/diy/ - Do It Yourself

>>1028473
How do you draw dick on oscilloscope?

>>1028485
Put scope in X-Y mode, send appropriate voltages move the beam to draw whatever.
You can see a microcontroller and two crude DACs in OP's pic.

I just got an Arduino. Can somebody explain how the fugg pin inputs work?

>wire goes from 5V to switch, two more wires go from other end of switch to analog and digital pins
>LCD displays inputs of these two pins
>while the switch is pressed all is well, analog = 1023, digital = 1
>when the switch is released analog flashes up seemingly random numbers, usually in the 5000s
>digital stays at 1 but will flicker to and from 0 if I wave my hand around it

Where do these numbers come from? Is there a way to stop it (so analog = 0 and digital = 0 as neither wire should be carrying any current)? I wanted to use a digital pin to tell if a switch is pressed, but obviously I can't do that right now.

>>1028534
Connect a resistor (e.g. 10k) between the switch/inputs and ground.

As it is, when the switch isn't pressed, the inputs are left floating, so they'll just be picking up noise.

Although it's more common to do it the other way around, i.e. one end of the switch connected to ground, the other via a 10k resistor to +5V. The inputs will read low when the switch is closed and high when it's open.

One reason for doing it that way is that the current sunk by a TTL input connected directly to +5V is much higher than the current sourced when it is connected to ground. For CMOS, it makes no difference.

Another reason is that it works if the switch is an open-collector NPN output, which are fairly common.

>>1028534
What you have is a floating pin when the button is released. Logic inputs have near infinite resistance and if left floating will pick up random stuff from the environment like an antenna.

Wire it as this:
VCC---switch---sample---resistor---GND

Your homework assignment will be to explain how this solves the problem.

>>1028488
Specifically this type > https://en.wikipedia.org/wiki/Resistor_ladder < for anyone who's interested.

I heard once that back in the day some people would attach one to a parallel port for a real basic sound card.

>>1028535
>the switch connected to ground, the other via a 10k resistor to +5V.
And if you connect the switch that way, Arduino (or at least its processor) offers internal pull-up resistors which can be enabled if needed. They're rather weak.

>the current sunk by a TTL input connected directly to +5V is much higher
It's the other way around. Bipolar 7400 (LS/S/AS/ALS/F) stuff has very low input current when pulled high. It's outdated and you can generally use CMOS 7400 families instead.

>>1028485
you need a degree to dickscope

What kind of circuit do I need to modulate the pitch of an audio signal?

Searching for frequency modulation only yields RF stuff and FM synthesizer is not quite what I'm looking for since that's generating audio from basic oscillators. What I want is when you play an audio signal you should be able to vary some pot and the frequency of the audio is either reduced or increased thereby distorting the audio.

>>1029043
>play an audio signal you should be able to vary some pot and the frequency of the audio is either reduced or increased
That's called a time machine.
The pitch of an audio tone is measured by the number of cycles over a period of time.
If the audio is recorded in some way to make it a signal you can play, the cycles:time is already fixed.
To change the cycles you have to change the time.
Look up how AutoTune works.

>>1029060
So how do you explain FM radio then?

What I want to do is literally no different than what is done with FM radio. All I want to do is basically make the audio signal the carrier wave and apply a variable signal that will allow me to shift the frequency of the carrier up and down within the audio spectrum.

I don't understand why this violates some law of physics.

>>1028473

>>1029063
>literally no different than what is done with FM radio
FM radio has a center frequency which it shifts around (up or down) by the frequency of the applied audio.
It doesn't just go UP (or down) and stay there.

How bad of an idea is it to get started with doing things with a USB port instead of a proper power supply?

On the one hand, I know it should be 5v off the line every time and it makes anything I create necessarily portable, but on the other, I'm afraid of drawing too much by accident because I'm new and dumb and either destroying whatever I'm drawing from or setting it on fire.

Also, are there any easy programs for reading inputs or outputs from a USB? Playing with fans and lights is one thing, but it'd be cool if I could have functional devices I can slot into my netbook but not have to learn straight up C to write drivers for.

>>1029071
Well then what options do I have to modulate the pitch of an audio signal without doing it digitally?

>>1029075
Where does this 'signal' come from?
What type of 'signal' is it?
If its and audio signal you are generating (tone) just change the frequency of the signal generator.
(turn the frequency knob)

>>1029076
It's music. It'll come from any music source so a phone a computer, etc. Anything with a 3.5mm jack really.

>>1029077
AFAIK there is no way without 'recording' the signal and 'playing it back' at a different speed.
The 'recorder' can be digital, tape, or other means and it can be played back almost instantly if it's digital.
This is pretty much how AutoTune works.

>>1029079
search for "phase vocoder"

>>1029091
That requires digital signal processing and FFT and stuff. I was looking for an analog way to shift frequency in real time.

>>1029092
any transmitter/ receiver pair with single sideband can do that by varying the clarifier. If you want to handle music then look into essb (expanded single sideband) which has a wider audio bandwidth.

>>1028473
is that DAC drawing a dick?

>>1029099
no

>>1029072
Get a USB to Serial/UART adapter, USB is not the sort of protocol you want to start with. USB is packet based and requires a specialized controller.

You can use a AC/DC plugpack/walwart adapter for a fixed voltage needs. Old cell phone charger can be used to get 5V up to a few hundred mA which is enough to get you started. I wouldn't use a USB port on a computer as a power supply, you could end up on relying on it's overcurrent protection.

>>1029043
> What kind of circuit do I need to modulate the pitch of an audio signal?
An audio signal doesn't, in general, have "a pitch".

Basic signal processing theory tends to start from the assumption that a signal is the sum of a finite number of sine waves, each with a specific frequency, phase, and amplitude. It then describes how various operations affect a single sine wave. If an operation F is linear, then F(x+y)=F(x)+F(y) and F(k*x)=k*F(x), which means that an analysis of how an operation affects one sine wave provides a meaningful representation of how it affects the sum of multiple sine waves.

Even so, real signals aren't actually the sum of sine waves, but that's typically a close enough approximation over a short interval of time.

So, one way to change "the pitch" of an audio signal is to split it into chunks time-wise, perform a discrete Fourier transform of the chunk to split it into sine waves, change their frequencies, then add them back together. Except, you need to operate upon overlapping chunks and blend them together so that you don't get discontinuities at the boundaries.

tbc...

>>1029063
> So how do you explain FM radio then?
Frequency modulation doesn't vary the frequency of an input signal. It varies the frequency of an oscillator, i.e. it *generates* an output signal whose (single) frequency is determined by an input signal.

>>1029092
> I was looking for an analog way to shift frequency in real time.
Phase-locked loop to track the dominant (fundamental) frequency and multiply it. Various frequency dividers generate harmonics. Fundamental and harmonics are modulated with the input signal and the results low-pass filtered. Each signal is the amplitude of a given harmonic in the input signal.

A second oscillator is tuned using the control signal for the PLL and the desired frequency shift, and again divided to generate the same set of harmonics but of the new frequency. Each is modulated by the strength of the corresponding harmonic in the input.

tl;dr: it's much easier (and more accurate) to do it digitally.

>>1029072
> I'm afraid of drawing too much by accident because I'm new and dumb and either destroying whatever I'm drawing from or setting it on fire.
So use a $2 USB-charger wall wart rather than a $1000 PC. If the device is correct, it will work with either. If it's wrong, you're out $2 rather than the cost of a new mobo (or even a new PC).

> not have to learn straight up C
You need to learn C, and probably some assembler as well (you don't typically need to program in assembler, but you need to be able to think in assembler and sometimes debug in it).

>>1029122
>real signals aren't actually the sum of sine waves
by that you mean there's also phase?

>>1029043

Jesus christ the responses you've gotten have been straight-up retarded. Like, shockingly so. I'd like to personally apologize for just how retarded these people are.

What you want to do is take an audio signal and move it up the frequency spectrum, like pic related, right? That's just straight-up frequency modulation, as you've identified. In electronic music the circuit that's typically used is called a Ring Modulator.

More specifically/helpfully, an ring modulator is a type of 'frequency mixer', which is a circuit that uses a non-linear device (in this case a quartet of diodes) to multiply two input signals.

If you remember your trig identities you'll probably immediately see where this is going:

sin(a)*sin(b) = (cos(a-b) - cos(a+b))/2

If one of the inputs is a pure sine wave (a 'carrier'), then in frequency domain it'll look like the signal has been cloned, with one copy being shifted up by the frequency of the carrier and the other being shifted *down* by that same frequency. Typically you then high-pass filter to get the up-shifted frequency when you're modulating and low-pass to get the down-shifted frequency when you're demodulating.

>>1029308
> by that you mean there's also phase?
No, I mean that real signals aren't periodic.

>>1029431
This only works if the shift frequency is outside the bandwidth of the signal. Otherwise the two copies of the spectrum will overlap and you can't eliminate the reflected copy with a filter.

IOW, it's of no use for pitch-shifting audio.

>>1029122

Not quite right. For starters, it's not an 'assumption' that signals are made from sine waves.

Mathematically, a signal is simply a function that is parameterized by time. The sine functions form an 'orthogonal basis' for all functions, which is just a very fancy way of saying that any function (even discontinuous ones like square waves, or aperiodic ones like the Dirac delta function) can be expressed by a summation of (potentially infinite) sine waves (the 'Fourier series expansion' of that function). Ergo, for any given signal, you can find a set of sine functions that can be added together to get that signal exactly.

