/diy/ - Do It Yourself

>>1158696

Old thread reached bump limit.

I have 15 UV LEDs and I want to make a UV flashlight. What is the best circuit to make them all work efficiently?
The forward voltage is 3.5 volts, so it would take 4xAA batteries and 15 resistors to make them work.

I'm looking for a circuit (maybe with a dedicated IC) that can step up the voltage and drive 15 UV LEDs from a low voltage like 2x AA batteries.

>>1167010
Personally I'd just run them all from as many Li-Ions in parallel as you wanted with a little dropper-resistor, probably just a single 18650, but if you want to use 2 AAs, maybe a joule-thief would work? It's probably the simplest circuit, but definitely not the most ideal. Since you're considering dedicated control circuits I can assume that you'll want something more sophisticated, to which I'd suggest buying one of those <$2 chinesium boost/buck/whatever converters from AliExpress or ebay. I'm not sure if dedicated LED drivers work at such voltages.

Have you done some wattage calculations to ensure two AAs can keep up? They should unless you've got particularly bright LEDs, but you never know.

Does anybody know a decent and not-expensive front panel fabrication service?

>>1167010

you can get these complete units for about $10 on ebay, or $2 at the thrift store. they work on 12V wall warts or 3 LiPos in series. great for showing people how nasty and dirty their sinks and counter-tops are (not to mention where the dog pees and the kid spreads his cum on the walls)

>>1167019
>Personally I'd just run them all from as many Li-Ions in parallel as you wanted with a little dropper-resistor
The problem is the voltage difference between the Li-ion and the LED voltage is not too high, only around 0.7 volts at best, this means the current will vary greatly as the battery discharges, that's one of the reasons I want to use an inductive circuit.
> I'm not sure if dedicated LED drivers work at such voltages.
The ones I've seen operate at 4 volts or higher. I'll probably use a 555 CMOS, and inductor and a transistor and tune the circuit to get the voltage I want.
>Have you done some wattage calculations to ensure two AAs can keep up?
Yup, 15 UV 20mA LEDs consume 0.3 amps in total, or 0.02 amps at 52 volts, with 2AA batteries feeding a boost converter the input current will probably be around 0.6 amps max, assuming some losses, which is reasonable.
>>1167036
I'd rather just /do it myself/ from things I already have.

>>1167044
http://ww1.microchip.com/downloads/en/DeviceDoc/20005408A.pdf

So i got my hands on some 24 volt hid flood lights that i would like to use in a 12 volt system. Is it possible to use a step up converter for this or do i need to replace the ballasts?

>>1167060
extra details required
though its mostly that i'm not sure what ballast means in this context

>>1167010
>>1167036
well, UV =/= UV. A/B/C??? also you need luminol for that to happen. only use for a uv flashlight with LEDs might be inspecting gems and bank notes.

>>1167074
>UV C
dat cancer machine

Aw fuck yeah. Got this Philips PM3350A combiscope for free.
Very pleased with this being my first scope.

Can anybody tell me what this adapter is called? I've got a busted power supply where this plugs into a bit wired to the board, and I would like to match it so I don't have to solder the new supply directly to the board.

>>1167398
Fuck, forgot image. Pic related.

>>1167342

>>1167480
Many keks, anon.

Does anyone know anything about this soldering iron? I kinda want to but it.

https://www.amazon.com/X-Tronic-3020-XTS-Digital-Display-Soldering/dp/B01DGZFSNE/ref=sr_1_1?ie=UTF8&qid=1493177875&sr=8-1&keywords=xtronic

Today I was gifted a variac from my school, it's a nice beefy 10 amp model and it's just the transformer with bare screw terminals (no power outlets or whatever) I could use some input on how to make this into a proper bench power supply, I want to make sure I cover most use cases so I can plan around integrating them into the project early or leave room for future expansion. I want this thing to suffer feature bloat.

Ideas so far
-Circuit breaker and power indicator
-Standard ac outlet as well as banana jacks to connect bench loads
-Rectifiers/caps for DC output
-An isolation transformer
-Surge/inrush suppression
-Voltage/current output measurement
-Possibly a 2:1 transformer output so if I am working with lower voltage stuff I can vary the voltage more precisely without fear of overvolting by moving a 10th of a turn

I already have a DC bench supply for low voltage stuff so that isn't the primary focus of this project.

>>1167044
Honestly the best course of action for you would be to power this from a USB power bank, you can get some that are barely bigger than the lion cell for a few bucks at five below, poundland or whatever your local Chinese crap purveyor is called. The charging, discharging and boost circuity is all included and the five volts provides plenty of overhead for your 3.7 volts. If you are lucky some versions of the chargers even have spring tabs holding the battery once you open it up (as opposed to most I've found that have the wires soldered), and that would allow for quick cell swapping.

>>1167507
>that average rating
>that rating distribution
looks good to me

Sorry for asking a lab/homework question, but I can't seem to get how PSPICE got this voltage over here.

>>1167613
There's 15V on the top end of R1. 114uA through a 22k resistor causes a 2.51V voltage drop, and the resulting voltage is what your simulator says.

>>1167613
15V - 22kΩ * 114.2μA = 12.49V

>>1167622
>>1167624
Oh nice I see it now, thanks.

I'm kinda struggling to get the design part of these stuff right, I get my resistor values for everything, but Rs (R7) keeps fucking me up, I ended up in triode mode in 2 labs and I have to redo them.

Does this amplifier look alright to you Anon? I'm trying to figure out a way to get a better Rs value, via Vgs, but I don't know where to go.

>>1167627
Actually I think I got this. I have different values now, but I assume a Vg value to get, then since I have the current Id from Rd, I do a Vg = Id*Rs and find Rs that way.

But the thing is, I need to get Vg from R1 and R2 because in this case, the input impedance is given, and therefore, I have to stick with a high (7.5V) Vg. I don't know if that makes a big difference or matters.

Does it make it go to triode mode? Or is that when Vd and Vs are the same?)

I'll go to the lab in a couple of hours and test it and see what's up. God, not having a power supply, function generator and an oscilloscope is a pain in the ass.

>>1167630
>Vg = Id*Rs
I mean Vs = Id*Rs, then Vgs = Vg + Vs, and Vg is gotten from voltage division via R1 and R2 with that 15V

>>1167630
>the input impedance is given, and therefore, I have to stick with a high (7.5V) Vg.
There's an infinite number of R3/R4 combinations with an equivalent parallel resistance of 120k. For example 360k and 180k. (I'm assuming you're not required to take the mosfet impedance into account.)

>Does it make it go to triode mode?
>Or is that when Vd and Vs are the same?
That too, but it's a more inclusive term. See https://en.wikipedia.org/wiki/MOSFET

>>1167638
>linear mode aka triode mode
Damn, so I need to be in saturation mode so I can get credit, I did put this in the falstad.com circuit simulator and hovering on the mosfet, it said linear. Assuming I failed it with my new resistor values again. But it went to saturation mode when I had a very low reaistor on Rd.

How do you check if it's in saturation or linear in a oscilloscope though? Do you have to conpare like it says in there?

For linear/triode
> VGS > Vth and VDS < VGS – Vth

For saturation
> VGS > Vth and VDS ≥ ( VGS – Vth)

For the project I'm doing with my arduino UNO, I need to make a zero crossing detector, I'm using pic related(with 10k resistor between VCC and output) and I'm trying to trigger interrupt with falling edge.

It should, and as far as I know it does generate short 5V spike on the output whenever the voltage crosses zero, but for whatever reason the interrupts won't trigger at all. I've tried to replace the detector with button to check whether it's my fault or the board is broken but obviously there's something wrong with me.

Any ideas?

>>1167700
try using the analog comparator. you've got a <0.1V/us slew rate going on and i don't know that the 328p has schmitt inputs.

I wanna hook up a fucking shit ton of leds. Not sure how much juice the pull bit will running 25amps worth to a 30 amp fuse be okay or shoukd i back it off some.

Also, ina car what guage wire should i use for a shit ton of 30 amp circuits?

My car is an antique Oldsmobile. I will be running a large windo unit ac in the trunk, a large audio system, maybe air ride, air compressor for train horns, and lights enough to see it from mars. If i did the math and measurements right my machineist is making a bracket for 8 alternators. I will simply add stock alternators untik i am full and upgrade to 240amp irragi. I dont give a fuck what i put in it. 8x240= near 2,000 fucking amps.

I may even put a breaker on it to plug into my house if the power goes.

Does anyone know where I can get a custom made coil wound on a one off basis eg a primary for a ferrite transformer?

Also has anyone use a reverse engineering service to generate a circuit diagram from a double sided SM board?

>>1167802
>custom made coil wound on a one off basis
Do it yourself.
>reverse engineering service
Do it yourself.

>>1167802
if you're using ferrite you surely don't need too many turns. why can't you just wind it yourself? magnet wire is kind of expensive but for smps xfmrs you can just get the smallest spool available and have plenty extra.

transformer tape is a meme btw. if you need robust isolation use a split bobbin.

Hey guys how would I go about downloading firmware or software off of card reader? Not info of card issuer just the software, I can hook it up to a PC, but I might be able to solder some wires to some pads, and maybe find a way, any ideas?

>>1167647
Falstad's thing most likely had a different mosfet model than your school's simulator.

It should be enough to check the situation at input waveform's positive peak. Measure Vds and Vgs (two probes or a differential probe), find out Vth and compare.
You could also check that 140uA (or better, Id at input signal's minimum) is above the current used to define Vth.

>>1167818
I would but I want it wound without lumps and kinks and to save time.

I would reverse engineer the board but am having trouble getting a decent picture to work from, i don’t have studio quality photography and the solder lumps mean it wont sit in the focus when if i try to scan the board plus how to scan the side with the components on.

>>1167834
These are 2 different projects. The primary I need is from an old Brandenberg HT psu, The split primary and secondary are one encapsulation. One of the pri winding is open cct so I want to just create a complete new separate pri side as the sec. is fine but I want a professional job though. I have analysed the good winding half of the pri. with and inductance meter and wire gauge tables and there are just to many turns to make it practical for me to do a good job, I may have to though but winding coils is so bloody boring!
Transformer tape does have its uses though if you really must have hi-pot isolation/opto coupling.

>>1167872
>having trouble getting a decent picture to work from
hold it in your hand
follow the traces and map them to paper

>>1167713
It's long since that post but I've tried it, with two different circuits. First was exactly the same as before, second got rid of graetz bridge. In both cases I've used lm 386(the only thing close to comparator I've had) in open loop configuration.

It still doesn't work. What's weird is that I've seen people on the internet doing it without comparator(just resistors to graetz to transopt to output) and it worked so what the fuck?

This might deserve its own thread, but I'll see how knowledgeable /ohm/ is on the subject first.

So soon I'll be wanting to start etching my own PCBs, partly because it gives me more freedom and experience then just getting it from overseas, and partly because shipping to NZ has got to be awful. I've seen instructables around showing how this can be done through electrolysis, which pulls the copper off the board and into solution, the only problem is that it still leaves you with a difficult to dispose of solution. Since this is because of the toxic transition metal ions, I was wondering if anyone here had any clues on how to electrolytically replace the copper with a far more harmless cation, both collecting a few cents of copper metal and allowing me to pour the waste down the sink.

