gate input or the Vref I am using.
Thanks for your help!
So long as the cap is sufficiently large it should give a sawtooth, the transistor forms a voltage follower, the cap+resistors form a high pass, which should have a corner frequency well below your desired switching frequency.

Current mode control is generally easier with those chips than voltage mode.
File: H-Bridge[1].jpg (22 KB, 639x417)
22 KB
Your +3/presumably -3 input on the left won't turn on your top pair of MOSFETs, the gate has to be a few volts above the source pin which you want to be sitting at 3V, so you'd need a +6 or higher supply for both the opamp and whatever's creating the +3/-3 input. To get around this, can use a P channel MOSFET on the high side instead of an N channel, that way you can use your -3V to turn it on.
>Current mode control is generally easier with those chips than voltage mode.
yes i read about that, but i'd like to experiment with both modes

>So long as the cap is sufficiently large it should give a sawtooth
i thought about that aswell, but wouldn't it eventually turn into something like i posted?
File: saw.jpg (32 KB, 584x444)
32 KB
May need a resistor on the emitter to ground to make sure it stays in voltage follower mode. The ESD diode on CS should provide clamping to set the negative limit on the sawtooth, should look something like this. Not starting from zero will limit your minimum usable duty cycle, if you want better, make a nicer sawtooth gen possibly with opamp, or just use current mode.
File: sepic.jpg (52 KB, 1106x464)
52 KB
I need a circuit that takes in 3(or less)-40 volts, and puts out 12v at a max current of 550 ma.

What is the cheapest circuit that can do this reliably?
I started with the LT8302 chip to make a fly back, but found its Nps with my voltage range value only allowed me a 1:1 transformer, which I'm thinking is pointless.
I then used TI webench to make circuit in pic related.
Anyone have any better ideas?

How do I prototype and build with surface mount chips? Do I need to special order a board with pre printed leads from a software like Eagle?
Firstly, why do you need such a low input voltage? That'll be your limiting factor for most topologies.

Something like this may work http://www.ti.com/lit/ds/symlink/lm5118.pdf as your controller, allows a pretty wide input range.

For relatively large surface mount stuff down to SOIC packages, 0805/0603 passives, using a veroboard + get creative with your knife can be good enough, for smaller packages you really want a custom printed board. If you're not in a hurry can get decent quality double sided boards from china for around $15 shipped for 10x10cm through PCBWay.
I'm building a charging circuit for a 12v battery, which powers a device that runs at 12v 550ma Max. The power is coming from a 2000:5 current transformer that ranges from 69k-500kv at 100-1000A. becasue of this I have to choose the lowest possible burden resistance(which looks to be 13 ohms) to minimize the power rating of that resistor.
That then goes through a full bridge and a smoothing cap into the converter.

The project needs to be build pretty soon, its due by may, but I want to have it build and tested way before then so I have time to put together my reports on the design and process taken to make it.
I should have read that over before sending.
The CT is clipped onto a High voltage power line with those voltage and current ranges.

I may just connect straight to the devise if I can get this all to work consistently enough.
File: 1510250629166.jpg (79 KB, 750x750)
79 KB
Things like Pic related work for some cases. I probably wouldn't try to use one for a switching power supply because they really need short, fat traces to work well.

>minimize the power rating of that resistor
y tho? It appears that power is free in this application.
How about building a dynamic shunt-type regulator with a hefty MOSFET and a voltage divider on the gate?

probably your onboard voltmeter is really an ammeter. or you removed the series current limiting resistor, which is what makes a voltmeter a voltmeter.


no good. generally speaking, batteries should not be connected in parallel coz they can have different potencies, and then power is wasted flowing from the stronger to the weaker. to fix, run 1 fan from 2 series batteries, and the other fan from 2 other batteries. adding 2 diodes would also do the job, but it also wastes some power.
Well actually A is at 5V always except when the 10ms peaks come through, since it's tied to 5V whenever the phototransistor isn't on. B is always 0V because it's grounded, obviously.

nope. when the AC waveform is near zero, the LEDs go off and the transistor goes off, producing a 5V pulse on A. for most of the cycle, the transistor is on so output is around 0V.
I wired a case fan to a wall wart so I could make a laptop fan.

However, it runs like shit. When I first plug it in, all the fan does is buzz, it won't spin until I start spinning it and even then it spins slowly in a jittery manner.

the wall wart is DC, outputs 12 volts and has sufficient amperage, so I don't think that's it

have I somehow burnt out the motor in the fan?

also if I wanted to insert a switch into the mix, is there some sort of voltage limitation to switches that I should know about, or should it be fine since it's outputting only 12 volts?
That's a much more interesting problem then. First thing to consider is your assumption that current transformer output across a resistor makes a voltage source won't really be valid unless your resistor power is much larger than load power.

Treating your input as current source gives more interesting options since you can effectively control your input voltage to be whatever you need. Battery as output also makes things easier as you've effectively got a voltage sink/source.

I'd do a boost converter, or if you need isolation for measurement, flyback.

your wallwart isnt happy driving an inductive load. try a different one, preferably an old heavy non-switching one.

just about any switch will handle 12V at 500mA unless it's a really tiny pushbutton one. anyway, they often have their ratings written on them.
>is there some sort of voltage limitation to switches
yes but a fan doesn't pull much current so anything bigger than a tiny microswitch should be fine
It's an ancient wall wart from the late 90's. I'll just have to dig through my box of bullshit and find another one then
Hmm, dried-up caps maybe.
File: 1509941875886.jpg (23 KB, 368x250)
23 KB
The burden resistor has to have a Watt rating 2 times higher than expected. So with a 2000:5 Current Transformer the output current range is .25 to 2.5.
The Max power through a 15 ohm burden is ~90W, so I need a burden at 200W rated.
If the Burden was only 10 ohms I wouldn't need such a high rated resistor.
However, I cannot find a converter, of any type, that has a range which is lower that 3 volts.