Of course since you can take sine waves of the same frequency (but different phase) and add them together, the simplest Fourier series expansion has precisely one sine function (of an arbitrary amplitude and phase) for each specific sine wave frequency that shows up in the signal. If you then add a sine function for every frequency that's not in the signal, (with an amplitude of zero so as not to affect the result of the summation), you get an infinite number of sine functions each corresponding to exactly one frequency and amplitude/phase. The Fourier Transform simply takes this infinite summation and re-expresses it as a function of amplitude/phase vs frequency.

Don't even get me started on the Discrete Fourier Transform and the Nyquist sampling theorem, which adds in some modular arithmetic and is just generally fucking *beautiful*.

>>1029439
>This only works if the shift frequency is outside the bandwidth of the signal. Otherwise the two copies of the spectrum will overlap and you can't eliminate the reflected copy with a filter.

Good point. What you could do is modulate up (say) 30 kHz, filter, then demodulate down by (30-x) kHz, and low-pass at 20 kHz. That'll give you an image that's shifted up x kHz without distortion.

>>1029439
>Otherwise the two copies of the spectrum will overlap
Yeah, if you use the most basic mixing. However, it is possible to do it so that the resulting spectrum contains only the upper (or lower) sideband. The phasing method (produce a 90 degree shifted version of the audio, produce a "carrier", use quadrature mixing) is one audio-compatible way to do it.
Google SSB for more info.

>>1028473

> tfw Dave Jones responds to your troll comment on a video and it gets put in the thumbnail

>>1029431
multiply by e^(-2pi i w) where w if the shift you want trust me i have a ph.d in fourier analysis

>>1029443
Specifically, you need to apply low-pass filter at each step, so that you're getting the reflected spectrum each time. The reflections cancel as does the 90-degree phase shift.

Anyone know what way round I would need to install the electrolytic capacitors here? There's this T on one pin each? Is that for the long(positive?) or the short(negative?) pin here?

The T pin on the left is connected to the 3V3 pin on the left, so I guess that's power?

The SMD solder points are for the voltage regulators. One 5V and one 3V3. I'm not sure which is which. Judging from the connections I believe the one on the left is the 5V one, though I'm not certain.

This is a breakout board for the BBC micro:bit in case anyone's interested.

>>1029492
That 'T' is trying to be a +.
Your capacitors have their negative pin marked.
See your documentation for regulator placement or see/measure where the wires are going.

>>1029492
Probably the T's are badly placed plus signs.

>>1029499
>That 'T' is trying to be a +.
>>1029500
>Probably the T's are badly placed plus signs.

Thanks! That's what I figured. Just didn't want to risk popping my capacitors over it.

We should really add KiCAD to the sticky for
>What software should I use to print circuits?

Just sayin'

>>1029434
Well, technically, you can get signals with infinitely long period, so basically non-periodic, from summing up sine waves. The problem is, the band over which you sum those waves would also have to be infinitely wide.
That's where the spectral leakage comes from when you sample a signal of finite duration.

Either I'm retarded or this is a 300 ohm resistor. Bands are from left to right brown, orange, 2 blacks and a blue. Put me out of my misery, am I a retard?

Answered my own question I think, the LED didn't pop at any rate

>>1029604
did you buy these off chinese sellers on ebay?
that could be it, or that your multimeter is running out of battery

>>1029604
your color code would be 130ohms, 0.25% tolerance. (or 600kohms 1%, in reversed order)
where did you buy them? never trust chink products

>>1029618
>>1029619
Bought them in a Malpins.

Not looking too good for the others that came in the pack either then?

>>1029431
> What you want to do is take an audio signal and move it up the frequency spectrum, like pic related, right?
I suspect that he doesn't (even if he thinks that he does).

Shifting an audio signal by a fixed frequency offset will change the ratios between the frequencies.

"Pitch-shifting" involves multiplying the frequencies by a constant rather than adding a constant offset, so that harmonics are preserved.

E.g. doubling every frequency will shift the pitch upward by an octave, while multiplying by ~1.06 will shift the pitch upward by a semitone (assuming an equal-temper scale).

You can't do that using modulation, and I have no idea how you'd do it via analogue means. Doing it digitally isn't exactly trivial: if you just do FFT, resample, inverse FFT, you'll end up speeding up the audio by the same factor.

>take things apart
>never put them back together

am I /ohm/ yet

>repairing a motherboard
>got the bad caps off
>can't get the rest of the solder out
The solder that was on it is fucking terrible and hard to melt. I've tried a solder sucker and some solder wick, but it barely melts. I've turned my iron to its highest setting and no luck, what do I do now?

>>1029658
First, are you using flux with the solder wick (or is it pre-fluxed)? If not, try that. Failing that, try adding a little fresh solder to what's there and wick/suck the combination.

>>1029658
I have been in your exact situation many times before.


Try poking a small component lead through the hole, heating and dicking about until enough solder is removed

But you really need a proper desoldering gun for this type of work.

>>1029662
No flux on the wick. I'll try that, thanks.

>>1029663
Had a cheap radio shack desoldering iron, it worked alright but it burnt out.

>>1029658

heat a stainless steel needle in the hole and push it thru. let it cool, then twist it off. no other kludge works as well as this.

Can I just smack a few diodes in series to drop down 5.8v to 5v at 3A load? Are resistor dividers viable at this current? (only have 1/4w resistors)

>>1029716
what sort of load? 5.8v might be fine. resistor dividers won't work. your diodes will need to dissipate 2.4 watts which will require a heatsink.

>>1029664
Do you have lead based solder or RoHS compliant shit solder? Add lead solder to the hole. The whole thing will get much easier to work with and melt at a lower temperature.

>>1029723
A very old powerbank, with a type B USB port for charging it, which came with a very chunky wall-wart.

I tried charging it through my pc with no luck but my bench supply set at 5v does it no problem within a couple of hours

>>1029619
Newbie with ohms, so they can be reversible, and have different resistance?

>>1029754
It will be the same resistance regardless of the direction of current, but you cannot immediately tell which order to read the bands: Brown-Orange-Black-Black-Blue and Blue-Black-Black-Orange-Brown happen to both be valid 5-band resistor codes. The former decodes to 130Ohms, the latter 600kOhms.

>>1029754
Nope, doesn't, you can connect them either way, resistance stays the same.

What >>1029619 meant was that if you read the colour code the wrong way around, it would give a false indication of its value.

On some resistors this could be a problem as some colours have both a tolerance indicator and number to them, check pic related, green is both the number 5 and a tolerance of 0.5%

Just take a few seconds with your multimeter if you are unsure.

>>1029755
Ah Ok, thanks!

>>1029756
Ok, thanks for the multimeter idea.

>>1029755
You're unlikely to find a blue tolerance band though, unless you're working on some fancy stuff.

>>1029642
>am I /ohm/ yet
the force is strong with this one

>>1028473
Is that the pickup pattern to a tele mic?

>>1029492
>those footprints for C1/C4

Jesus, why are awful footprints so common? Even footprint libraries that come with software like Altium are awful.

Has anyone ever made a decent footprint library? With packages pretty much standard nowadays it should be possible to just buy an off-the-shelf library with decent silk-screen art.

Pic unrelated (pretty sure it was laid out by hand, actually).

>>1029604
At least one of the bands in that pic looks like it's gold, from the way it shines slightly in the light. So I think you've read the bands wrong. Can you post a close-up of the resistor, in better lighting?

>>1029794
>Has anyone ever made a decent footprint library?
Yeah, I like my own.
I recommend that you make your own, too.

>>1029555
Not until you can do via stitching in KiCad and not have to relearn the retarded parts/footprint library everytime you start a new project.

It is so fucked that my way of starting a new project is copy-pasting the files of an old project into a new folder and then renaming the files because i can't be bothered with reimporting every parts and footprints and 3d models and libraries and shit.

I want to control 4-5 transistors, each with a ~1A load at 5v with arduino/whatever. I have the hardware but I'm fucking shit at making circuits.

My problem is not "the circuit"; I can kind of get something working. The problem is the fucking components.

I'd end up with 6 wires jumping from each transitor to 5v or something like that. How can I design a circuit but only considering the components?

Here's my idea done in 1 min (instead of leds i'm using various 5v devices)

TL;DR: HOW TO AVOID HAVING WIRES JUMPING EVERYWHERE IN A PCB

>>1030039
Also, in this case it'd be a wire jumping from base to microcontroller for each transistor.

The rotated version that has "straight" base<->microcontroller cables is a mess with the collector <->5v

>>1030039

There's some graph theory involved that says, basically, "you can't do that". You have to jump traces somewhere.

However, if you are using through-hole components, you can simply modify the pad pattern to be more spread-out and run traces underneath them.

Pic related.

>>1030051
>>1030039

Also, in this case, the transistors should switch on the low side of the LEDs, not the high side. Between their own diode voltage and the forward voltage of the LEDs, they might be difficult to turn on fully if they're on the high side.

The reverse is true if they're PNP/P-channel, however, however.

>>1030051
>>1030039

Also, in this case, the transistors should switch on the low side of the LEDs, not the high side. Between their own diode voltage and the forward voltage of the LEDs, they might be difficult to turn on fully if they're on the high side.

The reverse is true if they're PNP/P-channel, however.

I cant never get the 555 to work.

I've picked simple protoboard projects to see it at work, involving 3 resistors, a capacitor and a led in order to create a blinking light.

I tried this three times and I can never get the led to blink, it just stays on.

Tried this with a red led and a green led and followed clear instructions. Everything is connected properly and tried this with two NE555

So whats going on?

