/ohm/ - the electronics thread

File: 1860831442237758384.jpg (973KB, 3264x2448px) Image search: [Google]

973KB, 3264x2448px

So I recently came across a genuine hakko 936 soldering set when I was helping clean out a lab. The controller is busted which is probably why it was hiding in a box in the closet and not being used. The iron is still good and I'm thinking of making something like pic related which is a mobile soldering station, runs off any voltage like from a car or lab power supply and even RC batteries.

When I look up 907 controller on eBay I'm given dozens of variations, does anyone have any experience with these things or recommendations? Or even any info on diy versions (I've seen some done but with little documentation)

File: Untitled.png (33KB, 768x768px)

33KB, 768x768px

would this all be possible with a single arduino? i know very little about them and never used one but im willing to learn how to wire and program it. basically this stemmed from wanting to electronically control the throttle of my diy generator engine (yeah im that guy from before) to keep it in the efficient rpm range when the electrical load starts to slow the engine down. i thought that since im already going through the trouble with the throttle and im going to have the other sensors anyway i might as well hook them all up to a controller and have them output to a screen rather than have 5 different gauges and displays. im thinking of using an uno r3 with 14 digital i/o pins and 6 analog input pins plus whatever sensors and controllers ill need. on the pic i included what the expected types of readings on inputs should be. the only thing i am concerned with is trying to read alternator voltage and amperage because i dont know if i will accidently fry the board or not.
I should note I'm not really looking for a how to, I just want a "yes that will work" or "no you'll need to do this to make that work" or whatever in-between.

Absolute noob here. I made a small circuit on a breadboard by following a tutorial, but they didn't cover how to go from being on a breadboard to being soldered together. The board's grouped by rows, so do I just solder everything that's in the same row together? So for example, all the negative terminals are on row 16, so I'd wrap all those together and solder them, right?

File: muh circuit.png (749KB, 1032x774px) Image search: [Google]

749KB, 1032x774px

>>1070429
Forgot pic

Hi /ohm/ I recently found this compilation of schematics, most of them are guitar-related (amplifiers and effects) I thought it might interest you.

http://www.freeinfosociety.com/media_index.php?cat=13&start=0

>>1070429
google it

>>1070423
Yes, it could work, arduino has 6 analog inputs, but it also has 14 digital pins which can be configured as digital inputs as well. The only problem would be signal conditioning, this is, making all the signals into something arduino can interpret, but there are lots of modules available to make that easier.
Take into account arduino functions are pretty inefficient, if you need a faster code you'll have to use "true c", they're usually fast enough, but with so many things to control it might become noticeable.

>>1070433
>>1070429
Yes, anything connected on a breadboard should be soldered toghether. I recommend using tin plated protoboards to begin with.

>>1070371
Hmm.
seems to be the same old design with a triac, like my weller.
this means it needs AC if you want to base the build on the original design.

i guess you are going to need to make a new controller with pulsed DC.
a µC and some opamps to work the temp feedback and a fet to pulse the heating should be about all you need.

might be a bit fiddly, but doable.

http://dangerousprototypes.com/forum/download/file.php?id=4702&mode=view

>>1070371
since the iron is supposed to run on 24v ac, i dont think you can go much lower than 20v dc and still have decent heat.
so car batteries and such will be out if you want to use this iron.

kool idea thought, keep us posted.

File: Untitled.png (15KB, 768x768px) Image search: [Google]

15KB, 768x768px

i have a meter calibrated for a 100 amp 75mv shunt that cannot be recalibrated to 200 amps like i want. if i take two 200 amp 75mv shunts and wire them in series and then use the outside sensing lines for the meters sensing + and - lines, will it act like a 200a 150mv shunt and give me an accurate result? pic for clarity

>>1070594
Two shunts in series are equivalent to a 100A 150mV shunt. Two shunts in parallel are equivalent to a 200A 75mV shunt.

For 200A 150mV, you'd need 4 shunts (2x2).
But why would you want a 150mV shunt for a meter with 75mV FSD?

>>1070594
>accurate result
No. You need to use the given sensing connections properly to get accurate readings.
>will it act like a 200a 150mv shunt
Ignoring the rather significant problem above, yes.
This was to increase the range of a 100A meter to 200A, right? There are no guarantees that your meter is capable of displaying currents above 100A. Check if it will read 200A with 150mV input first.

>>1070598
It would be two 200 amp shunts not two 100 amp shunts.
Because the meter I want to use shows volts amps watts and kW hours but is only available with up to 100 amp configuration and is not reprogramable

>>1070604
So what if i run a jumper line between the two sensing points I left unconnected in the picture?
I'll go ahead and check if it works too

>>1070607
> So what if i run a jumper line between the two sensing points I left unconnected in the picture?
You'll be running 100A through the jumper.

If you want to put two shunts in series, you'll just have to live with the fact that the total resistance will be slightly more than twice the nominal resistance, meaning that the voltage reading will be slightly higher.

There's no way to connect 2 4-terminal (Kelvin) resistors in series to make a single 4-terminal resistor with exactly twice the resistance. You can't eliminate the extra resistance caused by the connection from the measurement.

File: 20161014_215658.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

made a new preamp today.
ax84's blue preamp.

its appealed to me because it had things i havent tried yet.
pentode input, cathode follower, led bias on the later stages and triodes in paralell.

i used shielded cables on the heater, making it silent form the get go.
still has the heater elevation from the previous set up.

has decent gain, but that must be because i once again postponed building a tonestack.

there is a "squish" pot from ground to the pentode's screen through a cap. pretty interesting.

I'll experiment with bias next.
after that i think I'll split the parallel tubes into two gainstages or drop the cathodefollower for a single gain stage.
dunno yet.

power supplyfag here from last thread. looks like there's no way to amplify DC voltage. all the amplifiers seem to say AC, and building up the circuis (they resemble class AB amplifiers, but i've also built class A type ones) i get roughly the same voltage out that i put in... so this kinda sucks now. lol

>>1070884
>there's no way to amplify DC voltage
>what is an op amp.

>>1070889
if you read the last thread, you'd realize this all began because an op amp cannot handle 40v supply rail differential. and i'm not buying a 5+ dollar op amp just for it either.

I was going to put up some LED lightning with PWM control, and I begun to think if it would be a good idea to try to avoid spewing EMI all over the place. The reason I'm somewhat concerned about EMI is that there is some 10ft of cable between the wallwart and the LEDs that could act like a huge antenna (maybe?).

So the idea would be to have some sort of input filter for the PWM controller so that the chain would be somthing like: wallwart->10ft cable->filter->pwm->leds. But I have no idea how to design such filter.. I suppose some inductors and capacitors might be involved.

What sort of filter would be best suited for this purpose, and any hints on component selection for it?

>>1070884
Sounds like you need a buck boost regulator. If you don't feel like building one from scratch there are plenty of premade ones online or car parts stores.

>>1070889
Fuck off. That's not what op-amps do. The output voltage can't be higher than the supply you're powering them with.

>>1070903
without any context, was op amps a bad answer?
one of their uses is literally DC coupled amplification.

if you want higher voltage than the supply, you are not 'amplifying' the voltage.
get a 34063 and calm yo tits.

>>1070902

Stick an inductor in series with the LEDs, and add a flyback diode to allow current to continue circulating during the off portion of the cycle. Value is relatively arbitrary; anything above, I don't know, like 50µH should work fine. Doing so effectively forms a rudimentary buck converter. This will have the added benefit of providing a more constant current to the LEDs.

Even with no filter, it's unlikely it would cause any problems as far as EMI goes in this case. Trying to keep it to a minimum is just good practice in general, though.

>>1070902
For the current drawn by the PWM unit, a capacitor will probably be sufficient. Either ceramic or low-ESR electrolytic depending upon the value.

For the current supplied by the PWM unit, an inductor and a diode will turn it into a constant-current buck converter.

>>1070903
I definitely don't want to build a buck or boost converter from scratch. I'm just building a linear power supply. I need to be able to supply an LM317's adjust pin with as low as -1V and as high as ~30V that's 31V range, and most op amps even if they have +/- 18V supplies, won't allow you to get to within ~3 V of either rail

So even if I had 0 - 36V available, i need -4 to 34V, or thereabouts, in order to push the required voltage into the voltage regulator. the transistor, class AB amplifier looked promising but wouldn't push out more than 5V. there's an article i found where the Op Amp's power rails change in response to the output , so i might give that a try

>>1070912
also funny you should mention, i'm boosting an old computer power supply's 12V to 35V (it's the max setting on this boost converter) then doing a linear supply off that. because fuck getting a transformer large enough to supply ~100W at ~15-0-15 VAC

>>1070912
>>1070918
was adding to my own post btw, not another person responding

>>1070912
Well, there's no "linear" way to generate a DC voltage higher than your input.

>>1070933
I'm not intending to produce higher voltage than the input...

File: grrrrr.png (48KB, 1185x508px)

48KB, 1185x508px

Anyone here have experience working with instrumentation amplifiers?

I'm trying to test this one (INA333) I want to amplify a 2mV signal to a useful level. 1V, 3V, 5V, anything really.

I've been looking over the datasheet and it seems like I'm doing everything correctly, but I can't amplify the signal beyond about 3.6mV.

I've tried using 1/2 VCC at the Vref pin. I realize that voltage divider isn't doing anything in the current pic.

Any ideas what I'm doing wrong?

(Software is TI- TINA. Its a SPICE-based program you can get free from TI's website.)

File: hnnnnng.png (49KB, 1294x553px) Image search: [Google]

49KB, 1294x553px

>>1071093
shit, I got it. Apparently you do have to actually use a dual supply with a dual supply IC

>>1071096

or, at least, the simulator thinks so. could be different in real life.

>>1071097
Yeah, it threw me off because all the reference designs I've seen just use it as a single supply amp.

>>1071097
The real life is described in INA333's datasheet, particularly in figures 20-23.

File: rtfm.jpg (113KB, 829x720px) Image search: [Google]

113KB, 829x720px

>>1071097
Always read the datasheet, otherwise you get burned (and sometimes you still get burned even if you read the datasheet, but it's less likely).

>>1071097
Would have thought bipolar supply was standard in these applications.
dealing with bias and decoupling has some pitfalls when considering signal integrity.

