/diy/ - Do It Yourself

This right here I need to go from sata to usb, only data no power. What's the circuit within that converter. is it easy to build for someone with sound basic knowledge in microelectronics?

>>1937869
It is probably not easy. It's not so much a circuit as it is a microcontroller, ASIC, or FPGA or something like that. The magic is in the software.
It would be better to buy a cheap one and then hack it up as necessary to meet your needs.

>>1937869
USB is a somewhat high-end communication protocol, and so is SATA. Converting from one to the other could probably be done with a relatively modern MCU. Converting from one to the other quickly would be a very intensive process, requiring either a dedicated data conversion IC or an FPGA.
If such a dedicated IC exists, and even if you can buy one and design a circuit with it, the PCB design will also be somewhat of a hurdle, requiring length-matched pairs and other high-frequency considerations, on a 4+ layer board. This would be an intermediate project.
If you need an FPGA (or fast MCU) to do it, then it will be even more difficult. This would be an advanced project.

>sound basic knowledge in microelectronics
I'd emphasise either a particularly broad and deep knowledge or significant experience in the subfield, preferably both.

>>1937859
There are LED driving calculators out there, but they basically work under the following principles.
LEDs have a nearly constant voltage drop, meaning a small increase in voltage will produce a large increase in current. An increase in temperature will also reduce this voltage drop. Hence, if we drive an LED with a constant voltage, as the LED heats up from operation it will draw more and more current until it explodes. Hence, we need to limit the current. By assuming the forward voltage to be constant (~3V for blue or white, ~2V for red, etc.), and using the recommended forward current (20mA should be safe for 90% of 5mm or 3mm LEDs, but 15mA might be better for those near/above 3V), we can calculate how much voltage needs to be dropped across the rest of the circuit and hence what resistance it needs.
>V_r = Vcc - Vled
>R = V_r / I
>R = (Vcc - Vled) / I
Though personally I'd lean towards using something like an LM317 as a constant-current driver.
But this simple picture only applies to a single LED, not many parallel LEDs with one resistor. If one LED has a forward voltage of 3V and the other has a forward voltage of 3.1V, and they're in parallel, then the 3V LED will run a lot brighter than the 3.1V LED. Because it will be pulling more current, it's highly likely that it gets more current than it can handle, as the resistor is specced to let two LED forward currents through. Again, this blows up the LED.
In reality, whether or not you can put LEDs in parallel without their own current limiting depends on how well matched they are and how close you're running them to their operative limit. Having the LEDs thermally coupled to each other may also help. It's common to see LEDs in parallel on COB LEDs with just one current limiter, while the parallel LEDs on an LED strip each have their own resistor. A single bag of chinky LEDs are probably all from the same batch and should be good enough to parallel without individual resistors.

>>1937931
cont.
But having multiple resistors can also help spread the load.
For example, if you've got 12 LEDs, each requiring 20mA at 3V, then that's 2V dropped across the resistor and 240mA through it. 8.3Ω or larger btw. That's 480mW of power for just one resistor, which is beyond the limit of the common garden 1/4W resistor (or 1/8 if you've got those tiny ones). A 1/2W resistor will handle that but still get relatively hot.
So personally I'd split it into 6x 50Ω 1/4W resistors, 4x 33.3Ω 1/4W resistors (kinda toasty), or 3x 25Ω 1/2W resistors. Distributing the resistors will also help prevent total cascading failures from destroying every single LED, assuming that's a possibility.

On the other hand, if using an LM317-based constant-current supply you'd be able to handle all that current with a single TO-220 package with relative ease, though I'm not sure how much of a beating the current-setting resistor will be under.
There's also other options that will make your current-sinking adjustable, which maybe a TL431 circuit can do. Or just any old op-amp + resistors + pass BJT + potentiometer.

>>1937846
Was that someone's idea of a simple VCO?

Building up my ute ('straya mate) to go camping and whatnot.

I've gonna put a solar panel on the roof, 300w, with one of these bad bois(pic) to manage it. It's going to be connected to a 190AH AGM battery, should be great.


Anyway, I had a thought on how to charge it while I'm driving just in case I don't get enough sun or whatever. Just want to hear what you guys thing if it will work or not.


If I get a CCCV boost board ($10-20) and connect that to switched power of the ute, and set the voltage to something like 25v. and wire that into the solar input, would that work okay? It might only pull 50-100w but that's better than nothing imo.


If that works, I also had an idea to use a break relay ( I think they are called), wire that in parallel with the panel to the controller so when the sun is out, the relay is open stopping the car charger current, when the sun is down the car is free to charge.


What do you guys think, could this work or am I crazy?

(I am going to fuse and breaker everything, gotta be safe frens)

Do you guys know anything about this kind of LED alarm clocks? The LED is too bright so I am trying to add a dimmer pot. I thought I would have access to all the segments, and the pin markings sound promising SEG1 SEG2 etc. but looks like it is just controls, a button board is connected to it. So looks I need to remove the board and take a look whats behind it? I am guess there may be some micro or some other controller board underneath that drives the segments?

>>1938133
Your LEDS are probably multiplexed so that they have a common ground (VSS). Stick a pot on that line.

>>1938133
>So looks I need to remove the board and take a look whats behind it?

you see how the display is attached? plastic is melted to act like a rivet. when you cut those off, you cant really reuse them after. so, it's gonna be wonky, you'll need to glue or tape it back in place.

i'd suggest you get some colored plastic <whatever> and put it in front of the display.
(i have some 200 sheets of red acetate, so come by the house and i'll give you some.)

>>1937996
just the saw generator, not the whole thing dummy. you can even skip the bjt matching/tempco if you want

I’m having some problems with a PCB I made. I’m trying add USB capability to the board using a FT230XS-R (USB to basic UART IC). I’m sending bytes from an MCU’s TX pin to the FT230X TXD pin, however I’m noticing some strange behavior. Namely, that I’m picking up a signal when probing the RESET pin as well as the 3V3 pin. I’m getting a signal out of USBDP, but not out of USBDM. The two LED indicators I’m using on the TX and RX CBUS pins are also very faintly lit, for whatever that’s worth. Could anyone provide insight into this problem? What could be some potential causes? I’ve checked my soldering and it looks okay, but don’t feel it’s appropriate to rule anything out. I’ve never implemented this before, and my background is not in embedded, so apologies for the amateurish question. Any help would be greatly appreciated. Thank you very much in advance

>>1937869
I disassembled one some time ago.
It used the chip SPIF301 LQFP-48 pins and not a lot of passive components.
As said before, the differential pairs for high speed data lines should be the hardest part of the design.
Maybe there is an OSH design out there? I think that is a kinda common project.

>>1938032
>If I get a CCCV boost board ($10-20) and connect that
you dont need anything like that ...you connect the charger to a voltage source(car battery) to the battery you want to charge(the solar battery) and connect the solar cell and it automatically charges your battery from a voltage source if you dont get enough juice from the panels (and shares load also)

all you have to do is connect the wires going into the voltage input over a relay thats switched when you turn on the car

>>1938179
> I’m sending bytes from an MCU’s TX pin to the FT230X TXD pin,
TX connects to RX
RX connects to TX
Pin names are what they do, not what the connect to.
> The two LED indicators I’m using on the TX and RX CBUS pins are also very faintly lit, for whatever that’s worth
Not too much, what's the voltage, and what should it be?

>>1938179
first of all all your decoupling caps and caps on high speed lines i would exchange for ceramics because they are more stable and have lower series resistance
also it is crucial to have the decoupling caps as close as possible to the IC-s

it sounds like yo ueither have noisy voltage lines and is causing interference and ghostin , or you have bugs in your code

>>1938185
Are you telling me that the MCU's TX pin connects to RXD on the FT230X?? I basically copied this schematic from elsewhere so I'm pretty shocked to discover that, if that's the case. I thought the 'D' in TXD stood for "data" and so that's where the TX signal should be connected.

I'll scope the LEDs and find out - not too important for now I suppose.

>>1938188
I don't think I have noisy voltage lines, I've scoped my two power nets and I get a pretty stable 5V/ 12V on both of them. As for bugs in my code, I have a TAG connect on the PCB and can see serial data coming in on a USB port, this is exactly what's going into the FT230X as well

>>1938194
usually you connect a Tx of an output to a Rx of an input since Tx means transmit and Rx means recieve ...but it should say in the IC-s datasheet what kind of signal it expects

>>1938202

>>1937846
is wood a good place to put your electronic projects in? I don't see the problem if it blocks you from grounding the circuit, it's better than metal

>>1938240

wood is mostly terrible coz it needs to be heavy and thick. if you want to install pots or switches, for example, the shaft is like 1/4-inch long, so you need a thin material to be able to bolt it on.

plastic is light but not sturdy. metal is best of all worlds.

>>1938240
as good as any i suppose ..the biggest problem for most uses is that wood is a bad heat conductor so the circuit might overheat if its not designed well

>>1938194
maybe your MCU is multiplexed and you swapped TX/RX in firmware to make it work, donno, but yes, as
>>1938202
and
>>1938209
noted, TX[D] is data OUT and RX[D] is data IN.

this is why i prefer miso and mosi

I was looking in the datasheet of a DAC0800 (a current output DAC) and saw this application.

Why does it use an amp instead of a regular load resistor? I remember reading about it somewhere, but I can't remember where, and I can't find it when I google.

>>1938345
Fuck, forgot image.

>>1938345
Presumably you'd use the amp to get a voltage relative to ground rather than a differential voltage.

>>1938138
no, anon, thats just button inputs to adjust settings etc., the display works fine with this cable disconnected.

>>1938149
>you see how the display is attached? plastic is melted
yeah, i will probably use some tiny self tapping screws.

>i'd suggest you get some colored plastic
heh that was my first thought when i first opened it. but i wanted to fix it "electronically" even though i agree it is probably the easiest and the most practical approach. but who cares about practical.

>i have some 200 sheets of red acetate
why thank you anon, the film on this one is blue so a red or two over blue would make a nice shade of purple. but i honestly would prefer replacing it with red, it probably wouldn't even be as annoyingly bright. i hate blue leds. But unfortunately all leds are also blue.

oh my, so i took it apart. holy shit, it is not a 7 segment thing, just a bunch of literal LEDs inside a plastic case with hollow slots shaped as segments. sounds obvious but i didn't even think about it. they are driven by a 48(?) channel driver( common anode, luckily). Luckily VDD and Vled are two separate lines. I think I found the Vled input. Anons, take a look at Q3, it is not a voltage regulator, it is a transistor, right? it outputs 2.5v to the LEDs' anodes but one of its inputs is connected to a micro pin via a resistor (the cyan square). So maybe thats how they implemented the two brightness levels. Perhaps the emitter is pulled to the ground via the micro pin. There is also some kind of a feedback resistor (120 ohm?) that is connected back to the input of Q3. I wonder if it is so that when Q3 is turned off, 3.3v is fed directly to the LEDs via that resistor??

Also the beefy 2R2 in the red square looks like a current limiting resistor. so maybe I replace that one with the pot? Or I could lift the trace after Q3 and put a pot there.

>>1938385
A pot isn’t that good an idea to carry a potentially significant amount of current, I’d lean towards using a switch and a couple resistors. If the existing resistor is 1k, then replace it with a 10k and switch another 10k and a 3k in parallel with it using an SPDT.

But I might be overreacting, depends on what the current actually is. If it’s 20mA then just go for a pot, hopefully one with a reasonable value.

>>1938416
>If the existing resistor is 1k,
there is none, i think the driver sets the current. i measured it to be 5mA.
there is one 2 ohm resistor on the path before Q3, not sure what its purpose is but i tested it and it seems to be a perfect spot to replace to cut it and replace it with a pot. but now there is suddenly a feature creep. i know want to adjust the brightness manually and also automatically with a photo resistor or a photo transistor. so that it is bright during the day and dimmed at night. and can also be adjusted manually. oh and maybe possibly also a switch to turn the auto mode on/off.

>>1938385
>>1938133
>>1938439
put pic related in line between the power and the leds.