>>1030085

The capacitor is also working, since when I disconnect the circuit the led stays on for a little bit.

The only difference from my circuit to the one from the guide is that the guy used a 33k resistor. That goes from 2 to 7 pin of the 555.

I used 3 10k resistors connected in series and the last pin I bridged it to the 7 pin. This is not wrong is it? Theoritically I will be having a 30k resistance right? does it make much difference from the 33k one? could this be the reason why the circuit doesnt work?

>>1029502

Is that a plus or is it a ground symbol?

Anyone think the spacing (or lack of) for the electro caps and heat synchs will be an issue? They are LM7812 LM7912 voltage regulators for a bipolar eurorack power supply powered by ac 12v 1000ma. On breadboard they don't get very hot at all, but I didn't use a very significant load.

I'm limited in horizontal spacing and like the symmetry of the layout, so I really don't want to reroute. I plan on etching this tomorrow. (This is actually the bottom of the board, the regs and synchs will be top with no copper underneath them)

>>1030140

Who cares it will be fine

>>1030087

SOILED IT

>>1030140
This is the full eagle layout. It's my first attempt at making a circuit as well as using eagle. I am just ripping off the frequency central co uk's supply which is based on the MFOS wallwart supply design.

>>1030140
with copper pour

>>1030085
We are not mind readers. There are no guarantees that your "clear instructions" were correct or that you connected everything "properly".
In other words, post pics.

>>1029449
Must feel good man. I sent him a letter with an item and at the top of the letter I tried to start an AVR/PIC flamewar (not that it matters anymore anyway lel), but the roody roo conveniently cropped the top of my letter.

My first internship starts in about a month and I'm going to have to write software for an FPGA. Is this difficult?

>>1030314

Depends how much experience in programming microcontrollers you have, plus from what I've read FPGAs are a whole different story.

Can anyone vouch for the quality of the stuff offered from this website?
http://www.omc-stepperonline.com/
I am looking for some high resolution (400 step/rev) nema 23 motors and this is pretty much the only place I have seen what I am looking for.

>>1030314

Not if you know VHDL

>>1030148
OK anon, I'm going to roll with it then.

>>1029618
Nice catch anon, the same battery was in the multimeter for over 10 years. Replaced it and it seems to be giving stable indications now

Using an arduino uno as a gamepad with switches connected between the IO pins and ground used to toggle joystick buttons on and off.

What I want though is for the switches to also control LEDs. I just found out about transistors and what they do so I want to use them instead of a SPDT switch as I had originally planned. What I need to know is the type of transistor I need (one that lets current flow when the base pin is unpowered) and what resistor to use between the switch and the base pin.
(I marked the LED resistor as ? But it turns out it needs to be a 33p ohm 1/4watt resistor)

>>1030489
*330 ohm

>>1030489
> What I need to know is the type of transistor I need (one that lets current flow when the base pin is unpowered)
A JFET or depletion-mode MOSFET works like that: They conduct when Vgs is zero and conduct less as Vgs goes negative (for n-channel) or positive (for p-channel).

But you're probably better off just using 2 NPN BJTs like pic related. Current to the base of the first pulls the collector low, turning off the second.

July's power bill will be approximately $300, with a tier 4 price of $0.44/kwh. It is 109F today, and will be 109F tomorrow. PG&E sucks.

Let's assume the author of this post is mechanically minded, has a solid background in DC electronics, is handy with a soldering iron, and has an engineering degree.

The posted picture is a 6"x6" solar cell that should produce ~4.3W in theory (0.5V, 8.5A ish). Ebay has posts where 112 of these are being sold for $137, which would be 400+ watts.

How hard would it be to tab, solder, and otherwise connect a quantity of these on say, a 4'x8' plywood sheet? I'm talking about the physical aspect of making a (very large) panel from individual cells.

48" x 96" should be good for 7x14=98 cells (42" x 84" actual cell size), which is ~48V and 8A for 390W.

Does anyone have any experience with building solar panels from cells?

I'm starting a "Introduction to Electrical Engineering" course in a month as part of my Manufacturing Technologies course, all I know about electronics is the stuff needed for getting a tech class amateur radio qualification, which is mainly basic "a hertz is a unit of frequency defined as a cycle per second" and "don't electrocute yourself" stuff. Anything I should keep in mind going into this?

>>1030574
From what I've read by the time you solder them all together and make a weather proof enclosure you are better off just buying a completed module from a store.
Also if you want to get solar cells even cheaper browse ebay for broken solar cells, these are factory mishaps/seconds that tend to go for way less.

>>1030577
Since it sounds like you might be involved with manufacturing/designing PCB's for manufacture, never forget to add fiducials (http://www.ladyada.net/library/pcb/fiducials.html)) to your design. This is like one of the most common things graduate engineers forged to add to their PCB.

>>1030574
I've done it before.

> say, a 4'x8' plywood sheet
The cells are extremely fragile. You should cover them with glass or Plexiglas or they'll be destroyed by a bird landing on them.


You're being overly optimistic with your calculations. Solar panels will only produce their advertised wattage when pointed directly at the sun on a cloudless day.

>>1030341
Hope you have an extractor for dat PCB fiberglass dust...

>>1030329
Some people on eevblog say they are good: http://www.eevblog.com/forum/beginners/specs-on-this-stepper-motor/msg816129/#msg816129

>>1030314
Never done any programming for FPGA's but hear that Digilent have some awesome kits for learning (http://www.theamphour.com/302-an-interview-with-clint-cole-of-digilent/))

>>1030039
Why not use a shift register since you are already using an arduino: https://www.arduino.cc/en/Tutorial/ShiftOut

>>1029831
>>1029794
I usually make my own with https://www.pcblibraries.com/
But I hear that https://www.snapeda.com/ is getting better
Also if you are after rough 3D models then http://www.3dcontentcentral.com/ is the place to go

>>1030314
enjoy troubleshooting timing faults

>>1029613
The output pins on Arduino have a current limit of like 20mA anyway, so if you didn't use a resistor you would have been fine anyway ;^)

>>1030704
> The output pins on Arduino have a current limit of like 20mA anyway,
That's a "limit" in the sense that you need to ensure that you don't draw more than 20mA if you want the chip to continue working. It doesn't mean that the chip itself limits the current to that figure.

What kind of wire do people usually use in headphones?
I was thinking some generic solid core 16 gauge copper.

>>1031186
*18 gauge, since that's what fits in the mini-XLR connector's solder cups.

>>1031186
Since it needs to be able to take lots of bending, the usual choices are litz wire or tinsel wire. Solid core is about the worst possible choice.

>>1031192
Soldering Tinsel makes me want to shoot myself.
This is for a connector inbetween the driver and the cable.
Since there isn't a whole lot of movement on the inside of headphones, there shouldn't be a problem using solid core

>>1031197
>Soldering Tinsel

Crimp it, don't solder it. Soldering tinsel is like shaving with a chainsaw.

>This is for a connector in-between the driver and the cable.

Anything will work. There's bugger all current to worry about and the effect of inductance on audio frequency signals is nil.

A couple of months ago, I replaced the dead lights in my old stereo receiver with LEDs. I turned on the receiver yesterday, and two out of three of the LEDs stopped glowing.

Where do I start troubleshooting this? The LEDs are in series, so if one of them stopped working, the other ones should too. The LEDs are replacing 12v 60mah bulbs, so I put a 470 ohm resistor.

>>1031301
>so if one of them stopped working, the other ones should too.
or all others die as well because the LED fails short :^)

What kind of cleaning would you do on these contacts? Are they fine the way they are? I'm concerned about the brown discoloration around the edges.

>>1031350
Use pic related.

>>1031350
isopropanol with a cotton swab

>>1031396
Worked great, thanks! Thought there might be something special needed because of how the contacts are constructed (film instead of a regular board)

>>1031387
Didn't need to use that this time, but thanks for the suggestion. I'll keep that in the back of my mind if anything similar comes up in the future

>>1031301
What is the total voltage across the string? 12V or 36V?. I.e. are the 3 LEDs replacing 3x12V bulbs in series (36V) or a single 12V bulb (12V) or 3x 12V bulbs in parallel (12V)?

3 red LEDs give you a 6V drop. If the voltage across the 3 is 36V, then 470 ohms would give you 30/470=64mA, which is way too high. If it's 12V, then it would be 13mA which should be fine.

Also, is the voltage AC or DC? Filament lamps will work fine with either, but an LED probably won't like being reverse-biased by 12V or more (you can't assume that reverse voltage will be divided equally).

And if it's AC, filament bulbs will only care about the RMS value, but LEDs will care (to an extent) about peak value (1/60th of a second is an instant for a filament, but a long time for a semiconductor).

I have hyperhidrosis, especially bad in my hands.

Is it ok to smear my hands all over my circuits and cover them in sweat, or do I need to wear rubber gloves?

Will I get laughed at if I show up for a uni course with rubber gloves? How bad will a electrician bully burn you with a soldering iron before he decides you've had enough?

>>1031438
>do I need to wear rubber gloves?
Pic related should be sufficient.
They are inexpensive and comfortable.

Whats type of adapter do i need to buy to be able to power this? I have no idea what I'm looking for on amazon

>>1031475
A 12VDC one...

Would this work as an ADC?
What am I overlooking?