I could design the burden to spit out any value voltage. But, because of the current being 100-1000A the output voltage is going to be huge. Always having a Vmax that is 10x the vin. So, 3-30, 3.5-35, 4-40.

Really, I'm running out of ideas.
Our range is huge.
I wish I could just buy a 4-50 volt buck boost that was reliable. Everything I see is Chinese garbage that is way overrated for what it actually can do
I'm not too sure how optocouplers react. I assumed that only near the peak they'd be producing enough light to be picked up by the phototransistor, but I guess you'd get a very distorted sinusoid as the output, with the phototransistor acting in the linear region for some part of it.

Use a capacitor to either smooth out the switching or change the power factor. If you've got a 2-channel oscilloscope it should be a piece of cake.
Well we only need 7W out at max, and with a 2000:5 CT and an Rburden of 13, I get a min power output across the burden of ~9W. So I'm good there.
As I said in here >>1278435, the range is always (x to 10*x).
There's voltage and amperage limits on all switches.

Maybe your wire connection sucks. Did you solder or use a wire connector? Or did you just twist it together

Are you sure its a dc case fan? It should be written on the fan too
I miscalculated that.
Its ZERO point 9 Watts. not 9 watts.

Which makes this even harder.
File: 1505379558529.png (7 KB, 754x443)
7 KB
Okay, how about something like this? R2 is a static load resistor. Q2, a small-signal NMOS, switches on Q1, a power PMOS, when the input voltage is too high according to R3, R6, and Q2's Vgs rating. Q1 shunts current through a secondary load resistor R5. R4 provides hysteresis to keep Q2 out of the linear region. That should get your input range down to a more manageable factor of 3, anyway. Add a resistor and/or zener on Q2's drain and/or gate if exceeding Vgs(max) due to spikes etc. is a risk.
(whoops forgot rectifier on CT)
is super glue electrically conductive?
I was doing maintenance on my laptop and as I was screwing it back up, the speaker cable got caught between a case screw and the case so when I screwed the screw in it pierced the cable clading and shorted through the screw

I used cyanoacrylate to patch the little cut/hole in the wire, but I want to make sure it's not going to spontaneously ignite or just short through it
0.25A into 13 ohm parallel with your load gives a maximum power into load of 0.2W. To get any decent power you need to be in series (or series-parallel hybrid) with your power resistor, not in parallel. Or a much higher value resistor.
also, at max power you're only getting 1.625V at the input.
I like this circuit. R2 is still way too low if they want 6.6W into load at 0.25A in, for that it'd need to be at least 422 ohm.
Wow, I obviously have along way to go in electronics. Even as a Senior in a decent university I only vaguely understand what you just described. I know what each of those components are, but am clueless on how they interact.

Since you posted this I've been souring my textbooks looking for what hysteresis is, and what the linear region is.
I get that R2 is the burden, but at a certian voltage the Q2 is turned on(regulated by the voltage division between r3 and r6?), this then shorts out R2 and runs the voltage through R5, making that the new burden resistor.

Like what is said here, the R2 value would have to be pretty high to get the power we need out at the low end.
I could use a 3000:5 or 1000:5 CT so it doesn't necessarily need to be 400, but it would be high.
I have Pspice at school, I'll play around with it on there tomorrow.

How do I go about choosing real life Mosfets to use in the circuit? I'd assume Q1 would have to have a high power rating.

What software did you use to model this?
How did you learn how to do this?
No, but smudging shit with CA glue does not really insulate it either, as the coating is typically thin and brittle.
Right, just an illustration based on original anon's numbers.

Almost. R2 remains in circuit. Q1 just adds R5+Rds(on)(Q1) in parallel once Q2 turns on. (remember Kirchhoff's current law?) R4 pulls the control voltage a little higher when Q1 is on, and a little lower when Q1 is off, in order to keep the circuit from oscillating or chattering. You'll need to choose your own values for all the Rs, of course.
Select Q1 for a low Rds(on), a Vgs(threshold) somewhat less than the lowest acceptable out voltage. If Vgs(max)(Q1) is less than 150% of max output, you should probably add a resistor in series with Q2's drain to avoid overstress. Q1 probably won't need massive power handling capabilities, as R5 should dissipate the vast majority of the power, but select for a high enough drain current to handle the minimum load condition.
Q2 is less critical, but I'd personally select for a Vgs(th) around 2-2.5V so that R6 can be fairly large and R4+R6 won't affect the load significantly.
I'm the guy who bought a handful of electret mics a while ago to mess with their electrets and never got around to it. Just soldering them to a TRRS cable now, and i've noticed they're much quieter than my TRRS earphones' mic is even when the earphone mic is three times further away than this one. The audio quality is also worse when I amplify it, but that's just due to noise. Is this because I have an inadequate bias voltage? I'll be building a rig to mess about with audio and electret microphones soon so that's something I'll test, probably with a standard adjustable voltage regulator or opamp.