>>1938439
>>1938545
Yes if that 2Ω resistor is a current sense resistor, then increasing its value will decrease the current. A 100Ω resistor shouldn't be too hard to find.

>2W
Is that necessary? Assuming 2Ω = 20mA, then the voltage across it will be 0.04V, which when evenly spread between the 2Ω resistor and the pot (at 2Ω) for maximum power wasted in the pot, will make it 200µW of power maximum. So the potentiometer could be much smaller. Unless you picked that size because it's multiturn or something. To which I say: lmao too expensive, guy is probably fine with 6 radians.

Also what does that transistor circuit do? I think I've seen similar things all over the place, including maybe inside op-amps.

>>1938551
>Also what does that transistor circuit do? I think I've seen similar things all over the place, including maybe inside op-amps.

when load current rises to a certain level, it generates enough voltage across the BE junction of T2 to turn it on. when that happens, it sucks current away from the base of T1 to shut it down somewhat, decreasing the current through T1 and the load. then it either starts to oscillate (bad), or it reaches an equilibrium point where the current becomes limited by the sense resistor (good).

>>1938579
So, the output current is hFE times the sense resistor current?
I that's the case then my power calculations might be off.

I assume it's more stable than a current-mirror? Less wasteful at the least.

Laptop motherboard burned out. I want to use the display for anything else. How do I figure out which $25 display controller to buy off Amazon? Are there display controllers which will let it be hooked up to a DSI connection, or only HDMI?

>>1938587 again
Related question, anyone know if ASUS laptops tend to use compatible displays with each other? Maybe I could just plug it into the other laptop I have which has a broken screen. They're different display resolutions, though, would a laptop autodetect that?

>>1938183
MPPT charger need more voltage than the battery to work efficiently, I don't really think 14.2 volts will be enough for it to work?

also current limiting would be nice to not have it kill my oem alternator

>>1937846
noob tinkerer here . i have a question. say i am making a tesla coil there are many approaches . from what i can tell from tesla coil implementations for the driver and voltage and amperes and taser circuits that use 2 transformers. a transformer is just a inductor with at least 2 sets of windings . a tesla coil is just a type of transformer. all the secondary knows is the strength of the magnetic field. not the volts or amps. generators use wire coils and permanent magnets with rotation applied to the magnets to make voltages for instance.

however charging a capacitor with a anemic single transformer 15kv arc lighter will not drive a transformer . it wont even make a neodynium magnet move when placed next to a primary. a dual transfromer arc lighter at 15 kv will. both have to charge a capacitor. so the input current matters for a capacitor? i knew input voltage does. charge a 200 uf 400 v cap with a 9 volt and its a joke. do it with a boost inverter and you might hurt yourself if your not careful. is the same true for primary coils? relays that use higher voltages tend to have thin wire for low input current. i have ever found a relay with thick wire but from what i can tell 3.7 v 4 a is way better at making a magnetic field than say 200 v 0.01 a or even 15 kv 0.0001 a

if its all about Q value. the power not the volts or amps individually will two 3.7 v 4 a batteries dumped in a 3 pin schotkey with 3 mini ferite beads on the leads to ensure they are in parallel out perform 30 v 4 a ? for reference the wonders of wtf is happening and why it do that youtube links to compare how low volt high current out perform their competitors some times

30 v 4 a SSTC
https://www.youtube.com/watch?v=lPHHvVTVbwM

russian slayer exciter 230v ? a

https://www.youtube.com/watch?v=Im6iXEeVaww

arc lighter driven
https://www.youtube.com/watch?v=4GRyxuRIPPo

>>1938605
The point is to drive the transformer resonantly. Just driving a transformer at any old frequency won't give you the result you want. Note that a slayer-exciter works by having a parallel capacitance from the end of the secondary to GND, which makes up the C of the resonant LC circuit.

>>1938587

it's silly to spend time and money on making a monitor from a laptop display when you can get monitors of the same size for $6 to $10 at the thrift store.

>>1938621
Those are frequently less than 1080p, if he's got a 1080p laptop screen or better I think it's still good value to purchase a display controller.
Laptop monitors are also really lightweight, meaning you could make what's basically a wired tablet. Batteries only make shit heavier.

>>1938607
resonate frequency only accounts for some of it and i was waiting to post any more vids or questions until some one responded.

im absolutely sure rimstar did the math as did greatscott (this guy does math on shit nobody does they just eyeball). both have pathetic tesla coils and this also allows me to show a h bridge tesla and a tiny spark gap.

so im pretty sure that while the frequency is important the different in one you would see in a class room and one that you use in a competition is q factor. so assuming all the same frequency and primary and secondary are used what wins? higher voltage and decent current or just more current to get a stronger magnetic field because the secondary doesnt know maths it just knows a big magnetic field hit it

tiny spark gap which given spark gap its probably 20 kv ??? a

https://www.youtube.com/watch?v=0SaEPW_iMF8

sad tesla coil by a guy with good maths

https://www.youtube.com/watch?v=-CY3aCDc3RY

>>1938675
What, you're that "is high voltage or high current better for a transformer" anon? My answer remains the same, the magnetic core simply doesn't care whether you have 10 turns at 2A or 2000 turns at 10mA. The current you push through your secondary will be determined by the switching hardware, somewhere below 60V is usually the sweet spot for optimum MOSFET efficiency, though with IGBTs it may well be higher.
Either way, I'm pretty sure that if your turns ratio is N, then Vout_max ≤ N*Vin_max, where Vin_max will be determined by the breakdown voltage of your switching transistors. Regardless of Q factor. Q factor is certainly important for building up energy in the tank circuit with relatively low switching currents, but the hard limits on the nature of the transformer remain.

A few deciding factors will determine the operation of a tesla coil, assuming it's working properly.
>voltage: how far it will jump; how strong its electric field will be
>inductance and capacitance of tank: how much energy is stored, and hence how much of an impact it can make when discharging that energy
>power: determines how long it takes for the tank to charge up, and to what degree losses from suboptimal Q are combatted
>frequency: higher makes inductive coupling more efficient (I think) but makes switching less efficient
>wall plug efficiency: determines how much heat you need to dump

Do you understand the engineering decisions behind a circuit like a slayer exciter example? I'd advise that you do some maths yourself to figure out how close they're running to the limitations of their parts.

>>1938363
>differential voltage
Sorry, but I'm no good with this kind of stuff. Shouldn't the current running through the resistor result in a voltage drop across the resistor that would give me +EO anyway? IO to a resistor, IO bar to ground.

>>1938711
The outputs are current sinks; they sink 0-2mA, with IO_bar being complementary (i.e. IO+IO_bar=2mA). The data sheet gives examples of tying them to ground via a 5k resistor to get a voltage from -10V to 0V, and typing them to +10V via a 10K resistor to get a voltage from -10V to +10V. You should be able to get 0V to +10V by tying them to +10V with a 5k resistor, but the zero point will vary depending upon tolerance and temperature. The example with the op-amp will probably be more accurate. With differential outputs, the differential voltage (or in this case current) is usually more accurate than either individual output (i.e. the common-mode voltage/current is relatively unpredictable).

>>1938757
I think I understand. Thanks for the help!

>>1938621
IDGAF and you have no idea what I'm planning on doing with it anyway. If you have nothing useful to post then don't.

>>1938780

your chances of success are not good, maybe 10%, but good luck anyway on your project.

I was reading a hackaday article on how a VTTL (3.3V) gate can link to a TTL (5V) gate but a TTL gate can't link to a VTTL gate because the input signal can't exceed the supply voltage. I thought that the highest TTL standard voltage was Voh at ~2.4V?

Hi. My DC motor for my CNC came, have it hooked to my knob controller but when everything is on there is only one constant speed, and the knob controller does nothing. Why is this? Do I have to get a different controller? I have seen similar ones adjust the speed as intended but mine does not.

>https://www.amazon.com/RioRand-7-80V-Motor-Controller-Switch/dp/B071NQ5G71/ref=sr_1_8?keywords=dc+knob+motor+speed+controller&qid=1584394890&s=hi&sr=1-8#ace-9766277718

>>1938885
What speed controller are you using?

>>1938885
>>1938890
Nvm, I'm retarded. What I meant to ask was how you hooked it up.

>>1938890
my intended speed controller is what I linked. does this not control the speed, and if so why not?

>>1938891
sorry my bad as well.

24V DC to the knob controller on the power end, 2 wire DC motor to the motor end. motor below.
>https://www.amazon.com/gp/product/B07YBXG82R/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

>>1938814
>I thought that the highest TTL standard voltage was Voh at ~2.4V?

naw, nigger, 2.4V is the min voltage considered a high. i.e. a TTL high = 2.4V to 5.5V.

>>1938896
>>1938885
see with oscilloscope if PWM is being controlled;
If not scope - see with multimeter if voltage changes when you rotate knob

>>1938896

so you're sticking 24V into a 12V motor...
pls show drawing of your setup, and pics of any smoke you may notice.

>>1938907
Makes more sense, thanks.

it's about setting up a motor
one minute
>>1938909
Voltage does not change with the knob, though I can change voltage given off on my power supply. Unfortunately this only varies between 23.2 and 28.6 volts.
>>1938915
Attached, no smoke or any signs of failure at the moment, has been standard use. Will this change if I have the motor on for, say, 6-8 hours at a time?

im a zoomer how do I learn to solder and fix electronics? i need money
I'm total tech illiterate

>>1938786
>your chances of success are not good, maybe 10%
Looks like a straightforward project to me, so long as the display controller is available.

>>1938923
If it doesn't overheat after 30 minutes, it should be fine for hours. Even so, I'd avoid running it at maximum power.

>>1939005
try tapping/hitting/jiggling the broken thing
20% of the time it will fix whatever problem its having.

Gimme a gauge for 4VDC, 2.5A, 5-10% loss over 60ft circuit, room temperature. Or better yet give me a table or formula. Don't want to get into the weeds deriving it.

I just tried to google "DC stranded cable ampacity" and there are absolutely zero relevant results. It's all AC tables which include skin effect, proximity effect, etc. so it doesn't apply to DC. Why is google so trash?

Trying to use these nippon America tweeters in a speaker design. I thought I could just run them parallel to 8 ohm main speaker, but doing that they are silent. When I hook them alone to same amp, still silent. Tried hooking them to headphone output to phone; they make noise so not dead.
Aside from hooking resistance across leads what am I missing? Thought main speaker would be enough to make it work now just feel retarded.

>>1939056
https://www.calculator.net/voltage-drop-calculator.html

Plug in your values and vary the gauge until you hit your 5-10% loss mark (0.02-0.04V)

>>1939056
skin effect is practically irrelevant for 60hz unless your conductors are much bigger than say 250kcmil.
proximity effect isn't a thing
etc. isn't a thing
Just use the AC tables if you don't want to think too hard.

>>1939008
would you possibly know why the knob controller does variably control the voltage though according to the chart? videos all show these sorts of setups with the knob controller acting like a knob controller would. Doesn't matter too much because the adjustable range on the power supply is very likely high enough that the operator i'm giving to won't care but it will still be weird trying to explain it in the current state
i did end up buying a 24V motor because i can't be fucked with a potential overheat

I´m trying to troubleshoot pic attached (amplifier stage for eletrostatic headphone) and running out of ideas.
I expect a full voltage swing on O+, O- of about 640V RMS with as low as 4V on I+, I- but my output saturates at about 40V and if I connect a headphone I hear sound but distorted and crackling when input is not present.

I suspect a failure through a earth issue which had the circuit loose ground reference for a few secs and prob fried some components.

My power rails on input are fine. I noticed that Q10 and Q9 are sending gate voltage straight to drain. I think that´s what I might be measuring.

Fml, that was so much work. Anyone else with an idea? I don´t want to measure and or calculate every value and voltage.

>>1939056
>I just tried to google "DC stranded cable ampacity" and there are absolutely zero relevant results

I doubt that, the problem is that you aren't looking for the right thing. "Ampacity" is highly variable, depending on the nature of the cable and ambient conditions. Something like cheap PVC hookup wire, of the same gauge, will not be of the same ampacity as THHN, simply because the insulation material is not tolerant of higher temperatures, and, therefore, cannot safely carry as much current. Whether or not this cable is in open air or buried in a foot of insulation also complicates matters.