>>1071093
> I've tried using 1/2 VCC at the Vref pin.
You also need to use it at the inverting input.

"Ground" can be whatever you want it to be, but it has to be consistent.

Can someone help me with
>>1071002
Thanks

>>1071142

30V is higher than 24V so you're definitely over the max in voltage. look for a diff controller unless you like to live dangerously.

>>1071145
so the 12/24v is talking about input and output?
So the range would be between 12v and 24v?
If my panels produce 18v would that be fine

I'm making a project that needs a decently powerful 12v electromagnet, 50-200lbs or so, I haven't decided. The only problem is, it's fairly heat-sensitive, and all the reviews I'm seeing on cheap magnets say they get crazy hot. Is that just a limitation of electromagnet, or do I just need to spend more?

>>1071324
>cheap
One of the most obvious ways to reduce the price is to use less copper. Less copper means more heat.
Maybe you could replace the magnet with something entirely different.

>>1071331
So a nicer one will probably be fine? Anyone have suggestions as to where to source it from?

>>1071093
EMG amplifier?

>>1071334
Bigger, heavier magnet heats less. Whether it's fine depends on you.

Depending on what you're trying to do, it might be possible to reduce the power dissipation by using full/high current only momentarily to pull shit in and then using much lower current for hold.

>>1071335
ECG actually, but that would be cool too I would like to make one of those. Maybe turn it into some kind of arm controller for a robot or something.

>>1071126
ohhhh, I didn't read that as carefully as I should have. Thanks for pointing that out! I'll try single supply again and offset my inputs at least 100mV.

File: SVF.png (56KB, 1440x1088px) Image search: [Google]

56KB, 1440x1088px

Hi again, so my circuit seems to be working well so far. I've set my instrumentation amplifier so a gain of 10, though I may increase this. So, 1mV in, 10mV out.

Now I just need to add some additional filtering post- INA before the final amplification stage and I was hoping to get some input.

It will be a band-pass filter, with the pass band ranging from 0.5Hz to 40kHz OR 150kHz switchable. I want this to be switchable so that the ECG has two "modes". One for basic patient monitoring, and one more detailed and possibly noisy one for diagnostic purposes.

Anyway, I'm wondering what the best way to accomplish this might be. I've been thinking of implementing a state-variable filter like I've designed in pic related. The bandwidth can be toggled by the SPDT (it adjusts the LPF gain). BUT when its toggled, there is a couple extra dB's of gain in the pass band.

here's the simulation. I'm pretty sure it will work but it may not if you don't own EveryCircuit http://everycircuit.com/circuit/5002482853347328/state-variable-filter--mtr-diag

I'm wondering if I should just go with a more basic sallen-key topology.

Sorry for the word-vomit.

>>1071389
>40kHz OR 150kHz
I guess you mean 40Hz or 150Hz. If you're going to feed the signal to an ADC, you could just go with 150Hz (or more) and filter more with software if/when needed. Or if you want switchable filters, then 50Hz or 60Hz notch filter would be worth considering.
>0.5Hz
Unless you have a rather fancy HP filter, this isn't really enough for ECG. It is enough for heart rate measurements, though.
>if I should just go with a more basic sallen-key topology
I would. Or with MFB.

>>1071402
alright, cool. When you say 0.5Hz isn't enough, do you mean that would be too low or too high?

I do also want to add a 60Hz notch in series after the BPF.

>>1071406
Well, depends on what you really want, but the phase error of simple HP filters distorts the low frequency components of the waveform. 0.05Hz is the usual recommendation, if there's no (much) movement and/or you're interested in the ST segment.
AFAIK this recommendation is based on the assumption that you have a basic 1st order filter. If you have a good enough filter, you can go up to 1 / lowest expected heart rate, meaning 0.5 ... 1Hz. 0.67Hz is one recommendation for such filters. "Good enough" filters are generally implemented in software.
Or maybe you simply aren't interested in the low frequency components, in which case 0.5Hz would be enough.

File: 3245.png (40KB, 1187x374px) Image search: [Google]

40KB, 1187x374px

>>1071415
ok thanks a lot, I'll definitely take that into consideration. I do want to make it as accurate as possible, so I will have to play around more with high pass filters. I likely will go for something closer to the 0.05Hz range.

You seem to be really knowledgeable about this stuff. I was just trying to design a driven right leg circuit for extra common-mode rejection.

In reference designs I've seen from TI, they have the non-inverting input of the comparator set to 1/2 VCC. This doesn't make any sense to me, and I can't even get a signal on the oscilloscope using that configuration.

Am I right in using a (very) small reference voltage at this pin? Maybe I should even include a potentiometer to adjust this Vref?

Green wave is simulating a 1mV common mode voltage.

>>1071427
It's not a comparator, but an inverting amplifier with "ground" at mid-supply. The basic idea of that circuit is to keep the common mode input voltage of the input amplifier close to mid-supply so that the input amplifier has as much headroom as possible (when using single supply). If you have dual supplies, then V3 would be 0.

File: 93245.png (34KB, 1076x396px) Image search: [Google]

34KB, 1076x396px

>>1071440
Ok, yes, that's about what I thought. I did realize that its an inverting amplifier. This makes a lot more sense now.

Gahhh, once again supply voltages messed me up! I tried supplying this op amp with dual supply and it works as expected now.

However, the datasheet for the OPA333 says that its single supply operation and doesn't mention anything about dual supply (despite it having V+ and V- supply pins). I guess I'll just have to see if it works in the real world when I'm prototyping lol.

>>1071447
"Dual-supply" basically means that there is some pin which must be kept mid-way between the other two supply pins.

"Single-supply" means that there isn't such a pin.

In either case, your signal "ground" reference should be somewhere between the two supply voltages, not at the most negative supply voltage (regardless of whether you label it "V-" or "0V").

File: 1474847935117.gif (24KB, 60x95px) Image search: [Google]

24KB, 60x95px

>>1071476
Oh, I just thought that it meant that it would be powered with (relative) positive and negative voltages. ... I'm learning a lot today. Thanks for helping a noob out

File: floating supply.png (3KB, 251x219px) Image search: [Google]

3KB, 251x219px

>>1071479
In either case, your signal "ground" reference should be somewhere between the two supply voltages, not at the most negative supply voltage (regardless of whether you label it "V-" or "0V").

I am aware of that at least :P. My power supply for this circuit will probably look something more like this in the final revision.

>>1070254
salvaged 2 electrolitic capacitors
1.5kV - 2uF

Any projects i ideas/projects i can use them for?

I think my tractor has an electrical short somewhere that is draining the battery - that or the stator is bad

if i connect an ameter between the positive lead and the positive terminal while the key is off, i should see no current, correct?

any current would indicate something is shorting out..

am i correct?

/diy/ What is required for precise control of a DC motor? I understand how DC motors work, but from what I've read, a feedback system in the event of the motor slipping or something like that (I'm planning on using the motor for wheels). Is that feedback system a PID controller, or am I googling the wrong things?

I've been looking at the arduino and found a shield on adafruit that allows for 4 dc motors to be driven, but I assume that means no feedback system.

Am I on the right track?

>>1071649
yeah

What gauge wire do you typically use and where do you get it? I'm looking at 22-22 ga on Amazon and pickings are slim

>>1071691
depends what you want to do with it. How much current are you driving through it? Also, amazon is shit for most electronics things (not talking about consumer electronics). Try your local electronics store if you have one. Otherwise, digikey, or any of the sources listed in OP.

>>1071692
4A. I'm going to be using one of https://www.amazon.com/gp/product/B00A71CMDU at some point in the circuit. I haven't had any physics since high school, and I don't know how that affects things. I've only been at this as a hobby for a few weeks

>>1071693
Mind if I ask what you are going to do with that DC/DC converter?

Like a lot of stuff on amazon, I've never heard of that make. A guy on youtube by the name of big clyde regularly does teardowns of stuff he's bought on amazon, and its almost never up to spec.

And depending on the application, the typical sources like digikey may have a better solution. For example, a linear regulator is a common chip which you can feed with a large voltage, and it will always output a nice, relatively smaller voltage. Common ones are the 74xx series, where the xx are the output voltage you want. eg, 7405 can be fed with 12 V easily, and will output a regulated 5V. And they're usually about a dollar!

As far as wire guages go, you could try to look at the charts to glean the amount of amps and volts that you can safely push through a given length https://en.wikipedia.org/wiki/American_wire_gauge. But honestly, as long as it doesn't get too hot and catch fire, you're ok.

22 AWG is quite thin though. I would recommend something larger for 4A. Probably more like a 20 or 18 AWG.

File: Untitled.png (8KB, 800x600px) Image search: [Google]

8KB, 800x600px

>>1071698
Here's what was going through my head when I bought it, although it seems kind of nonsensical now that I actually look at it

I mean to take more time to learn how all of this actually works, but I just kind of jumped into this project with no background knowledge

can a 110 volt 15 amp 60hz modified sine wave inverter be filtered to a pure sine wave with an rc circuit? how do i calculate what values are needed for the caps and resistors?

>>1071720

absolutely not. RC filters are not for high-power stuff coz they would consume half the power going to the load.

File: USB car charger.jpg (45KB, 500x500px) Image search: [Google]

45KB, 500x500px

>>1071703

that's pretty easily doable. see if you have one of these around that's rated for 5V @ 2.5A. if the plug is micro-USB then you can use it directly on the RasPi.

>>1071744
I'll go hunting for a linear regulator like that anon mentioned, then, unless there's a more elegant solution

I saw in an instructable that someone doing a similar project had soldered a USB cable onto a +5VDC connector on the circuit board for their monitor, but I don't really want to try that since I'm not very experienced with soldering, and it takes forever to ship them in from Hong Kong in case I ruin it

>>1071751

you're not paying attention, dude. i just gave you a better solution. a linear regulator is gonna get very very hot, generating about 17W at max load. the thing i showed you uses a switching regulator which only uses about a watt.

>>1071771
Oh! I didn't realize that's what you were saying to use it for. I think I understand now, use it as the "magical converter thing"?

I don't have one rated for 5v/2.5A, but I'm going to take apart one of the ones I don't use anymore so I can get an idea of what's inside of these things before I go looking for one that fits my needs

File: Capture.png (54KB, 602x511px) Image search: [Google]

54KB, 602x511px

This curriculum makes me want to commit sudoku sometimes.

