/ohm/ the electronics thread

Thoughts on BotFactory? https://www.youtube.com/watch?v=rZWw0BR6S1c

Is this really to electrical engineers what 3D printing is for mechanical engineers?

>>1177307
God yes. I'm hard as diamonds from watching that.

>>1177250
I got 200 from hyperphysics, the 5000 is for specifically pure iron. Stainless 304 is 1.002, so good luck detecting that. I'd recommend looking up how real metal detectors work and trying something like that.

>>1177323
Well I measured it to be just over 1V with a maximum of 2.6V. Different standard?

>>1177307
>0.03 ohms/square
One inch of 10mil track = 3 ohms.
Ok for many uses, but it's nowhere near a universal replacement for normal PCBs.
Well, the conductivity probably improves over time.

File: passive mixer.gif (5KB, 583x335px) Image search: [Google]

5KB, 583x335px

>>1177334
>Well I measured it to be just over 1V with a maximum of 2.6V. Different standard?

the other dude is sorta confused. the mics are usually biased from a 5V source in series with a resistor of around 2K. since the preamp inside the mic takes some current, it creates a voltage divider with the 2K resistor, resulting in approx 2Vdc at the output (with a tiny millivolt audio signal riding on top of that.)

as for your project, whatever the heck it's supposed to accomplish, you shouldnt need any external op-amps or batteries, you just need to create a passive mixer using resistors. the resistor in series with the music source should be way bigger so that you get an attenuation of 100 or more. also, putting a cap in series with the music source would eliminate a potential source of problems.

Stupid question since I'm new to electronics. I've got a car horn that plays sounds and music and all that, and I'm hooking it up to a 12 Volt battery instead of the car's fuse box. It's got it's own fuse. Stupid question: Do I need to ground it?

>>1177351
technically no but you still should. that sounds really inconvenient though.

>>1177350
No, it was 2.6V when the mic wasn't even plugged in, there was just a 5MΩ resistor between ground and the mic line instead. Just check my data: >>1177208

>>1177351

grounding is used as a kind of shielding around metal devices, so that 120Vac doesnt end up touching your body in case a wire gets loose. in cars, it's more of a convenience: you can run current thru the chassis instead of expensive copper wires.

neither case applies here, so grounding is neither needed nor useful.

>>1177357

yep, 5V is a typical value but it can go lower, the min being around 2.5V

>>1177293
>Is anybody reading this shit or will it get copied and pasted into the next thread?
Heh

>>1177334
I've been trying to find that out, but all I can find is explanations of the earliest metal detectors, e.g. early 1900s technology. I'll probably post on reddit or something.

>>1177392
Apparently they use a secondary coil that picks up the induced magnetic field of the first coil thanks to eddy currents in the target metal, or something along those lines. I just googled "how metal detectors work and got it.

Ok so it turns out kitchen sponge is not a replacement for soldering sponge, it blackens a little and leaves dark residue on the iron. Copper/brass wool is kinda expensive and my local shops don't stock soldering sponge. Is it better to use brass or copper? My local store sells only the brass for $18, while online brass is $3, copper is $1.50 and so is steel, but I'm pretty sure steel is a bad idea. Sponge is also $1 for a bunch, but is it better or worse than wool?

>>1177452
I prefer copper wool over sponges.

I'm guessing you use copper because it's a fairly soft metal so you don't damage the tip. Following that logic, brass would be worse because it's harder.

File: OACOM5j.png (119KB, 737x631px) Image search: [Google]

119KB, 737x631px

Right, I've got a device that will output a 12v for as long as the button is held, issue is that I've bult a 555 timer that will activate for 5 seconds, but I need it to provide a 5 second output with only a .5 of a second input

>>1177460
https://electronics.stackexchange.com/questions/180716/555-timer-one-shot-trigger

I am setting up a capacitor bank that I want to charge fully and then let it hook up to various other projects when I want to use them, so I have the capacitors selected and being attached in series, my question is I have 10 capacitors at 300V each, how do I figure out how to charge them with battery power if possible? I was looking into DC/DC converters which seem like what I am looking for but is there anything I should know that is really important before buying one that says it can take like 12V to 300V? And would that be enough to charge them all?

>>1177466
I've read that twice and I can't seem to make sense of it.


My mental age is like 4 when it comes to electronic stuff.

I have to analyse a circuit and I shouldn't fry it by no means.
Before testing for continuity, is there anything I have to do besides discharging the capacitors?

>>1177479
Why are you trying to charge things to 300V?

>>1177487
Well just to around there, I figured I was trying to charge them close to max for use in projects like a coilgun. This is the first time I have ever dealt with capacitors at all and I only like two threads ago got to asking how to select capacitors for this usage. So now I am on the next stage and figured I would ask for an opinion if I am on the right track or if I need to change course.

>>1177460
Try a capacitor in series with a resistor, and a transistor that triggers based off the voltage across the resistor should work. As the capacitor charges, the voltage across it will steadily increase, which means the voltage across the resistor will decrease. Once it drops below the transistor's threshold (as tuned by the base resistor of the transistor) it will turn off. You can tune the time with the resistor in series with the capacitor, and maybe the base resistor too.

>>1177307
Yeah, the PCB making is only good enough for simple things. The fact that it does solder pasting and pick-and-place are pretty interesting though.

But, "squink"? Wat?

>>1177491
>projects like a coilgun
http://www.nutsvolts.com/magazine/article/electromagnetic_coil_launcher_project

>>1177491
>charge them close to max
That's asking for problems. Under normal use you don't want to charge capacitors above 50% rated voltage, you can get away with more for a while but they wear out much faster and have lower capacitance values towards the rating.

https://www.sparkfun.com/news/1271

Are there any spectrum analyzers available for a reasonable price? I don't really care if it's used as long as it works. I would however like it to measure audio range frequencies up through RF though I'm more concerned about the lower end, at least at the moment. Tracking generator would be nice as well.

I'm on an $11/hr burger flipper budget. I can probably do up to $500 at the expense of not eating for a week. Used is fine but it HAS to be working. I can't afford to take gambles here. It can be a bit out of cal though.

Also on the lookout for a network analyzer too. Preferably vector but vector analyzer prices are insane so if I ended up getting anything it would almost certainly be scalar. Scalar network analyzers are the only ones I've ever seen under $1000.

Which ever piece of equipment I find an insane deal on first is the one I'll pull the trigger on.

>>1177633
>$11/hr
You earn less than me, and I'm in a third world country.
You should follow your American Dream, now that your lovely Trump is in charge, and become a millionaire, you lazy bum.

>>1177636
I'm trying to learn about electronics and become a (more) productive member of society but so far it's costing me a lot more money than I'm making so I'm not really sure when I start making the millions, or even just the tens of thousands for that matter.

oh and by the way $11/hr isn't even the minimum in my state. $10.10 I think is the minimum and that's pretty good compared to most red states which are at the national minimum which I think is still $7.25 or fairly close to it. Granted the cost of living is higher where I am so the money doesn't carry as far but as far as internet instrument purchases are concerned I'm better off. A $1000 spectrum analyzer is a $1000 spectrum analyzer regardless of what state you live in.

>>1177633
You could probably use an SDR, though they're meant for RF.

There's plenty of FFT libraries.

File: beavis and butthead i understand your frustration.jpg (63KB, 500x375px) Image search: [Google]

63KB, 500x375px

>>1177638
>Granted the cost of living is higher where I am so the money doesn't carry as far

>live in So-Cal
>cheap 2-bedroom apartment is like $1400/mo. minimum
>friend just moved to Las Vegas w/ friends
>same 2-bedroom apartment is $500/mo.

File: IMG_20170515_141748_1CS.jpg (2MB, 4305x2049px) Image search: [Google]

2MB, 4305x2049px

Can /ohm/ help me identify this component? It's some sort of electromechanical switch, when you apply current it actuates a little arm that moves a bigger arm. The circuit diagram is not very informative.

>>1177639
>SDR scope
Not him but this is fucking genius. Any hints as to what input protection I should use when measuring up to 5VDC? I guess it would be manually ranged with a rotary switch and some opamps/RF amps, and some sort of automatic voltage protection on the input. You could even use the amp's clipping to limit the voltage to the antenna plug, with the help of another decimal-gain amp afterwards. This is a really interesting idea.

>>1177643
>life in a commie state is worse than life in a based republican state
No shit.

>>1177660
Not a relay is it? Does the bigger arm make electrical contact, or is it for mechanical use? Maybe it's a solenoid valve or something. How does the component behave when a current is applied through the coil?

>>1177665

It's part of a turntable, when the current is applied the metal arm (just under my thumb above the part) is pulled towards the bottom of the component (at the end with the black plastic. This pivots the metal arm out a couple of millimetres which A) closes a contact giving power to the motor and B) moves a bigger, plastic arm out to flip a catch on a cog that runs off the turntable and controls some of the automated parts of the player. The clearance at the business end is pretty tight so I can't see if the metal arm is being actuated electromagnetically or mechanically.

>>1177512
I am trying to build this in sections that can be removed for other projects later so right now I am just trying to make a capacitor bank and charger, saying it was for a coilgun was just to say what its immediate future use is.
>>1177631
Alright, so assuming I have 10 of my capacitors hooked up in series and only go to 50% of their max voltage, that is 150 Volts per, which wouldnt that mean my power supply has to provide 1.5 kV? And if I did that correctly, is their a better way to provide that much voltage using batteries or is the dc/dc converter the best bet for getting it up that high?

what is the difference between ltspice and pspice?

>>1177696
Or you could hook them up in parallel instead, and charge them up with rectified mains power. I don't see why you want high voltage in the firet place, coilguns and other solenoid-based projects require less turns to get the same magnetic field out of more current, so using higher voltage is just more work for you. Anyway, you can always switch the capacitors from parallel to series after you've charged them, though I'm not sure what kind no of hardware you should use for the switch besides changing the terminals manually. A couple of boards with solid copper traces on them that you press on top of the capacitors to get the series or parallel configurations? Anyway, just make sure you don't feed 1.5kVDC into your wall socket.

>>1177841
LTspice is available free of charge.

Is there some software for plotting a circuit so it works in a breadboard with the most efficient layout?
>inb4 lazy
Yeah...

Is there a term that I can search for that is like an electronic switch? I want to put several switches throughout a circuit I am doing to control some buttons so that if like using an arduino I can send a high signal to it and have that switch then connect, where as like low would tell it to disconnect. Is there a part like this anywhere? most of my googling just keeps getting me normal by hand switches.