Anyone know how good sound quality picked up from a small 8Ω speaker is if amplified?
I'll be pulling 20W away from a TO-220 MOSFET into a 2.5W/C heat sink, and the transistor shouldn't be allowed to get above 60C. Should I have to use a fan?
Depends on your ambient temperature, of course, and whether the heatsink is in free air or is otherwise constricted.
You sure? Heatsinks are usually rated in degrees/watt. Assuming you mean 2.5 degreesC/W, yes, you probably will need a fan to keep below 60degC. Assuming you mean 0.4 degreesC/W, it's conceivable you might need a fan on a hot day.
Shit, yeah 2.5 C/W, it will be in open air indoors. Thanks for the input.
Fortunately, you really don't need much airflow to keep that thing cool. A very slow fan would easily be enough.
you'll have another ~2C/W of thermal resistance in the fet and the interface between the drain tab and the heatsink. it depends on the part and the compound/thermal pad/torque used but you won't get much lower than that. so (2.5C/W+2C/W)*20W+30C = 120C at the junction. that's okay for a 175C rated fet. if the airflow gets it down to 1C/W you'd still be looking at 90C junction which is a fine margin unless you have some special reliability concerns. did you mean the temperature at the tab?
Well, I have a few troubles.

I hav some bad SRAM memories, but I need to test them first to know how many purchase, but well, I have no idea how this memories (and every memory, I suck with digital stuff) works. I tried to make a code in arduino, and well, it does not work.

Somebody give me light to this please, or only explain how this memory works, in teach-to-a-dumbass mode. The model of the memories are TC5546P-25.

Yeah, my english sucks too
>The model of the memories are TC5546P-25.

do you have a datasheet showing the pinout and timing diagrams? all I can find online are TC55416P-25 which could be different.
Shit, I failed.

If your arduino has 23 GPIO pins you should be able to do this.
You sure it isn't TC55465, which looks like a rather normal 64k x 4 fast SRAM?
Considering they're in sockets, I'd check the sockets first. They're much less reliable than the memories.

Async ram works usually like this:
Write: pull OE, WE and CE high. Enter address, enter data, pull CE low, pulse WE low, return CE high.
Read: pull OE, WE and CE high. Enter address, pull CE low, pull OE low, read data, return OE and CE high.
File: 526847173280.png (23 KB, 859x526)
23 KB
Will this work for driving a dc motor rated 3V 1A?

and here's the timing for the write that you describe, which should also work with an arduino that has 5 volt GPIO pins.
Kinda badly. You should drive the transistor harder (with 1k or so resistor) and even then you can expect the transistor to waste quarter of your supply voltage.
I have an Arduino Mega.

I'll try to write something knowing that.

The entire board looks pretty well, and sockets too, but I'll clean all of them.

If anyone has curiosity, it is from an old Video Analyzer. A Tektronix VM-700A.
>I have an Arduino Mega.

And you don't have to use all 16 address lines unless you want to do a thorough test; you could tie as many as you like high or low or manually do that. Once you can read and write one memory location you have proved that the chip is not totally dead. If you cannot do that then that chip needs no further testing.
As a first test would be good, but I need to test it entirely. I don't know how this old analyzer would run with a few bad words on the RAM.

Then, I have another problem. In the board, with the SRAMs, are other ICs, with serial number MB81C75-25P. I don't know what they are, because i don't find any datasheet.
File: TIP120sat.png (30 KB, 400x295)
30 KB
You would need at least 4 mA base current. TIP120 is not optimal for your application, there are better solutions.
>, but I need to test it entirely.

so if you have 23 GPIO that will be easy. If you don't have that many then manually set one or more address lines and run your diagnostic as many times as needed to check the entire address space.

> MB81C75-25P

that number shows up but I can't find any information.
I've tried your solution, but it did not work.
Same goes for a higher input, but the highest I could get was ~5V, so that might be the problem.
But thanks anyway!
I think I'll leave it as it is (just turning in one way), which solves the problem, but not as nicely.
How have you deduced there's something wrong with the RAM chips?
It is also highly unlikely to have just few bad locations.

That's a 16k x 4 fast SRAM, very similar to the other chip. Cypress' 7C166-25 is equivalent.
The analyzer gives a RAM error when starts, and when it is running a minute more or less, it restarts itself. This SRAM is from an extra module, so it isn't absolutely necesary to the analyzer.

Also, thanks a lot for the equivalent. Now that I know it is a RAM too, and with the pinout, i'll try to test it too.
>it did not work
It cannot, the circuit is self-defeating.
File: 687468851775576752.jpg (96 KB, 853x1280)
96 KB
Hi, so I'm trying this lamp proyect, and I got this set of old led lights and I grabbed 25 of them, then I'm attach it directly to the electricity wire... I just want to know if there's anything missing between the lights and the electricity cord. I watched a few vids of people setting up led lights and they used some "diodes", but I'm really not wisey on the matter.

I also noticed, there are different type of led light wires. The lights on the picture use a "thicker" wire (8 copper thin wires) and there are cheaper onces (they use around 4 copper thin wires).

I mention this cus I tried the cheaper and had a hard time with the bulbs, they burned every 10 secs. And I would like to avoid that happening with the ones I'm going to use.

I will deeply appreciate your advise.
I tried to make a program, but it does something strange. I'm not a lot into programming, so this code could be a piece of shit.

So, when I read the output of the pin, it is always high, even if I remove the pins, which is something absurd.