>>1071781
Yeah, plug-and-chug is so boring, isn't it?

>>1071664
> /diy/ What is required for precise control of a DC motor? I understand how DC motors work, but from what I've read, a feedback system in the event of the motor slipping or something like that (I'm planning on using the motor for wheels). Is that feedback system a PID controller, or am I googling the wrong things?
What are you trying to control? Position or speed?

PID control is used to make the output of a "plant" match a control signal. A common example is a servomechanism, where you want to rotate a shaft to a specific angle based upon a control signal. This requires attaching some kind of sensor (e.g. a potentiometer or a shaft encoder) to the shaft and varying the input voltage based upon the difference between the desired shaft position and the current shaft position.

If you're trying to precisely control the speed of a motor, you can use the same approach but measuring the speed rather than the position.

The general reason for using PID control is to account for the fact that the plant's output (e.g. shaft position or rotation speed) doesn't correspond directly to the input (e.g. voltage or current), but is also affected by factors such as inertia, friction, etc.

But in many cases, a simpler approach will suffice. E.g. "bang-bang" control simply alternates between full power and no power depending upon whether the output is below or above the desired value. When it's about right, the system will naturally alternate between the two states. The most common example of this is a thermostat, which turns heating on and off in order to maintain a desired temperature.

>>1071720
> can a 110 volt 15 amp 60hz modified sine wave inverter be filtered to a pure sine wave with an rc circuit?
No. A passive RC filter can't do better than 6dB/octave. And as >>1071737 says, it will dissipate power.

An LC bandpass filter can produce a pure sine wave, but I wouldn't assume that a mains inverter will be happy driving such a load.

>>1071838
Friends and I were looking into making a cart with zero radius turning, but we don't know anything about the mechanics regarding vehicles. I've looked at a few implementations of zero radius turning and it seems like having 4 motorized wheels seems to be the most accurate in terms of control, opposed to letting wheels skid or whatever.

I read that the "bang-bang" approach doesn't allow to correct for wheels skidding or other problems.

So I'm looking for a rotary switch and I can't seem to find one to suit my purposes. I need 5 positions, each one being basically a triple-pole, triple throw. Anyone have any ideas on how to shop for something like this? I've been looking around digikey but no luck so far.

File: ideal rotary.png (198KB, 351x299px) Image search: [Google]

198KB, 351x299px

>>1071868
forgot pic

>>1071859
There's no difference between PID and bang-bang or whatever in that situation.

There are three basic levels of complexity:

1. Apply equal and opposite voltages to opposing motors, In theory, the motors should rotate at the same speed in opposite directions. In practice, there may be slight speed differences causing the turning axis to be off-centre.

2. Use feedback-based control to ensure that the wheels actually turn at the desired speed. This requires the ability to measure the motor speed. The control mechanism doesn't matter much so long as the control loop is stable, which basically means that it doesn't over-correct errors. This won't do anything about wheels slipping, though.

3. Use feedback-based control to ensure that the vehicle actually turns as desired, so if the turning axis drifts, you adjust the drive signals so that the vehicle ends up in the desired final position. This requires absolute position sensing, e.g. using cameras and computer vision techniques, or accelerometers and gyros and using a Kalman filter to correct for drift.

In all cases, the drive electronics are the same. The difference is in what sensors you have and the math and programming used to determine the drive signals.

>>1071873
Thanks a lot, this helps a ton!

>>1071868
how can a 5 position rotary switch be triple throw?
you want a 3 pole 5 position rotary switch?
you can have a 3 pole 4 position or 4 pole 3 position no problem
if you want 5 position you probably need to buy individual 'wafers' bits, one for each pole and stack them and put a rotor through them all.

>>1071799
>plug and chug
Senpai, you don't even know. That was actually a pic of my paper not the book. I had to figure those equations out.

>>1071781
>>1071906

How did you derive those equations? Electrophoresis/diffusion?

>>1071781
hey man learning how shit actually works is cool as fuck.

>>1071691

I use 30 gauge solid core kynar wire for ps1 modchips

I use 26 gauge solid core wire for short power leads

I use 14 gauge dual stranded wires for speakers

For video I use 75 ohm mini coax

File: work.jpg (222KB, 800x973px) Image search: [Google]

222KB, 800x973px

I intend to go into school soon for electrical and mechanical engineering, I also have a decent background in computers and light knowledge in
programming with python (and hopefully someday C and its derivatives). What exactly should an individual such as I specialize into? The whole field in general interests me, I just like to build and fix shit with both electronics and computers. I might drop the mechanical engineering part though and replace it with comp sci, since that does interest me more.

>>1071781
This is basically my program... Microelectronics and Electrotechnology. I love the circuitry, IC, uC and simulation stuff, but yeah, material science gets pretty fucking tedious at times. Though I at least know how those things really work deep down, which is great. No blackbox syndrome whatsoever.

>>1072347
see: >>1071781

I don't want to discourage you from going into engineering, but there's a lot more solving textbook problems with calculus than fun electronics and programming projects. The best way to prepare is to study calculus. If you go into engineering without knowing any and get too busy to thoroughly learn everything in Calc1 you're completely fucked.

>>1071868
>>1071870
it's going to be tough to find a mechanical switch to do this. what's your application? you may want to consider an array of FET switches.

>>1072445
He will almost certainly be able to get a 3P5T mechanical switch, but it will probably need to be a modular type, i.e. 3 wafers, each of which is single-pole, N-throw for N>=5, stacked on a common rotor, with a stop to block off unused positions.

File: filter.png (18KB, 1491x780px) Image search: [Google]

18KB, 1491x780px

Is there some common name for this type of low-pass filter? Does it even make any sense?

>>1072472
> Is there some common name for this type of low-pass filter?
Nope. It's just an L-C filter followed by an R-C filter.

> Does it even make any sense?
No idea.

I have two arduinos connected by about 10m of cable. The two arduinos will be powered from two separate switch mode power supplies. The cable is to send a simple signal from one to the other i.e., arduino 1 pulses a pin high for 1 seconds and arduino 2 reads that as the signal.

Should I just run a signal wire and a ground between the two arduino's, or should I use an optocoupler on the signal wire and keep the grounds isolated?

>>1072492
No need for opto-coupling.

File: Dream IDE Reborn.jpg (432KB, 2048x1152px)

432KB, 2048x1152px

What should I do if the datasheet for a chip I'm interested in is behind a NDA? Specifically, I'm talking about the Marvell 88SA8052.
I'm wanting to make something like the Dream IDE Reborn (see pic). It's a relatively clean and easy way of adding a hard drive to a Dreamcast as a slave on the IDE/PATA interface so that games and other software can be run from it by various means. But I want to make a version that has a PATA/SATA bridge on board so I'll have a smaller connector to install and just generally look nicer as a finished project.
I'm looking at the 88SA8052 because it seems like people have had the best luck with adapter boards based around it, but I'm willing to look at other chips if their shit isn't behind a NDA.

>>1072494
Ok thanks, I will try without.

File: iS2WacD.jpg (3MB, 4032x3024px)

3MB, 4032x3024px

I'm not sure if this goes here, but here goes

Is this a decent connection? I'm practicing soldering on a bunch of scrap wire. This is my second one and I want to catch any bad habits early

>>1072527

Its good enough. But here's a better one.

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

>>1072527
Soldering looks good. Hot enough for good solder flow, but cool enough to not melt the insulation.
Wire stripping not so hot. Shouldn't have broken strands.
I'd coat in solder to the end if the strands.

An easier/faster way to join wires: (or at least the way I do it)
-Strip short, about 1/4"
-twist the strands to get it tight
-pre tin both wires
-fix the iron, hold a wire in each hand, put the wires together how you want then, then touch to iron to melt solder. Lift off the iron and hold for a sec to let the solder solidify.
-clip anything sticking out with wire cutters.

>>1072527
>some stands of wire got broken while stripping
>loose ends of stranded wire not soldered on right side. They could fray.

6.5/10 not bad not great.

>>1071649
Eh, I think you might see a small current for dashboard LEDs, computers, stuff like that. But then again, I don't know much about tractors.

>>1072539
>>1072540
The first wire stripped properly, with no broken strands, but half of it was cold because I was still figuring out how you heat the wire up properly

What do you mean by pre tinning? Fixing the iron is just clipping it in place I assume

>>1072536
I'll check that out thanks

>>1072546
>pre-tin
coat both wires (seperately) with solder

>>1072549
Okay, thanks for the help

Is /ohm/ normally this slow? I'm new to the gen and am wondering if it's usually like this.

>>1071691

McMaster Carr has a great selection, though I've never used them personally on account of being on the wrong continent.

>>1072435
Are there any other prerequisite classes/things to know that I should get into like calc? I have been told to get into electronics engineering and embedded systems, one even said mechatronics.

>>1072451
i didn't realise that one could get modular rotary switches anymore. awesome! skip the FETs then

>>1072492
10m between them would make a hell of a ground loop. this would probably be fine for digital signals, but watch out for noise if you want to do anything really fast or in the analog domain.

>>1072632
yes

>>1072549
this. then you just touch the wires together and heat until solder flows and you're done. or add a bit more solder if the joint looks really dry

File: Western-union-straight-joint.jpg (16KB, 350x218px) Image search: [Google]

16KB, 350x218px

>>1072743

this kind of thing is fine for hobbyists, but in a professional setting it's completely unacceptable. you have to make the joint mechanically secure by twisting the wires in one of several ways before soldering. or you get fired. looks like OP was going for a Western Union joint, and got it about 50% right.

does the voltage rating on an ESC matter? some of the ones im seeing dont have a range listed

>>1072799
It matters. Everything has a maximum voltage.

>>1072761
but we're hobbyists here, not professionals seeking professional advice. so...

>buy some jumper cables on sparkfun because there literally are not any non-shitty ones available on amazon, and I need them to complete a project
>costs more to ship them than it does to buy the damned things
I can't even just go out and buy the fuckers. Somehow, there isn't a hobby electronics store in my entire metropolitan area

>>1072804

...so this is a perfect risk-free opportunity to learn you what a lineman's splice looks like and why it's good? There's a difference between 'this is the defacto standard for splicing wires because it works and requires no tools' and 'tin-plate your exposed strands or go home'.