In the real world, electricians go by the NEC rules (or equivalent), and everyone else either tests empirically, or does the math the long way on critical, high-dollar projects.

If all you're really concerned about is voltage drop, just look up a voltage drop calculator.


But because I've already spoon-fed you this much, I might as well just tell you that you're going to need 10AWG to get less than 400mV drop over 60ft. Or, if you actually mean 60ft total conductor length (30ft circuit), you'll need 14AWG.

>>1939110
I’d first recommend splitting the circuit into known segments, or at least easily character usable segments. With any luck you can troubleshoot the thing just be measuring the signal amplitude between each segment, without having to worry about measuring DC bias voltages and currents and such. For starters, I’d measure the output of that op-amp to see if it’s behaving ideally.

>>1939060
Would a crossover circuit help?

>>1939092
No clue. Can you elaborate on the "knob controller"? I'm assuming it's designed for a DC motor, right? Test each part of the circuit independently to figure out where the issue is.

>>1939110
Ok looking closer that circuit drawing is a bit of an abomination. I'd spend the time to neaten it up as well as splitting it into standalone blocks. Is that an optocoupler being used to transmit an analog circuit? LEDs in the audio circuit? Using depletion mode MOSFETs? This picture?

The 10m90s looks really useful for HV led driving though

So this is less of a technical question, more of a career strategy question.

Right now I'm stuck in a dead end mechanical engineering position with no room for movement in my company. Don't ask how I ended up here, I just did.
I have talked to a manager in the electronics department about projects in limbo they can't get done and he did mention a sizeable project that has no deadline but would be "good to have" for internal purposes.
My question is; should I take on this project and work on it on the weekends blowing off my current home projects? Or should I just focus on my home projects strictly?
The purpose of these electronics projects is to fluff my resume and get a job designing electronics. I'm thinking that taking on this project at work may look better on paper rather than building a device I am actually interested in but ultimately is still a hobby project with no oversight.
Thoughts?

>>1939519
I think doing a project for a stakeholder other than yourself will always look better on a resume than otherwise. Regardless of the presence or absence of oversight, you're going to have to consider the needs of the other party. And it's not like you have to put "no deadline" on the resume, right?
What sort of internal project is this? Is it within your skillset? How much room is there for creative freedom?

t. checkout operator

>>1939370
Crossovers in my mind just clean up the signal for the speaker, making them sound better. In this case it just doesn't work, but I'm going to try the crossover described here.
http://www.frugal-phile.com/piezo-XO.html
I'm thinking it's something with the amp that just doesn't work with these drivers. They are straight up Motorola KSN1005 knockoffs... there's nothing exotic here and they appear to be functional even if I can't get them working.
Whole thing is meant for a Halloween prop; looking for loud screeching sound so these are perfect but I'm running out of time. I've got 10 of them...

I'm looking to add some vactrols to a hex schmitt trigger oscillator but I only have 4 and the cheap ones take a month from China. I know I can create some with shrink tubing but I've never done it before. Is there a massive difference in quality?

>>1939519
you should have it 90% figured out before you offer to take it desu
>>1939608
probably, but neither will be linear, and both are just to make buzzy sounds right? Breadboard up a test setup and measure the vactrol's sweep range

I dont know if this is the right place to post this but anyway. Ive been trying to get a transfer function from a series DC motor because I want to make a controller that would maintain a constant torque at any speed. My problem is that im unable to get a transfer function with w(rotational speed) as the input and torque as the output that would consider both mechanical and electrical variables. Ive been reading some works and nobody tries to what im trying, the only transfer function you would find is w/Vin but never Torque/w. What im doing wrong bros? Should I just give up with transfer functions and embrace space states??.

>>1938923
>Voltage does not change with the knob
Even with shitty multimeters with device that claims to have variable PWM duty cycle between 1-100% you should get significant voltage changes
>buying electronics through amazon

Anyone have recommendations for my first lab bench power supply? I want something solid but affordable. Should I build my own?

>>1939543
Kind of what I'm thinking as well. I don't have to write that there isn't a deadline on it. Could use this manager as a reference if I can pull this off.
I don't even really know if this would fall within my skillset. It sounds like it might involve PLCs that interface to a touchscreen HMI. Haven't ever touched anything PLC related before.
Don't know about creativity. It's a very simple device that's just mean to test their products. Literally just a bunch of I/Os that would be programmed to send or recieve signals.
>>1939715
Given my reply above, I would say I'm like 50% sure of what's going on. I have a vague understanding of what is required.
Regardless I told the manager I would need more info before I would agree to take this on.

>>1939060
>>1939585
Put a low value (<100uF), non-polarized capacitor in series with the positive terminal (high pass).

>>1939251
Thanks, I´ll check the op-amp.

>>1939370
It comes from Kevin Gilmore who is considered somewhat of a Guru in the obscure field of Amps for E-Stats. After he gave up his hobby he kindly provided the information to the community. Unfortunately all that is left are this diagram and the gerber files.
Tried importing it into Altium which worked but I don´t know how to import the BOM and I don´t want to populate everything by hand.
Tbh it went a bit over my head, usually ppl build and sell these amps for 6-7 grands online but I though I can do it myself and safe a bit of money. But it turned out more compex than I thought, some exotic parts like the LK389 or the C2M1000170D (although now a bit more common is again sold out around here until at least March next year). Yeah, I will see if I can wrap my head around it.

>>1939608
Make a dozen and test hand-match them. Green LEDs are best IIRC.

Also wait you can buy them? I just bought 50 LDRs and 100 green LEDs from aliexpress.

>>1939772
My recommendation is to steer away from diying anything to do with mains power. Any kind of line switching IC is going to be less efficient than a DC-DC converter IC due to requiring that it bootstrap its own power supply, and any kind of mains linear power supply is going to have a really low maximum current and higher chance to hurt you.
Get a 12V, 24V, 36V etc. power brick or transformer, and use that as a baseline to use switching or linear converters on it as you see fit.
When it comes to designing the constant-voltage and constant-current feedback loops, it's actually quite an interesting process so I'd recommend doing so yourself. That is, if you're doing a linear design with op-amps and such. If not, just use a dedicated switching IC.

also that wierd charge pump to get the output of the op-amps closer to the rails looks like shit

>>1939826
From what I've seen on an occasional youtube video, PLCs themselves aren't too tough to work with. Ladder logic is designed to be relatively easy. But that touchscreen HMI might be difficult, really depends on what's driving it or what it needs to be driven.

>>1939912
Maybe someone else has done a deconstruction of gilmore's circuit?

>>1939923
You can buy them prebuilt in a plastic casing. The problem is that they're used a lot in audio circuit (synths especially). Audiofags are unbearable consumerists so I can't tell if versions being sold are actually better or "better".

>>1940002
they're so cheap to make that I'd never consider buying them

I want to find a good drawing program for creating single line diagrams, any ideas other than AutoCad?

>https://www.ebay.com/itm/Third-Hand-Station-Soldering-Iron-Magnetic-PCB-Board-Fixed-Clip-Flexible-4-Arms/113961118154
Anyone familiar with this thing? Is it trash or serviceable?

>>1940008
what do you mean "single line diagrams"?

>>1940023
Diagram that shows how subsystems work together at a more abstract level, not the electrical connections between physical pins
I think it's also a term used in electric power supply context, when the multiple phases and ground connections are implicit in the diagram

>>1940031
block diagrams?
>>1939772
either
a) use a desktop PSU (DC-DC converter bricks optional)
b) laptop power supply + lm317 (DC-DC converter bricks optional)
c) suck it up and buy a proper one
Don't fuck with the mains

Is there any way to use a potentiometer to act as a variable resistor on three different lines without either line affecting each other? If it was only two I'd use a stereo pot but I can't find any with three layers.

I hope this is the right place to ask this. Is it safe and possible to attenuate a rca audio signal to a rca video signal?

no memes allowed

>>1940031
Exactly this, I am trying to find new software as I am working from home and have drawing to do for work that I would like to do on computer instead of what I used to do which was pencil and hand to my drafter. Would like to just e-mail to drafter now. I have access to Auto Cad but I dislike it.

>>1940102
they aren't even close to the same specification, but as long as you limit the signal volts and amps to below the max of the input, it's safe to feed audio out to video in.

>>1940110
inkscape for general purpose
draw.io may have some useful built-in symbols

>>1940113
Thanks I appreciate it

>>1940096
There are some with three layers, but they’re rare and also somewhat expensive. There’s the option of mechanically connecting three normal pots together, using digital potentiometers, or using Vactrols/OTAs/JFETs/etc. to make voltage controlled variable resistors instead. I’d lean towards Vactrols myself, but maybe that’s because I’ve got enough shit to make over 50 of them.

How well matched do they need to be? Are they placed on the ground rail? Do you need both sides or just one? There might be some ways to spoof what you need them for.

>>1940110
Be nice if you could generate a flow-chart or block diagram from a netlist. I just end up drawing shit in GIMP, but it’s a pain when I make a mistake and have to redo something.

Is there a decent cheap frequency generator that can go up to ~30 MHz (5-volt square waves at least)?
Must have a readout. I want to test the maximum stable clock frequency of various circuits.

>>1940110

there are programs for drawing UML (unified modelling language) models, which would seem to fit the purpose. Lucidchart is one.

there was also a multi-purpose drawing thing that i used about 15 years ago, but cant remember a name.

>>1940350
>unified modelling language
Oh neat, that's very functional looking.

>>1940135
I was going to use the pot as a frequency control for three oscillators so that along with individual control I could also control all three. I've never heard of digipots before but they sound like a good shout.

I want to dirve one of those AE0203D04DF piezostacks. Whats a currently available HV opamp that can do +-150V? I know of the 3584JM but that one seems to be obsolete. Also throw in diferent ways to drive this if you know any.

>>1939898
Thanks, I’ll give it a shot and report back.

>>1940411

normally, you'd be working at low voltage, driving a small transformer to boost it into the hundreds of volts, but you can also use some small high-voltage NPN and PNP transistors in a push-pull up to +-300Vdc.

see https://www.edn.com/high-voltage-amplifier-drives-piezo-tubes/

Rosin arrived

Imagine the smell

>look up how to do 1 (one) seemingly simple thing in electronics
>it's not simple
>dear god it's not simple
>it's a recursive tree of bullshit you never knew you needed to know
every time

>>1940450
Are you gonna dissolve it in ethanol for some ghetto flux?

>>1940454
Maybe a small part.
Solid rosin is way better for hand soldering anyway.

Maan this shit is so fucking good. My iron tip was always black and oxidized, and oxizdes now are gone. Not going to (((flux))) ever again, because it is shit compared to rosin.

>>1940436
Interesting. I forgot to mention my goal here is to build a Slip-Stick Motor and so I want to drive the piezo with a sawtooth at up to 100kHz.
What do you think of Pic related, using the voltage drop across a large resistor?

>>1940450
That could make a lot of violinists happy.

Does anybody happen to know how to get access to the full version of Power Electronics by Barry Williams? The link on the book page doesnt seem to work. http://personal.strath.ac.uk/barry.williams/book.htm

>>1940494
libgen?

>>1940496
It doesn't have the latest version. Both are from 2006 but the index from the book page has about 200 additional pages.

>>1940494
Says to email if you’re having trouble so maybe give that a try if you’re not having luck finding the up to date version elsewhere

>>1940488
Nah, they seem to prefer overpriced rosin with relatively high solvent content (or whatever it is called, thing is also used as paint thinner).
This rosin I got from some shop that was selling it as hair removal wax-resin.

>>1940497
IIRC The GCU library is publically available so try its online portal. Try UoG as well. Failing that, trawl the torrent sites.

>>1940503
>>1940500
>>1940496

Found it now

>>1940281
clean square waves are not that easy. Usually cheap chink function generators are only really usable up to 20 MHz and the square waves are unusable above say 5.