>>1177921
I'm an electronics idiot, but relays that are linked up with something that senses whatever you're looking for sounds like what you're going for.

What can I do with an old Xperia Z1 components?

Any good videos on 3 phase power, specifically stuff with hi leg?

>>1177923
>>1177921
I'm a fucking idiot, relay was what I was looking for, thanks for kicking my brain back into working.

>>1177932
You can also use transistors. No mechanical parts means practically infinite cycles, plus they're cheaper.

https://www.youtube.com/watch?v=WRm2oUw4owE

>>1177664
Living in a blue state is definitely better than living in a red state. I'd rather stay in CT and pay my ridiculously high taxes than move to Alabama or Texas.

>>1177888 https://electronics.stackexchange.com/a/31585

>>1177930
I don't know of any videos, but I might be able to answer some questions.

>>1177984
>mentions significant disadvantage to living in blue state
>no mention of disadvantages to living in red state
Tell me more.

>>1178045
Blue states usually have nicer people, more robust economies, more services for people, and people generally have more personal freedom.

People always say people are nicer in the South but that's only true if you are ideologically identical to them. If you are different or hold different beliefs they hate you. In the North nobody gives a fuck. It's great. I don't know if Cali is the same though.

>>1178077
You poor benighted fool. You've drank the koolaid.

>>1177633
If you need a real, swept conversion spectrum analyzer, the 8568A is kinda a boat anchor but you can get one for cheap.

But if you don't need 80dB of dynamic range then an SDR will probably do.

>>1178081
pls no bully

>>1177633
>spectrum and network analyzers
>$11/hr burger flipper budget
>I'm trying to learn about electronics

This kinda does not compute. I mean, the most electronics labs don't have either of those, partly because they aren't usually needed and partly because they're expensive.
Besides, nicer digital scopes have FFT built-in and even shittier ones usually allow you to download the data and run FFT on your computer. Bandwidth and resolution are naturally limited to that of the scope's, though.
Not that you can't try your luck with RTL dongle or PC's audio input.

>>1178094
>viciously lays into a giant region of the country
>pls don't disagree with me

>>1178077
Cali fag here, if you disagree with the liberals you get fired, beaten, and your guns taken away.

File: sassy.png (241KB, 675x663px) Image search: [Google]

241KB, 675x663px

Has anyone built a high current H bridge before? I am wanting to add AC functionality(literally switched polarity square wave) to my welder which is an DC inverter design. Would MOSFETs be suitable to use as a second stage inverter?

Obvious modifications would be:
>move HF igniter down the line and add protection diodes across input
>implement a simple circuit to switch the bridge at 100Hz for AC and toggle between polarities for DC+/- modes

What annoys me is i can't find any existing implementations, they either use insanely expensive IGBT modules(200A full bridge is $600+) or don't document how the second stage works at all.

My main concerns are:
>OCV being 80v is close to 100V breakdown voltage(would a spike wreck my shit?)
>surge current is likely to be ~1kA, and since the fets are right on the output they are going to get the full force every time welding starts
>heat dissipation is going to be significant if i want to save costs and use few devices

>>1178140
You could use high-current automotive relays instead of transistors, that is if you don't mind the noise and 1hr lifetime.

If I have a full bridge rectifier and in the datasheet it says "Maximum reverse voltage: 200 V", does this mean that if I connect it to European mains (~230V) it will blow?
http://www.mouser.com/ds/2/366/J534S2VB-190942.pdf

>>1178168
Yes. Even worse, the peak voltage (not RMS) of european mains is 330V, and if you have a smoothing capacitor across the rectifier output you can get double the peak voltage across the bridge once the waveform reverses, but I think that at that point the double peak voltage will be across two diodes and not one, so you should be fine. Look for a >350V rectifier for Europe mains, 400Vs are common for the same reason that 200Vs are.

File: resistor for scale.jpg (565KB, 2049x2209px) Image search: [Google]

565KB, 2049x2209px

>>1178179
oh that sounds bad. Yeah i have two other rectifiers rated for 400V, but i'm going through a bunch of random parts that my uncle gave me.

Speaking of which, can anyone tell me what pic related is? Could it be an oscillator? All it says is S063.

Just in case you were wondering how compressed air compared to batteries as a way to store energy, I calculated the theoretical maximum for this.

E=5/2*P*V*(X-X^(5/7))

Where P is the outside absolute pressure in Pa, V is the volume of your container in m^3, and X is the ratio of internal to external absolute pressures. If your air tank is on Earth, P will typically be ~100000Pa.

This equation might be wrong, at least there's some inconsistency when X<1, but I think it holds up. I calculated the definite integral for adiabatic compression and did a bunch of substitutions with PV=nRT and k=PV^(-7/5). This equation only holds for diatomic molecules, also maybe CO2 for whatever reason.

At 30psi relative to the atmosphere, you'll be storing about 200J per Litre of tank, which is the muzzle energy of a somewhat slow, small bullet. At 3500psi relative to the atmosphere (such as in a high pressure scuba tank), you'll be storing around 47400J per Litre of tank. Comparatively, A Li-Ion cell has a maximum energy density of 3.6MJ per Litre, which is 18000 or 8 times greater respectively. To rival a Li-Ion's capacity, the tank would have to hold 15400atm, or 230000 psi.
I'm calling bullshit on 3D manoeuvre gear.

File: head[1].png (136KB, 688x735px) Image search: [Google]

136KB, 688x735px

Hey guys I hooked magnetic read head to arduino board to see if arduino can read what cards I swipe. I also have lm386 in the max to amp the magnetic head.

My question is, since its all binary, what pattern should I look for to find the data from the card that was swiped?

I read some places that there will be leading 0 (aprox 20 or 60) and at the end, and I did come across similar pattern (40-80 leading/trailing 0) and I went to convert it to decimal and ascii (in case the track had holder info.

Also whats the purpose of the head common pin on the magnetic read head?

>>1178185
It's a quartz crystal. That can size contains almost always a 32768Hz one. It requires other electronics to oscillate.

>>1178225
cheers

>>1178140
Just use a bunch of mosfets in parallel and a massive heatsink.

They also have solid-state relays. To be honest, you'd be better off DIYing your own transformer to go from high to low voltage. Switching +/- 170V or even 330V is much easier than low voltage high current. On top of it, since a welder is "isolated", you can go simple and use a basic bipolar design (i.e. the ground does not have to be neutral). With like a 40% duty cycle(10 off, 40 on, 10 off, 40 on) you can prevent shoot-through. Fixed frequency makes it even easier, because you could use a 555 with a fixed delay and a not gate to trigger the inverted signal. The only problem with mosfets is the higher the frequency the more amps you need (1-2A at 1MHz), but 100Hz is easy as piss and could probably be driven from the 555 directly (or throw in a 100mA buffer amp).

Logic level mosfets trigger more easily than normal ones, but a comparator/difference amplifier tied to +/- 10-12V also would work.

>>1178375
Oh, erm, forgot to add, to go from the square pulse to AC for the transformer, you can use an in-line passive filter. There's better and worse component values.

>>1178208
>what pattern should I look for
depends on the card, you need to know what to look for
https://en.wikipedia.org/wiki/ISO/IEC_7813
>Also whats the purpose of the head common pin on the magnetic read head?
nigga wtf you serious?
how are you connecting it now, just with one pin?

electronics has a pretty heavy reliance on voltage, maybe you heard of it? voltage describes a potential difference and requires a closed loop for current to flow.
admittedly its fucking retarded to have 3 individual common pins, the point of a common pin is it shares a common potential, (typically vdd or vss of the component) they should have called it +/- instead.

>>1177921
Relay or transistor.

Noobquestion: can, and should I, connect my inner circuit ground to wallplug-ground?

I have a NodeMCU Microcontroller (basically an arduino with WiFi) with a few sensors for environmental Data (all digital, using 3V: one bmp280 and 2 ds18b20). It works fine from USB power but to throw it in the garden I salvaged an old phone charger. It will be put in a Plastic or wooden housing so that does not have to be grounded.

I don't wanna blow my fuse trying it out though.

>>1178422

it's fine if you do it, but it's not necessary or useful in any way.

>>1178422
>>1178437
It can probably help with shielding if your circuit has a problem with interference, but I'd be careful that you don't create a ground loop between the wall-socket ground and the circuit power supply's neutral.

>>1178437
>>1178438

I assume a ground-loop occurs when two grounds of different potential are connected?
I've read somewhere that modern fuseboxes have one fuse in them that blows everything if something fucks with ground, would those blow in that case?

Either way, I will just leave it disconnected ; interference should not be an issue out here in bumfuck nowhere.

>>1178100
Not other guy, but I'm with you dude. The southeast isn't too bad. Its cheap and easy to live down here. Now job opportunities aren't great where I live but my needs are met very comfortably at 17/hr. If I wanted to make more I would have to move to another state though.

>>1178380
Look man I usually see vss/vcc labeled for voltage, but common head? Will I need a resister for this or straight 3.3v or 5v be ok?

File: driver.png (27KB, 2302x1567px) Image search: [Google]

27KB, 2302x1567px

>>1178375
>solid-state relays
>To be honest, you'd be better off DIYing your own transformer to go from high to low voltage
This was the original plan, pic related. But right as I'm about to pull the trigger on a complete BOM, someone listed a 200A miller inverter for $100(literally got a free torch, ground lead, stack of tungstens and torch components).

So right now I have a 200A current controlled DC supply with all the support circuitry already there. I'm only looking at a second stage to get AC now.

>bipolar design
It's my understanding that PNP transistors are hugely inferior, but at 100Hz this shouldn't be a problem right?
>With like a 40% duty cycle(10 off, 40 on, 10 off, 40 on) you can prevent shoot-through.
Thats a non-issue since I already have these IGBT gate drivers with a 10k gate resistor to prevent ringing or shorts.

>>1178378
From my research there is absolutely no gain to filter the AC into a sine wave, market value of transformer designs vs inverters(which all output square waves) would make you think unfiltered is actually better.

Is there a recommended book for learning Verilog? I'm taking my first class on it in August though I'd like to get a head start.
My professor likes to teach by example so he's not using a textbook.

>>1178505
If you're doing step-down though, you would need to output through a transformer, was the point.