>>1072527

Could use more solder, ideally the entire exposed wire should be coated.

Try practicing splices with 22AWG single-core wire (breadboard hookup wire), it's a lot easier to work with. For thick multistranded wire like what's shown in your picture it might be easier to work with if you do a rat-tail splice, but if you're doing a lot of them look into twist-on wire connectors/wire nuts or similar.

Greetings /ohm/.

I am planning on using a 750W 12v DC server supply to run some kit. It can deliver up to 58A, which is groovy and exciting.

I want to derive a 5v rail for a subcircuit that will peak out at about 5A during startup. I was planning on using a 6A or thereabouts regulator and having at it, but something seems.. dangerous, about hanging such a paltry regulator off a 58A pipe. The worst that could happen is a wrecked regulator right? Should I put a fuse upstream of it or something?

Does anyone know how to convert low voltage DC to high voltage AC that can handle 1600W 240V 15A for an air conditioner. I am just trying to get my power bill down and so far i can get "free" power but its low DC voltage. I cant think of any ways to get pure sine wave FROM DC to AC at the rated watts...

Modified inverters are shit and Pure inverters way too costly for a money saving idea.

>>1072811
and that's nice. but that's all it is

Anyone know where I'd find the tranmission power regulations in the UK for the 2.4 GHz spectrum? I've got one source telling me that it's +10 dBm because of the EU regulations but I've got another saying it can be up to +20 dBm in the UK. Any ideas would be greatly appreciated.

>>1072812
There's no point in adding a fuse. If you try to draw too much current from the regulator, it will fail faster than a fuse will.

The current capability of the PSU is irrelevant; there's no difference between running a 5A regulator from a 6A PSU or from a 1000A PSU.

>>1072814
> Does anyone know how to convert low voltage DC to high voltage AC
Inverter.

> Modified inverters are shit and Pure inverters way too costly for a money saving idea.
If there was something which did the same job as an inverter but was cheaper, no-one would use inverters.

>>1072830
that sucks. thanks for replying though.

>>1072814

i think your concerns are over-rated. i'm 99% certain an AC unit will be perfectly fine on a modified sine-wave. even a squarish wave. reason is the motor, being a heavy inductive load will tend to smooth out any irregularities. as for the electronic controls, they run from DC (rectified AC) so it wont make any difference.

File: 20161019_133457.jpg (1MB, 2048x1152px) Image search: [Google]

1MB, 2048x1152px

Hi /ohm/. I just finished making a bias probe for servicing tube guitar amps. It works with any octal with the cathode on pin 8.

File: 20161019_133616.jpg (1MB, 2048x1152px)

1MB, 2048x1152px

>>1072939
I used the 1ohm resistor method.
>set meter to mV
>reading is plate current in mA
Definitely safer than the shunt method

File: file.png (2MB, 1800x1350px) Image search: [Google]

2MB, 1800x1350px

Hello fellows,

i hope I ask in the right in the right thread:
Can someone recommend me a decent brand/distributor of switching power supplies in europe ? Everything I find is this MeanWell, but I need alternatives.

Thanks!

It's been about a day since I posted >>1072503
It looks like they don't have a way for hobbyists to sign the NDA, so am I fucked?

File: microsw_newkrnjwke.jpg (46KB, 465x423px) Image search: [Google]

46KB, 465x423px

I'm looking on aliexpress and trying to find crimp-on spade connectors that fit onto common microswitches.

(microswitches like pic related, that have spade connectors)

I have a few different brands of switches on hand, and all the male spade contacts that I measured with a vernier caliper--all measure about 3.34mm wide.

In China-land there doesn't seem to be any crimp-on female connectors to fit that width. There is 2.8 mm ones, and 4.8 mm ones. And wider ones, but none around 3.5mm...

There was some advertised as 3.5mm, but the picture showed bullet connectors, not spades. And other pages selling small bullet connectors said they were 3.5mm too.

I am just going to try using the 4.6 ones, they should work. But I still keep wondering what I'm missing here?

>>1072939
>>1072941
cool.
we need more tubebros around here.

I think you'll find /ohm/ isn't very interested in tube amps.
my last few posts hasn't had much of a response.
>>1070730

>>1072941
why the huge resistors thought?
what toleranse are these?

at one ohm, you need one full Ampere to dissipate only 1W.
those look like 5W too me.

New solderer again,

My station has adjustable wattage. Should I just be setting that to the lowest point, then gradually bringing it up to the point where I can melt solder, and then leaving it there until I change tips?

>>1072994
Pretty much, they probably won't give you the time of day.

>>1072965
what's wrong with MeanWell? they are pretty decent.

>>1073032
many folks will just solder wire to the contacts and then put shrink wrap over the joint. then you are free to connectorize as you see fit.

>>1072811

>>1072761
how do solder to something that has a loop like a potentiometer? ive always just stuck it through the hole and shoved solder on top

>>1073126
that's what I usually do and it's never failed. when I desolder commercial products they usually hook the wire around the loop and solder, but for a hobbyist that just makes it harder to undo if you want to change it later

>>1073075
does it have a temperature display or?

>>1073188
https://www.amazon.com/gp/product/B000AS28UC
Just wattage. It's worked well so far on wires using the '4' setting (not sure what wattage that's supposed to translate to), but I don't know if that's too high for PCBs

>>1073195
Ah okay. Well hard to say without temp. with temps, i would've said 300 - 350 celsius. i have a chinese one that has temp readout but it seems to work pretty good so far. reasonably accurate.. DMM temp probe said it was 345 when it was set for 350 so pretty satisfied.
but no you definitely want it hotter than just hot enough to melt solder. does it melt solder at the lowest setting? sounds like no based on what you're saying

>>1073198
I guess that would depend on the size of the tip I would be using. Definitely not enough to melt solder when using the giant flatheaded tip it comes with, but I also bought a smaller one that I'm not sure about

>>1073120
just have to consider some other possibilities

>>1071691
Ebay has the best prices for wire since it's surplus or old, but you have to know what you're looking for. New wire is pricey.

File: VJtBmGI.jpg (852KB, 3024x4032px)

852KB, 3024x4032px

>>1072539
Tried this while I was practicing pre-tinning and it turned out well (at least I think)

>>1072812
The point that how many amps the PSU can deliver is irrelevant as long as you don't short circuit anything was already made.

Your PSU can most likely not deliver any meaningful power over 5V.

>>1073126
I dont always do this, but a more 'correct' way is to pre-tin the wire and bend it into a hook with small pliers.
hook it onto the loop and squeeze it so stays in place, then solder it.

gives a good solid joint, and can make the soldering easier as you dont need to hold the wire while soldering.

hook or not thought, pre-tinning is always worth it.

>>1073242

the wires should be exposed 1/4-inch at most, and they should overlap completely, instead of just having the tips touch. it shouldnt look like too homos touching dick tips.

the solder is supposed to go all the way inside the insulation, coz when wires break, they most often break at the point where the tinning ends. by having that point INSIDE the insulation, it adds strength.

>>1073123
>many folks will just solder wire to the contacts and then put shrink wrap over the joint. then you are free to connectorize as you see fit.
Yea but I don't want the bother of adding another connector tho, I want to use the ones already there.
And I wanted to get them from China-land and not Digi-Key, because the China-land ones cost 2 cents and the Digi-Key ones cost at least several times as much.

I'm finding lots of other people online asking for these connectors for the same switches, usually for limit switches for 3D printers.
They usually get told to use the .110" size spade connectors, which is the 2.8mm ones.
I ordered some of those anyway so we shall see how that goes.
Of the couple dozen places I looked, there does not seem to be an intermediate size between 2.8mm and 4.8mm

>>1073273
i agree.

and DO NOT use shitty heat-shrink. use 3:1 or 2:1 ratio so it gets nice and tight and make sure the heat-shrink tube covers an inch of insulation to prevent the wire kinking since soldering makes the wire hard and brittle.

>>1072647
You basically just need to be good at math. Calc is one thing, but if you aren't super good at algebra that should be your first priority.

If you like tinkering more than theoretical work you should maybe look into an associate's degree or becoming a technician of some kind.

File: Capture.png (28KB, 786x550px) Image search: [Google]

28KB, 786x550px

Hey /ohm/,

Can someone help me?
I've got a relay module that gets stuck sometimes.
I used to power it with an Arduino, thought that it didn't use enough power. So then I used some 5v (I think) 200mA external power source instead, and only the Arduino to switch.

But now the same problems occurs, it sometimes get stuck. When I tick on the relay I can hear it click and then it will work again for a while. Is it remanent magnetism? How do I fix this?

Pic is a diagram of the module.

Thanks.

>>1073273
I had so much exposed because I was practicing tinning and just decided to try that I'll try this again properly though

>>1073326
>Is it remanent magnetism?
No. Either your relay driver leaks enough current to keep the relays closed (not very likely) or your Arduino crashes (much more likely).
You can check the first with a multimeter: there should be practically no voltage across the relay coil when they're not energized. If there is, you can add resistors from Q1/Q2 bases to ground to sink away opto's leakage current.
Check the latter by adding some suitable debug shit (print, led blink, whatever). Fixing it is not that straightforward, but you can start by taking the relays farther from Arduino and making sure the grounds are separate.
Well, or maybe your software is just buggered.

>>1073330
I will check both tomorrow, I'm pretty sure the code is ok tho. Will check to be sure. The relay module has a led to check if your turning it on or off. Until now I used the led to check if I was powering the relay.

The led would turn on, but the relay would not switch. only when I help it physically (ticking) it would switch.

thx for taking the time to answer.

>>1073341

you're not being super clear. in the first post, you implied the relay was getting stuck in the on position, but in the second, you're saying it's not activating until you hit it. i've seen this before, and it was caused by the relay just being a shitty no-name relay. i had to shake the fridge to get the compressor going. i replaced it with a proper relay and it's been doing fine for 4 years.

another possibility is you're applying too low a voltage. if it's rated for 5V, why not try 7V to see if that doesnt make it perk up.

File: 1410795789269.gif (873KB, 325x203px) Image search: [Google]

873KB, 325x203px

>>1072677
Yeah, I was worried about that and about the wire picking up noise and false triggering my circuit.

It's just digital "on-off" signal and I am holding the pulse high on the transmitter for 0.6 seconds and looking for a minimum pulse length of 0.5s on the receiver. I'm doing that to eliminate false triggers.