This is the .ino
File: IMG_20171115_211951.jpg (3.54 MB, 5248x2952)
3.54 MB
3.54 MB JPG
Maybe wrong place, but how do I make this Chinese spinning light silent? I'd happily just have it spin at 1% it's speed
cut either the long red or black wire in the middle, strip each end, connect a resistor in between
make it a potentiometer instead if you want to have a way to change the speed

what is a rar file. can you put the text file on pastebin?
>what is a rar file
>on 4chan
>in /ohm/
are you real?
>are you real?

nah, I just don't want to download unrar to read a text file.
File: IMG_20171115_215059.jpg (3.46 MB, 5248x2952)
3.46 MB
3.46 MB JPG
Good call anon! Let's hope I don't cause a housefire
Well, it turns out that I somehow messed up the wiring and connected the -3V supply to the output.
I kinda feel ashamed.

Good to know, thanks!
>linear potentiometer
Man those things are cool.
Okay wondering if one of you dudes can help me out.

I'm trying to develop a method re-using waste heat from a CPU (using a thermoelectric generator), to power the cooling mechanism such as a thermoelectric cooler or a standard fan.

My initial problem is testing typical CPU power consumption at various heats (creating a temperature vs watts), to either confirm or deny the idea that CPU's get less efficient (use more energy) when they are heated above around the 70'C mark.

How would I go about getting a precise reading for the CPU power consumption?
I butchered a pair of old busted headphones for them, I kinda regret it but I got 4 potentiometers
I say. I don't know how to write a program. This is a copy from a other project with a lot of modifications.
The idea is: I generate a binary number, that in each loop increase by 1, so I'm testing each address, but well, as I said, I'm not a lot into digital stuff. It's only a hobby.

awesome. unfortunately I have to go offline now. If no one replies we can look at this tomorrow.

thanks for the pastebin.
I don't know anything about arduino but this code is wrong because it's going out of bounds
for ( i = 29; i < 45; i++) pinMode(Address[i], OUTPUT);

you probably want this instead based upon the array at the top
for ( i = 30; i < 46; i++) pinMode(i, OUTPUT);
The problem is that by putting a peltier generator between the CPU and its heat sink you're essentially putting a thermal resistor in the way of the CPU's heat, so you won't be able to run it as fast without it overheating. To use the same power for heating will almost certainly be less efficient, though with a fan I'm not quite so sure.

Anyway, provided your thermoelectric generator can be used as a cooler as well, just put it between the heat sink and the CPU with its cold side against the CPU, and power it with different voltages, starting at the maximum. If you don't have a thermometer you'll need to get one, preferably not an IR/laser one because they can jump around a little. Though if you know your shit you can probably just get the temperature reading from the CPU itself. Set up some default task and wait for the CPU to reach thermal equilibrium, and note the temperature and performance by TFLOPS or whatever. Then change the voltage and repeat. Using different resistors in series with the TEG/TEC works instead of different voltages, but they might get a little hot. You should get a nice graph of some kind, but as far as I know the CPU only slows down at higher temperatures because it needs to, as a way to avoid the increasing propagation delays giving false logical results, rather than doing so naturally. This means you'll probably just get two straight lines on the graph, one flat and the other a steady decline as temperature increases.
Thanks man, yeah I get that completely about the thermal resistance. I need to model the relationship between temperature and increase in power consumption, then relate that to the potential power gain of a TEG from that deltaT.

It's for a dissertation so if I can get a good relationship modeled then work on ways of effectively implementing this TEG/TEC I hope that could give some useful results.
Also in general I kinda like (in theory) the idea of the TEG/TEC combo, you've essentially got a greater cooling power available as the temperature rises. If optimised this could be a self regulating cooling mechanism, ideally self sufficient but so far I doubt it as the potential power gain for this delta T is only a few mW
Well you could use a BJT to amplify the current from the TEG to power a TEC or two, not like power usage is an issue. But a TEG+TEC couple but itself is just going to make more heat.
Nice perpetual motion machine you got there.
Well you're getting the energy from the heat of the CPU vs the surroundings, and using this energy to cool the CPU

I don't really understand how it would produce more heat >>1279016
It's the thermodynamic principle of a heat engine/refrigerator. Have a look at this page:
A thermoelectric cooler is a heat pump that moves heat from one side of it to the other, and is at best 15% efficient. Meaning for every 1W of heat moved to its hot side, you're also pumping 6.7W of electrical heat into the hot side. As you might imagine, 7.7W will be harder for the heat sink to dissipate than the 1W that the CPU is making. Unless you have enough heat sink power capacity to spare, thermoelectric coolers aren't worth it. The way to increase the heat sink power capacity of course is with forced convection, i.e. a bunch of fans. But all in all using the TEG will stifle the flow of heat much more than the gains you'd get out of the extra cooling, so I'd just stick to the TEC and some fans, unless you can manage to get a TEG-powered fan to be more thermally efficient.
File: IMG_1200.jpg (143 KB, 785x1029)
143 KB
143 KB JPG
Some eyecandy for y'all.
File: 2n1302.jpg (105 KB, 800x483)
105 KB
105 KB JPG

i see your 2n3053, and raise you some 50-yo geraniums, still in working order.
That's damn old, in contrast I bought the 2N3053s new. No idea why they keep them in stock, but I figure there's nothing wrong with a few more obscure packages. I had a need for a linear regulator that runs at either Vcc or 12V, whichever is lower, and I'm not too sure if normal linear regulators can do that well, so I'll just use this and a 12V zener.
File: 1502304307796.png (154 KB, 858x897)
154 KB
154 KB PNG
>I'm not too sure if normal linear regulators can do that well
You'll be losing one Vbe with the usual emitter-output design, at least.
There are low-dropout (often less than one Vbe) linear regulators. ON's KA278R12 came up in a ddg search.
I know about low dropout regulators, I'm just not sure what a 12V regulator will output if you give it 12VDC, or even 8VDC input. I guess I could test that.
File: Mos Circ.jpg (394 KB, 3024x4032)
394 KB
394 KB JPG
If you are still around I've gotten a lot of progress and wanted to run it by you.
I set R2 at 120 ohms, which puts the minimum power outup ( @ 0.25A from CT) at 7.5W.
I then found that when R2 is operating at 50W the current out of the CT is 0.65A.
I then Assume R6 at 100 times larger that R2, making it 12k, which makes R3 360k in order to produce a 2.5V drop across R6 that would satisfy Q2's threshold voltage.