>>1940450

>>1940654
Shit is unironically good for all copper-containing metals, tin plated metals.

>>1940450
>dat hacktop

Any burned-out electrical engineers on /ohm/?
What did you do after burning out? Is there a way to recover, or is it fatal?
And what kind of work is there for a burnout engineer, other than foodservice and retail? I doubt I could pay the bills by repairing hi-fi equipment.

Hello boys,

my brother has a remote controlled seat post for his bike which went dead.
Looks like I found the component which died. pic related.
How do I find out which component this is without a schematic?

No higher res picture sorry. No scope here. Would have to use one in uni.

>>1940738
you read the markings on the component and look them up.

>>1940638
couldn't you just run a shit square through a high speed schmitt trigger to clean it up?

>>1940281
Si5351

On a related note, does anyone have any guides on modulating xtal oscillators? The chip I suggested is shit at fast sweeps, so I want to replace the 2pF load cap with a varicap and modulated it with a ramp or something

>>1940739
hmm ok. Can't read them with my eyes. Will look under the scope tomorrow

>>1940469
I'd consider using inductors instead of resistors, so long as you're going to be driving it at one frequency. Not sure if it will work with non-sinusoidal waveforms though. Push-pull stages are probably a better way to go in general.

>AE0203D04DF
The datasheet doesn't say anything about how much current they draw or how much power they produce. But considering that it's 2x3x5mm, I can assume that it won't be drawing more than 2W or so. Hence, some TO-220 BJTs in a class-B or class-AB configuration should suit you just fine. Heatsinks possibly required. An H-bridge (two push-pull amps) would be better if you only have a single rail power supply.

You want to drive it with a sawtooth waveform, hence you'll want something with negative feedback to keep it linear. >>1940436 should work, though the antiparallel diodes being there might not be perfect for protection purposes.

>>1940739
Figuring out part numbers from package markings is often difficult.
You might have to reverse-engineer it by seeing what other parts it's hooked up to, which might have more easily searchable part numbers.

>>1940738
You need to thoroughly clean up the board with isopropyl alcohol and a toothbrush to rule out corroded open traces/vias and cold joints first.

Currently trying to find a detailed explanation of how CEM3340 / ASS3340 / V3340s works. I can see the IC block diagram on the datasheet but I'm having a hard time deciphering it. All explanations tend to be how to set up a circuit to properly use it rather than how the chip actually works.

>>1940767
lmao, i am a retard.
I did try to clean it with isopropyl and a paintbrush, but that didn't work as good. Tried this a few times on other things and a toothbrush never came to my mind.
It's clean now and started blinking again. will check tomorrow if it actually works.
Thank you m8

>>1940762
This, reverse engineering it will give you a good idea what it is. It's one of those 5-pin guys (SOT-223-5?) so I'd guess something like a single op-amp, something to do with battery charging or management, a switching controller, maybe even an ADC or mosfet driver.

>>1940795
Good luck!

>>1940773
from what I can see, it starts with a triangle generator - charge to a certain point, then discharge to another, likely through a current source. I'm guessing the circuit which toggles charge/discharge generates a squarewave, with the rise/fall in sync with the triangle's peaks. Multiply these two together, you get a saw wave (rise*+1, fall*-1).
The triangle is then fed into a schmitt trigger to give a different squarewave (the one you actually hear), and PWM is done by changing the threshhold.
The 2nF cap at pin 4 is the timing capacitor, I think, and the 270k might set the reference current for the e^x antilog converter.
The rest is just amplifiers (pins 8-20), OTA's by the look of it (they look like op amps with a CV fed into the side)

>>1939519
>a sizeable project that has no deadline but would be "good to have" for internal purposes.
How much experience do you have in electrical engineering. A "project" that's just an idea and no requirements and minimal electrical engineering experience sounds like a disaster waiting to happen.

>>1941112
I wouldn't say it's an idea. They already have the "proof of concept" so to speak of it. It's just a Din rail full of terminal blocks that subsequently get wired up to relays that behave as logic gates (based on how thosr are wired up). The whole issue is when they get a tester to run a test on a product, it takes them hours to reconfigure the tester. They want to do the same thing but make it so that tye terminal blocks hook up to a PLC and is controlled via a touchscreen HMI.
As I stated, I have no experience with PLCs. I got my bachelors in electrical engineering but beyond some hobbyists circuit design projects and school assigned projects, I don't have any "work experience" in electrical engineering.

>>1941373
I've seen that mentality at work a whole lot. lots of people with little experience think they can just grab some COTS products and slap together a magic solution to any problem. Not saying you shouldn't do it but it won't be as easy as you think.
If your company wants a programmable IO tester get one of these.
https://www.ditmco.com/

Could anyone recommend a good book on power electronics?

Bonus points for something covering the type of power electronics used in wind turbines

What the hell is control theory and why should I learn it?

>>1941603
do you want to ever automate any system?

>>1941609
Yes

>>1941655
do you want to automate the automation of systems, so they do all the work, and you get all the money doing nothing?

>>1941673

>>1941673
Absolutely, but I’m not exactly in a position to get paid for anything of that sort, yet.

What the hell is the value of C when reading battery datasheets?

https://www.powerstream.com/lip/GMB402042-300mAh-PCM_A0.pdf

>>1941756
1 C is equal to the A-h capacity of the battery.

Charging / discharging at 2 C means 2x the battery capacity. 3C is 3x the capacity.
so a 2500 mAH battery, C=2500 mA.

Is there any kind of instrument which is effectively a variable wavelength light source?
I'm looking to expose a chemical to light frequencies ranging from IR to UVB to see the effect it has and it'd be nice to just sweep it across the spectrum and observe the effect. I have a feeling that either no such instrument exists or it is expensive well beyond my means.

I'm sure it's possible to cobble something together with LEDs but in that case rather than doing a proper continuous sweep it's more like you're moving in discrete steps since LED emission wavelengths are generally fixed and all you'd be doing is switching in different LEDs. Also you couldn't use RGB for the visible spectrum either since it it can't create true yellow or violet for instance. Discrete LEDs for every frequency of interest. It's a clunky solution.

>>1941806
full spectrum light source + refraction and filters will get you close, but it's not trivial to generate arbitrary light frequencies.

>>1941770
In the context of the battery I posted, it says:
Nominal Capacity is ≥300mAh@ 0.2C
Discharge
Min≥290mAh

300mAh would mean the battery lasts 1 hour when 300mA is drawn from it.
So is it saying that mAh rating is only valid if the current being drawn from it is 60mA?
Otherwise it's rated at 290mAh?

>>1941826
Yeah.
Min is what you can expect to get at a minimum. You might get more, but don't design around getting more.

>>1941806
Free electron laser is exactly that. Not actually that difficult to make compared to some lasers, would make a really cool project for an advanced operator, if a relatively dangerous one (low energy ionising radiation + high voltage + blindness hazard).

If not, best option is a blackbody (or other wideband) source + basically a spectrometer; a prism or refraction grating with which only a certain wavelength comes out at a certain angle. Probably quite difficult to get UV out of something like this, as opposed to a FEL.

I also know it’s possible to generate X-rays in a vacuum tube, and tune their wavelength somewhat I think, so maybe with a lower energy per electron / voltage you could make arbitrary UV.

You making a spectrometer?

>>1941806
white light + diffraction gradient, look up how UV/vis sources work. Should just be a lamp, diff gradient (cheap on ebay), and a fine tuned motor.
LED's change frequency with temperature, so you could use a resistor as a heating element. but you probably won't get very wide swing with that

>>1941877
Never mind, I think X-ray tubes use spontaneous emission from a particular metal on the target (anode?), hence all the photons produced have a very similar energy. Much thinner FWHM compared to a gas emission or LED.

Also you don’t want your bandwidth being too high, since you may run into higher-order harmonics through a diffraction grating. I’d instead switch a series of LEDs on (maybe 8 different wavelengths) and focus that through a diffraction grating to let you select the perfect line. Not sure if IR will go through a diffraction grating or a prism (not to mention whatever optics you use, maybe you’ll need a diffraction mirror), but for anything below silicon’s 0.7eV photon I'd use an incandescent lamp that you vary the power to. You’ll also need to change the diffraction grating a few times over the spectrum to keep it in a high-resolution region.

>>1941887
Does an incandescent lamp make enough UV to even use? Not to mention whether it penetrates the lamp’s glass envelope. If he needs smaller wavelengths than a 405nm LED can put out, then I’ve no idea what he could possibly use, besides an array of gas discharge tubes. Need a fair few of them because their FWHM is a lot narrower than an LEDs’. Phosphors, maybe? Variable angle frequency doubling crystal bathed with a black body?

>>1941855
Great, thanks for the help brother

>>1941895
are commercial uvb/c lamps broad spectrum, or line?

>>1941919
They're a mercury tube, so pretty narrow. A UVC LED should be wider, and they're not that expensive these days (2-10 USD each).

I have a 5m roll of addressable leds SK6812 waterproofed.

Can I plug the 5m roll into my arduino with a 5V 1A plug and start practicing?

or will I have to cut off a section?

>>1941955
The 5V power supply coming out of the arduino will be whatever power supply is going into the arduino (unless you're powering it off a barrel jack or other >6V supply). If that supply can handle the amount of current that the SK6812s want, then you'll be fine.

>>1941756
Followup question on this.
I'm looking to devise a circuit to monitor the battery percentage. On the datasheet, I see that:
Max Voltage = 4.2V
Nominal Voltage = 3.7V
Cutoff Voltage = 3.0V

So would it be safe to say:
4.2V = 100%
3.0V = 0%
3.7V = Something in between?
How am I supposed to interpolate the voltage level with the percentage of charge available?

I'm reading that most battery monitors are specific to the hardware the battery is running on, battery characteristics, and educated guesses based on test results on the battery itself. Which makes sense, given that max and min battery voltages isn't a linear path afaik.
What is the generally acceptable way to monitor battery level? Just test the shit out of it and make a lookup table?

I was watching a video to learn about inverters and it showed a full H bridge circuit using transistors as switches could produce an AC square wave. As far as I understand this process is simplified by using two transistors connected to one side of a transformer. Is that the basic operating principle of DC->AC conversion?

It also mentioned that a sin wave could be generated using PWM and a comparator to smooth it out. I've used integrators before to get other waves out of square waves before but does all wave generation begin with square waves? For example, is there any way to generate a sin wave that doesn't involve processing another wave?

>>1942010
>Is that the basic operating principle of DC->AC conversion?

it's just one way to do it. it's very far from accurate, but it's popular coz it's easy and cheap.

>is there any way to generate a sin wave that doesn't involve processing another wave?

there are 1001 ways to do it. one very easy, and surprisingly accurate way is to put a microphone in front of your face and say a vowel.

>>1942010
> Is that the basic operating principle of DC->AC conversion?
There are multiple topologies; the main ones are full bridge, half bridge, and flyback.

> does all wave generation begin with square waves? For example, is there any way to generate a sin wave that doesn't involve processing another wave?
There are plenty of sine wave oscillators (e.g. Colpitts, Hartley). The main components are an LC circuit or crystal to fix the resonant frequency and and amplifier; if you want a pure sine wave also you need some form of gain control.

But where the power involved is significant, there's a strong preference to using transistors as switches rather than as linear amplifiers. Power dissipation is minimised when the current is minimised ("open" switch) and when the voltage is minimised ("closed" switch), and maximised when both voltage and current are significant.

>>1941958
you could be lazy and instead of using a ic you could use a 3v relay and a resistor . the resistor would be enough that if the voltage drops to 3v and your current is low it wont let the relay connect the normally open path

you would need to check your battery / pack and see the ah / mah then do soem calculations or look them up online. current matters for relays. a 5v relay can run on 1 18650 which is lower voltage . same for a 9v one if your current is high enough. there is a blog some one made with tons of 18650 listings for passive draw current. found out my laptop batteries do 4 amps .

anyways google current limiting resistor and find one of the many vids of some one explaining how to reduce the current without just doing what i do and tac on a 1k resistor when things catch on fire or explode

how the FUCK do i uninsolate this cunt

i tried soldering it at 400C i tried burning it with a lighter, i tried threatening its family and nothing is fucking working, i can't get it to conduct any fucking electrons

>>1942080
try sandpaper, or scraping it with knife.