In case those delays didn't make it clear: it doesn't need to be very fast so I think I will be good. I'll wire it up and test this Saturday and let /diy/ know if it worked.

>>1073341
> The led would turn on, but the relay would not switch. only when I help it physically (ticking) it would switch.
It sounds like the coil isn't getting enough. The armature the farther it is from the electromagnet, the more voltage => current => flux it takes to move it.

So if there isn't quite enough flux, it won't be able to move the armature from its rest position, but once the armature moves a bit closer, then the flux will be enough to move it the rest of the way and hold it there.

what wire strippers does /ohm/ use?
my klein ones are starting to rip off strands every time I use them, and I figure it's time to lay then down to rest

>>1073468
i've got a pair of those blue DataShark ones at work that have been bretty gud. the yellow ones for bigger wire gauges seem shitty though.

File: Screen Shot 2016-10-20 at 9.02.10 PM.png (124KB, 1224x800px) Image search: [Google]

124KB, 1224x800px

Hey, do any of you guys have experience making adders in circuitLab?

I've been trying to build the schematic for a circuit I wanna build to teach middle school students about logic level addition, but I cant seem to get it to properly simulate turning the LEDs on and off when the switches are depressed.

https://www.circuitlab.com/circuit/5759r9/full-adder/

Thanks in advance.

>>1073576
I was expecting the voltage across the diodes when the switches are open to be zero, but I must be doing something wrong, as there is voltage drop and current flowing through both LEDs.

File: file.png (30KB, 1841x875px) Image search: [Google]

30KB, 1841x875px

Hey guys,
I've got an issue with DC bias in my circuit. The blocking cap for the microphone lets about 15mV DC through, and the op-amps amplify that at every stage. The supply rails are also around 15mV apart from what they should be, and I bet that doesn't help. Any suggestions on what to do?

The DC bias at the peak detect is basically always enough to have the diode be on, which I don't want. I'm thinking of throwing a few diodes in series with that one, but I'm hoping there's a better way. Any ideas?

File: 1471128996267.jpg (256KB, 1280x960px)

256KB, 1280x960px

I need some guidance on study progression.

I am very interested in learning everything there is to electricity, circuits and electronics.

Completed a couple months ago ''Getting Started in Electronics Forrest Mims III'' and also have the ''MAKE...'' one.

I followed that by learning the techniques required to solve for voltages and currents and some theorems like superposition, thevenin and norton equivalents.

On top of that, I also read some chapters from ''Electronic Principles'', mainly those related to diodes, transistors and amplifiers.

But, to be quite honest, I just feel confused here, as If I wasn't following a clear progression, or if my studying was all out of place and inefficient.

I was hoping that some of the people here that already went through the process (perhaps you self studied or have a major in it) would shed some light on this?

>>1073600
>I am very interested in learning everything there is to electricity, circuits and electronics.
Should probably go to uni for that

>>1073585
Well, you could start by moving C5 to the output of the unity-gain buffer.

Also, a typical Sallen-Key filter would use 21.5k and 33nF (i.e. 1000x more R and less C).

>>1073628
Alright, thanks. I'll try moving C5 tomorrow and update with the results. I'm going to try to avoid another digikey order if I can due to the shipping. What would changing out the resistors and caps in the filter do differently?

>>1073576
>>1073578
Try adding pulldowns to the gate inputs, dunno if they have them by default.

>>1073600
For me personally I learn best when I have a project to motivate the learning and such. For example, you might go:
1. How do I build an audio amp from a chip? (Don't learn a whole lot)
2. How do I build an audio amp from transistors? (Learn a lot)
3. How do I reduce the nonlinearity of an audio amp built with transistors? (Learn even more)

>>1071664

You mean a BLDC motor presumably, brushed DC is simply controlled with PWM.

For BLDC usually the controller usually just switches state based on the hall effect sensor states. All binary, no prediction or estimation necessary.

Sensorless BLDC seems like a huge PITA to me.

File: adder.gif (42KB, 828x486px)

42KB, 828x486px

>>1073576

you need pull down resistors on the switches, and current limiting resistors on the LEDs.

also please dont do stupid shit like put grounds pointing upwards. schematics are like a language, and you show your ignorance by speaking it ʍiʇɥ ɐ ɔɥiuʞ ɐɔɔǝuʇ.

File: ss+(2016-10-21+at+05.38.22).png (4KB, 236x235px) Image search: [Google]

4KB, 236x235px

Hey ohm, just a quick question, this circuit works fine as a LV cutoff, is their a similarly simple circuit using zeners that works as a HV cutoff?

>>1073600
>I am very interested in learning everything there is to electricity, circuits and electronics.
That won't happen. The field of study is massive and when you count in everything else you need to know as prerequisites, there's even more stuff to study.
In practice, you need to have some focus at some point, but if you're currently studying the basics (using Mims' book etc.), there's no hurry for that.

University curriculums are more or less public. You could check what they consider important. On the other hand, their choices can be very boring/demotivating for self-study, as it takes a while to get to the point where you can actually design something interesting. They also teach plenty of stuff which is good to know (at least in principle), but has no real value in practice.
An alternative would be what >>1073643 said. It is certainly more interesting way, but the drawback is that your knowledge will have pretty obvious holes in the basics, which hinders your learning of new stuff.

>>1073639
> What would changing out the resistors and caps in the filter do differently?
Increase the input impedance of the second stage. As it stands, it's loading the first stage quite heavily.

It also reduces the size and cost. Capacitors get larger and more expensive as the value increases, resistors don't. So for a given RC product, you want to maximise R and minimise C. The limiting factor is the op-amp's input current, which should be much smaller than the current in the passive components in order for the "ideal op-amp" approximation to hold.

>>1073681
>HV cutoff with zener
Nobody does that ever, because they just use the zener as a voltage regulator instead

What are you guys majoring in/what do you have a degree in?

Any good books on tape electronics? Got some ideas but I have never worked with tape heads before, wanna learn about em.

>>1073761
Biology

>>1073761
EE+Physics

Mustard race

>>1073761
Degree in EE

File: internet blasterimg_1384.jpg (405KB, 1536x1024px) Image search: [Google]

405KB, 1536x1024px

4G antenna.

File: 1472763659653.gif (3MB, 360x249px) Image search: [Google]

3MB, 360x249px

Do IPS panels have refresh rate limitations? Or even higher resolution TN panels?

I need to know what's stopping me from putting a 144Hz 2560x1600 controller board on a 2560x1600 ips panel, all matching connectors.

>>1073585
>that schematic
dear god, please fix the way this is drawn

>>1073842
ultimately you are limited by the mobility of the liquid crystals and their driver electrodes. the LCs can't change orientation instantly, and the electrodes have some capacitance that will slow them down. even if you manage to drive them faster than 60Hz or so with custom firmware or something, you'll need to re-invent the anti-ghosting trick for your given refresh rate

File: FJPYZX9GTGD4NE0.MEDIUM.gif (4KB, 475x200px) Image search: [Google]

4KB, 475x200px

Hi guys,

For my circuits class, I want to build a piano/music for the final project. I want a circuit that can play chords/multiple notes at once, obviously the circuit in pic related does not do that. How would I go about producing multiple sine waves at once? I don't want to have a 555 timer for each note. Would it be as easy as summing all the sine waves produced by the notes with an op amp?

Forgive my ignorance, I am inexperienced.

>>1073941
>Would it be as easy as summing all the sine waves produced by the notes with an op amp?

nope, if you want chords then you have to have multiple oscillators running at the same time, each keyed by it's own switch. then you can sum them. a dude on DIY built such a thing some time ago, using like 30 transistor oscillators, each one with an LDR so it could change the pitch.

i dont really know anything about electronics and i need something to charge this http://www.hobbyking.com/hobbyking/store/__90420__Turnigy_Graphene_950mAh_1S_65C_LiPo_Pack_w_JST_SYP_2P.html
without thinking i bought this http://www.hobbyking.com/hobbyking/store/__49450__Turnigy_Neutron_80W_DC_Touch_Screen_Charger.html
which needs a power supply. Since i only need to charge small lipo batteries would it be better and cheaper just to get something like this http://www.hobbyking.com/hobbyking/store/__58285__IMAX_B6AC_V2_Professional_Balance_Charger_Discharger.html instead?

>>1073957

if you have an old PC power supply around, you can pull 12V @ 20+A out of it to power the cheaper charger. otherwise, go for the more expensive noob-friendly one.

File: IC 4060 Pin Out.png (62KB, 718x994px) Image search: [Google]

62KB, 718x994px

>>1073941
>>1073946
You could use a binary counter and put a switch on each of the outputs. 4040 clocked with a 555 will work. 4060 will do it all in one chip because it's got a built in oscillator.

Output will be square waves not sin, but you could make them curvier by charging a capacitor through a resistor with each output.

>>1073959
ok ill just try to find some old power supplies
thanks

>>1073842

The electronics to control the column and line drivers is integrated into the panel and are generally refresh rate limited yes.

File: DI66Fig01.gif (5KB, 493x298px) Image search: [Google]

5KB, 493x298px

>>1073681

Google for overvoltage protection. Here, I did it for ya :

https://www.maximintegrated.com/en/app-notes/index.mvp/id/760

>>1073765
"magnetic recording"

>>1073941
if you don't mind square waves, you could have 12 "master" oscillators and then use flip flops to divide down the pitch for the other octaves of the keyboard.

to do a polyphonic keyboard, you can have a single switch under each key, then have a microcontroller repeatedly scan the keyboard for closed switches at a fast rate. you'll probably need to multiplex the keyboard, but you should have no problem scanning it fast enough.

if you're tricky, you may be able to get away with a single master oscillator.

>>1073966
Could you elaborate more on how that works?

File: cou3.gif (15KB, 568x437px) Image search: [Google]

15KB, 568x437px

>>1074005
A binary counter counts up in binary. The value is incremented with each clock pulse and there's an output pin for each bit. You'll get square waves of several different frequencies without multiple oscillators.

Binary counter chips are super common, but you can also build one with a set of D-flip flops if you don't have one.