>>1167957
just dump it in your neighbour's yard

>>1167959
Thanks, I knew someone around here would be lurking just ready to pounce with some positive feedback. I'm not hippy, but I don't want to dissolve my neighbourhood's pipes and neither my neighbours nor myself have a backyard, due to cramped city living. I'd consider disposing of the solution in homeless people's mugs while they're asleep, but I don't think it would be an efficient use of time.

>>1167966
its very reusable.

I've had the same liter of ferric chloride for ~10 years and done hundreds of square inches of boardds

>>1167974
I've heard bad things about fumes from ferric chloride, do I need to worry?

Also could I put iron and copper electrodes in the spent FeCl solution with a battery hooked up and turn it back into ferric chloride?

>>1167974
What method do you make your boards with and is it good enough for fine pitch SMD chips?
Because that's what I need
Differentfag btw

>>1167872
>without lumps and kinks
Let me guess: you have a roll of wire lying on the floor, you keep the coil former in one hand and use the other to wind it? That induces twist, leading to kinks. Rotate the coil former instead, put the roll of wire on a suitable stand and keep the wire taut. Use a lathe or an actual winding machine if possible.

Well, it's not like you can't order custom-wound transformers in small quantities, but an electric motor repair shop might be a better option for one piece.

>reverse engineer
See >>1167898. Or use better camera or take more pictures.

>>1167957
Americans pour large amounts of copper sulfate to drain to kill roots and ferric chloride is used in water purification plants. So, if the American way is good enough for you, you can just (neutralize the solution with soda and) pour it to drain with plenty of water.

But like the other anon said, it is very reusable. If you dislike the drain option, bring the old solution to a hazardous waste collection point when you finally want to get rid of it.

>>1167976
The primary problem with ferric chloride is spray and splashes, not fumes - at least if you don't add something nasty like hydrogen peroxide to it.

>>1167921
Okay, I've checked documentation and ATmega328P has schmitt triggers on inputs.

I've made it work differently though, I've triggered the interrupt on any logic change and used the second circuit but doubled on 817's which resulted in what was effectively 827. Nothing outside of it.

Just in case if somebody would stumble on the same problem.

>>1167010
Wire all the LEDs in parallel with a single current limiting resistor for the whole bank. The current through that resistor should be 15x20mA = 300mA. With three AAs, your voltage will be 3x1.5V =4.5V. 4.5V-3.5V =1V at 300mA means a 3.33 ohm resistor that will need to dissipate 300mW.

>tl;dr get a ½W 3 ohm power resistor. You only need three AAs assuming alkaline cells of 1.5V. Redo the calculations for rechargeable cells using their nominal voltage if you intend to use them

>>1168116
While that seems to work, its actually better to put an individual resistor for each LED, as LEDs dont' have identical characteristics, and even very slight changes in those characteristics means they draw more or less of the current and can end up with non uniform brightness in your LED bank

Just finished this little continuity detector because my $10 multimeter doesn't do the beep, and I put an LED on the thing too, both switchable independently. It's a simple little project, but I'll be happy with it as soon as it stops stinking of 15 year-old danish oil. I'll neck myself if you tell me I should have used banana sockets.

>>1168127
I was going to post this.
I've been experimenting with different inductors, but the current the inductors are able to supply at 50 volts is pretty low still. I might put series of 4 or 3 leds in series and those series in parallel so the voltage doesn't needs to be so high.

>>1168116

will this retarded meme ever stop?

image the batteries are brand new, at 1.7V. that means 480mA total current, or 32mA per LED, which is like 60% over their rated current.

whenever you have an LED problem that requires just a couple of ohms, it means you solved it completely fucking wrong. coz you're thinking in absolute terms instead of variances and tolerances.

>>1168144

how much voltage and current will that dump into a circuit? if it's done right, it should be like 10mV and 100uA so you dont damage even the most sensitive of semiconductors.

>>1168163
To be fair, your typical multimeter (even good one) uses several volts and often more than 100uA of current.

>will this retarded meme ever stop?
Nope.

>>1168156
the rule of thumb for practical, efficient boost converters is no more than 10x steps. so for your battery bank that'd be 30v. realistically you'd want less. consider a 9v battery even though they're kind of shitty.

I have a ventilation system installed in my apartment and it comes with this remote that is installed in the kitchen (what for easy control of the ventilation when cooking). Has 3 speed settings and an auto setting. It should detect moisture automatically and switch on in the auto state, but that doesn't work for shit. Since I'm not gonna walk to the kitchen to turn it on/off manually every time I take a shower or take a shit, I decided to open it up and see what I can do.
Turns out it's just a simple atmega and transmitter. Now, it transmits on 868MHz or something, which I don't have a transmitter for myself. But I do have a bunch of 433MHz transmitters and receivers. There's probably some protocol fuckery going on with how this remote communicates with the main unit, so I figure my best bet is to just control the switches directly, since activating a button can be done simply by pulling it's pin low (on the connector in the top right).

So now my idea is:
Solder wires to the keypad connector and connect those to my own mcu with an attached rf receiver. Then make a humidity sensor in the bathroom that will send a signal to the remote to turn on ventilation and turn it off when humidity drops below a certain level.
Similarly, a light sensor in the toilet to turn on ventilation after someone takes a shit.
Sound like a decent plan?

Alternatively, I could try to reverse engineer the protocol and buy a compatible transceiver, but that will be much more involved and since I'm just a bumblefuck that's probably a bit out of my league.

does someone has made a traffic light with just capacitors and inductors?

>>1168316
no, most of them also have lamps or lights of some sort too...

>>1168316
if you put a ferrous rod in the middle of the inductor and constrain it to one axis of motion you can.

Is this thing a fuse?
When I turn this psu on, the fan moves a little bit and then nothing. Don't see any obviously failed components and I can't find anything else that looks like a fuse

>>1168368
circle the thing. the fuse may be the standing axial component right where the mains wires come in.

if it blew there may be something else wrong that you need to replace like what i think is a MOV under the silicone rubber at the top of the mains filter.

>>1168368
>the fan moves a little bit
Fuse isn't the reason then.

>>1168376
I meant the yellow box thing. Now that you mention it, the standing tube does look like a fuse. I can desolder it out of there and remove it from the rubber wrapper to check it out

>>1168394
the box is a capacitor

>>1168395
also this guy is right >>1168390 so it won't be the fuse anyway.

smps generally include a startup circuit that provides the initial charge to power the control circuit and switch the transistors a few times until steady power is available from a tertiary winding on the main transformer. what you may be seeing is the starter working and then the main switcher failing to start. not all startup circuits will exhibit that behavior but there's an infinite variety of them. if that's your problem it means the issue is definitely not on the secondary side and is probably something to do with the gate drivers or switching transistors.

an alternative explanation is that the supply starts, detects an overcurrent fault on one of the secondaries, and stops. this would make sense in terms of being able to provide enough current to start the fan up. however i'd expect it to keep trying every few seconds or milliseconds. if that's the problem the issue will be on one of the secondaries.

i don't think the main transformer has shorted (unless because of overvoltage or physical degradation of the wire enamel) because the fuse should've blown first.

>>1168395
looks like a relay to me

>>1168403
>>1168394
The standing tube was indeed a fuse. I pulled it off and it has not blown after all. The thing died when I was trying to hook it up in parallel with a lab bench power supply (don't ask why).
So since the fuse wasn't the problem, I take it that I can't really fix this?

>>1168406
i've never seen a yellow relay, and i've never seen a relay in a consumer computer psu.

>>1168408
look for cracks on the pcb. if you have a multimeter measure the resistance across the switching transistors' pins (if drain-source or gate-drain aren't very high there's a problem). if you see any crap on the traces clean it off. check for any bulging electrolytic capacitors. also burn marks obviously. that's what you can reasonably do without having much knowledge about these things.

>>1168408
>when I was trying to hook it up in parallel with a lab bench power supply

>>1168414
All those things look fine, assuming this solder is how it's supposed to be (it worked fine before)
>>1168417
Mistakes were made

>>1168414
>>1168424
This being the solder that looks like it's a big mess. It wouldn't reflow and puddle out during use, would it?

>>1168426
looks fine in context. i didn't read what you did to it in my first response because i have aids so disregard that. i take it you don't have a multimeter but if you do you can measure the resistance between the output rails and ground. you should get between 1k and 100k. if you don't an overcurrent from your experiment may have fried a diode or something faster than the primary fuse could respond.

that could result in the controller detecting overcurrent or overvoltage (losing regulation on a quasiregulated rail due to no load) on a rail and shutting down until the next startup.

For a tracked Arduino base, are stepper motors strictly necessary or can I build something that will move in a straight line with normal DC motors?

>>1168429
Well, measuring between +12 (yellow) and 0 (black) output wires, I only get a couple ohms, and my meter beeps for connectivity. That's clearly bad. It sure would explain why it automatically shuts itself down. I'll try again to look for any shorts

>>1168431
a dc motor will fall out of calibration without any feedback. a limit switch or a coupled encoder/pot both work. the latter is basically a diy servo.

>>1168436
Sounds like it'll be easier to just use a continuous servo then. Alright, cheers.

>>1168431
>>1168436
Or slap a gyro/compass on it.

>>1166999

Stupid question:

If I were to need a negative power rail, is that the same as connecting a powersource backwards?

Or does it need to be negative relative to the same ground as a positive voltage

>>1168503
>Or does it need to be negative relative to the same ground as a positive voltage
Correct
Since you say "negative rail," it sorta implies that there's a positive rail referenced to a ground. So yeah, it has to be below that to be a 'negative' voltage'

>>1168163
>10mV and 100µA
Lol I got a 3V coin cell with no current limiting, but hey the LED turns on across >2MΩ.

So if I have this old IC, (where these are the only power inputs), what are these referenced to?

Am I supposed to just pick one to leave at 0v?

I think I'm confused because I don't get analog at all. I'm a digital person.

This is some kind of rhythm generator IC I've been struggling with for a long time. I've posted about it before. Coming back to it because I had a really great idea of using the outputs to trigger relays, using the click as the drum

>>1168532
>>1168524

>>1168532
post your reaction when after a year of agonizing over an obsolete chip you finally wire it up and find that it's broken

>>1168532
Is this MM5871?
The ground pin is VSS and those voltages are referred to it. The chip itself does not require positive voltages, so if your other circuitry does not need them either, you can feed the chip from -27V and use a regulator like 7915 to produce -15V. You can use a +27V supply fo produce the "negative" voltage simply by reversing it like you suggested, but you either need a floating power supply or you need to be careful with your instrument's grounding. Point being, many newer power supplies have their gnd/negative output pin grounded via the mains plug and this becomes a problem if you connect anything else to your instrument.

Or, if you like positive voltages more, you could ground the VGG pin, connect +27V to the GND pin and feed +13V to the VDD pin. This is bit problematic though, since the +13V supply needs to sink current (instead of sourcing like positive supplies usually do), your logic inputs are referred to +27V and you need to be careful with inputs so that you don't accidentally overvolt them. MM5817 is a digital chip so noise in +27V line isn't really a problem, but other chips might dislike it.