I then assume the Rds value of Q2 is essentially so low it acts like a wire grounding the gate of Q1.
From what I've real, grounding the gate of a PMOS is what allows source to drain flow. Correct me if I;m wrong.
So R1 is just a pull up resistor. I should set that to 10k? That's what I've always used for pull up resistors.

I Then found a Mosfet with a -2 to -4 Vgs(th) value and a Rds(on) value of 0.117.
I then chose R5 to be 20 ohms, which when added to 0.117 ohms and then in parallel with R2 gives and equivalent resistance of 17 ohms, which gives ~7W with an T on 0.65, and 107W at our max current of 2.5A.

I have no idea what value to give R4. And when I try to actually model this on Pspice it doesn't want to run it.
check out?
is this solder flux what I should be buying? if not what should I be using instead?


I usually just use flux core solder but everyone else seems to dip all their things in actual goop first in addition to that
>everyone else seems to dip all their things in actual goop

not everyone! the only time i've seen something like that is when hand-soldering SMD chips using a dragging motion and a big-ass tip.

>is this solder flux what I should be buying?

that's solder paste, i.e. liquid. solder. search aliexpress for ''Soldering Rosin Flux Pen'' if you like pens, or bottles if that's your pleasure. personally, i wouldnt buy something critical like that unless it had a brand name, like Kester. who knows what the chinese stuff is made from; could be crushed cockroach shells.
File: 1505923057391.png (34 KB, 1512x684)
34 KB
Typically, they'll output Vin - Vdropout.

Is your PMOS backwards or just a drafting error?
I get good results with 100k. Its main concern is pulling charge back off of Q2's gate fairly quickly.
Q1's drain will be at 0V when Q1 is off and close to the rail voltage when Q1 is on. You want (R6||R4) / ((R6||R4)+R3) to be slightly more than Q2 threshold when Q1 is off at max tolerable output voltage and R6 / (R6+(R3||R4)) to be slightly less than Q2 threshold when Q1 is on at min tolerable output voltage. You may need to adjust R6 and/or R3 to achieve that. See Pic related, voltage source is a triangle 0-100V, to sort of approximate your CT. Left scope is Q2 gate, middle is Q1 gate, right is output.
The turn-on and turn-off aren't as sharp as I expected. You can add a series-cap-and-resistor in parallel with R4 to sharpen the switching action. I used C=3n3 and R=220k and still got a reasonable simulation. You may wish to play with these values when breadboarding, or if your simulator can generate current waveforms.
The circuit can have trouble converging because simulators don't handle positive feedback perfectly well. Try establishing an initial state of off, with the node on the gate of Q1 set to the initial level of Vin. And/or decrease your time step.
Remember, the goal of this circuit is just to reduce input variation to something a higher-quality post-regulator can handle. With values in pic I get something like a 33V max before the second shunt activates.

Mechanic is a brand name, and apparently a fairly reputable one in the east. Who knows whether your Kester is actually Kester, gnomesayin?
I heard that pens aren't that good. Personally I just have a tub full of semi-solid (jelly-like and melts at like 80°C) rosin flux and apply it with whatever trimmed component lead I have lying about, but it's a bit like painting a model plane with a paint roller. Something like a flux pen might actually be what I need, but I'm kind of wanting to use no-clean flux instead since rosin leaves a sticky mess.
File: 1498960711017.jpg (37 KB, 1000x1000)
37 KB
Pens are god-tier.
why do they still teach people still use darlingtons for hard switching
Those things are no-clean? I need one.
I didn't mention, but this pen is refillable
The thing about no-clean is that it's not the most active flux. It's nice when you have fresh, clean components and boards. Otherwise, you might want something a bit more active which you could just clean off with alcohol and a totohbrush.
I do find the solder flows really beautifully when I slather the board in flux, but I haven't found the ideal cleaner yet. That canned stuff evaporates too quickly. A bottle of isopropyl alcohol is probably what I should be going for, maybe with a spray attachment.
The pro thing to do is to give it an ultrasonic bath, shit's cash.
They're a little expensive looking. Could you make your own with a handful of transducers and a bucket?
I get what you say. Anyway, I'll try some programs to only read the output without the array, and try to fix the problem of reading HIGH. Really, it has me worried. The arduino is a chinese one, so I don't trust a lot in it. I'll test it in some arduino UNO.

Really, thank you for your help.
File: usb_dick.jpg (369 KB, 2776x1160)
369 KB
369 KB JPG
I have this usb stick that fell of my pocket a couple times and this component broke from the pcb. I can solder it but i cannot find the component online and i do not know if there is any polarity that i should be aware of.
Can anyone help?
The component reads kts12.0A2J
>evaporates too quickly
The runoff is supposed to carry the dirt with it.
Also, toothbrushes.