>check if positive or negative
>amplifies previous signal by 2 and recenters it in the direction opposite the previously obtained bit
>repeats as many times as these are chained together

>muh bifurcation
>muh tail recursion
>muh one comparator per desired bit as opposed to the {2^n}-1 of a normal flash adc

Resistor ratios on the doubling amps would have to be pretty exact.
And the rails on the doubling amps and summing buffers would have to be twice as big as everything else.
One major issue with the amplification would be noise - But that might actually just end up working beautifully as natural dither.

Idk.
Am I crazy brilliant, or just crazy retarded?

>>1031475
what the hell is it? is there a datasheet? never seen a 2 prong 12v connector like that before. usually they're barrel connectors.

>>1031528
This is effectively a successive-approximation ADC which has been "unrolled".

>>1030693
No, should I?

Scared.

>>1031428
To be honest, I was under the impression that it would be 12v all across.


Here is the link to the schematic.
http://akdatabase.org/AKview/albums/userpics/10004/Yamaha%20CR-420%20Schematic.pdf

The lights are at the bottom, right above the plug and all the power sockets. It says 12v60mAx3 next to the bulbs. I used the 12v to determine that I would use a 150 ohm resistor. Should I have used the 32v instead? Does my problem have to do with the capacitor that is directly to the right of the lights on the circuit?

>>1031549
How accurate do you think I could get it to be?
Worth breadboarding for a test?

I really like this because it would spit out something I could put on a parallel bus.
Everything else today uses this retarded serial bullshit I just don't want to deal with.

>>1031207
The magicians at Beyerdynamic seem to have been able to produce a good result soldering tinsel though.
Makes me wonder how they did it.

>>1031669
They probably have around 40V across them.

The transformer winding say 32V. If that's the RMS voltage, the peak will be 32*sqrt(2)=45.25V. Rectification and filtering yields the peak voltage, not RMS. C509 and C521 are spec'd at 50V.

The lamps are spec'd at 12V each, but are in series so you'd expect 36V between L1 and L4. They're also spec'd at 60mA. R501 is 180 ohms, which would drop 180*0.06=10.8V at 60mA. Which is about right if the voltage across C509 is ~45V.

With LEDs, R501 would drop 1.8V at 10mA or 3.6V at 20mA.

To run 3 typical LEDs from 45V, you'd want a total series resistance of at least 2 KOhms, 1 Watt. Ideally a bit more; 20mA is a typical maximum current for indicator LEDs, but they last longer if you don't run at the limit.

>>1031675
You can still buy parallel-interface ADCs, they're just harder to find because no-one uses them in modern designs.

>>1031708
Link?
Maybe i just don't know what to search for or something, but I haven't been able to.

>>1031675
>How accurate do you think I could get it to be?
Similar converters with additional sample-and-holds between stages are available up to 16b resolution, so at least that much.
That is, assuming you make an IC with carefully matched/compensated everything. If not, you should get a rough idea by considering the available resistor tolerances/tempcos and comparator/amplifier offsets.
>Worth breadboarding for a test?
For your personal amusement? Maybe.
Btw, the speed advantage you get with this is pretty small, since the signal still has to go through all the comparators and amplifiers on one go. For a similar but faster idea see half-flash (or sub-ranging) converter. It's basically the same, but it converts more bits at a time.
>Everything else today uses this retarded serial bullshit
There's no need for parallel bus unless you have a very fast converter.
And it's not that parallel output converters aren't available.

>>1031711
Come on, you've been lazy.
Just go to TI or LT or whatever site and use their parametric search.
For example, TI has this 18b 500kHz converter ADS8383.
Maxim offers a 16b 100MHz converter MAX19588.
Then there are antiques like Analog's 12b 8µs AD774.
There are plenty of others as well.

>>1031721
> There's no need for parallel bus unless you have a very fast converter.
And at >GHz speeds, serial would again be preferable (although you're not going to get that from SPI/I2C, and PCIe is a bit heavyweight for an ADC).

>>1030489

Almost there /ohm/. When the button is pressed the LED lights brightly, but there are problems:
I want the LED to light when the button Isn't pressed and I don't fully understand >>1030557 .
Also, the joypad button doesn't work when the LED does. It's either have the pushbutton connected directly to ground and have the joypad button work but the LED not work, or to the base pin which makes the LED work but not the joypad button. Suggestions?

>>1031908
> Suggestions?
1. Explain the requirements more clearly.
2. Post a proper schematic.

Where could i find software called ACTWin for HITACHI? Cracked of course

I need _modern_ ways to prototype and test circuits, fast. I want to leave breadboards behind and use some kind of software sim and then print out circuits. Is this feasible in 2016?

>>1031920
Requirements: LED to be On while button is NOT pressed, currently it's the opposite.
LED to be ON while wire from button is connected to GND (currently I can either connect pin 5 to GND to get a joypad button set to pressed, or connect to the Base pin on the transistor to get functional LED, but not both at the same time

>>1030489
Somewhat proper schematic

>>1031996
EveryCircuit is nice for small on-the-go stuff
No sim can give you the fuckedupness of non-ideal components, however

>>1032018
This quick sketch should give you the idea
The LED can be run directly if it doesn't draw more than 20mA and if you have no other significant loads

>>1031438
>>1031456

Talk to your lecturer. I doubt they'll care, you're not working with anything that spins so gloves are fine.

If your classmates complain tell them you have hyperhidrosis and threaten to take your gloves off and moisten-up the (shared) iron. Make reference to that guy from Dr Horrible's Early 2010 Sing-A-Long.

>>1031996

PCB mill. If you can't afford one, toner transfer & etch, or pre-sensitized boards / UV nail curer / printed or plotted transparencies.

>>1032077
>paid software

>>1032098
>PCB mill
>If you can't afford one,
I can't afford not to have one tbqh. Looking at amazon I see one at $400, but that seems like the low end, no? How much does a good one cost?

>>1032100
I actually use the Pro APK you can find in literally thirty seconds on Google

>>1032134
>not open source and under GPL
>literally commercial shovelware

>>1032163
https://www.gnu.org/philosophy/words-to-avoid.en.html#Open

gnucap look promising

>>1032105
>How much does a good one cost?

>>1030574
>Does anyone have any experience with building solar panels from cells?
it's not fun. buy $15 of broken cells and tab wire before spending any real money. I probably broke half mine just from heat stress.

I just obtained this old burg system keypad. How can I test to see if it still works? What can I do with it?

>>1033197

that's a nice part. Does it have any logic on it or is it just a bunch of switches?

Any pcb mill suggestions for first rodeo?

>>1033197
>>1033267
Those typically use a serial bus (RS485, for example) to talk with the main unit.

While you could try your luck with reverse engineering or document hunting, your best bet is probably to use the keyboard only and ignore the PCB.

>>1030339
This. The syntax is similar to Ada if you ever used it

Anyone replaced laptop keys before? Did you buy online in singles or buy a donor computer?

Was planning on buying a single key, but read online that it's usually bad quality and only saves you $15 from buying an entire donor keyboard.

>>1033452
http://dirtypcbs.com/
or
https://www.oshpark.com/ if you're rich

>>1034089
that takes all the fun out of it mate. You can print a board out on your own mill in 2 hours while you're having sexy funtime.

>>1034432
Depends on your perspective, of course. I prefer the design and layout, while letting tiny asian hands make my board and tiny asian lungs breathe in all the assorted fumes from those processes.

>>1034472
There's no fumes from milling. It's just a drill bit removing copper.

>>1034474
Yeah I guess I had lumped in etching and all that. Anyway, to each their own, eh?

I'm trying to get a Harddrive motor to spin continuously, but obviously when I put power to it through the harddrive circuitry, it'll spin up for a minute or so, then turn the motor off to sit idle and not waste power.

Is there a way I can get the motor to spin continuously without major alterations to the circuitry? I've posted about this issue on a previous thread and I know how to control the motor in theory via a controller for a BLDC with back emf sensing, but it seems a little tedious and I can't exactly figure out the pinout leads for the motor.

>>1034545

just use an old drive, they never stop. i have like 20 old ones from 20M to 80G and you just apply power and they go forever.

>>1034550
Where do I look for these ancient harddrives? I took a few harddrives from some old comupters my school was throwing out, and they're all relatively new (they all stop spinning after some time).

Hardware beginner here: how do i safely use this to supply power from my phone charger to stepper motors? Do i need to put some resistors or capacitors between the adapter and my motors?

>>1034551

Try goodwill.

>>1034561

The USB port on your phone can't handle current like that, you'll blow your phone up before you move a decent stepper even 1 step.

>>1034570
I was talking about my phone CHARGER, not the phone itself

>>1034574

Sorry, missed that word.

>>1034089
http://pcbshopper.com/

My nixie clock setup so far. Any help/suggestions? I'm not completely done but this is the base of what I'm going with. I'm not an electrical/computer engineer just a hobbyist and I've only read one book on the subject.

>>1034639
74LS145 is rated for 15V, which isn't enough for driving Nixies. While you can expect it to be able to handle higher voltages, you can't really count on it.
A similar IC capable of driving Nixies (as long as you're careful) is 74141 or its Soviet equivalent.
An alternative would be to use external transistors to buffer the outputs.

Also, what's that 1n capacitor on the shift clock line?

>>1034561
Just connect Gnd and Vbus to the motor controller.

The charger will already include a filter capacitor, although you might want some smaller decoupling capacitors (tantalum or ceramic) close to the controller.

>>1034652

Strange. I thought I used the Russian high voltage drivers K155N*1 I don't know if I used that IC because it was same pins, but now that I notice it doesn't have a 5V line or a ground. Pic related is what I'm going to use.

As for the 1nf capacitor from the schematics I have seen it's optional, but if helps with signal integrity then I'm going to include it.