>pic related
No, you completely misunderstand. Going from wall voltage (120/240/208 whatever) is what I mean. 120/240V 20A is much easier to control than 12/24V 200A.

1) I'm curious as to why you'd want 100hz and not standard 60 or 50, considering you're obviously doing temperature welding, and not arc or anything.

2) I wonder if you could buy a high powered motor and AC generator, and just gear them up. You'd need to adjust field coupling to get a constant frequency, or just disregard it and let the generator develop variable frequency.

>transistors
Use fucking mosfets. Transistors a shit. You'd be more likely to have issues even with IGBTs, because transistor hotspots become more conductive. If you paralleled mosfets, any overheating in one increases it's resistance, shunting current to the others. I think you're really underestimating the power of modern mosfets.

>>1178527
Not him but
>MOSFET
>metal–oxide–semiconductor field-effect TRANSISTOR

>>>1178375
I don't think you quite understand duty cycles.
>D = On time / total time
The one you have would be 40 / (10 + 40) = 80%
You would want 40 on, 60 off. It's EZPZ if it adds up to 100.

File: 1384005083509.png (71KB, 554x436px) Image search: [Google]

71KB, 554x436px

>>1178534
>knowing exactly the context and what I was saying but being an autist about it

>>1178538
It'd be positive and negative. 40% postitive, 40% negative. If you read what I had put, the first 40% would be positive, and it would trigger a delayed 40% negative. Either way you look at it, that ain't 80%.

>>1178527
I'm not doing step down, it's already regulated DC. I literally just need to swap the polarity at the weld frequency(100Hz is very common, it's easier to control than 50-60Hz, requires less power for same heat input etc). The existing supply operates at 1kHz using IGBTs to switch the mains into a HF transformer(it's incredibly small for something that delivers 200A continuously), this is immediately rectified and smoothed with a large inductor. 1kHz AC straight from the transformer is useless for welding because the molten weld puddle needs a finite time to burn off oxides during the negative cycle(torch positive, electrons are flowing from the puddle to the torch pulling material off the surface in the process).

I really like the sound of an electro mechanical design but I'm in Australia so no such devices exist except in industry.

>Use fucking mosfets.
That property you brought up is pretty useful actually. I've never designed anything with high power MOSFETs, mostly looking for validation.
>>1178538
I think he is referring to dead time, by leaving a transistor in High-Z it will start discharging and you don't get as much ringing compared to immediately driving it from one state to the other.

>>1178542
I'd seriously question any transformer claiming 200A anything. The current is based on the flux you can get in/out, and saturating causes a short. There's a reason why people would use two microwave transformer cores in their ghetto welders. It could be used as some sort of feedback, but saturation's a physical limit of material. I don't know what transformer you mean, so maybe "small" to you is in welder terms.

File: Inverter-welding-machine-high-frequency-transformer-amorphous-transformers-UF70-transformer-skeleton.jpg (132KB, 1000x750px) Image search: [Google]

132KB, 1000x750px

>>1178547
see pic, unlike a traditional buzz box the transformer is a powdered core and has a higher efficiency than a 50-60Hz laminated core transformer. to get 200A out of a laminated core it would have to be at least 4x the size and 10x the weight.

>>1178451
I don't think ground loops will cause a current strong enough through the ground breaker, rather I think its the differential resistances between the existing loads that cause a minute voltage across ground and neutral that cause interference inside your circuit.

I just (cautiously) measured 2-5Ω of resistance between my ground and neutral pins, so in a worst case scenario when running a 240V 20A (12Ω) load, you could actually get a few 10s of volts across the two pins, so it's a good idea to not connect them, especially when you're using high loads nearby. That's a worst case couple of amps which would probably heat up your circuit a fair bit, but a GFCI will probably be rated in the 15A range so you shouldn't trip it anyways. I would be concerned if your house had old wiring however.

If your power supply has isolated HV and LV sides, i.e. it uses a transformer, then you won't have a problem, but if it uses a buck converter or something similar you might run into issues.

>>1177293
please add gbppr.org. It is truly an amazing site about radios and electronics design from like 20 years ago, in the 90s, when there wasn't so much javascript and pointless ads shitting up the internet.

>>1178522
By example is really the way to go. But look for course materials on "digital logic" or similar, like Berkeley CS150.

File: 71WTbaqeKOL._SL1500_.jpg (63KB, 1500x772px) Image search: [Google]

63KB, 1500x772px

Won some money off of a scratch ticket, so I'm taking the opportunity to overspend on some really nice items for a couple of hobbies. Who makes the very best wire flush cutters and round nosed pliers? I know you can't go wrong with Knipex, so that's what I'm currently eyeing, but I wasn't sure if there was a better option. I'm going to get a small and large set of the round nosed pliers(small for electronics stuff, large for rough electrical around the house).

File: MainsTriacSwitcher.png (49KB, 1479x840px) Image search: [Google]

49KB, 1479x840px

Can someone explained to me the purpose of diode bridge and moc3020 circuit part in schematic?
Unless it's something very obvious I'm missing, schematic works without specified parts

>>1178637

that schematic is a monstrosity: parts vomited all over, no respect for left-to-right flow, transistors turned every which way, wires routed like spaghetti, incomprehensible external connections.

as befitting such a retarded schematic, there's a glaring error on the MOC3020: pins 4 and 6 will always have the exact same voltage given that they both go to a common point. so, yes, as drawn, the whole bridge and triac driver will never do anything. probably an error in the drawing. also, driving the LED in the MOC3020 without current limiting is stupid.

File: Fujiya side-cutters.jpg (21KB, 490x435px) Image search: [Google]

21KB, 490x435px

>>1178599
>not immediately looking at Fujiya Nippon steel
It's not like I've ever bought anything from them or heard about their quality, but like Good they're one of the only exotic brands my local electronics shop stocks, so they can't be too bad.

>>1178655
*Goot

Alright so to power the electret mic's built-in preamp I require some voltage relative to the ground with a resistor in series to control the gain. The problem I am encountering is that by tying the top +4.5V rail through a resistor to the correct terminal on the mic capsule I am creating a path to ground through the mic with several times less resistance than the resistors of the voltage divider (~2kΩ compared to 10kΩ). This effects the opamps and prevents the mic from getting the required voltage. How would I go about creating a power supply that is not effected by drawing more current from one rail than the other? Is it Zeners? I get the feeling isolation transformers are going to be required, which will probably mess with my frequency balance, but I was going to boost the bass with a few capacitors anyway.

Oh and I plan on powering the whole assembly from a USB and some sort of voltage doubler, so I guess I'll have to create the -5V line out of thin air as the negative of the USB power will be tied to ground. I feel like a solution of leaving the ground line at +2.5V and using that to bias the the computer mic input line is staring me in the face, though how the other input line would work is beyond my sleep-deprived brain. There wouldn't happen to be <±2.5V opamp ICs out there, would there?

The project is essentially making a couple of circuits, one that converts standard earphone-amplitude audio into electret mic-amplitude, and another that converts back again, so it has to be pretty flexible and modular.

>>1178671

seems we've been through this already.
- you're gonna decouple the mic from the op-amp with a cap, so the DC voltage across it is irrelephant.
- lots of op-amps work single rail down to 5V, so you dont need a split supply. the 20-year old LM386 comes to mind instantly but i'm sure there are thousands of new op-amps that'll work even better.

>>1178682
The input device/computer has to think there's a mic on the other end by a few hundred µA travelling through the line to ground, and the signal it receives must have that voltage bias to it. Ideally I'd take the voltage bias directly from the computer and add the waveform to it with the opamp, though I'm unsure how I'd be feeding the signal back through the same input. Maybe one of the gain resistors could happen to be the ~2.4kΩ resistor needed to get the current flowing, but I'll have to start drawing this up.

I saw the LM386 datasheet, that 4V isn't ±4V but 4V total?

>>1178774
99% of jacks detect a plug present by a physical (dis)connection, not an ammeter. If you honestly think it's an ammeter and believe it's uA, then slap a resistor on it.

>controlling gain by changing the pre-amp voltage
No, this is very wrong. Control the gain by changing the op-amp feedback.

File: electret amp.gif (11KB, 600x508px) Image search: [Google]

11KB, 600x508px

>>1178774
>I'll have to start drawing this up

you better, coz you're confusing to read. i still dont get what you wanna do.

>The input device/computer has to think there's a mic on the other end by a few hundred µA travelling

so you put a load resistor on the output of the op-amp that'll suck out whatever current is needed by your sound card. see pic.

>the signal it receives must have that voltage bias to it

it only provides the bias voltage to power up the mic, it doesnt need to receive it back. in any case the cap on the output of the op-amp will block any DC level going to the computer. see pic.

>I saw the LM386 datasheet, that 4V isn't ±4V but 4V total?

it's 5V total (except if you get a special part, lm386n-4, which is 4V total)

>>1178781
The computer has to know when to switch to external mic or not, and does this by seeing if there's a current path with a resistance between about 12kΩ and 1kΩ. If the resistance is higher this current is ignored and the computer remains on internal mic mode, and if it's lower then no current flows at all due to some sort of short-circuit protection.

A normal electret mic has its gain determined by the resistor in series with the voltage that powers the FET, or so Wikipedia tells me.

>>1178859
That looks about what I'm looking for, no virtual ground required, though I want the input side and output side to be modular so I can combine this with normal audio signals as I please.


If anyone has any data regarding the normal amplitudes of earphone audio and electret mic audio that would be appreciated because I don't have a scope. I've heard 1-2V and 1-10mV respectively, but I've no concrete numbers on these.

>>1178864

little in-ear headphones are about 2V at maximum volume, which corresponds to about 1/4 watt.

mics vary a lot but a gain of 1000 is often necessary.

>>1178873
Sounds about what I was going for.

File: low_pass.png (20KB, 1510x478px) Image search: [Google]

20KB, 1510x478px

I'm shit at designing circuits.
Here is my low pass filter to convert a pwm signal to analog voltage.

So my last lab was about multi-stage amplifiers, we used one mosfet and one bjt and designed to get a voltage gain.

The last question is about writing down the current and power gain. What should I exactly do here to get at least the current gain? Because I know for the power gain, it's (v_oi_o)/(v_ii_i). I'm asking this here because we didn't go over it completely and the question is there. I'm not sure if I should skip it or not. But might as well solve for it.