>>1942080

use x-acto knife or box cutter. for each stand, place on top of your finger and scrape off a bit of the insulation, dont have to do it all around. since it's sitting on your finger, that guarantees you wont use too much force.

>>1942072
same anon after smoking a cig i forgot to put the n channel fet / pnp bjt on it to cut off bleed on the normally closed connection of the relay. for this circuit to work your relay must have both normally open and normally closed. every 3-5v relay i have seen has both.. this circuit is the opposite of the auto cut off battery charger . the diode is mandatory as it prevents a spike when cut off. the circle is a led idk falstad lists it as a circle

your resistor goes on the normally open connection and needs to be tuned to your circuit. you didnt post one so idk what it draws. this cannot be avoided as the battery has to power the coil . idk what relay you will use and i dont know what 18650 / cellphone battery you got. idk what your circuit will pull. but if its to low it cant make the relay connect

>>1942080
those wires are a bitch.
What I do is, first I separate the strands. sometimes there's plastic fibers in the middle, snip those off. Then bunch them together, hit them with a lighter flame until the enamel burns to ash. I scrape most of that off with a fingernail, then apply solder and flux which usually takes care of the rest.

Learning about transistors and as much as I know that electricity flows from the emitter to the collector, I keep double taking because circuits are laid out to suggest it flows from collector to emitter because the entire history of electronics & circuit drawing is fucked like that.

>>1942145
Current flows in the direction of the arrow on the emitter: collector to emitter for NPN, emitter to collector for PNP. Electrons, being negatively charged, flow in the opposite direction to current.

>>1942145
You need to get over that. Nobody who works with electronics ever thinks about it. If you are designing a new type of semiconductor then yeah, it's important, but if you are using transistors that someone else made, it's not something you need to ever think about.

>>1942145
Speaking of which, saturation mode says that both PN junctions become forward biased by connecting the base and the collector. I might be getting confused but surely connecting them together means that they're both equally positive? How can the collector junction be forward biased if both sides are positively charged?

>>1942010
>a comparator
you mean "LC filter", right?

>>1942165
try simulating it or breadboarding it, with resistors everywhere of course.

>>1942194
I meant a integrator. Filters are good but from my limited experience aren't as precise.

>>1942163
it helps to remember which side is +/- on an LED (the pointy bit is always -)
>>1942080
sandpaper or the edge of a razorblade (angle it away so it doesn't slice into the wire)
>>1942010
find the load line of a power transistor and build a classAB amplifier, lolgoodluck. I think there are sine-capable mosfets too

I'm trying to learn electronics for music (I wanna make a eurorack synth), so I started reading 'Handmade Electronic Music". I was too dumb for the other ones. I really like it though, it walks you through dozens of small projects and they're all battery powered so I won't die.

My question is about purchasing an appropriate kit for the first project, a low-watt amplifier. I looked on eBay and found a kit that looks similar to the picture the author provides:
>https://www.ebay.com/itm/RAINBOWKITS-AA-1D-ONE-WATT-1W-BIG-EAR-AUDIO-AMP-AMPLIFIER-KIT-w-MIC-MICROPHONE/400534025318?hash=item5d41b03466:g:MQAAAMXQqdhR6z0s

Seems a little bit pricey though doesn't it?

>>1942201
Well yeah, but you can't just slap a linear device like an integrator on it if it's high-power, and if it isn't high-power you should just use a resonant oscillator with an agc
unless you're making the pwm with an mcu instead of an analog sinusoid (or a digital sinusoid with an eeprom lookup table + a counter + DAC i guess)

>>1942072
>>1942084
Just to clarify, I am using the battery posted here:
>>1941756
>GMB402042
I get what you're trying to do, you're trying to make it so that the battery gets cutoff when it dips below the rated cutoff value of 3.0V. Though I don't know how small relays can get, all my circuitry needs to fit a small form factor.
But what I'm trying to do is to have a digital readout of the battery level. Just like on a cellphone, indicated with LEDs, 7-seg display, or LCD screen.

The circuit will be a BLE controller based around a CC2650 (haven't developed it yet because I intend to do that with my buddy who is handling the software/firmware aspects).
Literally it's just going to be a remote to send out signals everytime a certain button is pressed.
I'd do it myself but I know fuck all about programming.

>>1942213
>Seems a little bit pricey though doesn't it?

thrift store has dozens of amplified computer speakers with 10x better quality at 10x less cost.

>>1942213
Search for PAM8403.

https://www.ebay.com/itm/10PCS-Mini-Digital-Power-Amplifier-Board-PAM8403-Class-D-2-3W-2-5-USRZ/124276568062

>>1937846
So I posted about this last week, I'm trying to make a circuit that can adjust the current into a pin (IIn), with a coarse and fine adjustable gain via potentiometers. Received some advice to put some buffers in and use resistor values to determine gains.

Would this circuit do the job, or is it bad?

>>1942194
Oh right, the voltage drop across the base resistor. duh.

>>1942251
would a BJT current source not work? they're super simple

>>1942213
>$16
>pricey
lmao
if you wanna be cheap, look up LM386 johnaudiotech on yt, you can make a little standalone board for like $1.50

>>1942241
I'm following this book to learn, so it's more about following the project and building than it is getting a speaker.

>>1942249
Thanks, but this doesn't make sense to me. I'm looking for something I can plug a battery into.

>>1942269
I think I was actually looking at making something similar. Haven't seen the video yet but I found a schematic detailing the thing I wanna build, page 18 or pic related:
>https://www.wiley.com/college/tsividis/0471386952/lab/lab.pdf

I think out of all the options presented I'm going to breadboard the pic related. I don't really know how to read diagrams or shop for parts and should learn.

>>1942276
Just buy the individual components off a local supplier (or Alibay/LCSC if you have the patience).

>>1942276
>I'm looking for something I can plug a battery into.
PAM8403 runs on 5VDC 500mA-1.2A.

>>1942279
Now that I have a schematic I'm going to, I didn't when I first posted. That video that anon shared was perfect, btw. I don't know how to figure out what components I need but luckily the guys at the local shop are nice.

>>1942280
well it was cheap s I bought it, also bought some LM386s from that other video, whateva

Could you make something like this:
>https://www.sequentix.com/shop/cirklon-hardware-sequencer

I don't get why they're so expensive.

fusion 360 electronics is literal dog shit

>>1942284
yes, anything can be made. Generally what you are paying for is the time it takes someone to develop, ie the knowledge. Parts and materials aren't expensive. Skills, techniques, time and labor are.

>>1942285
>fusion 360 electronics
im sorry what? i thought that was a cad for solid modelling

>>1942284
>I don't get why they're so expensive.
It's not mass-produced and aimed at professionals, what do you expect?

There are some projects for MIDI sequencers if you search around.

>>1942286
Pretty sure I can program a 'sequencer enginer'. I'll set a Cirklon clone as my long term goal. That'd be friggin dope af

>>1942290
once you know what you're doing, it can't be that many man hours dumped into building it, right? the price just seems crazy. Not unusual for 'artist' though

>>1942292
pay people for their work or do it yourself.

>>1942293
I'm willing to do it myslef, just didn't really understand if it was feasible or not. obviously I'm way out from that project, but still.

>>1942284
The actual sequencing can easily be done with arduino and some ICs. The harder part is building good pushbuttons. You can get a cheap midi keyboard for ~$40 and build a midi input for your box.

>>1942294
If you want a basic sequencer in the mean time CD4017s are the 555s of sequencers. Super easy to set up; All they really need is a clock input and pots to control the CVs.

>>1942295
Why are the buttons so hard to build? seems weird, especially with how popular 3d printing is these days

>>1942296
>CD4017
thanks, I'm a huge noob but I'm wanting to get into production while minimizing screen time. Originally I was going to finish that book I mentioned above (Handmade Electronic Music) and try putting this together:
>https://www.milkcrate.com.au/_other/sea-moss/#seq

But I'll see if I can find a build using the CD4017 too. btw, what the hell is a pot?

>>1942303
Not necessarily hard to build but you'd want nice quality buttons that are both responsive and easy to quickly use in a final build. Also, midi output has a bunch of data; IIRC it has pitch, velocity (how hard you press the button) and a bunch of other shit. It's a case of how much you want to build.

https://www.youtube.com/watch?v=9oGlCfwCoCw

This guy's sequencer is pretty good for what little components it uses. It's got 1 channel but 2 independently controlled cv outputs.

>>1942306
a 4017 is a decade counter whereas the link you used is a binary counter. Really its all the same; A chip that alternates through a bunch of outputs based on a clock signal.

>Pot
Potentiometer. You set up each output as a voltage divider so that x volts goes to the oscillator to alter its frequency. Twisting the potentiometer changes the x value and makes the pitch go up or down.

I’m the additive synth anon from a few weeks ago, and I’m also interested in sequencer design. I’m aiming for something relatively simple to make, but because my synth designs are based off a PLL on the input stage, any sequencer will need to output a square wave at the correct frequency. Making a note-accurate discrete component variable oscillator might be possible just using combinatorial logic and an EEPROM, but it’s not as simple as I’d like. Hence, I’ll probably use a square wave DDS IC of some sort. I’d tune it with rotary encoders, with a 7 segment display reading out the notes. One encoder you advance/go back, and one to change the note at that position. Not adjustable on the fly, but good enough for starting off with.

Any suggestions for parallel input DDS modules that I can just use an EEPROM lookup table with?

I've just realised that my soldering iron is kinda shit.
What should I get?
https://eaglepoll.ga/poll.php?vote=4

>>1942346
Not chinkshit knockoff, but proper T12 station. Better than any brand name within the $50 range, and probably the $150 range too. If you’re going above that, don’t go chink.

What I’d look for isn’t brand name itself but rather the tip specs. Thermal conductivity, thermal mass, thermocouple accuracy, power output, size/shape variety, price, that sort of thing. And as always, it’s a matter of compromise in order to suit your use-case.

>>1940017
absolute junk

>>1942377
can you name something decent off hand that's less than $100?

>>1942401
they are fine and useful. >>1942377 is a retard

>>1942402
Not him, but they are pretty shitty. They can’t handle too much force or they deform. Ok for holding wires together or to your PCB, but I’d much rather do that by hand and hold the PCB with a vice. I’ve heard good things about the “panavice”, but any PCB vice should work ok so long as it isn’t plastic. Getting a couple of lightweight goosenecks (or coolant hoses) with alligator clips on their ends for holding wires is handy nonetheless, I think I’d also want some sort of spring loaded needle mounted on a gooseneck connected to my vice in order to hold SMDs to the board. Like a tiny version of those things made for welding.

so i'm in navy and going through navycool for electrical technician certs. i don't know much about the world of certifications and was wondering if certs through certTEC or ETA-1 are worthless or not. i googled but haven't found much reviews about them other then theyre accredited by ICAC

>>1942415
>I think I’d also want some sort of spring loaded needle mounted on a gooseneck connected to my vice in order to hold SMDs to the board
So it can shoot across the room? Normal people use tape.

I helped my uncle upgrade to LED lights in his shop. My uncle gave me 4 Mercury Halide lights (300V), 300V autotransformers, and 25uF capacitors. I already used one set to light up my garage (1kW, LOL). Anybody have ideas on what to use the rest of the stuff for besides lighting? I'd rather just use the rest of the bulbs/capacitors as spares and use the autotransformers for something else but I'm not creative...

>>1942369
Idk how to say this. I'm fed up with chinkshit.
It doesn't feel right in hand, idk.
Currently I have A-BF GS90 Hakko 900 based iron, and idk. It is shit. Thermal control does absolutely nothing, tips are hard to find (good ones, because every second one is a knock off that is made out of 100% steel and won't hold the coating for long, so you will end up going to scrape it, tin the bare steel - fuck that)... Like only good tips I got were from ali, $1 for set of 10. I tried more expensive ones ($2 per piece), etc... Only one set survived basically.
Like some $10 30W fixed temp iron is better.
Same goes for soldering station. All repairmen that have old 900-tip chinkshit (Lukey 702, etc) set their temp to full retard, because otherwise it is pretty much useless.