I appreciate the feedback by the way.

>>1034682
>I notice it doesn't have a 5V line or a ground
It is common that the power supply pins are not shown. Your schematic editor probably has an option for forcibly showing them, though.
>if helps with signal integrity then I'm going to include it.
Your layout needs to be quite horrible for that capacitor to be an improvement. Try to draw a nice PCB instead. Oh, and add some power supply bypass capacitors.

>>1034696

I don't understand. If the PCB layout is shit then the capacitor is needed, otherwise it's redundant?

By power supply bypass capacitors do you mean decoupling capacitors? I have quite a few and plan to do so for signal integrity and over voltage protection.

>>1034723
By default, such capacitor worsens the situation by heavily loading the clock signal.
However, if your layout is bad (overly long lines, bad grounding, etc.) resulting in heavy ringing in the clock signal, the capacitor might help to dampen it.
Thus, use good ground, don't use overly long wires, use adequate bypassing and avoid running noisy lines in parallel with the clock wire. If long lines are unavoidable, series termination (small resistor in series, near the controller) is one common way to cope with them.

>By power supply bypass capacitors do you mean decoupling capacitors?
Yes.

>>1034742

Thanks for the info fampai.

>>1034682
>K155N*1

K155ID1

И != N, even if it looks a little like one.

>>1034781

Yeah, I knew it was something like that. It's the same with the tubes. They have those silly vodka runes on them but instead of being something legible it's like IV or IN.

>>1034545
So does anyone actually know how to get a newer harddrive to spin continuously?

I don't feel like hunting goodwill and just want to work on the harddrives I have right now

I'm at the end of my rope. Is there any place that will fabricate custom watch PCBs/screens, with options for size, buttons, and face design? I have two dead seiko G757s, both of which have died repeatedly. I just want to get one of them looking like it used to.

I'm curious what people think about the arty board vs the snickerdoodle

not sure what I want to use them for. Maybe some home automation, or sound synthesis.

>>1034785
итc eacy фaм

>>1034785
итc eacы фaм

>>1034792
Make your own motor controller. Reuse the existing driver chips if possible and use tarduino or something to produce the 3-phase drive waveform.
Or try your luck with goodwill, like someone already suggested.

>>1034797
If you just need spare parts for that particular clock, your easiest and cheapest option should be to find another one.
LCD makers do custom displays, but they won't make just one and you can expect the price to be prohibitive.
Dunno about the rest of the clock, but I'd assume it's specialized enough that you'd need to contact some boutique watchmaker... Or not, if you don't care that much about mechanics, size and shit. Pic related.

That said, there's some reason why they have died repeatedly. Any idea what is wrong with them?

>>1034829
The problem with finding another one is that this model is notorious for bad PCBs. On one of them, the PCB was corroded, which I fixed, then a few resistors died, which I also fixed, then some other things went and I couldn't fix them. The other one was corroded to the point of uselessness.

>>1034836
>corroded
So it isn't watertight? Whatever you do, consider fixing that problem first.
If you're sure you can't revive the existing PCBs and you can't buy ready-made PCBs... If you're willing to spend time, it should be possible to replicate the functionality of the watch PCB, but fitting it inside the watch frame and reaching equally low power consumption is going to be a major pain in the ass. Hobbyist-made watches tend to be big, clumsy and power-hungry.

>>1034846
The o-ring wore out after 35 years, that was the first thing I fixed.

>>1034812

I don't read Vodka runes.

>>1034792
Get a CPU with a SATA interface and have it tell the drive not to go into standby when inactive. The easy way would be to get a CPU which can run a minimal Linux kernel and use "hdparm -S" to set the timeout. A more involved solution would be to figure out what you need to send it and how so that you don't need an entire OS just to set a hard drive parameter.

The other option is to analyse the PCB to see if the motor controller is a distinct chip (rather than being built into the main CPU). If it is, you can probably modify the PCB so that the motor is always on.

>>1028541
Yep. A ladder of 20kΩ resistors (10k were available at RadioShack) looped over with a 10k made a nice DAC for playing Mods, demos, and some games and stuff. I used to build them for sale until the Soundblaster dropped to like $10.

>>1028541
>>1034883
Somewhat theoretical question, might be a stupid thing to ask

When does the R–2R ladder break down? Obviously it works when there's a large discrepancy between the values, such as the common 10k/20k.

What about 100/200? 10/20? 1/2? 0.1/0.2?

>>1034886
In theory never
In practice resistors have tolerances

>>1034886
In theory, the absolute value doesn't matter, only the ratios.

In practice, if the values are too small, parasitic resistance in series (i.e. resistance of IC outputs, PCB tracks, solder joints, etc) will affect the result. If the values are too large, parasitic resistance in parallel (i.e. conductivity of air, PCB substrate, solder mask, etc) will affect the result.

>>1034902
Tolerance isn't an issue, as it's a proportion of the resistor's value. So 1% tolerance turns an R-2R ladder into a 0.99R-2.02R or 1.01R-1.98R ladder. But that's true regardless of whether R is 1 ohm or 1 megohm.

looking to get a new multimeter, looking at the reed instruments r5007, and the holdpeak 770e. which is better quality and accuracy wise?

>>1034886
Resistor tolerances have been mentioned, but another matter specifically related to the values are the resistances of your digital outputs feeding the DAC. This resistance is typically in range 10-100 ohms. It varies with temperature and supply voltage, it is usually different when driving high instead of low and there will be small differences between outputs.

A typical output resistance value for ATMega168 running at 5V is 25 ohms. If you aren't trying to compensate it (trim the 2R resistors 25 ohms down) and you are driving a 8b R-2R DAC, then a simplistically calculated lower limit for R would be 25 x 128 = 3200 ohms for 1/2LSb precision. Better, more careful calculation would give you a somewhat different value, but the order of magnitude is this.

>>1035105
Should have been "the lower limit for 2R".
The upper limit comes from leakage currents and leakage resistances, but typically the tolerances of high value resistors are a bigger problem.

>>1035054
https://www.pretzellogix.net/2016/05/13/best-entry-level-auto-ranging-digital-multimeters-dmms-compared-and-reviewed/

>>1035121
ill read that, but the specific ones im looking at are 80-90 bucks so outside the scope of that review

just picked up a very cool phone charger, a tomo v8-4. it runs on 1 to 4 18650 batteries and can charge your phone or charge the batteries inside the unit and they can be swapped so you can carry around spare cells (like phones used to be able to).
since it uses only removable batteries it doesnt have an amp hour rating, so im trying to figure out what it is capable of. since all 4 battery bays are all facing the same direction (pos/neg all next to each other on the same side) im assuming that theyre in parallel and inside has some boost circuitry to up the voltage from 3.7 to 5, also because it can function with only one cell. which would mean it would add up the individual batteries mah ratings correct? is there any way to easily tell if this is the case?

>>1035660
> it would add up the individual batteries mah ratings correct?
Correct.

But boosting from 3.7V to 5V will reduce the rating by the same proportion, i.e. 1000mAh@3.7V = 3.7Wh (=13320 J) = 740mAh@5V, minus conversion losses.

>>1035688
so if the average quality chinesium battery is about 2000 mah times 4, so 8000 mah@3.7v, which translates to about 5920 mah@5v, and assuming 85-90% efficiency the approximate capacity is roundabout 5000-5300 mah.
does that sound right?

>>1035653
A few percent accuracy is the best you can expect. And unless you have a bench DMM, how are you going to check it? I'd just get an Extech.

Besides those listed in the OP.

What other books did you guys use to understand the theory behind circuits and their components?

pls respond

>>1036067
fuck off. Or explain why you would rather not use the ones in the op.

Trying to find the hard drive and specs of other chips on a BDP-BX59 Sony "Blu-Ray 3D" box, was wondering if it was on the same device as the player disk drive, so far according to other sony RE forums it has embedded linux, allows for remote firmware updates, and can also be rooted from a USB hack, on the back there is a "powered by Java" logo

>A - Sony CXD90008G-AA processor
found specs for this

>B - Samsung "K4B2G1646C-HCH9"
?

>C - Texas Instruments "22T CRVN 2050G4"
?

>D - Toshiba J86409
?

>* - buses to disk player drive, trying to find out if OS hard disk is on there as well

>>1036074

Did I ever say I am not using them?

>>1036067
how technical do you want to get?

>>1036121

I want to learn theory, solve stuff in circuit diagrams, from basic stuff, to things that involve vectorial calculus.

>>1036154

look up a book called "Introduction to Circuit Analysis", though there's probably a bunch

>>1036154
Horowitz & Hill's "Art of Electronics" will take you very far in circuit analysis and design. It's an awesome book that I use all the time.

If you want to go deeper into device physics, Taur & Ning's "Fundamentals of Modern VLSI Devices" is excellent. If you must go deeper still, Ashcroft & Mermin's book "Solid State Physics" is another gem, but it's not for the feint of heart.

If you want to take an electromagnetism route, I can't help you too much. The respected classical physics standards are good enough - once you get the basics it's just doing harder and harder problems until you kill yourself.

>>1036214
I'd say AoE's strength is specifically the fact that it spends very few pages on circuit analysis, at least on term's normal sense.

>>1036154
I don't really have anything to recommend, but Nilsson's "Electric circuits" likes to analyze shit. It's pretty basic stuff, from batteries to simple opamp circuits. "MIcroelectronic circuits" by Sedra/Smith is another very well known introductory book. It goes farther than Nilsson's book, but has (IIRC) somewhat less focus on analysis.
Basically, see what books and materials your local university uses.