>>1179035

the thing about filters is they're designed for a given frequency. a filter for 10Khz would be useless at 1K. so, your design is less than useful if we dont know what the freq is.

also, there's various filter design tools online where you just choose a few options and it pulls up a perfect filter for you, without one single caress of your calculator. like this one http://www.analog.com/designtools/en/filterwizard/

I got this component the other day and I finally was able to test it.

https://www.amazon.com/dp/B010O1G1ES/

It worked fine, but when I was trying to build something to it, it always failed. Then I realized I was using a USB cable that didn't have any data wires, so after an hour of trying, I switched to another cable and was able to see the component in the devices.

I tried to build it again but it failed once more. Then I tried to use another USB, but after I did that, the blue integrated LED that blinks every second stopped blinking.

Now it just blinks once when connected to power and never blinks again anymore unless I unplug and re plug it in, or click the reset button on it.

Did I kill the component? What should I do here?

>>1179088
I need to convert a pwm signal to analog voltage. I need only the mean value of the signal, so the cutoff frequency could be as low as I want it to be

File: small signal equivalent bjt.jpg (9KB, 388x202px) Image search: [Google]

9KB, 388x202px

How is this picture even supposed to represent a bjt when you can't even get a collector-emitter voltage?

>>1179092

but PWM can be used for all kinds of things, like perhaps generating audio sound effects of the ''Atari 8-bit sound'' variety. in which case, you wouldnt wanna cut anything below, say, 10khz.

>>1179097

i dunno what your objection is. the model shows you current gain between Ib and Ic. if you know Ic and the load, you can figure out Vc. as for Ve, it's base voltage minus 0.6457V. subtract, and you have Vce.

>>1179103
But putting a voltage source across the current source and an external resistor (such as in an amplifier) gives you a set voltage across the resistor (V=I*R), so if this is not equal to the value of the source, you have a nonzero voltage across the current source, which has 0 resistance. This doesn't make sense to me.

>>1179106
>current source, which has 0 resistance
Ideal current sources have infinite resistance.

>>1179106
there is current across voltage sources, and there is voltage across current sources

>>1177293
>What circuit sim software do you use?
Can you add fritzing on the list?

>>1179147
Fritzing does simulation?

>>1179115
>>1179126
My bad, looks like I'm not cut out to be a physics student after all /:

Took me 4 pages of trial and error before I remembered that V+ ≈ V- for an Opamp too.

Do you have a recommended source of (chink) 18650s?
I'm trying to upgrade my bluetooth speaker's puny 1500mah battery pack and I don't really trust those $5 """5000"""mah ones from gearbest.

>>1179258
old laptop batteries, the usually go bad because of just one cell

File: tmp_31735-IMG_20170518_200854299082998.jpg (2MB, 4160x3120px) Image search: [Google]

2MB, 4160x3120px

Any hope with shit like this? i found it In a College giveaway dumpster. the image is viewable but it has only grayscale and bits and pieces of cmy/rgb

>>1179429
You could use the backlight as a diffuse light panel, assuming that's a backlit screen you're talking about.

I need a resistance controlled by an analog voltage. Looking at ordering a digital potentiometer, but I know nothing about I2Cs to convert analog to digital to control it. Is there another term for a VCR that would correspond to an analog version based on using a transistor? Alternatively, how would one go about configuring a transistor so that the gate voltage could have a reference without introducing a ground wherever the VCR is placed in the circuit?

Would anyone know what a small cylinder of Krypton-85 would be used for as an electrical component? It's label is EG&G 0GP-50L-8 8806-058

Very certain it's more than 20 years old. It's small, like a dessicant from a pill bottle, but it the primary cylinder is ceramic, it has shiny metal endcaps which connect to leads, and has a little radiation warning on the side.

I found this channel with some interesting videos on PCB design, I thought you guys might find it useful:
https://www.youtube.com/user/matarofe

>>1179471
How small does it need to be? Dumb way: Digital circuit has an ADC, and you convert the digital value to a position of a knob with a rotation sensor or something. Dumber way: A dual-ganged pot hooked up to a motor. If a comparator detects the present voltage (from one side of the pot) is above or below the reference, it runs the motor until the voltage is matched (with a small hysteresis of course). The second ganged pot is the one you use as the resistance for your circuit.

>>1179091
Have you tried a powered USB hub?

>>1179482
Those sound like much worse solutions. The application is in an audio synthesizer, patching LFO in to control oscillator frequency, amplifier gain, or filter cutoff frequency. I was hoping to just throw in a transistor but discovered that I don't know how to make it float, and documentation on the subject is extremely poor.

>>1179486
Voltage controlled resistors aren't done for a reason. They're not easy. Every transistor needs to be reference to function. Even voltage controlled chips usually only have a select of internal resistors you can choose amongst.

All the applications you mentioned would be far easiest with a variable capacitor than resistor. And there are circuits on voltage controlled capacitance circuits, but those are still referenced.

Even still, voltage controlled oscillators don't need variable resistance, amplifiers don't need variable resistance (use a DAC to apply bias to a class A or B pre-amp). Low pass filters is question, why would you be altering your low-pass filter? Presumably because you're wave-shaping? Two op-amps can output a sinusoidal signal, basic DIP circuit.

>>1179485
Like plugging it to a wall with the phone chargers?

I tried powering it on on my breadboard through arduino, but still, it just gave a single blink.

I did try powering it on with a powerbank though.

>>1179489
It's a project for a class, for which the requirements are somewhat limiting (analog, no ICs, as a rule). I've been given blessing to use a digital pot, but have no idea how to use an adc or i2c to control it.

I started out with VCRs in mind because I was under the impression that they were the standard way to do it and that it was just a matter of configuring a transistor correctly. As I delved deeper I learned of VCOs as their own thing but they seemed more complicated than I was prepared to learn about for the project. With a VCR in mind, it seemed sensible to do the same thing to the amplifier. For the filter, modulating the cutoff is essential for effects like wobble base and, given that I'm using a sawtooth wave, has a rather dramatic effect on the timbre.

I hadn't really given much thought to varactors on account of the usage of capacitors by the elements I'll have, but looking over everything, it seems that may actually work. The sawtooth oscillator can be controlled by its cap, as can the filter. It leaves me without tremolo, but I think I can live with that.

>>1179499
How can you make a sawtooth but not be able to make a sinusoid?

File: s-l300.jpg (17KB, 300x300px) Image search: [Google]

17KB, 300x300px

I am running a dozen DuPont cables between anArduino and a sensor I'm flying on a drone. What woule be the best way to bundle these cables together that would be lightweight and still maintain flexibility? The sensor will be on a gimbal, so whatever I use to bundle the wires can't be too rigid, but the wires need to be well-protected enough that they can potentially touch damp grass upon landing.

>>1179503
Hot glue them on the pins. Zip tie them together

>>1179503
>the wires need to be well-protected enough that they can potentially touch damp grass upon landing.

I should add that I am running two cables in a chain between the Arduino and sensor, as in

Arduino to male dupont --> female dupont to male dupont --> Female dupont to sensor pin

so I need a nice waterproof seal where the two Dupont cables connect.

>>1179505
Hot glue the male cable ends into the Arduino headers and hot glue the sensor pins into the female cable ends?

>>1179507
Yeah why not

If you need more flexible glue, they have this clear sealant stuff, silicone or otherwise.

>>1179502
I dont have a problem with the sinusoid. I'm using a phase shift oscillator. I never said anything about the sinusoid. The sinusoid is meant to be the controlling voltage for the VCR/whatever. So, now I need to learn about 3-pin varactors and how to get that working.

>>1179509
You have not described your circuit at all. What is the goal, why are you using circuits which clearly you're having trouble finding controls for. You really need to stop and say "there's probably a better way to do this"

>>1179512
>The application is in an audio synthesizer, patching LFO in to control oscillator frequency, amplifier gain, or filter cutoff frequency
Sawtooth oscillator passed through filter and amplifier stages. Separate sinusoidal oscillator acting as LFO to vary those three, causing vibrato, tremolo, or cutoff. The only problem I have is how to have the LFO control those elements, because I have no experience with these kinds of circuits.

>>1179508
Cool, thanks.

Also, I just realized I might want to insulate the backs of the curcuit boards in case of moisture - What's the standard way of doing that? My first thought is to slather on some of the clear sealant stuff, not a good idea?

>>1179513
Why would it need to be sawtooth in the first place? Sinusoidal biasing a phase = phase/fm modulation, sinusoidal on gain to change amplitude, and sinusoidal on the input to cause the beating you want (if I'm understanding tremolo)

>>1179514
As long as the circuit isn't going to overheat, go ahead and slap it on. If you do have a chip that could get hot, you could always extend it off the board.

>>1179513
For frequency, Opamp + passive resistor mixing going into input of your VCO could work.

For gain control, JFET pulling down input on opamp http://www.nutsvolts.com/magazine/article/fet_principles_and_circuits_part_2

For cutoff, not too sure, looks like a standard approach is transconductance amplifier going into a cap http://www.ti.com/lit/ds/symlink/lm13700.pdf, pg 17.

>>1179518
Do you know about electronic music? The sawtooth is the desired output waveform to the speaker.
>Sinusoidal biasing a phase
I'm not sure what that means. What I'm looking for is the frequency of the output waveform to vary sinusoidally.
>sinusoidal on gain to change amplitude
That's what I'm trying to do. How?
>sinusoidal on the input to cause the beating
Tremolo is amplitude modulation. Amplitude varies sinusoidally.

>>1179521
>For frequency
Part of the issue that a true VCO is a larger circuit than I wanted to deal with, especially when the oscillation is a sawtooth, but I guess it may be the only way. Either way, I'm not sure what you meant by that. Elaborate?
>For gain control
Perfect. Thank you.
>For cutoff
I discovered that and was avoiding it because I don't know a damn thing about transconductance amplifiers and I would have to explain it in addition to learning how to build it. Would it be conceivable to apply the same jfet solution as for the gain, the jfet in parallel with the cap of the filter? These would be instances in which the transistor as a vcr does not need to float and so building it shouldnt be too difficult.

Overall, I'm considering dropping the LFO as part of the project.

>>1179471
>I need a resistance controlled by an analog voltage. Looking at ordering a digital potentiometer

you can buy, or better yet make, a vactrol. about a year ago, a diy_anon made a project wilth like 50 of them for a synthesizer.