>>1168573
I've been waiting for this, too.

>>1166999
can two 12 awg wires carry about 20 amps continuously at 12-14 volts safely?

>>1168582
yes

Is there a place where you can learn about three phase installations and maybe simulate them?
I'm having a class in my uni about three phase, using motors, contactors, timers, wiring star-triangle starters, and it's quite unintuitive, mostly since I'm not used to the symbology. I want to find some place where I can learn about three phase installations without blowing myself or anything up.

I've got an old microwave-oven-transformer secondary coil that I've removed from the housing safely, or so I thought, but there's an open circuit somewhere. I can't find it by inspection since the surface is covered in bits of tape and there are so many turns, though I assume it is on the outside. But if I immerse one side of the windings in salty water I find a slight continuity between either terminal, so have narrowed it down to around a quarter of the wire. Other ways of isolating the problem I've though of have been putting high voltage across the terminals and looking for the sparks, but I don't have access to high voltage at the moment. Any other thoughts?

>>1168602
Maybe you can inmerese it in water with some salt and apply 30 volts or so to it, maybe you can locate the open circuit by finding the place where a stream of tiny bubbles are coming out as a result of electrolysis.

>>1168605
Just did it with a 9V battery, worked great, though the bubbles were a little elusive at first. Soldering this one's going to be a pain in the arse.

What sort of mains input safety / suppression devices should I put on the primary side of a toroidal transformer aside from a fuse of the appropriate rating?
There will be 100+ kHz of switching on the secondary side.
Do i need a ~0.1uF - 1uF low ESR capacitor across mains? do i need a PTC in series with the fuse too? or an MOV? Common mode choke? Other things I've forgotten?

>>1168639
Cut the wire and wrap it up with tape with a very small bit of air inbetween the ends to make a gas discharge tube

>>1168640
fgt

>>1168268
Hmm, I gave it another thought and I'm not sure what the best way is to pull down the output of a button.
One of the pins on that connector is ground and the other pins are normally high. When you press a button its pin goes low.
What would be the best way to pull those pins low with an external circuit? Just looking for a suitable type of electrically controlled switch. Something like a relay but then solid state and miniature.

>>1168653
Huh? Why not use your microcontroller to do that?

>>1168576

> Is this MM5871?

yes

I will get an 7915 regulator and find another for the other supply

>>1168573

it's been sealed in an airtight package

Where's a good way/place to start learning Arduino with C specifically?

>>1168680
is your goal here to become a proficient C programmer? if it is then you'd be better off asking /g/ and first learning their way. starting on microcontrollers means you don't flex a lot of programming muscles and the best you'll be able to do when a control or data processing algorithm comes up is copy some dumb student's code. also don't let them recommend you a fucking book.

i don't actually have any resources to link you though sorry.

>>1168639
because of the steinmetz effect in iron cores they're already pretty good filters. if your windings are split then there won't be much capacitance to couple noise across either. if you're not trying to make a commercially viable product then you're probably fine. if you're the guy i think you are then the real mains distortion will come from a lack of power factor correction. you can't incorporate that practically though with an iron core so if you care at all you're fucked regardless.

>>1168706
>PFC
no i don't think that's me

So that just leaves the safety aspect. fuse + what else? PTC? MOV?

>>1168706
Well, since I am more familiar with C, I overheard a senior doing his senior project on a circuit saying that he'd rather have started learning Arduino with C since he was doing it with C recently I believe. He also suggested starting with C anyway.

So I'm just wondering if I can get to doing it with C instead of Processing?

I'll ask /g/ soon.

>>1168714
Don't worry so much about which language or environment you start with when beginning programming. Just start doing it and learn as you go

How difficult would it be to create a valve control system for hot wax for automated candle pouring?

I'm currently thinking the candle tins could be on conveyor belt, the valve controller would open and start pouring hot liquid wax, close the valve when enough wax has been filled, and then advance the conveyor belt to move the next candle tin right below the valve.

Boombox guy from the previous thread here. I bought a 6S LiPo and a BMS, and have another problem. The balance plug of the battery pack has 7 leads, from B- to B+, but the balance connector for the BMS has only 6 slots(or whatever you call them). The BMS doesn't have a slot for the balance lead from B-.

Is there a way to connect this battery to this BMS without frying the BMS? Can I just nigger rig a balance connector with 6 leads by removing the balance lead from B-? The BMS came with a separate balance plug with 6 slots, I wonder if could solder that to the battery pack and just leave to lead from B- out?

>>1168913
>series batteries
You don't have to bother with balancing if the cells are in parallel, but in exchange you'll lose more power stepping the voltage up.

>>1168947
>You don't have to bother with balancing if the cells are in parallel
No, you do. If there is a significant voltage difference between cells in parallel, you can get a massive current surge between them when they equalise.

Is it normal for old-ish diodes (4003, 4005) to be brittle?

I just tried to pry one off a board with a screwdriver and it broke. I feel like modern diodes can take way more abuse than that.

I made a chronograph a while ago, using an IR emiter/detector pair. It measures reflective objects such as brass jacketed bullets up to 2400 fps (confirmed, designed ideally for 3k fps, but haven't tested), but it's a bit touchy. To get it to work, you first have to adjust some trimmers to compensate for the ambient light in the area. A comparator evaluates the raw phototransistor output against a version passed through a low pass filter to look for objects causing a sudden increase in reflected IR light.

I'd like to revisit this and come up with something better to help either
A. Remove the need for manual trimmer calibration to ambient light
B. Improve reliability and be able to monitor nonreflective objects such as lead bullets

Anyone have any thoughts on the subject?

>>1169012
>prying components
ever heard of a soldering iron?

>>1169022

Melting the solder isn't enough to make things fall out

especially if you don't have all day

>>1168993
But if they're always connected in series, the voltage across them by definition will always be the same. This may be a problem when initially assembling the pack, but never afterwards unless you want to start replacing cells.

>>1169015

Does it use a photomultiplier? just use that to measure ambient light accurately and then build a circuit to compensate

>>1169027
*parallel

>>1169026
Ever heard of pliers?

>>1169029
>>1169027
Indeed. Too bad it's not that simple in the real world.

>>1169031

Same difference

>>1169028
>photomultiplier

No; I didn't know what that was until you posted it. For reference, this was a project for a class on electronics for mech. engineers (which I am, but I have a hobby with electronics). That could be a good idea to look into, thanks.

But it's a lot more crude, just looking for the light reflected back from IR emitters into phototransistors in parallel

>>1169033
Please explain? As far as I can figure, as long as you're not pulling too much current, the load will pull current from whatever cells it needs to to keep the voltage across all cells the same, if not the differential voltage will push a current through another cell and rebalance the whole thing. Since this rebalancing happens live, there should never be a significant current. I've never had a USB power bank go bad from this at least.

This should only be an issue if there's a significant resistance in the wires between each cell, which will only happen with very thin/bad wires (in which case the wire resistance will throttle the balancing current surge anyways) or very high current draw (in which case you should be getting more cells anyways).

>>1169044
Did you even read what >>1168993
said?

>>1169034
>prying a diode forcefully
>the same as
>using pliers to pull out the lead being desoldered
Maybe /diy/ isn't for you anon

>>1169066

both levers

>>1169070
>one breaks the diode
>the other doesn't
maybe /diy/ isn't for you anon

>>1169076

> this is a reply

maybe suicide is a good option for you

>>1169078
>can't admit he doesn't know what the fuck he's doing
>now he has to attempt to save face
keked

>>1169070
>I could ride a car to work, but I prefer to use my vintage Tiger tank. So what if it tears up the streets, same difference, they're both metal boxes with wheels.

>>1168722
Yeah I guess, I need to learn what pin is for what first I think and what you can do with them.

>>1169044
It doesn't just pull load, it will actively try to "charge" lower voltage cells. So if you have two 20C discharge, 2C charge cells, one at 3.8V and the other at 3.4V, the charged one will discharge at 20C into the other one until their voltages are equal. For a small voltage difference it isn't a big problem, but it can harm the batteries (and possibly start a fire if worst comes to worst).

I know nothing about circuit design, so please excuse any ignorance.

Currently building this, have all the parts I need except for C3, 2.2 nF cap. Closest I could salvage is a 10 nF cap; what would happen if I replaced it with that? Should I also replace C4 with a 100 nF cap to maintain the ratio? Also, think this is a decent circuit? I picked it at random after googling tone control schematic. It'll go into a 12v portable boombox thing I'm building if that matters. Thanks for any help.

>>1169112
But that shouldn't be a problem when the cells are at use and start off in the battery pack at equal voltages.

I agree that >>1168993 happens if they don't start off at equal voltages, and the higher voltage cells would discharge quite quickly to charge the lower voltage cells until the voltage is equal (or the battery explodes).

Lets say you have two cells in parallel, one 18650 and one 10440 (AAA size). I am assuming this covers any asymmetric setup you can throw at me, but if not, please provide. If you start draining the battery at a couple of amps, the cells will discharge at the correct rates so that their voltages remain equal to each other, so the 18650 would provide much more current than the 10440. If for whatever reason you were draining the pair and the 10440 dropped to a slightly lower voltage than the 18650, it would have current forced back into it by the 18650 to charge it back to the 18650's voltage, but this reverse current will never happen except for at a minuscule (dV/dt) scale, because the 18650 updates the 10440's voltage live with little to no delay.

The voltages of two cells will never be at different levels unless they are wired together at different charge levels, and hence your example will never happen unless you try to pull a current too close to the maximum current value of the respective cell.

Hey /ohm/ can I use arduino board to interface as a card reader? I mean using magnetic head, with amplifier (im using a LM386N) and read cards? What is the purpose of head comms? what pin should I use for head comms?

Not my image, but I am doing something similar.

>>1169153
Battery cells are not made equal. I very much doubt most cells will discharge with exactly the same charge/voltage ratio.

>>1169156
Even if they don't, the cells won't pull the same currents, like in my example of an 18650 and a 10440. Did you even read my post?

>>1169159
Not him, but massive currents will go through the battery with lower voltage as current flows from higher to lower potential, in addition to trying to power the load. so yeah you're fucked if you don't balance your batteries. Please, people, balance your batteries. There's a reason that shit exists out there

>>1169155
Probably, but i dont know how. There are certainly guides and projects and info dumps about it out there i'm sure. People make all manner of things with arduinos

>>1169167
Most things I have come across is people hooking arduino to TTL on card reader. I am sure I can, just have to do alot of coding of filtering and what not.

>>1169166
The moment any cell in the parallel bank gets off voltage it is forced back in check by the other cells. Voltages in parallel are always equal.

Balancing makes sense in series, but I am arguing that it's almost entirely unnecessary in parallel.

>>1169178
>The moment any cell in the parallel bank gets off voltage it is forced back in check by the other cells. Voltages in parallel are always equal
Yes, causing massive currents to flow
Please don't listen to this anon. I'd rather nobody sets fire to their house cuz of this shitty advice

>>1169180
But this voltage will never exceed a few microAmperes if you're lucky, theoretically it's zero. Try it yourself. There is a wire with very low resistance between the cells (1s of mΩ), so unless the resistance of the load is comparable to this, the whole thing's a voltage divider with very very small voltage increments between them. At these low voltages, the resistance of the wire is more than enough to act as a current limiter.