>>1070254
I do think Julian Ilett deservers a spot in the youtube list

File: notes.gif (21KB, 453x666px)

21KB, 453x666px

>>1074018

which is not particularly useful coz musical notes dont go by 1, 2, 4, 8, 16... ratios

>>1074027
There's no right and wrong way to make music.

>>1074035

doing it wrong: Beatles - She Loves You - https://www.youtube.com/watch?v=T0YifXhm-Zc

doing it right: Boyd Rice - Total War - https://www.youtube.com/watch?v=E-YoY60BRsc

Is it worth to learn how to use an arduino as a beginner?

File: 50240.gif (35KB, 1373x1124px)

35KB, 1373x1124px

>>1074035
There's my way and then there are plenty of ways to do it wrong.

>>1074027
Well, after you get the top octave frequencies right, you can generate the lower octaves with binary dividers. This was a popular method in electronic organs decades ago.

Hey guys, especially EEs

In the situation where you are using a BJT like a 2n3055 to pass extra current and also voltage adjustment.. you can't replace that BJT with a mosfet right? or can you?

File: tps79301.png (35KB, 717x434px)

35KB, 717x434px

>>1074065
I'm not too sure what you're actually asking, but it is entirely possible to build linear regulators using mosfets as a pass element. Both N and P mosfets can be used.

>>1074067
yes that's what i'm asking, thanks. Wasn't very good at communicating that sorry.

Most schematics i've found online for linear supplies seem to use a BJT.. typically 2955 or 3055 and haven't seen any that use a mosfet, so it was curious.. since mosfets tend to have better resistance properties and don't have thermal runaway. Figured they'd show a mosfet in that case.

>>1074068
Most schematics you can find are old or based on old designs. The fact that they're using 2N3055s says, in itself, that they're pretty old.
That said, if you don't need low dropout voltage + low quiescent power consumption, bipolar transistors tend to be more convenient in that application. As in, you can get better performance easier.

>don't have thermal runaway
The modern mosfets optimized for switching are not immune to that.

>>1074068

if your question is ''can i take a 2N3055 design and just replace it with a FET" the answer is no. the circuit would have to be redesigned to accommodate the differing characteristics.

>>1074079
I didn't mean straight up replacement. What sort of modifications would be required?

>>1074075
>thermal runaway
gotcha

>better performance easier
how do you mean?

>>1074080
Let's say you have an emitter or source follower power stage and that the output current changes from nearly zero to the maximum value. The mosfet version will typically require the gate voltage to change several volts, while the bjt version requires one tenth or so of that.
This difference must be covered by the control circuit, requiring more gain, higher slew rate and more bandwidth. The non-linearity is also more visible due to the higher control voltage swing, requiring more care to limit the over/undershoot. And the gate capacitance can cause stability problems, depending on design.

>>1074082
Ah I think I understand thanks

File: ti.png (142KB, 838x843px) Image search: [Google]

142KB, 838x843px

TI currently offers free shipping & no minimum order quantities on all orders. This offer ends tomorrow. Which interesting parts would you recommend to buy? Currently thinking about getting these items, pic related.

>>1074087
I'm guessing this only applies to addresses in the continental united states?

File: Clipboard01.png (2KB, 386x365px) Image search: [Google]

2KB, 386x365px

>>1074080
The main difference is that a BJT emitter-follower need only drop ~0.7V, while a FET source-follower will drop ~5V. That makes the classic pic-related design far less desirable.

So a FET-based circuit will typically use a p-channel FET and a more complex feedback loop.

>>1073941
> How would I go about producing multiple sine waves at once?
Multiple oscillators.

You can either have as many oscillators as you can have notes in a chord (so 3 oscillators lets you play 3-note chords), or you can have one oscillator for each possible note.

The latter requires more oscillators but they're all fixed-frequency and the control is simple: each key enables a specific oscillator. The former requires fewer oscillators but they're variable frequency and you have to figure out how to make the keys control the oscillators.

Nowadays, the only sane way to implement this would be to use a microcontroller. But then it's essentially a software project rather than an electronics project.

tl;dr: polyphony is hard.

>>1070440
thanks

>>1070438
welcome to /diy/

>>1073761
Chemistry

>>1074133
Don't. Unless you want to go full phd or already have something you love lined up, just don't. I graduated and took 6 months to find a job. I've been working in a factory running the same fucking boring ass HPLC experiment that I learned how to to sophmore year every single day. Shoulda been a fucking machinist or an EE, or shoulda stayed in school.

File: schmitt-trigger-osc.png (10KB, 464x171px) Image search: [Google]

10KB, 464x171px

Hi /diy/

I need some simple square-wave oscillators in my project. I was thinking to use a single 74HC14 chip which contains 6 Schmitt-trigger NOT gates.

Is there any way to add an on/off control to this kind of Schmitt-trigger oscillator, so that when the control signal would be '0', the output would also stay at '0'?

>>1074152
A 74LS132 is a quad Schmitt-trigger NAND gate, so if one input is 0 the output will stay at 1, otherwise it will oscillate. You can use a second nand gate to invert that, so you'd get 2x oscillators per package. Similarly, the '232 is a quad Schmitt-trigger NOR gate, so if one input is 1 the output will stay at 0, otherwise it will oscillate.

But you can't get a single gate which meets your requirement of input 0 => output zero (which amounts to having one inverting input and one non-inverting input).

>>1074167
Thanks four your tip. I tried it on breadboard with a HEF4093 Schmitt-trigger NAND chip and managed to get it work well.

>>1074167
>But you can't get a single gate which meets your requirement
Schmitt input XOR gate would work and 74AHC86 has 4 of them, but I haven't seen any manufacturer to actually specify the switching thresholds, so there are no guarantees of the oscillation frequency.

>>1074194
Took me a while to realize that's not true. The output can stop in either state.

>>1074194
> Schmitt input XOR gate would work
Nope.

If the input is 1, the output will oscillate. If it's a 0, the output will remain in its current state, which could be either 0 or 1.

>>1074194
>I haven't seen any manufacturer to actually specify the switching thresholds

All schmitts are supposed to have thresholds of 2/3 VCC, 1/3 VCC.

>>1074194
> Schmitt input XOR gate would work and 74AHC86 has 4 of them
74xx86 isn't Schmitt.

> but I haven't seen any manufacturer to actually specify the switching thresholds
I've never seen a data sheet for a logic part which didn't specify threshold voltages.

>>1074232
> All schmitts are supposed to have thresholds of 2/3 VCC, 1/3 VCC.
This is incorrect. Schmitt-trigger just means that it has non-negligible hysteresis (i.e. the low-to-high threshold is higher than the high-to-low threshold, and the difference is enough to be worth actually stating it as a feature).

E.g. Motorola's 74LS132 data sheet gives the L->H threshold as 1.5-2.0V and the H->L threshold 0.6-1.1V, both specified at Vcc=5.0V.

The 555 uses 1/3 and 2/3 Vcc, using a voltage divider and two comparators to achieve that. But the 555 is specifically designed as a timer, so predicable thresholds are important. Whereas Schmitt-trigger logic inputs are mainly just so that you don't get bounce on signals with low slew rates.

>>1074242
>74xx86 isn't Schmitt.
Not generally, but for example NXP's 74AHC86 claims to have Schmitt trigger inputs. Same with Diodes' version. Neither specifies the Schmitt trigger thresholds (like, for example, 74HC14 datasheet does) and go with normal 74HC thresholds.

>>1074060
that 50240 chip doesn't exist anymore. I only want to simulate one octave. Is there a way to do it with a single 555 and binary counters?

>>1074274
> I only want to simulate one octave. Is there a way to do it with a single 555 and binary counters?
Sure, provided that you're willing to use one counter per semitone (i.e. 12 of them for a chromatic scale). Also, you'll need to use 16-bit counters if you want accurate intervals.

Note that generating the 12 semitones is the hard part; generating additional octaves is easier, because you're dividing the frequency by 2.

Also, digital logic will give you square waves. You'll need to add low-pass filters to remove the harmonics if you want sine waves.

File: Clipboard.png (57KB, 819x538px) Image search: [Google]

57KB, 819x538px

Pic related, I'm doing some experimentation with a monostable multivibrator IC MC14538B.
The duration of the output pulse of the monostable oscillators in the picture is set with Rx and Cx. Tw = Rx * Cx

Would it be possible to control the output pulse width by using 3 parallel-connected resistor + PNP transistor strings in place of Rx and driving the transistors' bases with three pushbuttons? Ideally I'd like to be able to select eight (2^3) different Rx resistance values (and pulse widths) by different pushbutton combinations (0 not pressed, 1 pressed):
>pushbuttons: 000 -> pw ~0 %
>pushbuttons: 001 -> pw ~13 %
...
>pushbuttons: 110 -> pw ~88 %
>pushbuttons: 111 -> pw ~100 %

It would be easier if only one of the transistors would be active at a time, however that kind of control scheme would have just four different possibilities for output pulse width.

How should I select the resistor values for aforementioned use? Or would I need some additional decoder chip for this kind of use?

>>1074290
You can't get a linear relationship between a 3-bit code and the pulse width using 3 parallel resistors, because the pulse width is proportional to resistance but with parallel resistors you're summing conductance (the reciprocal of resistance).

So I wanna get into PCB etching. Do you guys have any sources for simple/fun schematics?

File: Screen Shot 2016-10-22 at 10.11.38 PM.png (167KB, 1728x866px) Image search: [Google]

167KB, 1728x866px

>>1073660
>>1073641
I wasn't using grounds at all originally, I was pretty frustrated at this point and just throwing shit in to see if it would work.

Thanks all for the help though, I'll try it out.

They dont really teach us anything in ECE until we get to the 300 level classes, so I've been trying to learn on my own.


Gonna try an 8-bit adder soon, wanna get a prototype running so I can teach some middle school students about basic logic.

>>1074342
http://bigclive.com/

He's a top lad. Has a good YouTube channel as well.

File: 20161022_225329.jpg (1MB, 2560x1536px) Image search: [Google]

1MB, 2560x1536px

Just finished this one and wanted to show it

>>1074372
post vid senpai

File: DSCN4513.jpg (544KB, 1600x1200px)

544KB, 1600x1200px

>>1074372
I feel like posting my progress too.

I finally finished putting together the microcontroller boards for my 88 square foot LED wall project. All that's left now is wiring them up to the LED panes.