It's a crystal. Nonpolarized. It's possible it might have broken in the fall and needs to be replaced, but it's an easy part to find.
>It's a crystal
Tnx anon.
I found some traces in the old solder that were in the one side and not on the other. I did the soldering and it werks.
My laptop charger cable was very frayed near the plug that goes into the laptop, so I cut it shorter and resoldered the inside and outside of the coax onto the plug. I had to remove the ferrite choke because it was very close to the frayed part.

The charger seems to work fine now, but the transformer makes a buzzing sound like the noise from a computer working on something whenever I move the mouse or do anything else on the laptop. Should I be worried about this or is it some harmless feedback?
File: 400350000.jpg (16 KB, 367x388)
16 KB
>I had to remove the ferrite choke
>now, but the transformer makes a buzzing sound
gee, I wonder if those could be related...
Literally not what I was asking about. I'd like to know if I can keep using this thing.
It's harmless feedback.

first of all, things like IO1 are set to permanent values, 46 in the case of IO1, which presumably is the pin number you are using for IO1.

your digital read looks wrong:

digitalRead(IO1); this will read the pin and do nothing with what was read

val1 = digitalRead(IO1); this stores the state of the pin in "val1".

I have no idea what this does:

if((IO1, LOW) & (IO2, LOW) & (IO3, LOW) & (IO4, LOW))

C++ is not what I'm familiar with, but you should be able to do something like this:

If ((val1 == LOW) && (val2 == LOW) && (val3 == LOW) && (val4 == LOW)) {

then they are all low and it's good.

} else {

something is not correct


also it would be better to use variables for the four data bits that you write to the sram.

out1 = 1; (or 0)
out2, 3, 4 set to whatever

then after you read, compare to that: if (val1 == out1) etc.

this way you can easily write 0000 or 1111 or 0101 or whatever you like. some ram tests write all kinds of combinations, but for starters I'd just use 0000 and 1111 to confirm that every cell can hold high and low values.
File: 1507433569781.jpg (77 KB, 585x953)
77 KB
Great. I made it work.
I were all day testing other programs, and I finally find the error in de if with the == and &&. I put the "val" and "out" variables that you mencioned and it is plenty functional, coulding change the data without effort.

Now, it only left make it smart, and make it tell me what address is wrong, but as a start, it's very good.

Thanks a lot mate.
This is the code.

I'll make a post in some forum, and upload it when it is finnished.

I have a power supply that needs its capacitors 820μF 25V replaced.
The capacitors are barrels with a + on top.
I have several others laying around, e.g. some 2200μF 16V with a k on the top and I was wondering if I could replace them with those.
The psu is from a display, so it reaches 50w at most.
>if I could replace them with those
Of course you could if you want to experience a nice explosion..
820uF/25V is available and 1000uF/25V would do as well.
>The psu is from a display, so it reaches 50w at most.
Why would that matter?
File: CT-12V.png (5 KB, 535x175)
5 KB
I would keep is simple and robust.

The fancy switching action only shifts the problems to the output section and will need a lot of experimentation because of the poorly defined Vth (tolerances) of the MOSFETs.

A 30V shunt regulator is as simple as a 30V to 12V step-down converter. At the low end (0.25A) the shunt regulator barely loads the CT at all and at the high end (2.5A) it has to dissipate about 75W which is no problem for a few 3055s on a heat sink with some simple circuitry around.

The choice of 30V is just a quick guess from the power balance with 0,55A out from 0.25A in. In practice you will adjust this voltage according to the efficiency of the step-down converter. For 88% it would be 30V/75W and for 75% it would be 35V/88W max shunt power, but only at (and during) 1kA.

The optional crow bar serves as life insurance...
so what exactly does the maximum current specification on a charger (e.g. phone, laptop) mean?

a) when delivering more current, the desired output voltage can't be guaranteed
b) voltage can be supplied, but components might take damage
c) possibly both
Normally it means the maximum load current at the specified output voltage. With no load the output voltage stays almost the same. A good PSU has overload and short circuit protection which means the voltage decreases to safe values.

The situation is a bit different with super-cheap phone chargers, here it may mean output voltage at specified current. I had such a thing that said 5V at 500mA. Without load it showed 9V and 4.5V on a 10Ω load.
What is the purpose od the new 50k ohm resistor(R7)?
If R1 is a pull up resistor, why would that be needed.
Doesnt grounding the gate of the Pmos turn it on?
Or should I meet the threshold voltage using those two as a voltage divider.

I'm having a lot of issue with pspice.
Im going to look into a different program.

I'm using a DC sweep from a current source, it's giving linear outputs lol
Not quite right.
I think load switching is a faulty concept. What can you simulate when the threshold voltage is specified as 2V..4V, a tolerance of 100%? Read >>1279480 for an alternative solution.
can someone identify this broken part on my board? It needs to be replaced but I have no idea what it is or how to do it
It is a diode. I can read 942, so it could be this:


Chek if you can read better the model, and search it on google.
A relay is rated for 2A 220VAC, can i put 10A 12V through it?
never tried it, but i doubt it
File: relay.jpg (29 KB, 339x339)
29 KB
I think so too. This is from a different relay, around the same rating.
Probably a 1N942 12V zener diode, but it might have a letter after the 2 like an A or B. If you can desolder it and read off all characters on it that would help. If that broke while the thing was running, which is likely, then you might well have some other broken components on the board, namely the ICs (maybe relays?) since the lack of the zener will push the supply voltage up at least a couple of volts.