Happy? Has anyone here ever measured a difference in voltage between two cells in parallel, or one cell suddenly shooting a whale-cum-load of current through another? It happens, but not on the scale people here are talking about unless you're drawing waaay too much current.

>>1169187
>But this voltage will never exceed a few microAmperes if you're lucky, theoretically it's zero
>voltage
>mA
jesus you don't even know what your'e talking about.
What you're saying only works with cells that are known to have similar discharge profiles / similar internal resistances.
If One battery sags 0.25V under high load and the other sags 1V under that same load, do you know where the 0.75V difference in the cells is going to cause current to flow?
Jesus christ.

>>1169190
>microAmperes
>mA
Right back at you.
That's assuming the current drawn from each cell is the same. When the cells are in parallel, the only thing that determines the current from each battery is the total voltage of each parallel cell and the current being drawn from them. One cell will never sag a different amount of voltage than its parallel neighbour, even if they would individually. In reality, one cell will pull more current than the other, but it will never reach the point of current flowing backwards or even being 0 through one cell unless your battery pack is severely underpowered for the application.

>>1169193
One cell will never sag a different amount of voltage than its parallel neighbour, even if they would individually.
Yes they will. The only reason you think they aren't is because current is flowing from one cell into the other you dumb cunt

>>1169190
>>1169193
And let me just clarify that I'm assuming you're referring to a situation where the few cells in parallel were already sitting in parallel, and that the voltage difference suddenly appears when the load is attached.

>>1169195
The only situation where the current flowing through one cell would be relatively negative is where the current the load pulls is insignificant in comparison to the current being forced from one cell into the other. This will never happen because the voltage will never sag this much for such an insignificant current pulled by the load.

>>1169195
>One cell will never sag a different amount of voltage than its parallel neighbour, even if they would individually.
Is correct, though this is samefagging, because the sag of a cell is dependant on the current being drawn, which is a function of current being pulled from neighbouring cells, not just on the load. Kirchhoff's current law or something.

>>1169196
>insignificant current
Right. The only regime in which your stupidity won't get yourself hurt.

>>1169201
The point being that if the load pulls a significantly higher current than would need to flow between the cells to keep the voltages equal, the current will just be subtracted from the current already flowing out of the potentially lower voltage cell and added to that flowing out of the potentially higher voltage cell, and will never reach negative. This is so-called load pulling, and can be expected.

When the load is attached, the voltages don't suddenly drop because of the resistance, they drop because of the current being drawn from them, so when slightly less current starts flowing out of the smaller cell as a result of the bigger cell pulling more of the load, the voltage across the smaller cell increases and the voltage across the larger cell decreases until both cells reach the equilibrium voltage.

Two cells with the same chemistry in parallel == one larger cell.

>>1169202
>The point being that if the load pulls a significantly higher current than would need to flow between the cells to keep the voltages equal, the current will just be subtracted from the current already flowing out of the potentially lower voltage cell and added to that flowing out of the potentially higher voltage cell, and will never reach negative. This is so-called load pulling, and can be expected.
incorrect

>When the load is attached, the voltages don't suddenly drop because of the resistance, they drop because of the current being drawn from them
correct
>so when slightly less current starts flowing out of the smaller cell as a result of the bigger cell pulling more of the load, the voltage across the smaller cell increases and the voltage across the larger cell decreases until both cells reach the equilibrium voltage.
incorrect

>Two cells with the same chemistry in parallel == one larger cell.
incorrect

>>1169204
As far as I can tell, the only things that makes the second and last statements incorrect (the second-to-last one at the very least) is that the cells have finite ESR and are not ideal, but this will almost never have an effect on your system unless you're pulling stupid amounts of current in the first place. If you put a 50mΩ load across a single 18550, then maybe, but that's a little bit of an extreme case.

Can we clarify which situation we're talking about here, whether it be voltages dropping at different rates because of different capacity causing the voltage difference, or voltages sagging because of the attached load? Man I swear I'd hit the jackpot with my second-to-last statement.

>>1169204
Could this be simulated in software with a voltage source in series with a charged capacitor to simulate how the voltage drops with its contained energy, along with a current limiting resistor to simulate how the voltage sags with load? I'm getting way too invested in getting to the bottom of this.

>>1169209
Another question to add, does the voltage only sag because of the voltage-dividing effect of the cell's ESR, and hence can be simply represented by the resistor?

>>1169209
a capacitor charging through a resistor would just take a while to reach the voltage of the voltage source.
There's also the issue of if the internal resistance changes over time or with temperature and by how much. I imagine it decreases with temperature, hence the runaway effect of accidentally shorting 18650 cells. Though they tend to have protection circuits these days

>>1169210
Not sure what you mean, but I'm somewhat certain internal resistance decreases with temperature though I don't know that for a fact

>>1169208
>stupid amounts of current
lithium cells are definitely capable of it the way conventional alkaline cells are not. It's one of the reasons they're used, in addition to their impressive capacities.
http://www.candlepowerforums.com/vb/showthread.php?386728-Acceptable-resistance-for-good-quality-18650&s=7d6099393c19b5b5b2b3f4210d81b0ee&p=4800688&viewfull=1#post4800688
One of the small number of places I've found for this and they don't cite if this is true or not but i'll assume its roughly correct
And also lithium cells seem to be able to (cursory google again)
https://www.rcgroups.com/forums/showthread.php?2120900-How-do-you-know-how-many-amps-an-18650-LiPo-can-safely-discharge
discharge at 2C or 3C. so a 3400mAh can discharge at 6.8A - 10.2A tops as a safe upper margin

With individual 18650 cells, that's not a problem because 0.05+0.05 ohm = 0.1 ohm 0.1ohm x 10A = 1V so you'd need a difference of 1V between the cells to see potentially 10A flow from one battery into the other and 1V is basically on the threshold (or past) the point at which a battery is considered discharged.
This is probably one of the reasons why industry matches internal resistances / etc of cells intended to be in parallel in their applications like tesla cars' batteries. Thermal failures of individual batteries have still happened though.
tl;dr if you wann amake sure this doesn't happen then make sure you match age and internal resistances of batteries

>>1169217
>discharge at 2C or 3C. so a 3400mAh can discharge at 6.8A - 10.2A tops as a safe upper margin
Forgot to add load current draw onto that too.

>>1169217
10A is significant compared to the 3.7/0.2 draw from short-circuit of a "good" battery, but I just checked a YT video where some guy tested the sag voltage and found the ESR to be closer to 20mΩ than 200mΩ. I was excluding significantly high current-draw from my thought process because Elon Musk is a hack, but at that kind of current draw you'd definitely come up with some bad effects of some kind. I'm surprised you were able to find a source on Li-Ion cells that ISN'T a vaping website.

I'm useless at LTSpice, so if anyone can figure out how to have a voltage source charge a capacitor with 0-very little resistance, and then go open-circuit instead of short-circuit, that would be great. I'm trying voltage-controlled switches, but I'm getting missing model definitions and floating nodes for some reason.

>>1169230
I wouldn't say elon musk is a hack. he's a rich businessman with some tech expertise and he's surrounded himself with people who know what they're doing (sometimes they could be yesmen)
but regardless, matched cells are the most practical way to go about it. If you want absolute safety then yeah individual monitoring of each cell would be necessary. proper matching of cells is a good compromise between safety and practicality, I'd imagine

>>1169231
The hyperloop is economically unviable, and tesla cars are a burden on the US economy which would be better spent on battery, solar panel, and thorium power research. The man's a subsidy miner.

Of course you always want matched cells, for consistency's sake if not anything else, I was just using an extreme example under the eventuality that there will be an extreme situation where the differences in matched cells become significant.

Also I posted the wrong image.

>>1169233
>subsidy miner
I guess time will tell

If you wanna use an extreme example, just make the cell zero volts and leave the internal resistance there

>>1169235
That will tell me something, but I eventually want to try to acquire a positive test with plausible values.

I'm not even sure whether having a voltage source in series with the capacitor realistically represents the cell, I've just got it there to get the 4.2-3.2V discharge, but I'm guessing the ESR increases as the cell is discharged and I have no way to model that. All the cell model needs to do is hold a voltage proportional to the charge within, which I could easily do if LTspice had any sort of timed switch. Wait maybe a current source would work since when it goes to 0A it would go open circuit? What happens when you short a current source across a 0ESR cap?

>>1169240
As I said, it doesn't represent the cell because a capacitor will charge to the voltage of the cell through the capacitor. it doesn't reduce its max voltage.
You're modelling the battery here so figuring out the current across the internal resistances is what's important. You don't need LT spice for that. I already showed that a 1V differential (4.2V vs 3.2V) in parallel would roughly pass 10A of current through them assuming they are both 50mOhm internal resistances. with older batteries and higher Ri, it'll probably be lower. As you said this only really matters in high current applications. But imagine for a moment a battery passing 4A to another battery and 4 A to the load. That's 8A current being drawn from that one battery, not to speak of whatever current is being drawn from the other (if any)

>>1169241
I know that that's what a voltage difference would do, but I'm trying to see if I can even create such a voltage difference in the first place, and under what circumstances. If you're talking about the voltage sources across the caps, then those are just there to charge them up to full initially, after which their voltage will decrease as the energy contained decreases like a battery. if you're talking about the voltage source in series, then it's no different from putting two voltage sources or batteries in series.

>>1169242
I don't see why there need to be capacitors in the equation at all

>>1169245
A Li-Ion's voltage decreases as it is discharged, one of the issues I'm trying to simulate is if one cell has lower capacity than the other, will its voltage decrease at a different rate or will the cells balance each other by providing different amounts of current. LTspice's batteries are just reskinned ideal voltage sources and have a constant voltage. By using a charged capacitor to simulate a battery, I get the effect of the voltage decreasing just like a battery.

>>1169037

here's a couple of tricks that can help - you can combine them to get even better results.

1) use IR filters on the receivers. one good source is old film that's been developed but not exposed (i.e. the last couple of frames on a 24-frame roll that you couldnt bother with). use 2 layers for better results.

2) run your LEDs at hight frequency, like 500kHz, then use a low-pass filter on the receiver to eliminate all the low-frequency ambient light noise.

applying either of these allows you to increase the sensitivity of the amplifier.

to measure non-reflective items, make a photo-interrupter, where the object passes between the transmitter and receiver. a laser would provide a more exact path to interrupt than an LED, which sorta spurts in an arc.

>>1169247
A capacitor doesn't gain or lose charge the way a battery does though. You're complicating the model by trying to turn a capacitor into a battery.
Capacitors have a well documented charge and discharge profile - that is, exponential. Batteries do not.

>A Li-Ion's voltage decreases as it is discharged, one of the issues I'm trying to simulate is if one cell has lower capacity than the other, will its voltage decrease at a different rate or will the cells balance each other by providing different amounts of current.