>>1074372
They built bucketsloads of these as a group project at my uni. Are you studying in the Netherlands?

>>1074415
No, Argentina

>>1074368
>subscribe because interested in all the LED stuff he reviews
>get inbox message with video on how to make moonshine
>still interesting

>>1074431
He really is pretty fuckin amazing lmao

>>1074397
what are the electrets for?

As part of a local positioning system, I want to use ultrasound to measure distance between rover and base stations. They communicate via radio so only one-way ultrasound is necessary, with that I managed to get around 14m range out of the $1 rangefinders you can get on eBay. I want to improve that though, ideally get to 30+ meters, and I'm wondering what my options are to amplify the output
>add more of those cheap ultrasound transducers next to one another - apparently they have to be spaced accurately, also not sure what I can drive them with, would a regular amp work with a simple square wave?
>ultrasonic cleaners - they're cheap, powerful and readily available, but are they good for something like this? Does the transducer on the other end detect their sound the same as any other transducer's?

File: synchronous-detectors-facilitate-precision-FIG04.png (14KB, 947x381px) Image search: [Google]

14KB, 947x381px

>>1074449

Here's a schematic of of the HC-SR04.

http://uglyduck.ath.cx/ep/archive/2014/01/Making_a_better_HC_SR04_Echo_Locator.html

What I would try :

Mod the filter like he suggests, then remove R9 and tap the output of A2B and use two synchronous detectors like in pic with square waves at quadrature at the same frequency as the HC-SR04 uses (generated by your microcontroller). Check the frequency with a scope, don't just assume the datasheet value of 40 kHz is correct. This will give you a very low noise quadrature detector so you can use much lower thresholds. You can either ADC both outputs and do it in your microcontroller, or you can rectify and sum the outputs and do level detection to create a logic level trigger for the microcontroller.

That synchronous detector schematic is from Analog devices, but you can use other muxes and opamps. You don't absolutely have to use a bipolar power supply either, can halve and buffer the rail voltage for a virtual ground. Opamps used should have low offset voltage.

http://www.analog.com/en/analog-dialogue/articles/synchronous-detectors-facilitate-precision.html

>>1074481

I meant tap the output of U2B.

>>1074481
Thanks for the suggestions, I'm already tapping into the receiver board and reading the signal with the arduino, though it's simply via direct level detection, but it still seems to work better than the onboard circuitry. The sensor I'm currently using is a newer variant with a different layout, but I also have some SR04's around. Sadly I only have a really shitty oscilloscope so I can't measure the 40kHz with any decent accuracy.

Wouldn't it be easier to boost the output though? I (eventually) want this to be something that anyone can easily put together without too much custom circuitry, though I guess some customization will be inevitable to reach the desired range. Also, power consumption on the rover isn't a big concern, hence I'm thinking that it might go further if I output sound with ~4 times as much power, should raise the range to at least 40m. If I could just hook up 4 of the transducers of these boards to an arduino through an H-bridge, it'd be a rather cheap solution.

>>1074489

Wavelength is 8 mm, so no you can't really wire them in parallel (even at precise distances from each other you get interference patterns).

>>1074449
>ultrasonic cleaners - they're cheap, powerful and readily available, but are they good for something like this?

all the ones i've seen are nothing but vibrator motors. there's no actual piezos, and the frequencies are all over the place.

>>1074511
Yeah, scammers sell that shit, but real ultrasonic cleaners aren't exactly difficult to find either.

>>1074489

This looks interesting as a cheap 40 kHz transducer. Nice low impedance so you don't need very high voltages. I'd say drive it at 20V p2p and put a 100 Ohm 1W resistor in series to dampen potential resonance.

http://www.ebay.com/itm/1x40khz-35W-Ultrasonic-Piezoelectric-Clean-Transducer-Plate-Ceramic-Sheet-HighQ-/182210510472

>>1074420
Which province, senpai?

>>1074505
I see, I guess it's down to getting one larger transducer then
>>1074511
I've been looking at ones on eBay, just like what >>1074616 posted and they're certainly not made with motors; I guess I'll order one of those and see where it goes. I found some [probably] not-Chinese retailers selling ultrasound equipment specifically for ranging that was operating at higher power and usable range, but they charge $60+ for them so it's pretty much out of question.

>>1074448
Their connected to preamps and an MSGEQ7 chip. It's for making LED patterns that respond to sound.

>>1070254
Is there an open source, cross platform (Windows and Linux) circuit sim with a nice GUI?

I've got tina spice on my windows machine, but I'd like something that supports both windows and linux either with a common file type or by running on both machines.

>>1074669
gEDA isn't a "circuit sim", but includes a schematic editor and an interface to SPICE.

>>1072814
>so far i can get "free" power but its low DC voltage

Tell us where you're stealing it from, jamal.

>>1072824

EU regulations don't mean shit, Ofcom is the UK's RF regulator and thus the only people that matter. I think they only report power limits in watts though.

>>1072503
>What should I do if the datasheet for a chip I'm interested in is behind a NDA?

If you're only building one or two, just steal it, or hoover up enough ancillary information that you don't need to.

For the Marvell 88SA8052, this page has a pinout (http://www.psdevwiki.com/ps3/88SA8040-TBC1) and example schematics, and I'm sure you can find other schematics for adapter boards if you look hard enough. Looks like configuration is just pin strapping on cnfg1-3.

Could someone please give me some information on wire wrapping and wire wrapping tools? I've recently been getting into working on old audio gear, and I've noticed that it is used often to connect stuff.

I've already watched a few tutorials on how it is done using hand tools, but I want to know more about the tools themselves before I actually go buy one.

From all the hand tools that I have looked at, it seems that they are limited to a specific gauge of wire. Is this correct, or am I wrong? If I am correct, are there any that can work with multiple gauges?

Also, does anyone have any recommendations for a hand wire wrapping tool?

File: Breadboard set up 4.jpg (354KB, 988x1334px)

354KB, 988x1334px

I'm a complete idiot, mock me as you desire but I have a question along the lines as >>1070429

I'm trying to turn pic related into a wearable-type thing that fits inside of a helmet and need to take it off of the bread board. If I went the soldering route, I would just need to solder all of the wires that are in a row together, right? For the button I would need to strip the wires connecting to it and solder them to the correct pins.
Am I missing anything? Forgive my ignorance, I just want lights to blink.

>>1073585
Update, for anyone interested:

I threw in another cap after the unity-gain buffer, and another after the sallen-key. This got the circuit to work intermittently; it would be good for a few minutes, then fuck up, and when I'd try to fix it by probing components with my scope, it would start working again. Eventually I realized that this was because of the cap after the sallen-key; it was accumulating a charge, and the scope probes let it drain slowly to 0V DC. I added a 200k resistor to ground to get the same effect and now it's working pretty well. If anyone's interested I'll put up a video of it in operation.

>>1073924
What would you do differently?

I've got a big boiler outside at my motel that we currently just turn on with a switch.
How do I make this into something that is automatic and not shit? It heats 10 discrete rooms.

>>1074800
Correct. Honestly, I don't know if I would solder anything to an arduino- you could fuck up the Arduino, and if you want to use it for something else, you'd have to remove the connection.

>>1074812
Well, nothing would be soldered to the Arduino board. I'm mostly talking about everything coming from the arduino that is going through the breadboard and to the LED matrix (The stuff in focus in the top picture).
The jumper wires seem to be stable enough for my liking, might throw some light adhesive just to ensure they stay connected while wearing it.
Thank you.

>>1074779

wire wrapping tools come essentially in 2 flavors: 30AWG for wiring chips, and 24AWG for wiring backplanes and making audio connections. you dont have to use precisely that size wire, but it has to be close. there are tools that will 'strip' the wire along it's length while you're wrapping. i dont trust those so i always strip manually. electric-powered tools exist if you're doing thousands of connections.

File: 90-deg IDE cable.jpg (102KB, 1000x1000px)

102KB, 1000x1000px

>>1074814

if you dont solder to the arduino, then those jumpers add a lot of height to your gadget. if you can find another right-angle connector, like the one for the LEDs, that would make it much slimmer. i suppose you can find some right-angle IDE cables from an old computer that fit. cut it and solder to the wires.

>>1074851
I just threw a bunch of those away.

>>1074851
The Arduino I have only has female ports though, I don't see how that would help? Again I'm a complete idiot.
I do appreciate the advice though. I'm planning on putting the wires+arduino into a plastic encasing, with holes drilled out for the power cables (and maybe the button if I extend it via a chain of male to female jumper wires).
All things disclosed, this is simply for a Halloween costume so I can stick a computer monitor on my head.

>>1074853
>The Arduino I have only has female ports though

ah, just noticed that. i was thinking it was like my Raspis and Odroids.

File: Arduino-Nano-und-1.8-Zoll-TFT-Display-mit-SPI-Schnittstelle.jpg (845KB, 1280x960px)

845KB, 1280x960px

>>1074800
why dont you use a TFT display?

File: PixelV2.jpg (31KB, 600x215px)

31KB, 600x215px

>>1074881
Display is going to behind a tinted screen. I picked the LED matrix because I figured the simple images would be the best seen through the black. Next year I plan on going all out once the "pixelboard for dummies" is available on Amazon instead of having to buy directly from the chinamen.

>>1074910
voltage controlled wet dry mixer I whipped up in falstad as a part of a reverb module I'm working on.

rewriting this post because 4chan thinks its spam for some reason.

noob EE student here, we've been learning about mosfets and opamp configuration in class for gain and such, but i was wondering how applicable is this?
For example, if i pump audio signals through some op amp filters and a summers, will i get comprehensible sound from a speaker?

>>1074989
Op-amps generally can't supply enough power to drive a speaker directly. But it's fairly common to build filters, pre-amps etc from op-amps.

>>1075017
What makes the speaker hard to drive? Is the voltage output of an opamp not enough or is something else the factor.
I've looked some things up online and they don't seem to mention this.
for example
http://www.instructables.com/id/DIY-Amplified-Speakers-for-your-MP3-player/?ALLSTEPS

>>1075138
> What makes the speaker hard to drive? Is the voltage output of an opamp not enough or is something else the factor.
It's more the current than the voltage. Typical maximum output current for an op-amp is around 10mA. At 10V, that would be 0.1W.