On the assumption that the zener is being used as a voltage regulator for the power supply, you might want to consider bodging a 3-terminal linear voltage regulator in there instead, but it would require a bit of reverse engineering and probably isn't necessary.

Also that's a nice capacitor there.
I'd be worried about a a crowbar or fuse on the cut it with the extremely high possible voltage across the CT when the line is at 500kv.

Would the shunt be able to deliver a constant power that is at least 7W?

Does 30v mean it outputs a constant 30, or if that is a rated max
That's why the crowbar and not a fuse. In case of malfunction all you can do is shortcircuit it all. The shunt regulator is a dynamic load that maintains 30V no matter what input current. It acts like a giant Z-diode but much more precise and easily adjustable. Didn't I outline my calculation? Just check it on your own. Don't forget that simulators are GIGO machines, garbage in - garbage out, but to two decimals.
>high possible voltage across the CT when the line is at 500kv
Wow. A current transformer doesn't care about voltage, only current. But without load the voltage can get very high and possibly destroy the transformer. It is still isolated from the HV side. Do you know how a clamp meter works? That's a current transformer.
Sorry to ruin it, but it almost certainly isn't 1N942. Those are pretty rare, expensive and come in a glass package.
It could be a bi-directional TVS.

Like the other anon said, try to read the remaining text on it.
Found a relay that burns only 0.5W, 40mA, and the LM311 can source 50mA. Gonna try that instead.
Oh yeah, DO-35s. Still, a similar 12V Zener near a bridge rectifier and big filter cap with a resistor next to it isn't unlikely. Not that I've bothered to look up the datasheets to the ICs or consider why there's an inductor there.
whats a good brand for solar panels ? i need to setup a little solar backup system 1kw maybe as a demo for a bigger project, money is not a problem
>why there's an inductor there
because of the 34063 next to it.

All of this is OK, but it needs to be dumbed down.
High voltage and solid state circuits don't mix.
What you want is a two winding 5 to 30 volt type transformer.
Connected to the CT you will operate it in saturation zone.
Then just buck the voltage down to 13.8v and that will change the battery.

I believe the buck will cut off once the battery reaches that level, so you might not need over charging protection.

Put a MOS in parallel with the buck that cuts out voltage spikes.
photovoltaic cells, arrays, panels or whatever you call it, should i just buy amazon best rated ones ? or look for a specific manufacturer?
MOSFET gates fail destructively under too much voltage. R7 keeps the gate voltage of Q1 within its rated Vgs(max).
>dynamic load
If you want to go this route, simply move R4 from the drain of Q1 to the drain of Q2, thus establishing negative feedback instead of positive feedback. You'll need a bigger Q1, possibly several.
Any audio electronic anons in these threads?
A few.
I’m looking at getting into designing my own audio stuff. I’ve made a few kits and I’m reading some electronics books, but I’m trying to find some projects that go into detail or break down the circuits showing what the components of each part of the circuit do. I’m thinking maybe something like a low pass filter or oscillator...anything really that will help me learn.
File: diode VI curve.jpg (12 KB, 526x421)
12 KB

good luck with that. you'll have to figure it out yourself like the rest of us did from reading and experimentation. you'll find that just doing a voltage-vs-current graph on individual components and simple circuits will reveal many of the the secrets that men (there's no women in electronics coz you cant apply the results to your face) have discovered in the last century.
I'll be messing around with some audio stuff too, in particular seeing what happens when you put an analogue signal into digital components. Namely flip-flops to half their frequency, and a schmitt-trigger feeding a differentiator, maybe some other ideas. Mainly to mess around with using sounds from all around me to make surreal music.
"Audio electronics" is stuck in a low-knowledge, cargo cult mindset. You need straight electronics/circuits knowledge, up through some basic frequency response analysis.

What are your current resources?
File: limor-fried-adafruit.jpg (21 KB, 250x350)
21 KB
>(there's no women in electronics
That girl only talks about the new products on their website or puts some prebuild neopixel scripts together
File: maxresdefault.jpg (100 KB, 1280x720)
100 KB
100 KB JPG
File: 1489647931901.jpg (64 KB, 580x386)
64 KB
When you pose for shutterstock

she is an engineer.
>good luck with that. you'll have to figure it out yourself like the rest of us did from reading and experimentation.

So no man had any guidance in electronics, it was all self discovery...is it ok to read books or should I only read manufacturers data sheets?

>you'll find that just doing a voltage-vs-current graph on individual components and simple circuits will reveal many of the the secrets that men have discovered in the last century.

Thanks I feel this might be helpful even though I’m guesssing you are trying not to be.

>I'll be messing around with some audio stuff too, in particular seeing what happens when you put an analogue signal into digital components. Namely flip-flops to half their frequency, and a schmitt-trigger feeding a differentiator, maybe some other ideas. Mainly to mess around with using sounds from all around me to make surreal music.
This sounds interesting. I’d be interested in results even if it’s was just hearing the audio.

>"Audio electronics" is stuck in a low-knowledge, cargo cult mindset. You need straight electronics/circuits knowledge, up through some basic frequency response analysis.
Ok I was hoping I could find some really basic audio circuits to get me going. I don’t really plan on skipping the whole learning/circuits thing was just wanting to learn some basics of how electronics work in audio. I’m interested in audio so thought audio electronics would keep me interested and get me experimenting a bit more. You know simple circuit designs or ic’s that a beginner could understand and build experiment with.