>>1179534
I learned of those too. The idea of it puts me off as absurdly hackish but it is something to consider. I didnt see any reference to using them for sinusoidal voltages, though. Does it work reasonably, given that linearity isnt a big deal?

>>1179537

the output is a pure resistance so it works fine for AC or DC unlike photo-transistors.

>>1179529
>Part of the issue that a true VCO is a larger circuit than I wanted to deal with, especially when the oscillation is a sawtooth, but I guess it may be the only way. Either way, I'm not sure what you meant by that. Elaborate?

You presumaby have some voltage going in for requested base pitch, then your VCO has a sinusoidal voltage out (hopefully without DC bias to make things easier), you can just sum them together with resistors + an opamp to buffer and/or amplify again to get your pitch + vibrato going into the oscillator.

>Would it be conceivable to apply the same jfet solution as for the gain, the jfet in parallel with the cap of the filter?

JFET paralleling the cap with a regular opamp and not transconductance would pull down gain, don't think it would change cutoff. For the standard RC low pass you'd need the JFET in series with the signal. I don't really get transconductance amplifiers either, just googled a circuit that looks like it might work. I'd drop the filter section, but the vibrato/tremolo doesn't seem too tricky.

>>1179534
That looks incredible, I'll have to add that to the list of obscure circuit components I need to fit into some projects.
>>1179537
The wikipedia page shows a picture of a legit amplifier, so you can assume that it is neither terribly hackish nor non-linear, provided you're using the right kind of light.

>>1179479
Picture would be nice, but it sounds like an overvoltage protection device. The purpose of the radioactive gas is to maintain a low, predictable level of ionization so that the device reacts rapidly. The amount of Kr85 is really tiny.

>>1179265
Great idea, I'm doing this too. Have too many dead thinkpad batteries. Also, >>1179258 I assume yours used 18650 before right? What if I were to replace a different pack? Wouldn't that cause problems with charging?

>>1179035
bear in mind that LM358 is absolute shit
it has a class B output and is going to produce a shitload of crossover distortion, introducing new frequency components into your signal

File: 1495122531334.jpg (26KB, 429x410px) Image search: [Google]

26KB, 429x410px

>>1178539
>calling something shit and suggesting the exact same thing as a viable replacement

File: aouidfg.jpg (1MB, 3464x1680px) Image search: [Google]

1MB, 3464x1680px

Sup /ohm/, I clearly have a gap in my knowledge because this just happened.

Pic related is one of two boards from inside an early 80's era SMPS, specifically, the PSU out of an 8086, which is dead as fuck. I have tried to capture both sides in such a manner that you can view the circuit.

Tracing my way through from the AC side to see where power drops I came to the bridge rectifier. I measured it's DC side on the highest range of my DMM, 1000V, and it exceeded the top of the scale. I pulled out an analogue and made the same check and the needle hit full scale like a hooker hits coke. It was at this time I pulled the plug and noped the fuck out before I killed myself over some apparently wrong assumptions.

As far as I understood, a FW bridge outputs can never exceed the AC voltage coming in, it sums the peak amplitude of the positive and negative side components of an AC source, giving you close to, but not exactly the voltage you put in (because circuit losses). Plugging appropriate values into http://www.falstad.com/circuit/e-fullrectf.html seems to back this.

Is it possible for a rectifier to fail in such a way that it outputs a stupidly high voltage? Are big fuck off filter caps (2x 400V/150uF in parallel) somehow skewing my measurements? Again, AFAIK electrocaps in parallel become one big cap (C.total = C1 + C2 etc), meaning they should be act as a 400V 300uF cap. I'm not touching this thing again till I'm clear in this, as I say, everything I thought I knew seems to be wrong here, and considering the shit that got left on the board when I shorted the caps (track side, right cap) I don't want to die fucking with this thing.

>>1179633
>it sums the peak amplitude of the positive and negative side components
Sorry, that would result in 0V. It "lifts" the negative component of the wave to fill the gaps between the positive component. Hurp.

>>1179633

what you're saying is impossible. you should get 180Vdc from 120 or 360Vdc from 230. even if somehow the bridge and caps turned themselves into a voltage doubler, it'd still be way less than 1000V

i'm guessing operator error. you have your meters on ohms, or some such thing. maybe the range switches are dirty.

anyway, to eliminate the possibility of seeing phantom voltages, try soldering 100K temporarily across the caps (after you discharge them with a wet kleenex)

File: IMG_20170519_215248.jpg (2MB, 4160x3120px) Image search: [Google]

2MB, 4160x3120px

>>1179641
Fairly confident I would have blown the guts out of both meters if I had done that, and this is far from my first foray into meter based troubleshooting, ranges are not the issue here. But I agree, Shit makes no sense. Meanwhile, on remeasure, I'm only getting a measly 750V of the rectifier. We 240V, so this is still wildly high. Yes, it's a cheap meter, but if I swap them I get the same results.

> 100K resistor
1/4W? 5W? >9000W? Does it matter?

> wet kleenex
I was considering a 500W lamp on some flyleads. This seems safer than touching that shit with a wet object.

>>1179664

does look like operator error, still.

the digital meter is on the AC voltage range instead of DC. the analog meter, i cant tell but it seems to be on 250V range input, not the 1000V input.

>>1179668
> 100K resistor

at 360V, you'd need 1.3W, assuming the voltage was continuous, which it isnt, it'll drop fast, so it wont get time to get hot. so, maybe go for 200K 1/2W

>This seems safer than touching that shit with a wet object.

the idea is you roll it up like a cigarette, and touch with one end while holding it safely at the other end. as long as your other hand isn't touching the circuit, there can be no shock.

>>1179576
nope, it uses a generic li-ion shrinkwrapped pack.
as long as your voltage is correct (it's 3.7v in these things mostly) it will be fine

File: wut.jpg (45KB, 300x562px) Image search: [Google]

45KB, 300x562px

>>1179668
Well, that's just fucking embarrassing, 365V as predicted. Thanks for the catch. Onward forth.

>>1179670
How did you calculate that with only a volt value?

>>1179678

power = V squared, divided by R

Am i okay bros?

this is right next to the counter that i work at, the wires lead into what i assume is the lights, it it dangerous? they seem live, even though 2 of them are capped

https://imgur.com/a/4gRlz

here is a close up + another box with wires as well

How do you solder smd parts with plastic like PLCC sockets? If I go at it with hot air, the plastic melts before the solder does.

File: PLCC_Socket_SMD_Mount.gif (66KB, 411x308px) Image search: [Google]

66KB, 411x308px

>>1179797
Use less heat, use an oven/hot plate or get sockets which don't melt.
Or solder it manually from inside. Cut off the bottom part if it gets in your way.

File: citcuit diagram.jpg (21KB, 449x620px) Image search: [Google]

21KB, 449x620px

I want to build my own stereo 4-channel audio mixer as an easy novice project.
I was going to use pic related plan but somebody said that I should put an op amp on each circuit so one circuit doesn't change the tone of the others. Does this really matter though if I'm just mixing volumes?
The finished project will actually be used to mix audio in my entertainment centre so I don't want it to sound like shit.

Thanks for any help.

>>1178592
Thanks but I've taken several classes in digital logic etc.
Good advice for other beginners though!

i find digital pcb routing to be very boring. are autorouters completely worthless?

>>1179857
Develop your own autorouter. Be the hero we need.

What are the "standard parts" for different kinds of transistors, diodes, etc, like the 2n2222, that one might keep in a starter electronics toolbox?

>>1179817
If your output is going to amplified speakers, just the resistors to mix works fine, although your signal will be 12dB quieter than it started, if you want to drive headphones or want to get rid of the 12dB drop you'd need an opamp on the output.

>>1179857
If you're smart with part placement and manually routing some of the critical traces first, going autorouter can give acceptable results. Never going to look as nice as a well done manual layout though.

>>1179946
General purpose NPN and PNP like the 2N2222/2N2907 or BC546/556 or anything like that are useful.

For diodes it'd be useful to have one for signals, one for high power, and some zeners of a few different voltages. 1N4148 and 1N4004 are common. A few Schottkys like 1N5158 would be nice too.

If you're interested in doing power stuff some generic MOSFETs would be good, FQP30N06L is common for N channel, P channel isn't as commonly used but can make simple circuitry easier, FQP27P06.

>>1179784
So long as you don't go poking it should be fine. Sticking it in a box would be safer.

File: 7.jpg (344KB, 1249x715px) Image search: [Google]

344KB, 1249x715px

>>1179857
There are plenty of autorouters, some better than the others. The general rule of thumb is that you route the critical stuff manually and let the autorouter to do the rest.
And no, they aren't worthless. They can help quite a lot if you have a large board with plenty non-critical routes.

>>1179961
The mixer is only for an (amplified) speaker system, so that's just what I needed to hear. Thanks a lot

Is this a good deal? http://www.befr.ebay.be/itm/Lavolta-BPS305-Variable-Linear-Adjustable-Lab-DC-Bench-Power-Supply-0-30V-0-5A-/191766020330?hash=item2ca62540ea:g:d7sAAOSwCQNWerxA

File: o silly scope.png (7KB, 800x600px) Image search: [Google]

7KB, 800x600px

20Hz-20kHz oscilloscope using $3 of parts and some sound software? Has anybody done this yet?

>>1180118
oh my god why

>>1180119
Why not?

File: CAM01328.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

How do I solder this without fucking it up? I have no experience repairing a flat flex.
If I can do this Ill have a new phone for free, but its a battery connection so I dont want the thing exploding in my pocket.

>>1180120
You can connect the signal directly to the sound card, just use a capacitor for coupling. There are a lots of videos on how to do that.

Using leds and photorresistors is a laughable idea, since they're not linear and have a slow response.

>>1180122

Maybe you can scrape the coating and bridge it with some enameled copper wire. I would try to connect it directly to the place where the flat flex starts and ends.

>>1180122
Gently scrape off the coating in a small patch on both the positive (small, probably) and negative (large, on either side, probably) traces on both halves, then use some small silicone-coated stranded copper wire and bridge them, then cover the exposed joints with a couple wraps of kapton tape.

That, or just buy a new flat flex assembly. They're available for most phones. Probably have to search something along the lines of "diy battery replacement [phone model]" on ali

>>1180130
>attach random voltage source to sound card with no protection
No thank you. I think the linearity wouldn't be that big of a problem, especially for evaluating digital signals or just frequencies, but it looks like the latency could definitely prove a problem. I guess using a high-linearity analogue optocoupler is a better idea, now that I know what one of those is, but they do have a minimum current. Maybe I could bias the input slightly, but then I'd have to bias the output too, and that makes it ten times as complicated (which isn't saying much).