> I've looked some things up online and they don't seem to mention this.
> for example
The TDA2822M is a power amplifier.

File: received_1311277665570635_1.jpg (128KB, 960x720px) Image search: [Google]

128KB, 960x720px

>>1070254
What is this component called?

>>1075180
Dip switches

File: received_1311279588903776.jpg (21KB, 700x220px) Image search: [Google]

21KB, 700x220px

>>1075180
We're making a H-bridge using 555 timers using this diagram and some guy already made this thing and it shows the compinent encircled in the pic

>>1075183
Thanks man, this is a big help

>>1075184

if you follow the wires, you can tell the DIP switch isnt actually doing anything. both switches 1 and 2, for example, are only connected on one side. the opposite side goes nowhere, so the switch is irrelevant. it was probably for testing.

>>1075184

whattaminit, how much current can a 555 sink?

>>1074912
why not use a couple of VCAs for this task? you would get better performance.

File: Buttons and Wires.jpg (257KB, 990x1452px) Image search: [Google]

257KB, 990x1452px

This idiot (>>1074800) is back again.
Long story short, I managed to get the project off of the breadboard using a bunch of f/f jumper wires. This worked great because I actually wanted the button to be attached to a cord in the final design. I have two options:
1. Wrap all of the connections in electrical tape, hope none of the connections come undone via jostling/movement.
2. Hotglue the m/f connections. Run the "cable" through some straws/still eletape them all together.
Pic related is what I'm talking about. I would heatwrap/heatshrink if I had the money/tools at my disposal, but my family workshop is limited in the way of tools that I can use.

>>1075345
FFS. Solder the wires to your switch instead of this tape and glue nonsense.

File: 20161024_193305.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

I tried using a tda2030 as a driver for a reverb tank yesterday.
its not really fond of the lower gain settings and i had to struggle quite a bit to get rid of the oscillations.
but it works now.

I think I'll see how current is actually needed and maybe shunt it down so i can use something else.

what has the most current capability of the common opamps? TL07x ,lm324 , 4558, 5532 etc?

the tank is in the feedback of the opamp, making it current controlled opposed to voltage controlled.
this is to counteract the inductance reducing gain at higher frequencies.

File: test.jpg (1MB, 3264x1836px) Image search: [Google]

1MB, 3264x1836px

Can someone help out an electronics analphabet out for a second?

Saw this (https://youtu.be/AkL72XVtKew) and thought i'd build one myself. Never done anything related to electronics at all. Ordered all the cheap parts via ebay from china, filaments aren't here yet. Thought i'd test the rest, stitched it together a few minutes ago. Will pic related work to test the circuit (and is the circuit itself correct)? I basically did what i saw in the video, and added two 100K resistors and a blue LED instead of the 4 LED COB filaments (i think the filaments have a forward voltage of around 70V each).

Also is it possible to fuck up resistors by soldering to close to the body? The 1M Ohm discharge resistor was reading about 1M before i put it in, now my shitty meter says its somewhere in the ballpark of 800K.

File: IMG_20161025_124249291.jpg (1MB, 3264x1836px) Image search: [Google]

1MB, 3264x1836px

>>1075390
here's a close up of the rectifier/smoothing capacitor.

Is it just practise to solder shit like that together or is there some sort of trick? My joints always end up with way to much solder or won't stay together.

>>1075390
>possible to fuck up resistors by soldering to close to the body?
not really.

>1M Ohm discharge resistor was reading about 1M before i put it in, now my shitty meter says its somewhere in the ballpark of 800K.
its reading low because its in parallel with the two 10k's in series with the capacitor. disregard the reading, its fine.

didn't watch the youtube vid, post schematic if you want me to look at it.

>>1075391
I'd say you need to pay more attention to cleaning the leads then tin them then solder them together. Try and use a lead based solder and maybe try dropping the temp a liitle as I think your flux may be burning off too quickly.

Otherwise I like the way you've done it.

>>1075370
If you look at their datasheets, they tell you.

File: stitched.png (530KB, 929x1083px) Image search: [Google]

530KB, 929x1083px

>>1075393
i think pic related would be the diagram. Never drew one, so i hope i didn't fuck it up. Added the original diagram and what i think how i modified it.

Kind of unrelated, why is "falstads circuit simulator" showing me two bad connections for this?
Anyways, my main concern (apart from wrong connections): The rectifier should spit out about 350V smoothed DC (no idea how, the video says so). Will this work with a single LED and 220K worth or resistance? As i understood it 350V/220000=1.5mA , is that correct for this kind of situation or are there other things i need to consider? Won't the LED get damaged by basically connecting it to 350V or is using resistors to limit the current to something within the LEDs nominal operating range all that is needed to prevent damage? 350V on a 3V device somehow doesn't sound right to me.


>>1075395
didn't know that i need to clean the leads, will do so in the future. Can't reduce the temperature of my soldering iron, it's one of those cheap out of the wall versions. I should probably get some flux, don't have any yet.

Thanks for the tips and replys.

>>1075405
> The rectifier should spit out about 350V smoothed DC (no idea how, the video says so).
"240V" mains is 240 V RMS, 340 V peak. A rectifier and capacitor will charge to the peak voltage.

> Will this work with a single LED and 220K worth or resistance?
Yes. But if the LED ever fails, there will be 340V across it, which could result in combustion or a shock hazard.

>>1075414
thanks, i even understood that after googling what RMS means.

I don't intend to touch the thing, it's supposed to be inside a light fixture with a (glas) test tube over it when it's done. For testing it will be on a long wire far away from me hitting the switch. As long as it doesn'tset anything on fire in case of failure i should be fine, hopefully.

Thanks for the explaination.

>>1075405
>dont have any flux

Yea, there's ur problem right there.

>>1075429
thought the flux core of the solder would be enough

>>1075430
Sometimes it is, but I find that at higher temperatures having more flux is a must.

And since you dont have a temperature adjustable iron, you're gonna need a lot more flux.

File: 1464076108631.jpg (27KB, 491x463px) Image search: [Google]

27KB, 491x463px

Dear /ohm/,

I want to control the rounds per minute of a 12V DC motor. I stumbled upon PWM as a simple way to do this.
1. Does the motor need any particular properties to make this work?
2. Do i have to implement a flyback diode(because i essentially switch the thing on and off really fast)?

Yours sincerely,

Anon

P.S.: You're awesome.

>>1075444
> 1. Does the motor need any particular properties to make this work?
It will work with any DC motor. But note that some "DC" motors are AC motors with a built-in driver circuit.

> 2. Do i have to implement a flyback diode
Yes.

File: IMG_20161025_171955533.jpg (2MB, 3264x1836px) Image search: [Google]

2MB, 3264x1836px

>>1075431
hm, it certainly stops smoking before i'm done with the connection most of the time.

Also it works. Now i only need to wait for the LED filaments. Thanks for the tips, this stuff is really fun, i'll sure be able to make use of them in the next few weeks.

>>1075460
Thank you. Will the tech sheet hold the info wether or not its an AC or not?

>>1075497
It's usually fairly obvious. Like, if it's just a metal cylinder, it's a DC motor.

AC motors tend to be built into things. E.g. PC fans use AC motors (but you can drive them with PWM; the controller senses the PWM duty cycle and regulates the speed accordingly).

If it has more than 2 wires, it's likely to be an AC motor which needs an external driver (commonly termed "brushless DC" or "BLDC" motors).

File: halo.png (568KB, 819x800px) Image search: [Google]

568KB, 819x800px

Hello /ohm/

I want to be an angel for halloween and use one of these ultrabright halo lights as my halo. How do I go about powering it for a whole night?

Thanks,
Vincent

>>1075514
link to product? What voltage does it operate on, how much current does it draw? Inside accesible? SMD resistors for current limiting or something you can change?

File: 1422545020360.jpg (94KB, 600x388px) Image search: [Google]

94KB, 600x388px

>>1075513
ok, thank you.
>mfw i thought for a moment modern engineering was able to hide a AC motor within a 3 inch cylinder full of copper wiring

File: 14808665_627819384046876_1665123280_o.jpg (157KB, 1152x2048px)

157KB, 1152x2048px

>>1075444
PID control and a microcontroller will be useful for controlling speed and RPM.

Oh of course also feedback, motor feedback from a sensor which in some way can measure your RPM so that your controller can adjust PWM period accordingly.

Feedback can already be incorporated into the motor such as a hall effect sensor, or optical encoder. Or you can build your own.

Pic related, an infrared diode and receiver will go high when the blade passes over the diode. Pulses/time in the microcontroller will give you RPM or RPS which will be input to your PID algorithm.

>>1075525
Thanks for the tip!

>>1075534
No problem. Of course this is for very precise speed control so you can say "yep its turning at 4000 RPS", if you don't need that much accuracy you can eye ball it and build a 555 timer circuit with a trimpot for duty cycle control.

Really depends on whatchu got and whatchu need. First question of all control problems.

>>1075537
Well to be honest family i doesn tmatter how fast it turns as long as it turns consistently at the same speed. (Imagine grilling a whole pig)

>>1075543
Welp without feedback control its less likely to move at the same speed. Load, disturbances, and motor characteristics may introduce some error to the rotation. Of course this error may be really small, so small you don't have to consider it.

I'd try it without any feedback control at first and gauge performance, if it doesn't perform as needed then consider more sophisticated solutions

>>1075519
Probably this one? I wish it were a bigger but it's way easier than making a ring of ultrabright leds myself
https://www.amazon.com/dp/B01CHZ7422/ref=cm_sw_r_sms_awd_x_4F6dyb34XFA19

>>1075548
Will do so, thx

>>1075551
I'd just hook it up to two USB powerbanks in series, 10V should be enough to power it sufficiently bright. Otherwise use 2x4 AA batteries.

>>1075559
What's the best way to hook up the two power banks without them exploding on me?

File: 1476711096144.jpg (41KB, 532x343px) Image search: [Google]

41KB, 532x343px

>>1070254
Good website for buying cheap pcbs? Looking for ultrasonic stuff (aeroponics, ultrasonic cleaning, mist maker)

new thread

>>1075677

>>1075677

>>1075677

>>1075677

>>1075677

>>1075649
Just put them in series, their charging circuit is enough to prevent damage as long as you don't fuck up the polarity.

>>1075657
ebay, aliexpress

>>1073761
CompE.