The voltage will absolutely reduce at a different rate, assuming there's no balancing circuitry, and the only effect of having them in parallel is that one cell basically charges the other. Charging circuits apply current / voltage to the cell in question as if in a parallel configuration.
Charging circuits also charge batteries (specifically lithium cells) in very specific ways. constant current until a certain criteria is reached the constant voltage etc. Having mismatched cells in parallel under high current draw is absolutely one of the ways in which you can damage cells.
The only thing you've got going is the correct presumption that batteries with the same chemistry have roughly the same discharge profiles

>>1169252
if i'm right though the constant current regulation is purely to prevent the overcurrent constant voltage charging would cause due to the differential between charge and cell voltage and the low esr. one cell charging another would be a continuous process so currents wouldn't ever get very high as long as the cells aren't hugely different.

>>1169253
Indeed.
But it'd be only continuous as far as the load allows. You cant guarantee (under high load) that cells will behave the same, which could cause potentially hazardous currents to flow within their own isolated, unregulated loop if one cell's voltage drops more precipitously. People do match cells for this reason. As the other guy said, under light loads its inconsequential

>>1169252
The actual curve (quadratic not exponential) of the capacitor discharge doesn't matter in the slightest, all that matters is that it represents a situation where once "cell" can push current or receive it. There is no other way to model anything close to a chargeable cell in LTspice without a shit-ton of semiconductors and probably a microcontroller with all the discharge data I'm trying to figure out already programmed into it. I don't see anybody having to match capacitors in their parallel capacitor banks, and you'd think with the high ESR of supercapacitors that this might be more important than with batteries.

I still argue that
>so when slightly less current starts flowing out of the smaller cell as a result of the bigger cell pulling more of the load, the voltage across the smaller cell increases and the voltage across the larger cell decreases until both cells reach the equilibrium voltage.
is true, but that's what I'm trying to model at the moment anyways. Surely charged capacitors are good enough for this purpose, unless you're trying to argue that the discharge curves of two cells might be different enough to have steep, sudden drops in voltage at different times to cause a surge, in which case you win. I actually got distracted for an hour modelling a CW multiplier in LTspice and looking for a relationship between frequency, load resistance, and output voltage, which is pretty fun.

As we all seem to be agreeing, as the load current increases and the load resistance enters the realm of significance of the ESRs of components in the circuit, it becomes less predictable and may well form significant voltages between the cells. But is it possible for the current going in/out of any one cell in this arrangement to be higher than the total current being absorbed by the load? Food for thought.

>>1169264
>Sudden drops
no, not under light load no. And definitely not as step increase in load or anything
But as for capacitors, they're definitely exponential. The voltage goes as a power of e (the time constant) unless you have a constant current source in which case they can charge linearily, which is a technique sometimes used to create a voltage ramp with a capacitor.

>>1169266
I don't care about voltage over time (exponential), but rather voltage as a function of the power output, energy remaining (quadratic), etc. Either way the function doesn't matter as long as it's a downward trend that can be reversed.

>>1169281
Good luck

>>1166999

Can I program an lpc810 directly from a USB cable?

Adafruit sells this one kit that comes with a cable and I'm not sure if it has some hidden FTDI or not

>>1169417
the ftdi/ch340/whatever and supporting circuitry will be in the fat part of the cable. the chip only takes uart and friends, not usb.

I thought this was pretty cool. I bought my grandmother's house out of the estate and I found this. It's the remote from a vcr, circa 1997. It blew my mind back then, mostly cause I had never seen it before. The big gimmick knob in the middle was temporary fastforward/rewind. The position I'm holding it in now would be the fastforward. It has some kind of spring or recoil in it so that when I take my thumb off it spins back to play. If I remember correclty it was supposed to be able to show frame by frame, but we could never get it to work. I also believe the speed was affected by how far the knob was rotated.

I just thought it was interesting, never saw anything like it before or after on a vcr. It mystified me as a kid, like those old rca stereo receivers where the knob moved when you turned up the volume with the remote. Now I realize things like this aren't hard to accomplish, but back then it fueled my interest in electronics.

>>1168714
You can do everything the Arduino libraries do but from scratch. Look at the avr-gcc docs. If you're coming from a 'big computer' environment though, I'd say a lot of it is more about learning the microcontroller's "API" and what sort of capabilities it can handle for you.

i have some mystery subwoofers i got for free, owner had no idea of their wattage and impedance and there is nothing printed on the speaker at all, not even brand. impedance should be as easy as just checking ohms, but is there any way to figure out the wattage?

>>1169622
Is speaker impedance equal to speaker DC resistance? Speakers are inductive loads and so have more impedance the higher the frequency you put through them, so I'm not sure how that's measured. H-have you tried google?

>>1167480
You've done well here

I have a university project for a over and under voltage protection adapter, and plan to put an lcd using a pic16f877a to print whether the relay triggers or not. No one in my team really knows how to code for the pic connected to the lcd, so I kinda need help for that part.

>>1169622
>impedance should be as easy as just checking ohms, but is there any way to figure out the wattage?

impedance is resistance to AC, so it's different than DC resistance. you need to measure it indirectly: put a small pot in series with the speaker, apply 1Khz sine wave across the pair and vary the pot until the voltage across the speaker is the same as across the pot. take out the pot and measure its DC resistance, which is now the same as the speaker impedance.

a quicker method is just to assume it's 8 ohms like the vast majority of speakers.

as for power rating, you can guesstimate it from the size of the magnet and comparing against other known speakers. another way is just to increase the power until it starts to distort, or the movement of the coil appears too extreme, or the coil is getting warm, then measure the AC voltage and current at the level, and multiply: P = V * I

a quicker method is NOT to play your music at nigger levels, and just assume that it can handle human-compatible volumes.

>>1169669
>you need to measure it indirectly
I'd say that's a rather direct method.
Generally speaker's DC resistance is around 70% or so of its nominal impedance.

Heya,
I've got a shitload of 3W LED's I wanna drive, and all I've got on hand is a bunch of 10W led drivers and some variable voltage limiters.
I'd usually say "Bugger it" and just use suitable LED's for what I have or buy some LED supplies, but if these things are single diode I was wondering if there's a way to run them en masse off one supply.
Or maybe hook three together in series and run it off a 10W supply?

>>1169706
You should be able to connect one, two, or three LEDs to the 10W driver and be fine. It won't just drive it at 10W unless you set it to push too much current through it.

>inb4 connect four

>>1169708
I thought the point of LED drivers was to maintain a constant current by varying the voltage, so a 10w power supply, giving out 700mA, will crank the voltage up to ~14v to keep it running at 10w, which would fry a single 3W, 700mA LED. Thus, connect three to get 9W, 700mA, so you're only running each LED at 1/3W more than they're rated to, which won't hurt them too much.
The problem here is you're still using a LED driver for every three diodes, which is a pain if you wanna run like 10-50 of the things.

>>1169709
See if you can get your hands on one of the modules listed here:
https://www.youtube.com/watch?v=piET0Biqo0I
Rather than needing a specific supply for each type of LED and configuration, you can use whatever you want.
Enjoy your weed plantation dude.

>>1169663
>pic
When it comes to microcontrollers, anything Arduino is cheaper and more powerful than an obsolete pic. It sounds like you're overcomplicating your situation, why even need to use an LCD in the first place when an indicator LED or two can tell you all you'll need, maybe a 7-seg display if you need to display an actual value. It will probably take less time wiring an ADC to the 7-seg (with the help of a few dozen Karnaugh maps) than it will to comprehensively learn the ins and outs of a microcontroller, though it will be a tad more expensive. I honestly don't see why the relay having been triggered can't be displayed by a single indicator LED, or a neon if you're running at mains voltage.

'Course you could just copy-paste some shitty code from a website after you follow a tutorial on how to plug the thing into your computer. Tell us how you let the team down later, ok?

>>1169709
10W is the maximum, and the driver will provide as much power as the source draws up to the 10W limit, at which point it either shuts down or explodes. Just like a wall wart's current and power ratings, they are maximum values. I suspect your 10W 700mA LED driver also has a maximum voltage value. An LED has a resistance that changes depending on the voltage or current it's used with, but it has a set relationship between voltage and current, the gradient of which gives a single value of resistance at any point along one of the axes. When a constant current driver supplies an LED with a constant current, 700mA, this corresponds with a particular voltage across the LED and hence a particular resistance. If the power supply cranked the voltage higher than this, the current would increase, which the constant current LED driver will not allow.

Constant-current drivers are good for LEDs and things with a diode/exponential I/V curve, constant-voltage drivers are good for anything that has a linear I/V curve because they're simpler, but I don't believe constant-power drivers exist for any practical uses save possible critical heating applications where the heat delivered remains constant.

>>1169717
Oh shit, now I understand.
Thank you very much, that was a great explanation!

>>1169722
This often comes up with new-friends asking whether a 12V wall wart will always supply the rated current, often enough that we aught to have a copypasta dedicated to explaining this but don't (at least not that I've seen), but I guess it should include a constant current supply explanation in it too. Glad I could help, but It's all I can do as someone who bothers the thread/board with my own noob questions.

>>1169722
Not the guy who responded to you >>1169708

But >>1169717
Is correct yeah
I should add that you can usually (i think?) set the output current on LED drivers with a resistor or two, so its not like its 10W or nothing, which is basically what he said too. So yeah your 10W driver should be fine. read the datasheet on it and see if it tells you how to set the current. And also know how much current your LEDs use too.

And about your second response >>1169709
its up to you what sort of configuration to drive your LEDs in. You can run them in a series string, or run them all parallel. The combined power usage of the LEDs will be the same in both cases. Parallel is popular because lower voltage is easier to find, and with the advent of lithium cells, high current at low voltage is also easy to get.
>run 10 - 50 of the things
In a case like that you'll wanna get yourself a higher wattage driver, or just make your own constant current source so that you won't need to run 10+ drivers and you'll be set. That is of course unless there's some redundancy reason to have multiple drivers running their own sets of LEDs etc etc.

>>1169742
>in a case like that
It's also what >>1169710
Said too

>>1168653
>>1168654
Microcontroller should work. The digital input/output pins can most likely sink/source enough current to drive the pin low or high. What's the receiver interfacing with if not a microcontroller?

>>1168503
>>1168524
You are right about connecting a power source backwards, though. If you put the - side on the circuit's ground, the + side is your positive rail. Backwards, with + on ground, the - side is the negative rail.

>>1168912
Check out solenoid valves.

>>1169129
What are the VRs? Potentiometers?

>>1169851
Of course. It's a tone control circuit.
He got his answer in other thread already, though.

>>1169846
that's correct, but an exclusively negative supply is not very common, i think

>>1169856
Not anymore, but it was the normal way in germanium-era electronics.

>>1169865
Ah, okay. Didn't know that.

>>1168409
I have one at work with a relay in it. First time I saw that too but apparently it happens.


It is indeed a capacitor by the way

>>1169129
looks like a standard tone stack. changing the value of that cap will change the frequency response, but hey, it might come out okay

>>1169840
So, I can connect an arduino output pin in parallel with a (normally high) keypad pin and just write a low value to it?

I was thinking to keep it completely electrically isolated (perhaps using an optocoupler as a switch) and have it run from a separate power source so it doesn't drain the remote as fast. I'd just bolt on my circuit and keep it easily removable without any modifications to the original remote (because reasons).