And this scares me, because T12 is so common now, you don't know what are you buying. Used, but original? New but chinkshit?

Euro irons, ERSA for example, don't have much knock offs. Idk how it feels in hand, but couple times I handled their irons, they felt like 2x more powerful that turd I have. But they are real expensive, and old tech (i.e. heater is separate from the tip), but hopefully German precision makes it acceptable... And I can get 100% genuine Ersa in my shithole, though they are expensive as fuck.
So I just don't know. I want a nice iron, I don't really care about price at this point...

any of you living in auckland new zealand, there's some old electronics shit for sale outside the heavy metal tshirt shop in eden terrace

mainly old television station shit, like that tv camera or tv-specialised oscilloscope thing

>>1942462
got me some cdips, double-gang + switch potentiometers, opamps, and varicaps.

>>1942439
No, something like a pogo pin. Could do it with a weight too, but that would make it bulkier.

Any good resources for practical self taught electronics?

>>1942502
Read the OP

what is the difference between an EE degree and screwing around building what you want?

>>1942544
Employers are supposed to take you more seriously if you have an EE degree. If you dick around with projects yourself, it depends on what kind of projects you've done that prove you know your stuff. You technically want both if you want to get hired anywhere (or so I hear).
Really depends on what your end goal is though.
t. electrical engineer working as a mechanical engineer

>>1942549
not really looking for work in the field, just wondering what the differences in knowledge and skillset are from a hobbyist project guy vs a degree holding jobbie.

>>1942462
>>1942468
Ok, now that I can take count, I got a PCB with:
6 whole socketed CDIP D27128 16k x 8b UV erasable PROMs
10 soldered TC5516 2k x 8b SRAM
1 soldered 74HCT138 3-8 demultiplexers
2 soldered 74HCT139 dual 2-4 demultiplexers

There's also maybe 20 of the potentiometers, about 50 BA4560 dual op-amps, three varicaps (the square ~plastic 10-150pF ones with like 4 extra mini varicaps on top).
Then there's ~10 of these black boxes, some with 10 pins, some with 14 (the extra 4 are just chassis mounting pins). Looks like theres two MELF packages in the bottom, and two adjustable screws on top. They have the writing "184006" on one line and "E282" on the second. Pic related. Any clues?

Oh and most of the tech shit in >>1942462 is for TV broadcasting stuff, hence the TV camera, TV-specialist oscilloscope, TV screen bank for mixing, control racks, even a background music looping tape. A handful of vacuum tubes too. While I could buy one of those just for the curiosity or the parts (the scope is $80), it would probably be a waste of space.

>>1942550
If you're looking for the difference between a strictly book smart designer and a strictly street smart designer, it's the same as any industry as far as I can tell. Street smarts tend to struggle on the theory while the book smarts tend to struggle on application. That's just generally speaking of course. Plenty of hobbyists can handle theory and plenty of degree holders can handle application.

Ideally, you'd do both; pursue a degree while monkeying around on your free time to build your own shit.

>>1942557
what's the 'theory' they struggle with?

I'm finishing up my degree in math and cs, just wondering as I'm getting into electronics heavily lately. when i declared my major, i thought I'd enjoy working as a math or cs guy more than an EE on a day to day basis.

>>1942544
EE students don't spend all day in an electronics lab doing projects like you probably imagine. There's one project course per year in most schools. The other 90% of the courses revolve around cramming calculus and physics textbooks.

Just about anyone can get an arduino and put together a thing these days. However if a company is making a million of those things cutting the price by a dollar with a highly optimized design turns into a lot of money. It's not enough to be able to cludge something together that works. You need to be an expert on the physics and math to make the best system possible.

>>1942568
So they'll actually have you analyzing pricing for scalable designs in uni?

I just want to focus on making musical hardware and automating several processes I'm responsible for (related to small scale farming, mixing/packaging, brewing, etc). Big noob so my plan is to finish the book I'm on (Handmade Electronics) then continue with something less baby tier, but if you have recommendations for what to read next, given my interest, please lmk. After getting my hands comfortably dirty I'd like to learn more theory.

>>1942570
There are courses on that, but what I was trying to convey is that if a hobbyist and EE are both asked to make a dildo controller the EE can make on that costs significantly less. What a hobbyist can do with an arduino and $20 of other parts an EE can do with $1 of OP-amps, resistors, and capacitors. I was a hobbyist before I was an EE and this is how I feel looking at my old projects.

does anybody know an european digikey/mouser/etc source for the spring pin header on the right in pic?

>>1942549
how much anti-credentialism is there in EE? Is it anything like programming?

>>1942639

nothing like programming. there's only a couple of initials you can add after your name but nobody really cares about them.

In terms of synthesizers, does a DC wave sound any different from an AC wave so long as they're the same waveform (e.g. square)? I know pitch is frequency based so I'd imagine they do.

>>1942734
>DC wave
Isn't it called a pulse if it doesn't swing negative?

>>1942734

square waves sound so bad either way that you prob cant tell the diff. but, with AC, the speaker is going both in and out of it's default position, whereas with DC, it only goes one way, so that'll affect the sound somewhat.

Intermittent problems are the fucking worst.
Almost 2 years ago I put together a diy home automation setup with a few arduinos linked together on a long, slow-clocked I2C bus. It breaks the rules, but it works reliably. Most of the time.
Every couple of months this system gets into a failure mode where one or more of the arduinos just drive the I2C bus high and then halt so nothing else can use the bus.
I've heard of an I2C bus becoming stuck low, but stuck driven high is a weird one. An I2C line should never be driven high at all.
It happens so rarely that I've never been able to figure it out.

>>1942760
Can all connected arduinos acts as both master and slave or only one is designated as master?
For long data lines for I2C I just add external circuitry to increase drive current for signals and add some filtering

>>1942771
Any node can master the bus if it has something to report. That's one reason I chose I2C, that and not needing any external circuitry to drive it like CAN bus does.
I clocked the speed down to 10KHz so there wouldn't be problems with reflections etc with long, chaotic bus topology. It works pretty well ordinarily, except for these rare failures.
In the end I might just try and use the watchdog timer to reset the arduinos if they freeze. And they do completely freeze up, it's not just an I2C issue, a frozen node won't respond to local events either.
What do you use for drive and filtering circuitry?

>>1942745
Now that's got me thinking. I understand the basic difference between AC and DC (DC only moves in one direction, only has one polarity). However, lets say a DC square wave oscillates between zero and 5V. If I draw a line through that signal at 2.5V then can't it be said that the DC square wave is an AC square wave with a 2.5V amplitude? Or am I just being a colossal retard?

>>1942549
>>1942568
Not the one that you two replied to but I'm quite curious about the application of physics and math in the field. When I read physics, I default to thinking classical mechanics, but when people who are in the field of EE talk about physics, do you guys mean electrodynamics/electromagnetism? I did pretty bad in classical mechanics and I hope that I won't have to use much of it ever again but I'm completely fine with EM. For math, is there any reccomendation on which subjects to get "real good" with?

Forgive me for being dumb here but, what exactly is the difference between here and /mcg/? Is /ohm/ more 'general' in the application of electronics while /mcg/ focuses more on the deeper layer like gates? I'm not trying to cause drama or anything but whenever I have questions, I have no idea whether or not I should ask it on /mcg/ or /ohm/, I feel like a bunch of people that use the two frequent between the two.

>>1942799
>can't it be said that the DC square wave is an AC square wave with a 2.5V amplitude?

no. it's not alternating, so the ''A'' in AC is unmet. however, if you just stick a cap in series with the signal, the DC component disappears, and you have AC. like magic.

>>1942826

the tag on /mcg/ is ''Arduino, Teensy, Atmel, ESP32, FPGAs, bluepill, Raspberry... All microcontrollers welcome!''

ohm is general, whereas mcg is specific to satan-inspired controller boards.

>>1942833
Oh... I think that my original thought is actually reversed then! So it's /ohm/ that's deeper in detail while /mcg/ gets microcontrollers off the shelf and immediately uses them for a specific goal. Thank you for dealing with my 2IQ moment. I guess I should've expected it when the OP image is literally as 'deep' as it can probably get.

>>1942791
>And they do completely freeze up, it's not just an I2C issue, a frozen node won't respond to local events either
One of the issues I had with multiple masters on I2C line was some of them trying to act as master at the same time, completely fucks up everything and depending on how code is written can cause complete crash of MCUs where it gets stuck in some function.
So internal avr watchdog would help

As for long i2c data lines there are all kinds of alternatives, I mostly stick to mosfet buffers for driving and opamps for reading but probably using dedicated chips is simpler
If you lines are very long I might even suggest some optocoupler-based solutions

>>1942828
it's not not ac, it's ac superimposed with dc, it's both

>>1942842

>>1942828
That's pretty neat. I remember reading that capacitors block DC and let AC pass but I didn't realise that could be used to get AC from a DC signal. I've only really dealt with AC oscillators for tone generation but I've never seen the use of a capacitor to generate an AC signal from a DC oscillator. I assume it's not as reliable as an oscillator circuit specifically built for AC?

>>1942578
OK, and how would the hobbyist gain that frugality skill the EE has?

>>1942851
Look up capacitive reactance.

>>1942854
>Capacitive reactance
You learn something new every day.

>>1942853
not him, but probably by getting a job.

Does anyone have any advice on how to find out which lead is connected to the outer foil on film capacitors? I have tried using my oscilloscope, where I set the voltage/div setting to 5mV on the channel that I'm using for testing, and played around with various time base settings, but there is no discernable difference between the two different ways that I hook up the capacitor. Alternatively, I have tried using a guitar amplifier to identify which lead is which, but once again there is no discernable difference. I have done the test in a room which has LED overhead lighting while having my finger on the outside of the capacitor to induce some noise in it. What am I doing wrong? I have watched some videos of people using these two techniques to identify which lead is which, and they have a night and day difference between the two orientations. I have also tried putting the capacitor near a dimmable halogen desk lamp that has audible 60hz noise when set to a dimmer brightness to no avail. Are there any other techniques that I could try?

>>1942873
disassemble one

>>1942873
Hold a piece of alfoil against the outside of the capacitor, then measure to which pin it has the higher capacitance.

No clue why you'd want to know that though.

>>1942873
If you're trying to figure out which way to wire the capacitor for lowest noise, then you've pretty much proven that with the capacitors you have, the noise performance is the same either way.
I don't know what exactly the differences are between modern film capacitors and midcentury caps, but since modern caps usually don't have the outer film marked, I think it just doesn't matter so much with the way modern caps are built.

>>1942873
https://youtu.be/BnR_DLd1PDI

I scavenged some thrown-out vapes, and they all had this component wired to the battery, near a hole in the bottom.
Is it some kind of succ sensor?

Can anyone recommend a senstive EMF meter circuit that I could connect to an analog input of an Arduino?

>>1942993
If it's a vape that you have to suck to turn on then yes.

Lets say i had my own 200 amp main sbreaker meter panel, something like https://www.homedepot.com/p/Square-D-Homeline-200-Amp-20-Space-40-Circuit-Outdoor-Ring-Type-Overhead-Service-Surface-Mount-Main-Breaker-CSED-Value-Pack-SO2040VP/100160711

can i power that thing as a s subpanel, and then measure it by buying my own meter with 4--20ma or some such output? is there such a thing? or would intruducing another meter into a residential situation piss off the utility company for some reason?

>>1943160
Look up WAGO signal conditioner or better yet call customer service talk to a tech.

>>1943106
What’s an arduino’s analog input impedance? I’d tend towards either an open-gate (plus TVS diodes) MOSFET or JFET, or a 3-transistor Darlington array. If you’re doing darlingtons, you could have both PNP and NPN ones, so you can detect both positive and negative pulses.

If I had a dollar for every time I've caused a short circuit by forgetting I just used the 10A setting on a multimeter...