File: 1491304330569.png (101KB, 1533x539px) Image search: [Google]

101KB, 1533x539px

do you prefer fewer components or a smaller bom?

I thought you guys might know: I need to fix an electric guitar's input socket but as it happens I don't own a soldering iron. Can I substitute this with something like duct tape or will it catch fire or something? It's just two wires that need to touch the socket.

>>1180150
it won't catch on fire and if you wrap the wires around the leads and get decent contact then taping them on should be reliable enough as long as there's no strain on the wires.

File: connect.jpg (8KB, 351x276px) Image search: [Google]

8KB, 351x276px

>>1180137
Then add some protection. No need to go with full galvanic isolation for that.

>>1180149
Depends, but generally I rather go with lower component count.

>>1180154
Sigh couldn't get the tape to hold, just too little space to work with. Guess I'll have to get a soldering pen or something.

>>1180122
If the flex is part of the battery just buy a new battery.

File: acpl790block.jpg (9KB, 298x209px) Image search: [Google]

9KB, 298x209px

>>1180137
Use an isolation amplifier with gain set to 1 then (maybe with variable gain so you can scale up small signals).
https://en.wikipedia.org/wiki/Isolation_amplifier

I'm building a diy synth at the moment and want to put a small fan in the case to keep the voltage regulators from getting too toasty.
I picked up a 40mm dc brushless fan and I'm just wondering if it could cause much noise on the audio signals.
Someone mentioned putting a capacitor on the fan's supply.
Does that just mean putting a cap between the ground and 5V cables on the fan?

>>1180215
http://www.powerampdesign.net/images/AN-24_Eliminating_Circuit_Noise_From_Cooling_Fans.pdf

Your capacitor idea will work. Use any electrolytic and a 100n ceramic.

>>1180218
Cheers man.
I've a 100 micro farad electrolytic cap.
I just solder that and the ceramic in parallel?

>>1180219
Yup. Make sure you get the electrolytic polarity right.

>>1180221
Also try to keep the wires from the caps to the fan fairly short, or otherwise don't run them near any analog stuff.

>>1180221
Cool, thanks a lot.
I've had enough electrolytic caps explode in my face when testing boards in work.

>>1180223
So ideally, I'll solder them on at the fan rather than the connector on the PCB?

>>1180225
Yes. if it's impractical to do that then it's not mandatory. It's more important for the ceramic to be close than the electrolytic.

File: s.jpg (2MB, 3264x1836px) Image search: [Google]

2MB, 3264x1836px

>>1180227
Done.
Thanks again

I just picked up a midi controller but it only has a USB connection.
The synth I want to use only has a 5 pin midi connector.
How do I go about connecting them?
Can it be done directly or will I need to go through a PC?

File: CAM01327.jpg (3MB, 3264x2448px) Image search: [Google]

3MB, 3264x2448px

>>1180130
>>1180133
>>1180184

The other side also has only two traces so I expect the flat flex to be doublesided. Also I cant seem to get to where the cable runs, its just weirdly assembled together.
Should I be worried about the briding wires having enough meat to run all the power? Also how sensitive is the ribbon to heat? How likely em I to just burn through the thing?

File: 1487145978118.png (846KB, 1117x1779px) Image search: [Google]

846KB, 1117x1779px

just finished designing a new project. it's a bunch of slider pots with a usb i/o controller. the idea is to be able to use it to adjust volume and the brightness on my monitors. the gimmick is that there's no microcontroller to program because i hate programming microcontrollers.

Any good idea projects for an old Android phone?

>>1177293
Why isn't Bigclive in the list of youtube channels?

Sorry to be a bit of a refugee, but where is the home for Amateur radio threads?
I can't find a hamfag general anywhere.

>>1180307
You can always make your own, but this board is pretty slow so I'm not sure if it's worth it. There are even separate generals for arduino and raspberry pi but I honestly don't understand why. It's all electronics, why not keep it in one thread?

>>1180314
i remember people used to get autistic about arduinos. i agree they belong in the thread. rpis tend to be home server bullshit so they might not belong but i don't mind them.

>>1180305
>why isn't this or that included
Because OP is at maximum length and no-one bothers to update pastebin.

>>1180307
Yeah, you can't. We have those occasionally, though, and they go reasonably well. Start your own if you can't be arsed to wait the next one.

>>1180334
He would have space without the "is anybody reading this?"

>>1180273
Those look like ground traces to me. Scratch a patch of coating off of each side and check for continuity.

If it's right on the battery, it might get a bit warm but nothing serious, you could use PVC wire if you had to but I wouldn't recommend it because it's shitty.

Check a wire gauge chart with current ratings for each size, and pick the smallest one that can transmit like 3amps.

>>1180307
There was one for a while. Can't be bothered to check the catalog but it could come and go.

>>1180137
Sound cards are generally AC-coupled anyway since they only have to deal with audio.

>>1180188
Those look like exactly what I had in mind.

>>1180443
I guess that means lower frequencies or uneven duty cycles would face some attenuation, huh.

File: IMG_20170520_194824.jpg (3MB, 4048x3036px) Image search: [Google]

3MB, 4048x3036px

What size connector should I look for on probes for this multimeter?

>>1180466
>take calipers
>do inside measurement
>find banana plugs with that diameter
>?????
>profit

You're not expecting people to tell you what size something is with no frame of reference, are you?

>>1180470
That's exactly what I expected.

>>1180466
4mm
i mean the hole won't be 4mm if its imperial, but thats all your going to get close to it.
worst comes to worst replace the sockets if necessary.

How would I go about temperature compensating a log amplifier? The diode connected BJT variety.

>>1180487
You could surely do it easily with a µC, but analogue? You'd have to find something that varies at the same rate with respect to temperature as the critical component in the amplifier (probably the same component), and feed a voltage divided across that into an ADC then into a digital pot to control your gain. Or you could use one of >>1179534, no promises.

>>1180502
I would use an MCU but I am terrible with programming. Like really bad. It took weeks, no exaggeration, weeks for me to write a program to read an analog voltage and output the value to four 7-segment displays. Another week to figure out how to convert the display from reading voltage to reading dBV after toggling a switch. Another week to get it to run standalone off the chipkit dev board. In the process two PIC32s were fried.

Part of the issue was most of my beginner programing reference material was for the Arduino. Anything else for for a PIC16 or something that used old MPLAB instead of MPIDE or MPLAB X which is what works with the PIC32. Some of the Arduino stuff was helpful but a lot of it just muddled things and made it even more confusing.

>>1180506
Why not use an AVR then?

File: clock division.png (3KB, 484x255px) Image search: [Google]

3KB, 484x255px

Is this how kHz/MHz clock frequencies are divided for things like RTCs and such? Just wondering if it takes more silicon than a bunch of chained T-flipflops.

>>1180534

nope. usually involves a series of toggle flip-flops, or D-type flip flops, that divide by 2 for each stage, then maybe one specialized stage to divide by 5 or some other weird number.

typically, you wanna use as few stages as possible for low power, which is why 32kHz crystals are a popular value for RTCs.

>>1180466
that thing uses ''standard'' banana plugs you find in all kinds of measuring equipment, which is about 4mm.

How does one create an EMP the size of the planet? Sounds like an interesting project.

>>1180554
Short out an inductor the size of the planet.

>>1180558
Need feasible way, thanks.

>>1180561
zap the iron core of the planet super good

>>1180561
Acquire over 9000 nukes and carpet bomb the planet.

File: logamp.png (59KB, 980x452px) Image search: [Google]

59KB, 980x452px

>>1180487
How come you are aware of this, but not the methods to overcome it?
Why not use a ready-made log amp IC with built-in temperature compensation?

>>1180506
After that much practice, the next round should be much easier... Also, >>1180512.

>>1180546
They use 32.768kHz because 32768 is 2^15. This means you only need to divide it by 2 a bunch to get to 1 second, and then I guess you'd use a shift register to get the /15s you need to get 60 and 3600 seconds.

File: Screen Shot 2017-05-21 at 9.41.27 PM.png (18KB, 1162x138px) Image search: [Google]

18KB, 1162x138px

>>1180696
Also holy shit how do I search for these on aliexpress without pulling up jandals.

>>1180702
Just use JK-FFs.

>>1180702
Look up a list of 7400 series (or 4000 series) to find what you want
>https://en.wikipedia.org/wiki/List_of_7400_series_integrated_circuits
then find it by number.

If you're just dividing by 2 a bunch of times, you could use a binary counter like 7469

File: stepup+converter+dc+oscillator2.jpg (46KB, 979x463px) Image search: [Google]

46KB, 979x463px

Is there a way to amplify 0.9-1.7 volts into 5 volts or more without any ICs? I've this circuit but I have no idea what value L1 is.

>>1180708
Actually I want each individual divided output (all 15+ of them) available to use as a preset frequency generator of some sort. Add a 16 bit+ parallel-out shift register or two for custom divisions and you've got essentially any frequency you could ever want, with a little fiddling.

>>1180735
>15¢ ea for 200 flip-flops
>2^200 ≈ 10^60
Now all I need is a bunch of 7-seg displays and a nice sign saying "TIME TILL UNIVERSE ENDS".

Is there a relay rated for 200A@36v? Or is there a MOSFET for it? I need it to switch on for like 1 minute.

>>1180709
An IC is just a complete circuit on a single piece of silicon. Any circuit implemented as an IC can also be implemented as discrete components (well, except for fast, complex digital logic; you wouldn't be able to maintain the clock speed with discretes because everything would be too spread out).

For voltage step-up, the main options are:
* diode-capacitor voltage multipliers
* boost converter
* DC-AC converter + transformer

The circuit in your picture is an unregulated boost converter using a classic astable multivibrator as the clock. The value of L1 depends upon the desired output voltage and current.

A better design would add feedback so that the PWM duty cycle decreases as the output voltage increases.

>>1180766
I thought voltage multipliers for DC require a 555?
The boost converters are ok, but there is none here and when I tried looking up circuits they had ICs like MAX756 which isn't available over here unfortunate, not even the lm2621-5.
Would DC-AC conversion really work with low voltage input of maximum 1.7 volts?
Also >>1180686

>>1180763
You can get relays rated for practically anything, but 200A isn't going to be cheap.