>>1036154
griffiths intro to electrodynamics is a popular E&M textbook but you must know vector calculus very well and be clever.

Sup /ohm/ I made a post on /adv/ and was redirected here. I'm 18 and have to make a choice for majors. I was always fascinated by electricity and how it worked from a young age and I have extremely good maths/physics skills (although I only know calc 2 at this point). Is electrical engineering a good choice for major? Thanks for all input. Live in Australia btw.

>>1036305
Yes. But what do you want to do in life?

>>1036311
I'd like to have a job that challenges me and allows me to solve difficult problems and that I get payed well for. Seems like electrical engineering also gives you awesome skills so you can do /ohm/ stuff in your spare time which is something I'm interested in doing also.

>>1036312
yeah but you'll be competing with pajeet and rui-chen for a job. their designs will be terrible but cheap and your boss won't know the difference. you will be fired and stuck jobless as this is happening to every company - the third world is flooding the planet with cheap shitty electronic "solutions" and engineers. then you'll learn to code an enjoy that sweet $100k+ starting salary.

maybe i'm bitter

>>1036400
>competing with pajeet
Pajeet was supposed to take over all the software jobs even earlier.
Grocery store clerk is your only option.

>>1036400
Wow that's crazy. So would you recommend against majoring it? Or would you say to go through with it but make sure my programming skills are good as a fail-safe?

>>1036441
Most of the world's electronics industry has moved to China. There's still plenty happening elsewhere, the question is whether it's enough to require new blood or whether existing EEs will fill the demand for the foreseeable future.

But then most people end up in a job which bears little resemblance to their degree title. E.g. most physics majors aren't career physicists.

In that regard, EE is a reasonably good degree for any kind of technical career which doesn't strictly require a specific degree. It also has the strongest programming component of anything that isn't actually a "computer programming" degree (CompSci/CompEng/SoftEng).

>>1036441

>>1036449
Doesn't know what he's talking about.
EEs are retiring in droves just like every other baby boomer profession.
6 of my friends just graduated in May with BSEE
5 are employed at: GM, a power company, Uni research, DoD stuff
If you're really concerned with pajeet or ching replacing you go into something that requires citizenship or clearance.

>>1036400
>yeah but you'll be competing with pajeet and rui-chen for a job. their designs will be terrible but cheap and your boss won't know the difference. you will be fired and stuck jobless as this is happening to every company - the third world is flooding the planet with cheap shitty electronic "solutions" and engineers.
you won't be "competing" with those guise.... your company will hire them for one-half to one-third what they're paying you, and then tell you that to get your severance pay you have to train them to do your job

>then you'll learn to code an enjoy that sweet $100k+ starting salary.
maybe but probably not. depends on the country you are in.
in the US, starting programmers is around $50K and ending salaries (retiring) is around $60K.

If you REALLY want to have a job in EE or comp sci, the mot reliable employment left is defense companies, at least in the USA.
You must be a permanent citizen, so no foreigners allowed. BUT,,,, you gotta have a security clearance, so you must live a pretty clean life. No arrests, no driving tickets, no drugs, no shitty credit (!).
And keep a decent and upstanding online presence, cause they check that too now.
You don't want ANYTHING that makes you look bad.

>>1036470
>Doesn't know what he's talking about.
>EEs are retiring in droves just like every other baby boomer profession.
>6 of my friends just graduated in May with BSEE
>5 are employed at: GM, a power company, Uni research, DoD stuff
yea but they will never get the pay+benefits that the retiring guys are getting

in the USA, programming was the first tech profession to be offshored
now it's everything else, all kinds of 'engineers'. US colleges don't talk about it, us news doesn't talk about it, but it's happening.

with product manufacturing, it just makes sense to be able to work closer with the manufacturers--who are mostly now not in the USA

if you know what the factories can do, you can still find things that they can't do. you will be running a one-man show tho

>>1036474
>with product manufacturing, it just makes sense to be able to work closer with the manufacturers--who are mostly now not in the USA
AMD- American
Intel- American
Xilinx- American
Altera- now American
IBM- American
Nvidia -American

There are plenty of manufacturers still here. I don't feel like finding the source for you.
Get out of here with your defeatist misinformation.

H1B's are a fucking problem though, i will yield you that

>>1036474
>baby boomer EEs are retiring
people have been saying this for many years. it's true, but they are not being replaced by american graduates if at all. it's why EEs have such a high unemployment rate compared to other engineering disciplines.

>>1034639
What is the 1V5 line used for?

>>1036563
>EEs
>high unemployment rate
That's why fewer EE degrees are being awarded than Mech (first time in at least like, 25 years), although EE has a larger employed body than most Engineers (like 316k vs MechEs 278k). Even if you take the 0% job growth for EE as true, the 14.6k new ME jobs would STILL be almost 25k less than EEs.

EE is one of the most employable degrees, as you can go into a shitton of different areas (power distribution, programming/hardware, research for quantum/electromagnetic related shit), and don't try to tell me internet of shit isn't going to continue growing while somehow vehicular testing is.

>>1034864
oy vey delete this

>>1036519
IBM is the only one of those manufacturing "products" (as opposed to components). And IBM's business is as much services as hardware. None of the others have any significant demand for people whose skill set isn't specifically digital IC design.

There is still a fair amount of EE happening in the US (and the West generally), but most of it is making stuff you don't notice for companies you've never heard of. The high-volume, low-margin stuff targeted at consumers has gone overseas, but that's far from the entire industry.

For products which contain electronics but which aren't specifically consumer electronic products, the manufacturer is likely to look closer to home for design (even if the PCBs end up being manufactured in China).

>>1036563
Many boomers are working longer / not retiring. The place I work at is very top-heavy, yet hiring is focused on new college grads.

>>1036677
I know a milk bottle producer that drives their 50+ year old works (at like $20/hour+benefit wages) to 60+ hours a week. For months. The young people they make stack the milk bottles for probably an entire shift or something straight, or make them do something else without training and are surprised the kids don't come back. One of their two technical guys on a shift was fired because he was drunk on the job. His job of maintaining the machines, including the power to change the ingredients and furnace settings for the bottles.

So many businesses are going to collapse when their employees just straight up die on the job.

>>1036683
>I know a ______ that drives their 50+ year old workers (at like $20/hour+benefit wages) to 60+ hours a week. For months. The young people they make _________ for probably an entire shift or something straight, or make them do something else without training and are surprised the kids don't come back.
that's everywhere I've worked. well, not 60 hours a week, but 55.

How often do you use your oscilloscope?

>find a tv by the dumpster
>heavy as god himself
>it's a 42" plasma, only analog ports, has some static but otherwise ok picture
>open it up, some bad caps, but since it came up properly it should work fine
>check the specs
>852x480 resolution
>pic related

>>1036887
Depends on what I'm doing. I have a single channel scope, it's ok for examining things but not easily usable (if you want data you have to use this weird usb-serial communication software). I used to to check the voltage level of capacitors (when replacing bad caps with old ones). I've used it as a low-accuracy voltmeter (if my multimeter was busy or too cumbersome). If I was working on radio/analog shit I'd use it all the time, otherwise it's less valuable than an extra multimeter.

>>1036901
What oscilloscope is it, how much did you spend and do you regret buying it?

>>1036901
I should have been clear, I used to the oscilloscope to watch the voltage rise when charging old caps. You don't just want to put old caps into something because the electrolyte dries out or crystallizes or some shit if it's not stressed.

>>1036887
analog circuits? constantly
microcontroller stuff? since it doesn't have storage nor high bandwidth, only as display sometimes. pic related. the graph has 4096 points resolution

>>1036904
https://www.youtube.com/watch?v=aVAgMjQjF6Q

DSO 112, Not my video. I paid $50. It would have been better in the long run to save up for a dual-channel scope with built in memory, but those are like $150+. USB scopes can be pretty cheap, but their software doesn't usually get updated and you obviously can't use the scope without a computer.

I don't regret buying it, but it has the biggest issue: there is technically software for it, but the company is Chinese and never updated. The serial software is ok but there's usually issues with the data. It comes out otherwise clean, but it's low sample rate and single-channel so you're not going to do complicated things with it (goes up to 100khz I think, beyond loses tracking so the signal can't be observed but can be paused/saved). The probes can be replaced, and I got adapters for normal 10x probes. I was watching capacitors charge up to ~170V using a resistor, breadboard, and a full rectifier. It was a little nerve-wracking because I was using mains voltage of course.

The nice thing is the rechargeable battery. I can just keep it in a bag with the probes and if I want to fuck around with it, the battery keeps a charge and I don't have to use a computer or anything.

>>1036914
Here's an example of a graph I made from my scope data. The graph depicts the output of my USB sound adapter ($12 44.1k 16bit adapter). I forgot what the sine wave frequency was, obviously at the limit of the sound chip.

>>1036645

Opps. Going to use 12V to the anode instead of 15V. I should probably change that. It's going to be used to power an IV-15 for the seconds indicator.

>>1036673
Pic related

Can i use radial caps in place of axials? The price difference is pretty big ($3.00 for axial vs 0.80 for same radial), is it just an issue of making sure the leads can reach both sides?

>>1036963
No, no, I meant the 1.5V regulator on the central top of the chart
I don't see any 1V5 line going anywhere

>>1037035
excel related (left is arranged by value, right is arranged by cap #) green caps are ones I was able to find cheap nichicon axials, the orange ones are the more expensive sprague axials.