>>1943195
lel I'd have one or two dollars
*snap*
>>1943160
nothing wrong with subpanels, those are something that one would see for a suite or a home extension. also nothing wrong with adding meters.
RV parks will use a meter on each unit to calculate the power bill for each customer, for example, instead of having 20 separate bills with the utility company.
with a clamp style meter you don't have to put your subpanel in series with the measuring device, either.

>>1943106
wtf is emf? you a 5G fag?
build an S-meter out of germaniums, random wire antenna, read the output with a voltmeter

>>1943193
Well someone on the arduino forum recommended an instrumentation amplifier and amazon sells them as modules.
>>1943211
I dont know if they have germanium diode anymore like last time I checked for them they were marked as obsolete. That and voltage controled capacitors.

EMF...is EletroMagnetic Feild...and the aim of the project is for debunking ghost hunting...also think about this if ojuia boards are real...why has anyone adapted one that gets displayed on TV and the supposed ghost interacts with antennas to move a cursor over a letter.

>>1943211
Oh yeah I know RF is non ionizing so 5G causing any type of damage I dont buy but I do understand that it is very real for people to develop a senstivity to certian RF frequencies. Like apparently Popular Science ran an article about some engineer that had to live off the gird after years of working with telecommunications. He would get nauseous when a cell phone would be receiving a call.

The thing I have yet to see answered, if RF is not a directly ionizing form of radiation could it cause something esle to give off ionizing radiation.

Hey, hey, strategically speaking: would it be a good or terrible idea to teach 5G/ghostfags about germanium diodes?
Technically schottkeys would be better, but if the idiots on fb start posting about germaniums, could we get a major production run of them? Could we even meme germanium transistors out of obsolescence?

>>1943228
I second this...get all remaining supplies...put diode in plstic baggie with cardboard stapled seal...instructions...connect to 1.5 volt battery...cyrstal will radiate ghost repling rays and also gives immunities to reply to this post messages.

>>1943225
You've got your work cut out for you if you want to debunk ghost hunting. Like most delusions, the belief is usually strong enough to weather any rational argument.
>>1934215

Im looking to see if anyone knows of a specific kind of IC that does what i need.

Im looking to make a low voltage detect circuit that when it detects undervoltage it shuts off another circuit(a charger) but the voltage detector can be reset only by pressing a momentary switch connected to it and the undervoltage detector should keep its off state if detected even after you reset everything by removing the voltage source and reconnecting it (so the only reset of the voltage detector state is only by pressing the momentary switch)

i know it can be made with a small MCU quite trivially but im trying to make it as passive as possible to both take as little quiescent current and to be robust

>>1943329
A mosfet with a bypass switch?

>>1943329
soft latch switch?

>>1943329
You're looking for a Brown Out Supervisory IC. TI makes the TL7702A and other related chips that may fit your design.

>>1943228
>>1943233
Sounds like a reddit idea. I like it!

Is digital design (gates, etc) a fundamental or vital skill in electronics? I'm so bad at this

>>1943555
Yeah. Pretty much most of your computer is composed of digital circuits (uses HIGH and LOW/1&0) rather than analogue circuits (uses a range e.g. 0<=x<=1).

>>1943560
Fuck. Okay.
I'm going to restudy this entire textbook then because I'm so bad at this.

>>1943555
Not directly, for the most part you'll never be dealing with circuits at such a low level. Even with computer architecture, the lowest you'll probably be encountering will be at the level of registers, full adders, that kind of stuff.
See: Ben Eater's breadboard computer.

>>1943310
The deal is I can believe in some aspects of the paranormal but the methods they use in ghost hunting shows seems sketchy to me. Like for instance this whole thing with an EMF meter detecting when a ghost is present...do they ever account for the right hand rule? Like you could get an idea of where the ghost is. But then it enters new issues like are ghosts radio active then?

About the right hand rule thing is the commerical EMF device I got changes as I rotate it through a horizontal position through to a vertical postion. Like think take the antenna wire hold it horizontal to the ground and then pass a current through another nearby wire that is also horizontal.Then turn the antenna to a vertical position. I may have it mixed up at the moment but the signal you got in the first position of the antenna is going to extremely different in the next position.

Additionally the theories they present as to why haunting is never fully connects to the occult but of course that could be a self censoring thing.

>>1943584
They don't account for the right hand rule because there's no formula / principle / gold standard for ghosts interacting with the electromagnetic spectrum. It reminds me of something I watched.
>"How do you kill a vampire?"
>"Any way you want because vampires don't fucking exist".
Maxwell's equations didn't factor in ghosts because he wasn't developmentally challenged.

I’m about to upgrade my work bench and get a bit more serious into fixing old electronic music gear and building diy Eurorack module kits to sell on, I dipped into this a few years ago but feel like with the C19 virus and our britbong government giving out unemployment grants now might be a good time to get back into it.
My problem is I keep looking at Hakko gear as feel like although it’s expensive it’ll last me a lifetime but then I see mixed (mostly good) opinions about much cheaper Chinese gear.
What is /ohm/s opinion on buying cheap chines gear? Should I just “buy quality and buy once“ or should I consider the Chinese stuff?
I’m looking at getting a Solder station, desoldering gun, hot air gun and maybe a digital microscope. Recommendations and advice would be highly appreciated, cheers.

>>1943602
ive had a shitty 20$ maplin 40W iron for 4 years, never let me down. And if it did, i would just go and buy another one.

>>1943602
Assuming you already have an o-scope and an AC supply, it's handy to have a good quality bench VCO solely to test other modules on (filters, VCAs etc.).

>>1943562
>I'm going to restudy this entire textbook then because I'm so bad at this.

if you're doing for academic purposes, fine, but if it's to learn design, then 95% of that textbook is gonna be useless filler exercises with no application in real-life uses.

like the simplification in pic related. nobody but actual chip designers would ever need to do that, and if they did, they'd use software to do it which is 1 billion times faster and 1 million times less likely to contain errors. same for karnaugh maps, half-adders, full adders, shift registers or counters or muxers and demuxers using individual flip-flops, etc.

those topics exist today because the people who wrote the books had to slog thru them, and are making you do the same. it's like a hazing ceremony.

of course, you'll use individual gates, and individual flip-flops, but usually all by themselves to provide a simple necessary functionality, but not in large complicated functions.

>>1943608
I have a £25 Antex iron from Maplins which I actually quite like, but wouldn’t one with a thermostat control be better? Also maplins has been liquidated here unfortunately.
>>1943610
I don’t have a bench scope or vco at the moment although do have both those in modular. I’m more wanting advice on Soldering/de-soldering irons and hot air guns right now but will also appropriate scope/power/vco advice or recommendations

>>1943616
oh yeah thats what i meant, my one is just a 40w with variable temp control. it doesn't show the temperature, and honestly i find such feature not very useful - just set the dial until its hot enough. One setting for soldering, another for removing laminate.

>>1943622
Sounds like the same as mine. Am I wrong in thinking that a solder station with a temp setting would keep the iron at more of a constant temperature though?(pic related)

>>1943596
The only thing that reignited an interest in the paranormal was a carl sagan video about the 4th dimension so then I went into a hmmm maybe that.

what the hell is SOP-8L? I want to use this part in a design but I can't figure out what the footprint is. Should I just pick a different part or what? Why would they not show the mechanical drawings of the part at the bottom when they're using a weird package subset without clarifying what it is?

https://lcsc.com/product-detail/DC-DC-Converters_HGSEMI-HG1509-5-0M-TR_C316675.html

>>1943735
I think it’s just a normal sop8 footprint, with the L referring to a particular height or length or plastic or RoHS status. Might also be a SOP8 with the heat pad on the bottom, but I don’t think so.

>>1942553
Wait why would a PCB have EPROMs and SRAMs on the same bus with demuxes feeding the chip-select pins? Not useless, but I'd have expected it to be easier to transfer data from RAM to ROM if they were on seperate chip-select cycles. Maybe they still are, if the inputs to the demuxers are independant on the mother board, since this is just a daughter board with a header.

I really wonder what this thing came from, since it's got ~100kB of EPROM and 20kB of SRAM. Some sort of computer that uses pin headers like this for its memory board? I may see if I can get other parts from the same machine.

How do I make a circuit that gets the charge level from a battery and outputs it to a percentage via graphical display?
I've tried hunting around online and most circuits are just using op amp comparators for some predermined percentage value. Which is fine, if I only wanted to show like 5 states. I want it to be accurate to down to at least 1%.
I know charge level has to do with the voltage output of a battery. Do I just find the max & min voltage level and extrapolate the rest in between?
I would like to make something with discrete components if the footprint doesn't end up being too large. Otherwise if there's an IC that can do this.

>>1943986
>Do I just find the max & min voltage level and extrapolate the rest in between
Depends entirely on the battery chemistry. In most cases, no it won't be a purely linear relationship between remaining energy (or charge) and voltage. The most accurate way would be to have an MCU with a built-in 8-bit ADC read the battery voltage, get the nearest value from a lookup table referring to the percentage that such a voltage signifies, and print that on a display of some kind. The lookup table is something you should be able to craft from battery datasheets.

>>1943986
This was something I was interested in with a solar charger circuit I made once...but it isnt as simple as building a voltage divider and writing a code that tells you a predicted vaule rather than the actual value. You have to do some kind of polling algorithm and you have do some kind transfer function between when the battery is underload and then no load all the while factoring the mAh of the battery.

Like the voltage divider thing can work but thats if you are just gonna give a low battery signal.

I cant give a direct answer but look at how this was achieved with analog instrumentation then adapt that whole set up into a code algorithm.
>>1943992
This.

>>1943992
>The most accurate way would be to have an MCU with a built-in 8-bit ADC read the battery voltage, get the nearest value from a lookup table referring to the percentage that such a voltage signifies, and print that on a display of some kind. The lookup table is something you should be able to craft from battery datasheets.
I think I will do this.
Question is the lookup table. Most Li-ion battery datasheets only ever list max, nominal, and min values. I don't know how I would get a voltage value for each decrement in percentage if I cna only work off of 3 datasheet values.
>>1944003
All the analog I've looked up just use comparators to compare the battery voltage to a fixed voltage via a resistor. Same shit as a lookup table and ADC really.

I think the main issue is developing this lookup table.

>>1944018
Say have you considered locating how your cellphone does it? Like download a custom android rom and locate the whole power management program files.

I get the impression though that a cellphone has a dedicated chip that only sends some kind of serial data stream to the main chip of the cellphone.

>>1944020
I forgot to add that while I suggest that...it might make you more confused.

Also when I am talking analogue I am talking like an analog dial meter.

Problem is analoge dail meters weren't how people checked the charge on batteries...it was specfic gravity.

>transducer
>datasheet said "minimum load to function = kX" and gave no further explanation
>ok this is the maximum impedance I can put it (maybe it is an open collector or something,they did not give any details on the output stage of the device)
>actually mean minimum impedance so it was the opposite
Weeew.. there goes two hours of my life to nothing

>>1943966
>transfer data from RAM to ROM
first of all, you can't do this, ROM is read only. also, it makes perfect sense to have them on the same bus because you are reading/writing data in both cases of ram/rom, so why would you have separate busses? it would double the amount of IO required on the CPU for not real gain. By using demux with chip-select, you can have up to 2^n memory devices on your bus, where N is the amount of IO pins you're willing to use for selecting data. so for even something like just 3 IO pins and an additional (cheap) demux IC, you can have 8 chips to read/write from

>>1944018
That whole thing with using a resistor as the reference signal sounds wrong. Think about if your entire circuit is drawing power from the battery that whole system would never work...why? Because that reference signal is dependent on that battery voltage signal meaning as the battery voltage drops the reference signal moves with it.

Like if you ever looked at how the linear voltage regulators work they use a constant voltage reference signal and to do this practically would require RF shielding in an analog circuit.

Something about a wheatstone bridge is poping in my head right now.

I cant make an image but basically.

The left side is the battery with a Rth resistor in series with it taking a voltage reading inbetween the battery and the Rth.