>>1167700
Post your code anon.

>>1169712
Well, thanks for the suggestions I guess.

What is the simplest way to get 5v down to 3.3v?

I just need it to work so I can run my micro, I have plent of power to spare since I'm running off 5 watts when the chip only takes a milliamp.

Can't I just use two diodes to drop it 1.4v?

>>1170189
DC to DC, I presume? Just buy a regulator. They cost nothing and are the most reliable way of accomplishing it.
http://www.ebay.com/itm/New-10pcs-AMS1117-LM1117-3-3V-1A-SOT-223-Voltage-Regulator-NT00206-Ship-Today-/112049991237

>>1170193

I don't have 3-5 days to wait in order to pay 5 bucks for a 20cent part

>>1170195
That's 70 cents for ten 20 cent parts, but I can understand if you think the wait's annoying. You can drop voltage with diodes in series, but the voltage drop is going to be dependent on the current draw. If your circuit draws a constant current and you have a constant voltage input, your idea works.

>>1170197

It won't vary by more than a 50mA

>>1170199
Alright. You can use diodes, but you may prefer a resistor voltage divider. If you do use a diode, if you're expecting to draw over around 150mA, put some form of heatsink on the diode. They heat up quickly. Just gluing a nail head or something onto it should suffice.

>>1170200

ty daddy

>>1170200
I'd use a 3.3V Zener if your local electronics store/your parts bin has one like the 1W 1N4728. Not sure if you'll get the precise 3.3V from simple diodes, and it's a little messy, but still the easiest way to go otherwise.

>>1170261
You certainly won't get the precise 3.3V from a diode, which is why I was recommending just buying regulators instead. Any change in current draw leads to a change in voltage output, but with just a 50mA difference the voltage output shouldn't change enough to kill ICs with overvolting.

I want to create a PCB with an individually addressable RGB LED matrix, 64 LEDs in total.

Unfortunately I can't use those handy LEDs with controller ICs built in, and I must use regular old RGB LEDs (picrelated).
What are my options of controlling them?

I know of 2 options:
>chain 3 MAX7221 controllers, one for each channel
pros:
>simple as fuck
>3 resistors. count 'em, three!
cons:
>no "true" individual addressing: can only set one color for all LEDs (through brightness of channels) and set individual LEDs to be turned on or off, since MAX7221 was meant to be a used in a digital display
>not a lot of colors, since digital brightness controls of MAX7221 only have 16 steps

>put WS2801 next to each led.
pros:
>full functionality
cons:
>so much fuss with soldering and programming
>60 fucking SMD chips holy balls

I am electronics-retarded, so I most likely got something wrong and there must be other options.
Any suggestions?

>>1170331
PWM, have each LED addressed as three individual LEDs from an IC, and transmit digitally, for instance, "(4,6,255), which would be (x,y,intensity).

>>1170331
You could use a WS2803 instead of ws2801. Same functionality but it can drive six RGB LEDs.

You could build your own matrix control circuits instead of using a MAX7221. Using regular 74HC164 shift registers I've been able to get individual dimming in several steps. Cycle through the matrix at 600Hz; have an LED on 1/10th of the cycles for 10% brightness etc.

>>1170331
Build an enormous grid of transistor switches.

>>1170333
Are you suggesting I get an MCU with 192 pins? Or do PWM through Shift registers?

>>1170339
WS2803 would definitely be a better choice, but it's still the same idea and the same pain in the ass.
MAX7221's main advantage is that it's so simple, does everything for you, and just werks. Also, THREE RESISTORS!
Maybe it would be possible to somehow hack together individual addressing through software somehow, I dunno.

Has anybody here ever tried to use mechanical relays instead of transistors to make some sort of primitive computer? I just want to know what it sounds like.

>>1170353
Never seen anyone say so, but several such computers exist.
>sounds like
https://www.youtube.com/watch?v=N5TUaejWNbE

>>1170346
Yes.

>>1170346
You'd be better off just using several Arduinos than trying to find a 192 pin.

>>1170356
>5Hz clock speed
And here I was wondering if it was practical. All those relays and the guy still has silicon ICs at the bottom, probably the shift registers and 7-seg controllers. Just one 4-input shift register would probably use at least twice the amount of relays he's showing there by itself.

>>1170367
There's a reason computers from that era where measured not in ATX or m-ATX size, but in amount of rooms.

Does anyone use these? Motorola Service Monitors

http://www.ebay.com/bhp/motorola-service-monitor

>>1170368
I guess the relays of the time were smaller than valves, but much slower.

>>1170364
Well, FPGAs exist. But it's a retarded idea anyway.

>>1170379
Well, take this thing for an example: https://www.youtube.com/watch?v=8RUoAekYkRA
That's huge, for a computer, and it's less capable than, say, the ATMega microncontroller on an Arduino, which is by no means a miracle of science.

Guess this explains why my mcu worked fine on battery but didn't like the $1 chinesium usb wall charger. Random triggerings of sensors and such.
I knew they are shit but I just had to try it.

>>1170557
>batteries are DC
well blow me down!

>>1170557
Isn't having an oscilloscope great, anon?

>>1170572
Amazing isn't it.
But I was talking about the huge amount of ripple that throws off the mcu.

hello i recently replaced a USB mini port on a handheld device i own, i guess i kinda botched it though cause when i plug in the wire it sorta wiggles around and that makes the connection unstable, the PORT i replaced itself does not move, just the usb wire when its plugged in. ive tried other wires same result, is there a way to make it stationary while inside the port? ive tried using a small strip of electrical tape over the top of the usb plug but it either wont fit or doesnt do anything to change it

thanks

>>1170572
Edison would be proud.

>>1170573
Yeah I love it. Measuring things gets my dick hard.

Hello, so I'm trying to start a project of turning a hp mini laptop 2011 into a well working laptop, that can record and stream ps3 quality games at the least. I couldn't find any info on doing so, so i came here. If someone could list the specs for doing such a project, that would help a lot, thanks.

>>1170610
t. Pajeet

I'm having a bit of an issue here.
I have a comparator that goes low in the dark and high in the light using an ldr. The output goes into an mcu, but to give a visual indicator that the pot is calibrated correctly I also want an led connected to the output of the comparator.
However, this setup doesn't seem to work very well. The comparator doesn't give a clean on/off anymore but something in between instead.
Any idea how I could wire the led?

Also it could theoretically be a crappy comparator as it's a dirt cheap chinese one, but without the led it seems to perform fine.

>>1170657

google Comparator with Hysteresis Reference Design - Texas Instruments

>>1170381
FPGAs are, in fact, the only way to create a decent dot matrix display.

>>1167700
Put a resistor between VCC and that BJT's collector

Like 100k

>>1170785
You've probably burnt out the BJT in that PC817 by not limiting the current in its off state. It's probably constantly letting current pass through to ground, never triggering your input pin. I've done exactly this, with the exact same piece of hardware

Does anyone have an idea for a circuit that acts like a zener, but with a reversed voltage profile? (conducts at high positive) Or conversely, a positive device which is conductive at low voltage differential but resistant at high? The voltages across it will be fluctuating so it would not take well to biasing. If I had some idea of what to go with I could look up more robust components for what I need.

>>1170904
Why don't you just use an op amp / comparator to turn on or off a mosfet when it detects voltage is low or high

>>1170904
>conducts at high positive
What is this supposed to mean? Just turn the zener around? You can also put a diode in series with it to make it unidirectional.

>conductive at low voltage differential but resistant at high
That would mean negative resistance. Such things tend to oscillate easily. Also, you might need auxiliary power supplies depending on your requirements.

>>1170916
I have mosfets in a configuration which would exceed their breakdown rating. This is for an actual project, and I can try multiple designs but this one is so damn close. Basically I need single-ended output which can drive +/- voltage referenced to neutral. The problem is the voltage is desired to be +/- 500V at max and at around 1MHz. THAT can be withstood by a mosfet, referenced to ground, but the single ended output fucks it up because I'd have up to +500V on the D/S and -500 on the S/D, exceeding breakdown.

>>1170917
I can have access to whatever voltages I need, though high current at more than +/- 15V would be a problem.

If I had "reverse" zeners, diodes that conducted at an extremely high voltage, and in reverse conducted at a low voltage, I could get the blocking working.

>>1170923
Maybe post your schematic?

>>1170923
The simplest, most obvious option would be to use high voltage mosfets. Or, if you hate yourself, several fets in series.
If the rest of the circuit allows it, you could also use a transformer so that the fets don't need to handle so high voltages.
There's also a chance that your circuit allows some other kind of overvoltage protection.

I don't really see your fast 500V anti-zener as a solution.

>>1170927
His anti-zener thing really sounds like it could be solved by other means. Not to mention I still don't get what he wants. His first description kinda contridicts what he just said.

>>1170904
You could try a MOV or other varistor (the difference being MOVs have a zener-style breakdown voltage while standard varistors have a diode-like curve, I think) if you want it to be able to have a breakdown voltage at either + or - voltage. I'm not really sure what you mean when you say you want a zener with reverse profile so it conducts at high positive, couldn't you just reverse the Zener?

Anyway, I'd suggest making a not-gate sort of thing to invert the signal of the zener/varistor, but instead of using BJTs you could use relays if they would hive higher voltage resistance. You'll also basically be shorting the voltage to ground through a resistor, so pack some power resistors. This would trigger off current though, so if the load in this instance varies, this won't work. To get around this you could have a look at voltage-triggered switches based on relays that mimic a CMOS inverter, which might be interesting and wouldn't waste as much power.

>>1170929
It'd also be easier if he just posted his schematic.
I suspect he knows he'd be torn apart once we saw what he was trying to do though

>>1170935
Maybe he's long-distance charging anon from that other thread? Or maybe he's the railgun anon from that other other thread. But what's the worse that he could be doing? Maybe he's building some massive RF setup and is that anon who was trying to waste as much power as possible?

>>1170936
>long distance charging
what, like being able to charge your phone with a device at home while you're at work? lol. Wouldnt you basically need to emit dangerous amounts of energy to do that? lol. think of the distance from a tesla coil that a fluorescent tube stops glowing at. now imagine extending that range to 10km or something similarly ludicrous.

He can't be the railgun anon though because railguns generally don't require 1MHz for anything except possibly a power supply, and ther's basically no reason for him to use any reverse zener crap for a power supply

>waste as much power as possible anon
Don't know what this was. enlighten me?

>>1170945
>long distance charging
If he's just radiating energy omnidirectionally, he'd lose extreme amounts of energy. If he's transmitting 1kW, and, say, trying to hit an area 1 degree wide in both latitude and longitude, he'd receive only (1/360)^2 of his output power, ie. 7.7mw. Maybe he's using a directed antennae? Fair enough, let's say he can narrow the beam down to thirty degrees and still be likely to hit the one degree his receiver is actually in. That's (1/30)*2, which at 1kW transmitted would mean 1.1W received. For the amount of power he's wasting on this setup, he'd be getting off cheaper by just leaving his truck running and charging his phone from the lighter plug.

>>1170972
Man, I fucked that up. 7.7mW, and (1/30)^2.