>>1037035
It's an issue of making sure that the caps actually have the same specs (capacitance, voltage, ESR, ESL, power dissipation).

Other than that, axial vs radial is just a question of what fits better.

>>1037071
ok just making sure. spec-wise the replacements for the 50 year old ones are at-or greater. some of the weird ones like 10uf 3v I bump up to 25v.

>>1037049

Ah! My bad. The 1.5V is actually going to .6V. I'm thinking about making an biasing circuit but I don't know how. Any advice? If I can cut the voltage in half I would be golden.

>>1037081
Thought of a diode voltage dropper already?

>>1034639
I just built my own nixie tube clock so I feel like I have to comment. I'm not well versed on the microcontroller you're using but if the unterminated pins are outputs I would suggest using them to drive the K155ID1's individually instead of multiplexing them through the shift registers. I'm also not well versed In C. I programmed mine in assembly. So this suggestion may end up being harder than using the shift registers, especially is you're using a pre-written library.
My best guess is the 1nF capacitor is to hold the logic level long enough for the chip to read it.
I don't know what tubes you're using but 200 V seems slightly high. You could probably get by with a little less and increase the lifespan of your tubes.
Not sure what the seconds tube is. Some kind of dekatron maybe?
It's always a good idea to put decoupling capacitors from VCC to ground on your IC's. Odds are you won't need them but for the small amount of money they cost it's worth the extra protection.
Are you including a timing algorithm in the microcontroller or is that the duty of one of the other chips?
Good luck. I hope you have as much fun as I did.

>>1037083

Nope. I am an electronics noob. Have any good recommendation for me?

>>1037088
Well, how much of a noob are you?

>>1034639
I notice there is no input besides the reset switch, how do you plan to set/change the time?

>>1037084

Post pics; I love all things Nixie.I am programming it in C++ because I don't like assembly and I'm not looking to get the bleeding edge of speed and save that much space.

The shift registers are absolutely needed. If I didn't have them I would need 16 digital/analog pins and I just don't have enough of the Amtel to get the job done. Yeah, and since I have a library for the shift register and the real time clock I'm going to stick with that.

The power supply I'm using will supply up to 220V which the tubes can handle easily, but I'm probably going to run it around 180-200V. The second tube is some sort of indicator tube. Pic related.

I'm adding decoupling capacitors as suggested right now. I'm going to post an updated schematic sometime tomorrow.

>>1037090

On a scale from 1-10 with electronics probably a 3.

>>1037096

I'm putting in an IR port. I had a lot of success with that (and they're cheap) with my LED stuff.

>>1037099
IR sounds nice and cheap, are you going to use a standard remote or mobile/custom signal?

>>1037102

For my LEDs I just used a spare JVC entertainment system remote I had from a junked system. I have about 3 remotes that are just garbage at the moment.

>>1037103
Since you have experience with that it sounds pretty nice, have fun and don't forget to post any pic!

>>1036078
D is almost surely a standard NAND flash chip which would be the primary OS storage device. The form factor for those chips is standardized and they're all pretty much the same aside from capacity. I'm not sure if by "hard drive" you mean an actual hard drive which clearly there is none in that picture, but that would be the closest thing. B are probably RAM. C looks like some kind of EPROM, assuming that's a UV window. I also see a few 8 pin chips which could be SPI flash and contain the bootloader and/or kernel but it could also boot directly from the NAND.

>>1036887
>>1036904
Rigol DS1054Z, 50MHz, 4 channel DSO 1G/s (250M/s with 3-4 channels active) I hacked it to 100MHz + extra triggering and decode options.

399USD, which for what it is, is a great deal.

I use it basically daily.

>>1037097
Here ya go. Ignore the extra stuff. All components are contained within the PCB. This was my first attempt and it's basically a prototype. I would have made some changes before ordering the PCB but it had to be done before the end of the Summer semester. I used it as a school project. I plan to redesign it and get it to around half the size it is now.

>>1037118

> tfw tek tds360 300mhz 1gs/s 200 bucks shipped

>>1037145
>tek tds360 Nice! (I assume you mean 400MHz model)
but…
>2 Channels
>1K Record Length
Not so nice :( Also as far as I can tell no segmented memory (though I admit the Rigol 1000's implementation leaves a bit to be desired) or intensity graded display.

It is a nice scope, but it hasn't aged well.

Hey guys. Never knew about this thread, but I like it. Electronics are the shit. I've got a few questions for you.

How much should I expect to spend on basic lab equipment like a scope, function generator, and variable voltage supply?

This one's more in depth. How can I use an audio waveform to control LEDs without a microcontroller? I've got an active LPF planned to filter everything but bass, with the intent that a strip of LEDs will briefly turn off on bass notes. I'm planning on running the LED supply through a transistor which will be controlled by the filtered audio.

What I'm worried about right now is that the peaks are too quick to set them as a threshold, because if I did that the lights would flicker too briefly. If I don't have a high threshold, they'll turn on and off too rapidly, whereas I want it to be more like one cycle per bass note. What's the alternative? I was thinking I could use a 555 and have it output for a little while after a high peak, covering a typical length of a beat, but that seems like overkill. Maybe something simple with a capacitor and diode or something? I'm not sure.

As an example, this is the waveform of Kanye West's "Fade" around 33 seconds. Check it out on youtube or whatever and skip to that time - it's very bass-heavy, but the waveform is still too complicated to simplify have it gate a transistor. The lights would be flickering too quickly to even see, instead of a nice pulse on each beat.

>>1037175
Final question that I forgot to put in there, what's a good program for creating schematics that allows you to quickly define the pins of an IC?

>>1037175
you can try a peak detector with a resistor parallel to the capacitor. the resistance defines the falloff speed

>>1037084
> My best guess is the 1nF capacitor is to hold the logic level long enough for the chip to read it.
The 1nF is a low-pass filter so that noise spikes don't become clock pulses.

>>1037180
Ah, thanks. That sounds good.

Would this work? I was looking up peak detectors and found this on stackexchange. The only issue is that may be too slow to charge. It should be quick with no resistor in one direction, but I don't know enough about acoustics to be sure that's it quick enough.

I don't really care much about the voltage drop across the diode, as this circuit doesn't have to be at all precise. But if it would be significantly better with op-amps, I'll find another.

>>1037186
Resistor goes across the capacitor not the diode.

>>1037196
What's the difference? The cap discharges through the resistor either way. Is it just convention or am I missing something?

I need a modern replacement for 2N5303 in a sane/friendly (not TO-3) package.
Please help.

>>1037200
To begin with, the circuit shown can produce negative output voltages. If the resistor were in parallel with the capacitor, this wouldn't happen.

>>1037209
The only thing Mouser is showing for 2n5303 is the TO-3 package. I'm guessing you're building something from a schematic instead of replacing a faulty transistor. If you can use a MOSFET the IRF730 will probably be a suitable replacement. Post a little more info and somebody might be able to give you some better advice.

>>1037175
>"basic" lap equipment
$500+ to start. The oscilloscope would be the most necessary, it'd be $200+. Function generator would probably be another $200+, and the multimeter might be the most expensive.

If you're not going radio, you can go much cheaper. For $150 or less you can get portable dual-channel oscilloscopes, new little gadgets. USB scopes are less than $100 and look good (easy data collection). You can get stupidly cheap function generators off ebay, little gizmos for square, sawtooth, sine, maybe triangle. Unless you need 6-digit sub-microamp/millivolt precision, you don't need a lab-quality multimeter.

Can I replace a 5uF 25v with a 10uF or that too much of a jump

>>1037336
Depends on what it's used for.

>>1037352
There's 7 of them. They are on the amp and driver board for the speakers of the radio unit

>>1037357
Got the schematics?

>>1037359
Yeah I'll post them in a second >>1037057 is my excel sheet. I labeled each cap with where along the line it is. The reason I ask is because 10uf caps are cheap nichicons under $1 ea but 5uf are expensive $3 ones

>>1029862
But anon, I learned the parts/footprint library once and created my own parts. Then I added my library and a couple other libraries from github to kicad's global search paths and they are there every time I start a new project. You're right that your way of starting a new project is fucked :/

You got me on the clunky via stitching but as a workaround, you can make a footprint with a single plated through hole and then plop that down wherever you want on the board.

Overall it's really good software IMO. Way better than eagle or geda.

>>1033197
I bet if you chop the keypad off the board at that 8 pin connector you'll find that the keypad is a 4x4 switch matrix. You could interface it with an arduino or whatever.

Here's the datasheet for one I just bought. I bet yours is the same or EXTREMELY similar.
https://media.digikey.com/pdf/Data%20Sheets/Grayhill%20PDFs/96%20Series.pdf

>>1037372
If you put two 10uF caps in series that's equivalent to a 5uF cap.

>>1037209
> I need a modern replacement for 2N5303
You can't determine a replacement based solely off the part number. You need to figure out the requirements and choose an alternative based upon those.

>>1037123

Noice. I want to design my own power supply but I don't have an oscilloscope so I don't want to attempt it without one.

How did you print on your solder mask? I haven't really looked that up yet.

>>1037501
I was originally going to use the machine at my university to mill the board out of a copper sheet but it was too far out of tolerance and wouldn't do plated through holes. I sent the files to a company called Oshpark and they mailed me the board. Not sure what their process is but they did a great job.

I bought something on digikey. It was an evaluation kit for a neural network ASIC called the CM1K.

They didn't have any in stock, but I could still order it. Did I fuck up? I have no clue when to expect it to arrive

>>1037503

I was thinking about having an outside company make my board, but I kind of want to take it from design to production all on my own.