>What are your current resources?
Currently reading the forest mims-getting started in electronics, Make:analog synthesisers and lessons in electronics along with these threads and various other bits from the internet.
I have quite a varied set of electronic components, soldering iron etc now from building guitar pedals and synth modules.
I don’t own a scope yet
>I don’t own a scope yet

I'm sure you know it already, but you pretty much cannot do analog electronics without a scope.

I finally upgraded to the Rigol ds1054z at about $400 and I love it. 4 channels used to seem overkill, but I don't think that anymore.

My only complaint is this: Does anyone know how to save Rigol images without all that menu crap on the sides, or how to hide all that crap when using the scope so it's closer to full-screen?
>I'm sure you know it already, but you pretty much cannot do analog electronics without a scope.
I didn’t know that, desu other than testing circuits some how that is all I know about scopes. What frequency would I need for my application would the ds1054z be ok? I nearly bought that model yesterday but couldn’t work out if it would be ok...kinda guessed £370 worth of scope should probably be fine for what I want, but I’d be gutted if I needed a function for audio circuit testing that it was no good for!
I finally upgraded to the Rigol ds1054z at about $400 and I love it. 4 channels used to seem overkill, but I don't think that anymore.

>My only complaint is this: Does anyone know how to save Rigol images without all that menu crap on the sides, or how to hide all that crap when using the scope so it's closer to full-screen?

I think I read somewhere while reading reviews that unfortunately the menus on the side can’t be hidden...I think it was one of the cons of this scope
Oh I also noticed a lot of places had stopped stocking the ds1054z and was wondering if Rigol were about to release a new scope to replace it?
>What frequency would I need for my application

I haven't been following but if you are doing audio any scope in the world will work. The ds1054z comes standard as 50 Mhz and there used to be a hack where you could unlock twice the ram, 100 Mhz, and some other extra-cost features, but I'm happy with mine as is so I haven't risked it.

>was wondering if Rigol were about to release a new scope to replace it?

I don't know, but someone else probably will.
>I haven't been following but if you are doing audio any scope in the world will work.
Do you mean any scope will work for audio waves analysis through the scope? If so then I understand this or do you mean any scope in the world will be fine to test audio circuits and their components? I’m guessing the former.
>any scope

and by that I mean any serious scope. I think they sell some jokes that only have about 5 khz bandwidth.

>r audio waves analysis

lol I was already clarifying because what I said isn't really accurate. I don't know what you mean by analysis, but audio is pretty much limited to 20 khz so if your scope has 100 khz bandwidth you should be able to do any audio work. But then I'm not an audio expert, or an expert of any type, so maybe someone with more experience will chime in here.
>lol I was already clarifying because what I said isn't really accurate. I don't know what you mean by analysis, but audio is pretty much limited to 20 khz so if your scope has 100 khz bandwidth you should be able to do any audio work. But then I'm not an audio expert, or an expert of any type, so maybe someone with more experience will chime in here.
What I mean by audio analysis (probably not the correct term) is analysis of audio waves so showing audio waveforms on a scope as opposed to analysis of audio circuits and components (electronics).
So just to clarify I’m NOT looking for a scope to view or analise audio waveforms. I’m looking for a scope to test, analise , troubleshoot electronic circuits and components that make up circuits that produce audio. Eg the circuits inside a synth oscillator (pic related)

If it has digital components like DACs and whatnot, you should go ahead and get a "real" scope, which to me means at least 50 Mhz. A year or two from now you might be interested in something that involves higher frequencies.

But as far as I know, you can design and troubleshoot audio circuitry with a 100 khz scope, if they actually make such a thing. I'd avoid low frequency scopes unless they have great reviews because they are mostly toys and not intended for serious work. Lots of 20 Mhz used scopes can be found, and they'd be fine.

However, dealing with transients is way easier with a storage scope, which some analog scopes are, but all digital scopes are inherently capable of capturing a signal and allowing you to zoom in or whatever.
>But as far as I know, you can design and troubleshoot audio circuitry with a 100 khz scope, if they actually make such a thing. I'd avoid low frequency scopes unless they have great reviews because they are mostly toys and not intended for serious work. Lots of 20 Mhz used scopes can be found, and they'd be fine.
Thank you this is helpful
>I’d be interested in results even if it’s was just hearing the audio.
I plan on putting it on youtube as background music, so if anything I create does sound good it should be there. I think my first order of business is to make an "electric guitar" with an old bike gear cable and a diy magnetic pickup, since getting a clear tone that isn't mains hum will be tough without something like this.
>Currently reading the forest mims-getting started in electronics, Make:analog synthesisers and lessons in electronics along with these threads and various other bits from the internet.

That's cool. As long as you are ambitious, you will keep expanding your horizons and pick stuff up organically. At the same time, having a formal framework for knowledge can be very useful, say if you want to analyze VCOs. So I would suggest following along with some college course materials at some point.
>So I would suggest following along with some college course materials at some point.
This is kinda what I was asking for. Rather than just trying to self learn and keep interested from books. I thought maybe some basic practical stuff involving audio generators/filters/lfo’s etc might be good, but obviously knowing very little I’ve come here for advice.

Keep us posted anon
Remind me how to tell what resistor I need for a given LED for a power source of two AAA batteries
I'm building a head tracker.
I meant AA, fucking phone
An AA battery is usually 1.5 volt allthough it might be save to use 1.6 in the calculation to make sure you don't break the led with new batteries. Also 2 AA batteries in series or parallel
Standard battery holders are generally series, so series.

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.