>>1170974
Eh, I just told him inverse square, and assumed that if it took X watts at 2cm it would take 65200X watts at 5m, which I think is correct. I don't think you can stop RF from spreading out anyways, especially if you're using a big solenoid as a transmitter with an elongated ferrite core.

>>1171000
Solenoid would be the omnidirectional thing I mentioned. You could get slightly better with an irregular pattern, such as a dipole, but transmitting power wirelessly is stupid ineffective. There's a reason only toothbrushes and other things that you don't want exposed power plugs on do it, and their range is still limited to something like two centimetres. If you can somehow strap a parabolic disc onto a Tesla coil, you could get my second example, but you can't, and the improvement is still not enough to make it worthwhile. You'd be getting off cheaper buying several kilometres of extension cord, for fuck sake, even after the hobos cut it up to steal power from you.

>>1171006
you could use microwaves

>>1171000
I think you meant 6.25X

>>1171000

>>1171017
Oops 2cm vs 5m. misread

>>1171006
I think getting a dipole that can handle that kind of wattage would be tough but I don't know much about RF. I didn't know a parabolic dish would work with RF, but I can say with certainty that I wouldn't want to put my cellphone next to a tesla coil to charge it. Already seen one person do damage to some computer electronics with a plasma ball PCB.

>>1171008
Microwaves are blocked by metal too well, and I think getting them to be received by a phone or something would be a right pain, but otherwise they are ideal for directional power transfer.

>>1170928
>>1170927
You guys simply aren't listening. The voltage I need to withstand is at minimum, 800V, at up to 1.5MHz, and 2 amps. They don't make positive channel mosfets good enough.

Two MOS in series won't work because the source voltage would move, on either side. The point is to prevent voltage from reaching the drain of one mosfet, which means the voltage at the source of the other would swing between 0 and |400V|, and the gate would always have to be swinging in tandem because Vgs max is 20-30V for any modern power transistor. And again, I can't use low-side switching because I need +/- on the hot out. My other design uses a tuned transformer, but (my power needs are at the end of my post) I'm probably going to have issues getting it tuned at the power I need without having kV or hecto-amp surges

>>1170929
MOV might be what I'm looking for it has the high-side "zener" effect.

What people aren't getting: ALL zeners have an ANODE TO CATHODE, POSITIVE TO NEGATIVE forward voltage like that of a diode, and a CATHODE TO ANODE, NEGATIVE TO POSITIVE voltage for breakdown. I need the other way around.

>>1170935
It's a class B amplifier driven into full conduction. Literally it. The PMOS and NMOS have to be high-side switched because of the aforementioned hot and neutral connections.

>>1170936
I need to drive as little as 100 ohms at 1000Vpp max with a neutral ground. This project is a bitch.

>>1171050
Alternatively, you could just post what your project schematic is and we could help you from there

>>1171052
It's a class B amplfier with mosfets on the high side you fuck. I can't finish the schematic because there'd either be shorts, or the circuit wouldn't work. At voltages below breakdown I can use standard components (i.e. clamp some voltages, bias, and call it a day). At the high voltage I need a lot of things which inhibit voltage

I know the zeners need a sink too, the problem is if I hook up the signal in any way with standard parts I'd have high resistance at the input or a short on the side of the function generator.

The pulse can go either way, either pushing more towards zero to turn off the mosfets, or pulling to enable, depending on how I plan to bias.

>>1171021
>wattage
Also known as "power"

>>1171061
>I can't finish the schematic because there'd either be shorts, or the circuit wouldn't work
wat
your circuit wouldnt work? maybe that's your problem?
they also sell mosfets with Vds in excess of 1kV so i don't know what you're bitching about. get mosfet drivers.
and yeah finish the rest of the schematic you fuck.

>>1171064
Look at the schematic you nigger. MONOPOLAR OUTPUT. Find a PMOS with vbreakdown > -600V and a current capability of >-1 A. You can't. They don't exist.

The circuit does work, I have literally done it on a breadboard in a lab. The problem is the voltages required.

>m-muh schematic
Stop being a fucking retard and read my posts, read them again, and then ask someone else until you understand what I have communicated to the thread and what only one guy seems to understand when I have described it on the fucking current-voltage curve and voltage transfer curve level. I'll draw fucking pictures later because it seems people have reading disabilities.

>>1171069
>muh schematic
So post it and prove that what I think about you is wrong.
I'll wait.

>>1171069
>>1171061
I can't fathom what you're trying to do that nobody has a solution online. Say what you're trying to do and stop being an evasive cunt about it. There are surely better ways to do what you're trying to do.

>>1171050
>What people aren't getting

What you aren't getting is that you're not explaining what you need clearly enough. Plus you don't seem very open to advice for someone who's asking for advice.

So you start out with some kind of "reverse zener" nonsense and then it turns out to be a class B amp that you can't finish the schematic for for whatever reason? You built the damn thing in the lab and somehow you can't draw the schematic? Don't shitpost on /diy/ please.

>>1171076
>post the schematic
>get the patent stolen
Yeah, fuck that.

>>1171095
Then I hope he suffers an unfortunate accident whilst dealing with high voltage. If he's smart enough to make something patentable, he wouldn't be so stupid as to be befuddled by a class B amplifier. or generating high voltages in another fashion

>>1171097
99% of patent submissions aren't patentable.

>>1171110
So he gets hospitalized... and he doesn't get a patent. Perfect.

>>1171091
>Plus you don't seem very open to advice for someone who's asking for advice.
It's the same shit with these donut steel guys every time.

I've accidentally snapped the antennae cable on my old RC receiver. The break is too close to the plug to just cut the right length of isolator off, so I'll need to replace it. Since it's a 2.4GHz DSM antennae, would it be possible to replace it with a wifi antennae?

>>1171076
>>1171071
The design is at least 400Vpp, for a load of 100-400 ohms, no exceptions.

The unlabeled boxes are the elements I would need to prevent mosfet breakdown. I have another design I will probably have to use because of time constraints.

>>1171160
Forgot to write: the frequency is around 1MHz. Time on is ms so I don't need to worry about power consumption.

I have worked out every combination of mosfets in series and orientation wrt each voltage and on off, and the voltage fluctuations would require near 400 volts and amps of current for each transition, which would crush my driving signal and circuit or exceed Vgs

>>1171162
Toss the p-fet and modify the drive circuit to live with two n-fets. That's what people do when p-fets or PNP transistors suck too much.
Or find an entirely different arrangement for driving your load.

I found this motor that (I believe) is from an old blender. There's three wires coming out of it red, orange and blue. Which one is ground etc? How do I wire this thing so I can run it? Only worked with small dc motors before so all I know is red and black. Can I power this thing from a 12V battery?

>>1171192
>check labels
>"Voltage: 230V"
Sure, you could power it with 12V. Typically, red is positive and black is negative, so the remaining might be ground.
This information is all either available on the very lable of your motor, or with google.

Oh for fucks sake. Wrote a long post but it disappeared!

Long story short:

I have a 3~welder that has just phase to pahse wired to it, meaning 400V.

It is water cooled and needs a steady floww of water. I have replaced the old leaky pump with a small 12V 2A pump that I'm running on external power.

the welder has a 25V (it's labeled as 25V but it seems to be running at approx. 25V) transformer for the wire feed.

I'd like to leach the pump on this transformer, meaning that I need some simple system to rectify and step down the voltage. Has anybody some idea about which parts I should purchase and how to wire them? It's probably a trivial problem for someone but as a power tool guy I fail at this.

Can anybody explain to me why this 1n5819 schottky diode still manages to light up the LED (albeit dimly) with reverse bias?
Multimeter shows no current flowing.
Did I get chinked or is this how a schottky diode works?

>>1171204
negative feedback?

>>1171197
Don't worry about your post, you probably posted it to /d/ inflation instead, happens all the time.

>>1171204
Mate I can get an LED to light up dimly using my body or a 5M as a series resistor, LEDs are the shit.

>>1171197

simplest solution is power the pump from a 120V extension cord using a common 12V 3A charger.

second simplest is to put a load in series with the pump of equal resistance to the pump. this would divide the voltage to the pump by 2. one choice for such a load is a 12V 25W halogen bulb from a car, or from home lighting. you'll also need to throw in a 3A 50V diode in series with the 2 loads to make it DC. when you have DC, then you gotta make sure to get the polarity right.

third simplest is to buy a suitable supply from ebay:http://www.ebay.com/bhp/24vac-to-12vdc
they sell lots of them but most tend to be around 1.5A.

least simple is to make your own from 4 diodes, a 7812 and 1 or more pass transistors.

>>1171160
Where's the load?
why are you certain there's no exceptions?
there's more practical ways to generate and control 400Vpp

>>1171320
They're specifications for a product you fuck. I can't just go cutting the voltage in half when it is the spec.

The load is at Vo, as it literally always is for every circuit that has a Vo.

>more practical ways
And like I said, 100-400 ohms, AM frequency (variable), voltage is actually variable (500-1000Vpp at user discretion), and the load is at lowest ~80ohms, then goes up to 400 ohms, which could go up to 100 min and 1k max if you can make a pretty clean sinusoid at 1000Vpp. I did make the mistake of writing 400 Vpp, it actually should be at least 800. 400 wouldn't be an issue at all.

>>1171390
Just step up lower voltage to higher voltage
There are these things called transformers

>>1171402
That's the design I have to go with, we're still talking 120RMS to 330RMS ~3 amp at best (if I put both 1.5A taps in parallel AND push 120 on both primaries), and having proper variable tuning on BOTH sides of the transformer for best results.The transformer cost $150 without shipping, and it's meant for a goddamn tube amp.

>>1171390
>>1171160

So it boils down to an inverter gate? I don't see why this was such an ordeal. Anyway, it's all about level-shifting the gate drive. Or use IGBTs.

>>1171410
120VRMS is doable. computer power supplies deal with it all the time. Now just step up instead of stepping down. selecting the transformer will be the hard part. which absolutely simplifies this 'antizener' business you've been trying to figure out

>>1171413
IGBTs don't work at MHz. At all. The highest can go up to a few hundred KHz. The switching times go to shit AND you will consume more power in the damn IGBT than your load.

You still aren't understanding: PMOS can't withstand the breakdown voltage, and if you wanted to switch ~400V at the gate, you'd literally need hundreds of amps. You're only pretending not to pay any attention and have no idea what you're talking about right?

>>1171415
800-1000Vpp = 282-350RMS and if I'm stepping up I have to use the current live, so yeah, 1-2A peak out. I think TVS diodes would work, but I don't know how they'd hold up to regular RF use and switching amps on and off over and over.

>>1171416
Call it 2 MHz, that's 500 ns. IGBTs are faster than that.

How long would you need "hundreds of amps" for?

hey /g/ asks a question.

>>>/g/60184794

>>1171420
Dude, look up ZVS drivers. unless your'e dealing with electrical grids and national transmission lines, i can't fathom that there aren't more practical ways to get what you want.