>>1037515
It probably says somewhere on their website how long it will take them to get stock.

What do you search for when you want to buy breadboard jumper wire?

"Make: Electronics" is telling me I really ought to use spools of single-core 24 AWG in a variety of colors that I cut to length and strip myself. Having tried to work with flexible dupont-plug jumpers and precut single core, this makes total sense to me.

Yet when I search on ebay or aliexpress, I get few results for 24 AWG single-core wire. The results that look appropriate seem to be American suppliers with typical high prices and absurd shipping costs.

30 AWG for wirewrap seems to be commonly available at good prices. That seems good for jumpers on universal PCBs (and of course, for wire-wrap), but I expect it's too thin for breadboards. Does that seem like an accurate assessment?

>>1037515

There should be a lead time on the product page. Otherwise, manufacturer standard lead time is usually 6 weeks or thereabouts.

>>1037426
>>1037359

theres the schematics, highlighted in red. C1-27 are electrolytics.

the point of my question is if 10uf is an ok subsititue for 5uf or do I have to stay with the 5uF

This is me: >>1037524
>"Make: Electronics" is telling me I really ought to use spools of single-core 24 AWG
I'm an idiot. I meant 22 AWG.

Even so, it's hard to find the correct wire in a variety of colors at a good price.

>>1037540
30 ft

>>1037536
For C13, C15, C16, C21, C23, C24, 10uF should be fine. They're DC-blocking caps with a centre frequency of a few Hz.

Note that C19 and C27 are specifically marked as non-polarized, so you can't use normal electrolytics there.

>>1037541
https://www.amazon.com/Elenco-Hook-Up-Colors-dispenser-WK-106/dp/B008L3QJAS/ref=sr_1_1?ie=UTF8&qid=1471041897&sr=8-1&keywords=hookup+wire

>>1037545
Right 19 and 27 are NP for the crossover circuit. But for C7 in the AM circuit, I should leave it at 5uF?

I'm just trying to consolidate as many caps as I can, right now to recap this radio will cost me almost as much as i bought it for, due to the expensive sprague axial caps

>>1037549
Ah, thanks.

Look at the bullshit we have to put up with in Canada:
https://www.amazon.ca/Elenco-Hook-Up-Colors-Dispenser-WK-106/dp/B008L3QJAS

Over 50% higher price compared to America, after the exchange rate, and that's probably the best price to be had domestically in Canada. You can literally buy stuff at retail in the USA, truck it back to Canada, and have a healthy mark-up at the typical prices in Canada. This is totally common. Makes no fucking sense.

I don't want to pay $35 for some breadboard wires. I just want a few short coils of this stuff:
http://www.aliexpress.com/item/Free-Shipping-140pcs-in-one-package-convenient-New-Solderless-Flexible-Breadboard-Jumper-wires-Cables-HOT-Sale/1621833882.html

>>1036067

This >>1036255

Also look around here:
http://4chan-science.wikia.com/wiki/Electrical_and_Electronics_Engineering

You can DL books from here:
http://gen.lib.rus.ec/
This links to bookzz too, but you can directly dl from gen.lib too, which you can also use: http://bookzz.org/

So I recently bought a bunch of CASIO pocket TVs to use as a (shitty) terminal for my NTC C.H.I.P.S.

I selected a EV-500 because it looks like it has the most scanlines.
Problem is tough either I can't identify the video signal pin on the bareboard at the A/V connector or the unit itself is fubar.

Found a service manual (https://a.pomf.cat/jasile.pdf)) tough, thankfully.

Ultimately I also want to shoehorn a keypad matrix and a joystick into it.
Since there's only one USB on the C.H.I.P, I would go with IIC since I already have a few GPIO ICs for that.
On the other hand UART would also seem like a good choice, especially since it's the most trivial serial there is.

What else would there be to consider?

>>1037550
C7 could also be 10uF.

http://www.ti.com/lit/ds/symlink/cd4017b.pdf

Does this mean that the output doesn't have a definite max current? I want to have around 300mA max at 12V.

>>1037660
thanks m8. gonna order these soon

>>1037661
The short answer is no.

>>1037661
Yes, it guarantees only a certain minimum value.
Sure as hell it won't supply 300mA.

>>1037674
Is there any chance at finding a counter or multiplexer that will or will I need to use MOSFETs?

>>1037675
If it's 40xx or 74xx series, no.
There might be others, but at least I'm not aware of decade counters capable of driving 300mA. There are some shift registers with high current outputs though, like ST's L9822.

>>1037681
Maybe anon could couple the shift register to a 2000-series darlington array to drive heavy loads?

>>1037689
Well, he can do the same with 4017 as well.
I assumed he wanted a single-chip solution.

>>1037681
Shift register sounds like it'll work. I just need to output on one channel at a time and increment the channel with a square wave so I think that hits the nail on the head. I'm too tired to pore through datasheets right now but if it's easy to set the channels initially then this hits the nail right on the head.

>>1037610
Nevermind that, I gutted the unit after trying at various testpoints with an NES for signal. Useless!
Can't read squat on those 90's LCDs anyway.

Still considering a TV-970 for a gameboy-type device, since for emulating gameboy and SNES games it don't matter anyway.

>>1037693
>Shift registers
I actually meant whatever logic ic was needed in my post, but I can vouch for shift registers, they are great!
I recommend the classic 595er, since you can latch your outputs and also daisy chain them nicely.

I'm going to start using a large (48v) li-ion battery with an e-bike. I plan on being present while charging but am worried about storing it in the house. I am thinking about pouring a concrete box with lid to keep it in while i'm away. Is this a good idea? Are there alternatives?

>>1037977
Don't worry about that. Even the Chinese 48V models do have multiple safety precautions built in. From simple bimetal thermal switches to temperature sensors and current/voltage limiting.

The low-quality packs however are not recommended for a different reason.
Cheap li-ion packs are prone to failure, coupled with a crapshoot mains adapter you won't get many charge cycles out of your system.

I'd recommend buying a quality LiFePO4 pack as it is heavy-duty on the electrochemical level, being tolerant to high discharge, having a great power density and great loading cycle life.

Also by putting it into a box you basically insulate the whole pack even more, which can lead to meltdown in the worst possible scenario.
Just trust your local retailer of choice and stick to the manual, buy a smoke detector if you need to.
There's nothing to worry about.

>>1031706
Thanks. I got around putting the lights in a week ago. Everything seems to be working just fine now. I've been using it quite a bit, and their have been no problems yet.

I'm having this really bizarre problem. I have a chipkit DP32 which comes with a PIC32MX250F128B I'm trying to learn to program. I wanted to try using the chip standalone so I programmed it with blink program that lights an LED. Hooked it up according to the data sheet minimum recommendations and the I/O I was using. Anyway the LED wouldn't light.

However as I was using my nixie tube multimeter the LED suddenly began to blink. I've since discovered that after applying the 3.3V to the chip if I cycle my multimeter on and off 2-4 times the LED will always start to blink. This is weird as fuck and I have no idea why it's happening. Any insight?

See my shitty webm

>>1038271
Quick question, did you connect all supply pins?
Because it sounds like you didn't.

>>1038273
I have:

MCLR PIN1: to 3.3V through resistors
Vss PIN8: to GND
Vdd PIN13: to 3.3V
Vbus: PIN15: to 3.3V
Vss PIN19: to GND
Vcap PIN20: to GND through 2.2uF cap
Vusb3v3 PIN 23: to 3.3V
AVss PIN 27: to GND
AVdd PIN 28: to 3.3V

120nF decoupling caps between for all the rest cause I didn't have 100nF readily available. Resistor network on MCLR is identical to that pictured except 120nF cap.

>>1038276
addendum:

Vbus is actually floating, not 3.3V. My mistake.

>>1038276
Well I'm drunk now, so take my posts with salt.

Looks like the connections should be good,
initially I was suspecting shoddy power rail connections because the MCU would most likely meet power requirements through the pin protection circuitry.

Did you try applying a battery pack or floating supply?

>>1038279
ad: since you mentioned vbus, did you try to debug? maybe it's stuck in an interrupt handle or some sht,

>>1038279
DC switched mode PSU. 30V 5A. Set to constant voltage 3.3V. Shouldn't have an issues with power. It should be regulated. I probably have a few adjustable regulators lying around but I can't imagine it'd really make a difference.

>>1038280
I'm still new to the software side of things. I haven't used the debugger yet. I will say that the thing works fine when it's on the chipkit board. When I hook it up to the usb it automatically turns on and executes the code no problem.

>>1038280
It has something to do with the OSC1 PIN. When I connected my oscilloscope probe to it the circuit would suddenly work whenever power was applied. Putting the probe on OSC2 had no effect.

I thought I'd configured the chip to use the internal oscillator since I don't have any crystals on hand. I either didn't do that right or I was supposed to do something with the unused OSC inputs instead of leaving them floating. I'm not really sure.

Not sure what turning the multimeter on and off did to start the chip working. I know the nixie tubes have high strike voltages and I don't know if that can induce current in the micro that started the internal oscillator or something. Weird as fuck.

Not even sure why the scope probe on the OSC1 pin got it working either. Worked on 1x and 10x.

>>1034652
shouldn't matter much.
the nixies internal impedance and the 10k or so series resistors drops all of the voltage

i've driven nixies with ul3803s no problem

>>1028473

new thread
>>1038379
>>1038379
>>1038379
>>1038379

>>1038184
Thanks.

>>1030314
Time to learn the ropes in ModelSIM. Questa too if you bother pirating