The main issue with a 200A MOSFET is that the package either won't handle 200A or it won't handle the I^2*R power dissipation at that current. Data sheets are notorious for citing impractical specs, e.g. with the junction at 25C (if you're running 100A through a MOSFET, keeping the junction at 25C requires removing the die from the package and immersing it in liquid helium).

But you can put several MOSFETs in parallel and run them under realistic conditions, and that'll be cheaper than a 200A relay.

E.g. PSMN5R8-40YS Vgs[max]=40V, Id[max]=90A, Rds[on]=4.4mohm, £0.31 for singles.

>>1180771
>require a 555
You have much to learn. A 555 is just one fairly common oscillator used to turn DC into a square wave, and all voltage step-up modules have to use AC one way or another.

>>1180774
Alright, is there a way to do that without any external power? In other words, can I turn a small ~1.5 volt input into an ac output without using any external sources?

>>1180771
> I thought voltage multipliers for DC require a 555?
A 555 is just an oscillator. A 2-transistor discrete oscillator will work just as well.

> Would DC-AC conversion really work with low voltage input of maximum 1.7 volts?
Yes. 1V is enough to switch a BJT, and you can work with even lower voltages using JFETs.

>>1180778
That's amazing thank you very much! Is there any specific transformer to increase the voltage or any step up would do it? Electronic components are really scarce over here.

>>1180775
That's what the left-hand side of the circuit you posted does. The voltage on the collector of the BD139 will oscillate between the supply voltage and ~0.5V (which is actually fairly high; look for something with a lower Vce(sat) figure if you only have 1.7V to play with).

>>1180781
At low power, the boost converter is preferable to a transformer. A transformer needs a push-pull drive rather than open-collector.

In mains PSUs, bridge topologies are normally only used above ~1kW. Their main advantage is that they can handle more power for a given magnetic (transformer/inductor) core size, but they need more power transistors (2 for a half bridge, 4 for a full bridge, as opposed to 1 for boost or flyback converters). It's only when the transformer starts getting large that its cost becomes a bigger factor than the transistors.

>>1180788
Unfortunately, boost converters aren't an option, but thanks for the info.

>>1180794
Why not?

>>1180797
1. No one sells them here.
2. I can't make my own because the ICs used in the most common booster circuits like MAX756 and lm2621-5.
3. I technically can order one online but that would either require me to wait for 2-3 months or get it in 3 days but pay at least $50 to get what probably costs around $10, and that's ridiculous for a project on this size.

>>1180799
Where are you? I can't believe you can get discrete components but have such a hard time getting any IC.

>>1180801
Lebanon.
http://www.ekt2.com/
This is where everyone gets their electronics including electronics shops of course, i searched their site and asked them personally by phone, nothing.

File: file.png (1KB, 800x297px) Image search: [Google]

1KB, 800x297px

>>1180802
Huh, that sucks. Well, anyway.
>>1180709
This schematic is a boost converter. Compare to the image. BD139 is the switch. It's being controlled by the oscillator made up of everything to its left. It looks like you can calculate the inductor value based on this info:
http://www.ti.com/lit/an/slva372c/slva372c.pdf

>>1180803
Image totally failed. It's the one at
https://upload.wikimedia.org/wikipedia/commons/3/39/Boost_conventions.svg

>>1180803
Interesting, thanks for sharing.

>>1180799
You keep insisting that you can't make one without an IC even though you posted a schematic for a boost converter which doesn't use an IC.

>>1180772
If I want to go relay route, is there even an exact 200A@36V relay or is there closer value?

How about IRLB3034 in parallels? It says it could dissipate 375W each, so I need 20 of them?

>>1180810
Ah yes I worded it wrong, the one I have doesn't have an IC indeed, should've asked instead for a similar circuit, welp, things are more clear now anyways.

File: RTL-SDR-RTL2832U-e1471375714199-1.jpg (44KB, 732x594px) Image search: [Google]

44KB, 732x594px

I have an RTL-SDR and I've been trying to capture signals with unknown frequencies. It's quite difficult to do so since I only have a 2MHz window on SDRsharp and I have to go searching trough all the possible frequencies until I find it.

Is there any way to find which frequency a nearby (30 meter max) signal has? In theory the signal should be stronger than the rest and easily differentiated, but the SDR can't detect it unless I tune to that particular frequency.

I have a rigol dz1054 scope, but it only goes up to 100Mhz, and the signals I'm looking for are in the range of 300 to 900Mhz.

What I'm looking for is some kind of spectrum analyzer / frequency detector that can tell me the frequency of a nearby signal.

>>1180811
it can dissipate 375W into an infinitely large heatsink. it can dissipate 1W by itself in standing air.

>>1180654
>Why not use a ready-made log amp IC with built-in temperature compensation?

Because they are expensive as shit. Like $20 a pop. I'll use a discrete IC if I have to but for prototyping I'd rather build something use off the shelf parts. Besides I'll probably learn more that way anyway.

>>1180811
That's not how it goes. Your fets won't be dissipating anywhere near 7.2kW.
The nominal maximum resistance of a single fet is 1.7mohm with 10V gate drive. The power dissipation using one fet would be 200^2*0.0017 = 68W. Well, the bond wires can't take 200A (and a 40V fet is a bad choice for 36V supply), but if you use 4 fets in parallel, the power dissipation is now one quarter, or 17W for all the fets, each fet dissipating a bit over 4W. You can increase the number of fets to reduce the size of the heatsink even further. With 20 fets you wouldn't even need a heatsink.

Note that you need to use bigger heat sink than a simple calculation based on the thermal resistances and maximum junction temperatures would suggest to avoid thermal runaway.

>knocked over solar garden light
>LED came on and wouldn't go off
>take it apart to repair
>corroded wire going to the panel had broken free
>melt the hot glue sealing it in
>shaky hands from fucked up nervous system
>ended up fucking up the panel even more
>solder pads ripped off
>panel is now dead

Anyone else with really shaky hands and back problems? I can't do SMD work at all because of it.

>>1177479
Use a ZVS driver and be sure to use biasing resistors across the capacitors to ensure they charge equally.

>>1180954
a robotic arm is a nice diy project

>>1180967
kek, that would probably enhance my life vastly.

>>1179429
Clean it up with some alcohol and a toothbrush. Solder new components around the corroded area.

File: Screen Shot 2017-05-22 at 2.32.29 PM.png (320KB, 2328x494px) Image search: [Google]

320KB, 2328x494px

Anybody care to have a guess (or some inside radio knowledge) as to the inductance of this?

>>1181093
>guess the inductance

350μH

>>1180859
oh ok

>>1180882
so 10 of them with a heatsink is enough?

>>1180954
I got a really shaky hands I could masturbate automatically.

The key is to stabilize your hand with your other hand. You need something or someone to hold the object still though.

>when your physics lecturer pulls out "The Art of Electronics"
>I am ready

File: LAUGHING.gif (1MB, 214x124px) Image search: [Google]

1MB, 214x124px

>>1180118
>the o'silly scope
my sides

>>1181520
The name is what I was actually proud of.

File: Untitled.png (28KB, 768x768px) Image search: [Google]

28KB, 768x768px

converting a gas air compressor to electric. its a 200 psi model and i cant seem to find any affordable 200 psi compressor switches, however there are plenty of 200 psi pressure switches for semi horns or some shit so i figure i can make my own compressor switch. does the included high quality professional grade pictogram look like everything is set up correctly? note that the pressure switch turns on when pressure is under 175 psi and off once it hits 200 psi.
the tank will still have its mechanical overpressure valve installed so if the switch fails the tank will still never exceed 200 psi.

>>1181193
Yeah. You'd probably manage without heatsink if you didn't put the fets too close to each other, but better have it.
You should use fets with higher voltage rating than 40V, though.

>>1181610
Looks great. What pressure does the overpressure switch kick in at? Because if it isn't a good couple of atm higher than the pressure switch pressure then you'll have some annoyingly oscillating hardware on your hands.

>>1181614
200 psi. it shouldnt be an issue because the electric switch doesnt kick back on til 175 so if for some reason the mechanical switch does open exactly at 200 psi and vent down to 190 or so it shouldnt be an issue.

>>1181617
If the mechanical overpressure valve starts opening slightly before the electric switch (quite likely if they're both rated for 200psi), then your motor would run constantly while the overpressure valve cycles between 190 and 200psi.

>>1181627
youre quite right. ill go ahead and try the 200 psi switch but if that starts happening ill just switch to a 175-190 psi switch

File: spurdocap.png (51KB, 1000x1800px) Image search: [Google]

51KB, 1000x1800px

Did I do good?

>>1181613
What is your voltage suggestion?

Also, can I use voltage divider to get 0v-5v signal into 0v-3.3v?

>>1181634
You can, but if you've got a significant load across the divider (~10* the resistance of the divider or less) then you're essentially putting a load resistor in parallel with the resistor on the voltage divider, which will decrease the voltage. Voltage dividers are usually only used in OpAmp circuits where you need a ground 1/2 way in between the voltage rails, but you don't draw much/any current through it.

>>1181636
The signal will be between 7-10mA at 0v-5v, my microcontroller is working in 0v-3.3v and I am trying to read it as analog value.
Will the voltage still be linear?

>>1181637
Depends on the resistance of your divider. 3.3V at 10mA is 330Ω load, so if you use a divider with a total resistance of about 50Ω you'll get about 3.2V across the µC, but you'll be constantly sinking 1/2W through the divider, which is pretty bad. Whether you can adjust the divider to get a voltage droop and drain current you can handle is up to your tolerances, but I'd just use a Zener diode voltage regulator instead. 3.3V Zeners are pretty common.

>>1181638
Wait, I could use a zener diode for that? Is that a better way than this resistor voltage divider?

File: zener supply.gif (7KB, 364x242px) Image search: [Google]

7KB, 364x242px

>>1181641
Most definitely. If you didn't already know, Zener diodes have a reverse breakdown voltage at a precisely fixed level, meaning if you put more voltage then that voltage, the current through the Zener increases more than exponentially. If it decreases, the current through the Zener goes very low. This means that when we put a Zener diode in series with a resistor, regardless of the value of the resistor and voltage of the source (up to a certain point), the voltage across the zener will remain constant.