Driving lower voltage high current to get your high voltage and (hopefully) necessary current is way more practical no? Unless there's some reason you have to be driving the high voltage directly. Surely stepping up from 120VRMS isn't that much of a stretch

>>1171422
1) IGBTs do not switch that fast you fuck. Here's an example of an IGBT: https://www.fairchildsemi.com/datasheets/FG/FGH40T100SMD.pdf

It's turn on time is good, <50ns. It's turn off delay is fucking atrocious, 300ns. For 1Mhz signal, your circuit would be completely open for nearly 80% of your cycle before turning off at maximum speed (takes 300ns to turn off, while the other turns on 100ns delay.. the crossover would be atrocious). Let's also count that it'd take 3.5mJ per turn on and turn off, so 1 MHz would require at least 3.5 fucking kilowatts to turn the damn thing on and off at 1MHz. With a total gate charge of 265 nC, you'd need 2.65A to turn it on or off in 100ns, which is 20% of your cycle. Shit all around at rf.

Since you're being an idiot, let me show you:

>>1171426
ZVS are designed for fixed loads, fixed frequencies, and fixed voltage 98% of the time (50 Ohm antennas). A power amp could work, but there's only papers on getting them to work with varying resistances or frequencies, nothing anyone has done at high voltage. I did go through a power amp already, a class E kind, and would go through kilovolt spikes across everything and wasn't flexible at all (nearly all the parameters are dependent). I'm probably going to stick with the transformer for now, but there is the option of ones that can handle variable resistance (not sure about frequency). To tune the "voltage" I could just change the power, but a crapton of component values depend on the power intended to be delivered.

>>1171422
Forgot 2: Mosfets require amps to switch voltage when all you're dealing with is 0-15V. To switch up to 0-395 in the same rise/fall time would take around 26x the current.

>>1171432
You didn't give any power requirements, or RF quality requirements. State your specs.

There are faster IGBTs out there, and you can parallel them.

>>1171265
>simplest solution is power the pump from a 120V extension cord using a common 12V 3A charger.

Technically yes. But but. The old 3~ cables had 3xphase + ground. The original idea was to switch the cable to incorporate neutral. Then I would have 230V inside the machine. But in old installations there is no neutral coming from the wall... And my 3~ extension is longer than my 230 extensions. Eventually it would have been a problem.

>second simplest is to put a load in series with the pump of equal resistance to the pump. this would divide the voltage to the pump by 2.
I was thinking about this. To add a second pump to up the pressure. But for a perfect system the other would have been an inline pump. There was no available with fitting specs and it would have meant an extra level of potential leaks (the machine is old and leaks are already a problem).

>least simple is to make your own from 4 diodes, a 7812 and 1 or more pass transistors.
Originally I was aiming for this

>third simplest is to buy a suitable supply from ebay:http://www.ebay.com/bhp/24vac-to-12vdc
>they sell lots of them but most tend to be around 1.5A.
But these things were abundant and easily found! Ordered a 5A one from the krauz Amazon and all I'd good again!

Thanks for the help. I feel enlightened.

>>1171213
Might be. I checked the thread you mentioned though but it wasn't there. Maybe the post got deleted as there was nothing freakish in it? It's something for the philosophers to ponder about...

>>1171434
You can always use feedback and optocouplers to regulate the secondary. You can parallel mosfets to drive higher currents through the primary though you might have issues with timing or whatever (not sure), but that's still monumentally more practical than whatever 'anti zener' thing you thought you had to deal with
People use zvs drivers to drive their diy induction shit (to be fair i haven't read up a lot on ZVS drivers) or generate tens of KV arcs to play with converting lower voltage to slightly higher voltage (~300VRMS) should be relatively easy

http://kralciplak.tripod.com/Electronic/smps.html#pushpull
Look at push-pull, half bridge, and full bridge (H bridge)
but instead of stepping down, step up

Just got redirected here - is this the right place to ask a few questions about building coils and general vape stuff?

>>1171445
You can also inductively couple the secondary to generate a much smaller voltage back to the primary with which you use to regulate the secondary, if optocouplers aren't fast enough.

>>1171446
>vaping
leave
but in all seriousness vaping is just some lithium cells, nichrome wire, and some basic current regulation (or at least they build on those basic blocks) so I'm sure someone can help you

>>1171449
>leave
Pretty much what I expected. I don't want to pollute your general if that's the consensus - I'm new to it, is all. Maybe I should just make a new thread

>>1171453
No, I honestly don't know anything about how vaping setups work, so i specifically can't help you. Not to mention you haven't really asked about something more specific

>>1171454
Oh okay. I made a thread but if I don't get any help there, I'll try back here. Thanks anon.

>>1171437
1) It's not for RF. It's just going to deliver voltage to a load at that requirement.

2) I gave the voltage and current requirements, can you seriously not do the math? if I round down the power requirements, it'd be 800pp on a 400ohm load, sinusoidal (because I guess I have to spell that for you too) . That's 200W average over the duration of the pulse, and only needs to be on about 5ms out of a second. At peak the resistance can be 100ohms or so, which is 800W peak.

>>1171445
None of the designs are for sinusoidal, or at least 50/50 square output.

>>1171444
I was actually implying that you frequented /d/ inflation, sorry.

>>1171498
The point is, step up your voltages. use any of those popular topologies or variations ofthem. if you have sinusoidal inputs already, it shouldn't be too hard for you to adapt it to a lower voltage to drive the primary of a step up transformer to get the voltages you need

Trying to figure out the concept of parallel resistive loads dropping effective resistance. So let's say for example I have a 20 ohm heating element and At 120v you could only push about 6 amps through it. If I take that element and cut it exactly in half making two 10 ohm elements and wire them in parallel, the overall resistance is now equal to 5 ohms and a 120v circuit can now push 12 amps through the elements, 6 In each.
Am I getting this right?

>>1171515
yes
provided your elements could handle the current / wattage

>>1171515
Resistances are right, but it's 12A per element (120V/10R), 24A total (120V/5R).

>>1171508
>"shouldn't be"
Yeah no shit. Try making 1000pp 1MHz sinusoidal with tunable output, variable voltage, and peak current capacity of 5A. There's a reason the machine I'm making hasn't come to market yet.

Building something that needs to read if a 240v AC solenoid is active or not.
Finding a relay that would run at 240v and putting it in parallel with the solenoid was my first thought, but something like an opto-isolator seems like a better idea. Sadly I'm not particularly familiar with them.
I just need a 5V DC signal completely isolated from the solenoid I'm reading off.
I'm half thinking a 240v neon bulb and a photoresistor but if that's what a opto-isolator already is then I don't need to bother.
Anyone care to help?
Thanks for your time.

Why is it difficult to find power transistors rated higher than 15 amps, while MOSFETs with ratings of 10, 20, 50, or even 100 amps can easily be obtained?

Which kind of switch should I use for PWM driving a brushed electric motor for a bicycle?

So I have this electromagnet I made real quick with 60 feet of solid core copper wire. Let's say I want to power it using an outlet. I imagine an AC/DC converter would be used, right? What sort of output should I look for, and well, what kind of stuff in general? Right now I power it just by touching the two exposed copper ends to either pole on the battery pack in the picture.

>>1171630
because MOSFETS have very low resistance when they are fully on. so when you pass current through the, (I^2)*R tells you how much power dissipation happens across them, while for BJTs its never quite so low
Another way to think about it is that the voltage drop across the transistor is very low for mosfets
as for the motor i'm not certain

The local electronic parts shop managed to run out of 2n3904s so I got a couple of 2n2222s instead. Now Im reading up on the part abd it says that the 3904 teaches peak gain at 10 mA, whereas the 2222 gets there at 150mA. I'm making a second stage in a preamp that already has a 3904 in the first stage. The input is coming from a piezo microphone clip, so what amount of amperes could it be producing?

>>1171501
And trust me, I did get it. Had to check out this inflation though. Did expect to see something about insertions but it was just big bellies. This site sure has some weird boards.

>>1171647
Fuck mate, just plug it straight into the wall!

>>1171630
Use a MOSFET for the motor, you'll probably want to make sure you've got a protection diode too if you haven't already. I don't suppose regenerative breaking on a bike is worthwhile?

>>1171647
Just make an AC/DC converter with a full bridge rectifier. I'd suggest putting some beefy filter caps on that too, but I'm not sure whether you'd get a resonance effect or not. If we assume your inductor is 5mH or so, then for resonance at 60Hz you'd need capacitance of 1400uF, so don't let your filter cap be too big.

>>1171647
You can run an electromagnet from AC.

The main issue is that it's going to have a very low DC resistance (60' is a rather short piece of wire), so you'd be better off with a constant-current supply. Or failing that, powering it with AC via a ballast inductor (this limits the current without dissipating power as heat).

If you want to power it from a constant-voltage DC supply, you'll need to measure its resistance.

Either way, you need to know the current rating for the wire.

>>1171779
>I don't suppose regenerative breaking on a bike is worthwhile?
If you have the budget/time for it, there's really no reason to not go for it. You'll still have ordinary brakes for emergencies, but for less hasty braking you'll regain some energy and extend your battery time a tiny bit.

>>1170041
Yeah. You might want to look up the arduino's output pin specs but that's how it should work. The digital pin is basically just going to act like the switch shorting the keypad pin to ground. When the arduino writes a low a value on that output, assuming it's open drain, it turns on a FET that connects the keypad pin to ground.

Look up a pull-up resistor. That's probably what's setting the keypad pin high when the switch is open.

>>1171204
Schottky diodes have really leakage current. I think on the order of mA.

>>1171423
>/g/
anyone?

>>1171839
No, it wouldn't, because those tiny transformers cost very little, while rewiring all houses everywhere and replacing all the appliances everywhere, is quite a bit more expensive.

>>1171839
>>1171423
Main problem with that is how flexible a DC system can be, you've got to pick a single voltage, anything that needs something different will need to have a buck or boost converter anyway to get what it needs, which now isn't that much cheaper than a mains voltage AC converter, so going highish voltage DC isn't sensible.

If you go for a lower voltage you then need far thicker cabling to carry the same power with the same efficiency, if you go with say a 12V bus driving a 600W load over 10m distance you'd need cables nearly 100mm^2 to get a 98% efficiency just from the cable, vs 0.5mm^2 for 230VAC or 1.5mm^2 for 115VAC.

>>1171802
>>1171779
Thanks guys. I'll google all the stuff I didn't understand, just needed a starting point

>>1171617
They make contactless ammeters. If you can deal with it, just put an ammeter in series and leave it on while the solenoid is running. Since the solenoid has to be switched on, you could also run a diode (rated properly) to a capacitor, and then use a voltage divider to drop the roughly 339 V signal (rectified 240) down to digital level. You could put a 100k ohm pot across the solenoid and turn the pot until you get a signal level you like. Feed that into your detector (or an op-amp as a buffer) and you're good. If you think the voltage would vary too much, just use the op-amp with a voltage reference, so if it's above say 2V, you output your "on" signal, and below is off.

>>1170923
How about post your schematic?

also,
New Thread >>>>1171979

>>1171498
You've been given lots of clues but given your dismissive attitude I dunno how far you'll get. If you actually knew anything you'd have designed this shit already and not needed to cry about how hard it hurts your brain.

>>1171983
This pretty much. Sounds easy but he's just dumb as a sack of hard drives

ohm i want to make a 2MW class D amplifier give me a schematic please also i only want to use tubes no mosfets