Note, if the resistance is too high, not enough current will flow through the Zener and the voltage across it will decrease to the point of no longer being in the Zener region, thus decreasing the voltage across it sharply. If the resistance is too low, the Zener diode will be conducting too much current, and will therefore be dissipating too much power. Always keep in mind the power rating for your Zener diode, and make the maximum current through your diode enough to get to half this power value for safety's sake.

When designing a Zener power supply such as pic related, your main concern will be with the power dissipated by the diode. If it's a 5W zener diode, then you limit the power through it to be 2.5W, which through P=I*V gives you a maximum current of 2.5/3.3 = 0.76A, or 760mA. Since the voltage across the series resistor will by the law of voltage dividers be the value of the Zener subtract the voltage of the power supply, or 5-3.3 = 1.7V, we can calculate its resistance through R=V/I. Now here's where you decide whether your load is running 24/7. Since it's a µC and can draw a variety of currents, let's say it isn't on 24/7, and take the series resistor current to be the only current when it's off. If you were using an LED or something you would add the current through the LED to the current through the Zener to get the current through the series resistor, but we won't do so here because the Zener diode would dissipate more power.

cont.

>>1181641

a zener is better coz the load will not be as significant a factor in determining the output voltage. make sure you dump enough current into the zener to make it work. you need about 5-10mA for a low-power (quarter watt) zener.

>>1181637
>signal will be between 7-10mA at 0v-5v,
>microcontroller is working in 0v-3.3v
>read it as analog value.

Just use a divider. A 0-5V signal through a 1.2K then a 2K to ground will give 0-3.1V between the resistors. Lower value resistors would be more immune to noise, higher would use less power.
Tweak the ratio with more resistors if you want a range closer to 0-3.3V. (3.3V/5V = .66, so you'd want the larger resistor to be just a hair under twice the smaller one's value)
>Will the voltage still be linear?
Yes, you're drawing practically no current through the uC.(disable pullup/pulldown resistors if that applies)

Not sure why other anon seems to be talking power curcuits and regulation when you seem to just be trying to interface 2 different voltage levels, unless I misunderstand.

File: transistor zener.gif (2KB, 325x150px) Image search: [Google]

2KB, 325x150px

>>1181658
cont.

Since R = V/I, the series resistance is 1.7/0.76, or 2.24Ω. This is quite low, and the power across it will be above 1W. Here you're really starting to pull power, with a total dissipation of the circuit in standby mode sitting at almost 4W. This is the minimum value for our resistor.

Try larger resistors until the voltage across the Zener is no longer 3.3V, then go back again to a value where it is 3.3V under load and idle.

>>1181634
Typical recommendations for mosfet derating are between 75% and 80%. Based on this, you'd use a part with at least 45...48V rating. Well, assuming that 36V is really the maximum voltage seen by the fets. If it's actually 3 car batteries in series, the fets might see around 44V, giving a 75% derated value of about 58V.
So, 50V or 60V mosfets would be suitable, depending on the actual maximum voltage.
Additional spike protection is probably still needed.

>>1181633
Saved

New idea: EL wire headphone cables. By putting the audio signal itself through the EL wire, you get a wearable VU. Also it needs to be at 100V, so I guess you'd put a big resistor before each driver.

>>1180150
A soldering iron + solder is less than 20 usd at a hardware store

File: w.png (32KB, 1024x768px) Image search: [Google]

32KB, 1024x768px

Right, I need to hook up some shit in a hurry and just want to make sure this is acceptable.
Basically I'm in Europe and need to power two power supplies that take 120V.
I don't have a proper 240V to 120V transformer but I have a transformer with two 120V primaries and two 40V secondaries.

Is it okay to connect the primaries in series and take 115V from each primary?
The secondaries won't be used.
Obviously one live will still be 230V in relation to earth so I was going to put varistors to Earth just in case.

File: servo-gearbox-illustration.jpg (20KB, 454x340px) Image search: [Google]

20KB, 454x340px

Guys, is there such a thing as a "ninety degree gearbox" for servos? Something like pic. I've been searching ebay with every term I can think of but can find nothing.

>>1182510
a wormwheel?

>>1182472
kind of but not really.
try to balance the load as much as possible you might get away with it.
hope it's not expensive!

>>1182512
No, just something so that I can mount the servo perpendicular to the thing that needs to rotate. You know that little knob you place the servo horn on? I need that to be placed on the side of the servo, rather than on top. I was hoping there might be a "gearbox" or something.

>>1182510
try servocity actobotics

>>1182513
Both 115V supplies will be powering the same model of power supply

>>1182514
like with bevel gears?

>>1182514
>>1182539
Bevel gears, man. You could easily make one yourself, but I'm sure there's a commercial product out there as well.

This is a stupid question, but can I use a 1/2 resistor in place of a 1/4 resistor? I'm assembling a Z80 computer kit and I'm missing the last resistor. They are both 2.2k Ohms

>>1182572
You can always use higher wattage resistors. It's called derating. The 1/2 should have a longer lifespan than the 1/4.

>>1182572
link to kit?

>>1181658
>precisely fixed level
Zeners are as bad as diodes. Not using them at least at like 20mA puts them at 10-20% below their nominal voltage, and higher current puts them above.

>>1182472
>>1182518
Think of effectively what you're doing. Assuming the transformer has near infinite impedance when there's no connection on the other sde, you're making a voltage divider with 2, let's call them, 10k resistors. Additionally, even if you could use the power supplies with what little current is available, the load would go unbalanced, just as a resistor divider would.

Just get a fucking adapter m8. You could even slap two "power" transformers back to back, so across one it steps down, and the other steps back up, but while your 240 is across both primaries, you tap 120 off of only 1.

There's is literally no way to hook up a single transformer to do what you want to do.

>>1182539
>>1182541
That's it, thanks a lot.
I'm clearly an idiot, bevel gears were what I was thinking of all along, I just didn't know that they were called that and didn't even think of just building something myself. I'll just bend a sheet of aluminium, drill some holes in it, and use that.
>>1182516
Thanks, but they only seem to have reduction boxes. I'll just use bevel gears and a homemade hinge.

>>1182602
Pretty sure a current at the standard 1/2 power rating should be about dead on with the Zener's rating, go any higher and you're asking for trouble. 20mA sounds like a good lower limit to set, thanks.

Hey guys could i get help with this problem please?

I need a replacement for a 5w 220ohm resistance. There is no way to get a 5w resistance in my city and buying one/waiting for it to arrive would take a lot of time in my country.
What resistance would be a good replacement for the one i can't get?

>>1180828
You can write your own scripts for SDR# to search around its available spectrum and find peak amplitude. One probably already exists.
415 and 915 MHz have some interesting stuff on them. Found a NBFM baby monitor once. kinda creepy

>>1182579
https://www.tindie.com/products/Semachthemonkey/rc2014-homebrew-z80-computer-kit/

Your options for parallel equivalents:
2x 440 2.5W
3x 660 1.67W
4x 880 1.25W
5x 1.1k 1W
6x 1.3k 0.83W (approx, 217)
7x 1.5k 0.71W (approx, 214)
8x 1.8k 0.63 W (approx, 225)
9x 2k 0.56 W (approx, 222)
10x 2.2k 0.5W

>>1182678
Love you anon, thanks!

>>1182658
Can you easily get 2 or 1W power resistors?

For 1Ws you could use 5 1K in parallel (I'm assuming the value isn't too critical), or 5 47ohm in series

>>1182669
Was the baby calling Skykings?

>>1182605
If your logic were correct, transformers wouldn't work. The two windings are magnetically coupled, so it's basically the same setup as in a variac. The output impedance is determined by the winding resistances and leakage inductance.

That said, it might still be a bad idea, requiring a grossly oversized transformer to attain reasonable balancing.

>>1182837
I wired it back to back with and identical transformer so the 240 is being stepped down to 40 and then back up to two 120s. All's working good. Cheers for the help

>>1182472
You realise you've just shorted the live and neutral together through two varistors, right? Not to mention shorting them both to ground through another varistor.

This is a fairly beginner question, but im making a digital clock using a 555 and a few 7490s, 4511s and 7408s in series, with a display (HH:MM:SS). When I've got the 555 timer set up on a 1 second delay with values:
>R1 = 1k
>R2 = 330k
>C1= 2.2uF
I get an extra second every 128 pulses. How do I hold the 555 timer high or low for that extra second? I'm thinking of using some kind of capacitor that starts discharging when the 128th pulse goes off but im not sure how to implement it

>>1182658

another option is light bulbs from a car, or christmas lights. i just pulled out a 7W night-lite bulb and it measure 210 ohms.

>>1182938

u best be joking nigga. nobody builds a 555-based clock that needs to be corrected all the time. to avoid mass frustration you need a crystal and frequency divider chip. if memory serves, last time i did a retarded project like this, i used an LM5369.

>>1182993
or use the mains frequency

>>1183010
That varies. I'd use a crystal and a RTC.

What's the worst that can happen if I use a normal 12V power supply instead of a proper LED driver board?

>>1183089
flicker

>>1183089
Depends what you're trying to power. If it's a 12V LED module with its own drive inside, it's fine. If it's just the bare LED package, you'll most likely push too much current and fry it.

>>1183094
It's a single traffic light I've had knocking around for a while. The drive board got wet and fried some resistors and I can't find specs for either the LED board or the drive board to A) replace the resistors on the drive or B) replace the drive board.

I'm just going to wing it, if it fries the LEDs, hell, it didn't work anyway.

>>1183099
Have a look at the board with the LEDs, if it's got a bunch of resistors or some chips, it should be designed for constant voltage supply.

Haven't seen many that use 12V though, most are higher.

>>1182837
When transformers aren't attached to anything, they can't couple, and only their parasitic components account for drain. If you wire them the wrong way on dual inputs they can short, but that's all you can do. without another connection being used.

>>1182856
Do be careful, if you pull too many amps, the core saturates and basically looks like a short.

>>1182938
>I get an extra second every 128
You make your clock 128/129ths slower.

>>1182658
a 10-15W hair curler has a heating element that's 700-900 ohms. Cut to use, solder ends on it, wrap around something, then tape it with electrical tape. (in layers)

new thread:

>>1183261
>>1183261
>>1183261

>>1183248
The two primaries are inductively coupled and transformers don't really differentiate between primaries, secondaries or tertiaries. You can leave some windings unconnected without it affecting the operation of the others.
Ask your favorite simulator if you find this somehow difficult to accept.