Then the right side is the reference signal.

So then you just take the difference between the left and the right.

>>1944020
>>1944023
Hmm, the ultimate goal at this point would be to see what the lookup table for a cellphone battery would be. Though, I'm reading that those lookup tables are estimates specific for the hardware and battery. I don't know if you can just rip those off.
>Analog dial meteres
Ah those things.
We pluck those onto our gear for customers (at their request)
>>1944033
That didn't even occur to me, you're right. The reference needs to be constant, and if a battery is outupping a decreasing voltage
I looked up that wheatstone bridge. I believe I've seen that sometime in my second year of EE.
A voltage regulator would give a solid reference though, but a tad bit inefficient no?

How do these cellphone companies create their lookup tables?
Surely I can just copy their procedure?

>>1944018
>Most Li-ion battery datasheets only ever list max, nominal, and min values
They often have a charging curve, which you should be able to do some mathematics on to figure out a proper energy curve. Might need to use the ESR in the mathematics, but excluding ESR (or just using a slow enough charge curve) most batteries should be approximately the same if they have the same chemistry (lithium ion, lipo (slightly different IIRC), NiMH, NiCd, PbH+, etc.)

>>1944026
>first of all, you can't do this, ROM is read only
shit, meant the other way around.
>why would you have it in seperate buses
Not necessarily seperate busses, but seperate boards so you can more easily swap out the EPROMs for ones with different programming, or upgrade the RAM, or being on seperate busses (with registers in between) so you can do RAM and ROM actions simultaneously. The first two mainly. They went to the effort to have memory on a seperate board that's easily removed, so I can only wonder why. But I've got very little hints as to what it's from.

On the back it has stickers saying:
>warranty expires: JAN 1988
>Software 103.06.00.04 (over the top of "103.04")
>Mitel Firmware Copyright 1980
>DB09478 (printed oddly, probably a serial number)
>12/3E 1988
And the PCB says:
>1096022-01 RB

>>1944046
>create their lookup tables
Trickle discharge your battery (over 24 hours or whatever such that the voltage drop across the ESR is negligible) and measure the current and voltage while you do so, ensuring the data spans from the maximum to the minimum voltages. Then chuck the data in python or matlab or whatever and integrate it to get an energy-voltage graph.
Should work with charging it too, in fact you could do both and take the average if you're really keen.

>>1944046
Keep in mind that the battery they are using is tailored for the cellphone so it's not like you could take their look up table and apply it to your design you would have figure out their methodology and apply that to your battery circuit. Like what I imagine what you actually need does require this high level. Unless your doing this as an exercise then have it.

But the thing too is that ladder comparator circuit you could not only have those indicator leds just shine a light for power level you could maybe have an otopcoupler in series with the Led and take the transistor side to control an interrupt on a MCU to change the power mode. You know turn on sleep mode when this led comes on.

Also if you know the cool thing about JFETs and how to do the constant current source trick you could make a circuit with a JFET in series with resistor and tap in the inbetween.

Again though from what I have read Constant voltage sources are in a milivolt level so sheilding is abosultely a must if your doing like a through hole thing.

>We found him hanging by his antistatic wrist strap, with the charred remains of his other hand sitting on some exposed wiring
There wouldn't be anything wrong with putting a 5Mohm resistor in my antistatic wrist strap would there?

>>1944058
just take your shoes off lmao

>>1944058
...I have no idea how this situation happens...like it seems like when take IT person and ask them to handle 5000voltage power lines and telling him that the wrist band is good enough.

>>1944052
>They often have a charging curve
These ones don't:
https://www.powerstream.com/li-pol.htm
Unless you know of a better battery vendor? My battery can be no more than 4mmx45mmx35mm due to space constraints, mainly why I'm sourcing from here.

>Trickle discharge your battery
What load should I put on it to discharge? I see you mentioned:
>over 24 hours or whatever such that the voltage drop across the ESR is negligible
Do you mean to have a load match the ESR of the battery?
But aside that, from what I'm understanding is that I would just record the current+voltage output overtime until I hit the minumum voltage level.
Then, would I scale the (current+voltage)/time graph to the rated mAh?
I don't see how you could apply this to figuring out charging graphs. Seems like that would be more complicated given you need to do CCCV charging?

>>1944056
>Keep in mind that the battery they are using is tailored for the cellphone so it's not like you could take their look up table and apply it to your design
Yes, I have come to discover this. Which is why I would just need to understand what the hell their methodology is for creating these tables.
I figure it's a good exercise as well for my own knowledge.
The optocoupler idea is neat, though I don't think I want to go the route of LED indication. I do want the battery level to be indicated on a LCD.
>JFETs
I skipped looking into how those work.. I only really have some working knowledge of BJTs and MOSFETs.
Perhaps it's time I brush up on my component knowledge if I ever want to make it in the design industry.

I do think I need to go to bed now, will check back on this thread tomorrow.

>>1944064
>have a load match the ESR of the battery
No, have the load resistance be much greater than the ESR of the electrochemical cell
>just record the current+voltage output overtime until I hit the minumum voltage level
Yes
>scale the (current+voltage)/time graph to the rated mAh
It's a bit more complicated than that, as the scalefactor changes as a function of voltage. For example, at 3.2V left, you don't have 3.2/4.2 = 76% battery left, you have basically 0% battery left. I'm too tired for maths now, but you need to do a definite integral or two. It's a really simple process to do with the collected data, can even do it using microsoft excel. It should look somewhat similar to just using a scalefactor and changing a y-intercept, but it won't match it completely.
The result of the integration will be an (energy+voltage)/time graph, which you use to make an energy/voltage graph, which is basically your lookup table.
>Seems like that would be more complicated given you need to do CCCV charging
Doesn't matter what the charging/discharging curve looks like, because ultimately the same amount of energy will have gone into or out of the cell. The integration process is just summing up the energy of each little section of time. Conservation of energy and what.

Good night!

>>1944064
>Do you mean to have a load match the ESR of the battery?
This releases magic fairies from the battery

>>1944064
Yeah BTJs and Mosfets is all I ever got into...but because I kept the electronics text book they made us buy you flip to the Jfet section and basically if you connect (bare in mind this may be wrong) if you connect the drain straight into the source like a feedback signal...the jfet will always maintian the same current draw which would be its Is max..atleast I think it maintians to Is max.

Reason why I dont know much about jfets is I dont know what a general purpose equivalent for jfets like how the 2n2904 and the 2n2222 (2N2222a I think is a premium 200ma verison) are to BJTs. And with mosfets it's the same thing cause it's the IF 510 N channel and 530 P channel...though I think those mosfets are obsolete now so the one you buy now are the remaining supply and also the low grade type.
Oh yeah this is a general comment but, I cant believe that they actually make counterfeit electronic components. Also if you build an RF circuit dont buy RF transistors until you got a protoype with General purpose...

If a lead is rated 1.25 (+/- .05) mm I should make the through hole 1.3 mm right?

>>1943986
You have to measure both the voltage and load. For most battery chemistries, the no-load voltage barely changes as the cell is fully depleted, so if you want to measure the capacity without load you have to manufacture load (in brief pulses, to avoid draining the battery). In effect, you measure the internal resistance and use a lookup table (or a formula) to convert that to charge state.

If the battery remains fitted to the device over many charge/discharge cycles you have the advantage that you can monitor its history which gives you a more accurate table/formula that takes into account product variation and ageing.

But bear in mind that measuring charge% has a substantial subjective element to it. You have to decide when it's "empty", which depends upon what voltage it needs to supply and at what current draw. You also have to decide whether the remaining charge figure should be based upon a typical discharge rate or some predicted discharge rate based upon recent usage (the higher the discharge rate, the lower the efficiency and the less energy you get out of it).

>>1944058
>anything wrong with putting a 5Mohm resistor in my antistatic wrist strap

all sanely made straps use a 1M resistor. not just for avoiding death, but also so you dont get a static zap (like when you touch your fridge in winter) when you touch a statically charged object.

Can anyone suggest what a minimum size for C2 would be if the delta fans in this circuit draw 2.2A at max speed and I'm running the PWM at 10kHz? This is to be connected to a raspberry pi and used to control the temperature of a quite hot enclosed "growing area" with a DHT22 temperature/humidity sensor

>>1944149

the size of the cap makes no difference given that the the motors have momentum, and 10Kz is such a fast rate that there will be no stuttering at all. given that any cap will do, no cap is the best cap.

but you'll need a reverse biased diode across then fans coz they'll cause a spike when you shout down.

>>1944149
220uf 35v will do.

>>1944149

>>1944149
no need for capacitor or resistor, but you need a diode. Since it is a NPN transistor you should be putting the fan in the collector and the emitter in the ground, with a diode from the collector to the V+

>>1944167
>Since it is a NPN transistor

look again, it's PNP.

>>1944082
>Reason why I dont know much about jfets is I dont know what a general purpose equivalent for jfets
J112's are dirt cheap, plentiful, and I made a 30-40MHz oscillator out of them so they have reasonable bandwidth. If they're not the standard yet, I'd vote to nominate them as such.

J201's and MPF102's were the go-to for guitar pedal buffers and radio recievers, respectively, but both are getting pricey and hard to find.

>>1944191
Oh shit sorry, the arrow was kinda thin so I missed it.

>>1944216
Thank you for that...can they be pushed into the VHF range? like talking pirate FM station boardcast area?

On radio receivers where you able to do it without this elusive tunable transformers or does this mythological "synthetic inductor" made from opamps work as a sub to them?

>>1944232
>tunable transformers or does this mythological "synthetic inductor" made from opamps work as a sub to them?
lolwut
AM is just antenna -> LC tank w/ trimmer cap -> preamp -> RC LPF -> audio amp
IDK about vhf though, simulate it and find out

My multimeter crapped out finally. I could fix it probably, but I'm at the point I want to just upgrade.
Anyone have good recommendations for electronics design?

>>1944235
I'm thinking about old school superhet recievers then...and honestly if you are getting AM that range of frequencies tend to be the most interesting. Have you ever pulled off side band demodulation circuit? You know so you could listen to the 40m voice sec5ion of the hamband.

>mfw learning Altium

>>1944236
What price bracket are you looking for? If you want something legit, then a Fluke or similarly reputable brand is the way to go, but they're 8 times more expensive than a reasonable chink meter with similar specs (ANENG AN8009).

>>1944134
I had to whip out my multimeter to test my two after that comment... goodness, they both measure .97M

>>1944161
Why the high voltage?
>>1944157
>>1944162 (Resistors on PWM line to the NPN transistor (I used 560R) and between the top PNP transistor and ground (I used 140R) noted)
>>1944167
I had two 4700uF caps lying around that I'm using for C1 and C2, I thought they are so huge that the reverse diode becomes irrelevant as when I power off the circuit it will shut down so slowly as C2 drains that there'll be no back EMF issue? Well... I'll put one in anyway. Do computer fans actually have back EMF issues, because they're brushless anyway so all the power to them is routed through a electronic speed control, and they can not be powered in reverse.

>>1944668
Wait you are running a BLDC? You just wrote "3 delta fan" and I had no idea of what that was.

>>1937846
Im trying to use an LMP7721 but I can't find it on multisim.
Is there a way to download additional components?

>>1944669
"3 delta fans" means three Delta brand 80x80x37mm fans pulled from a scrapped server. They're 1.8A at 12V which is stupidly powerful, like vacuum cleaner/jet engine powerful, but with the resistors I'm using in the circuit they'll run at a max of 7.2V which is 2.2A between all 3 of them, as I have a 2.5A 12V power supply. I actually had to jimmy them quickly into my setup as I stupidly tried to run a 12V 200mm computer fan as my sole extractor fan at 24V.... it seemed to work, but woke up the next day and it was dead.

It's funny because I was certain I was going to retire this setup when New Zealand legalised marijuana this year... I was sure it would be legalised. The final results of all the extra votes for the referendum aren't in, and it's close (46% in favour to 53% opposed) but it probably won't be, so I'll probably upgrade this setup to hydro (instead of retiring it or using it to grow poppies and vegetables in winter).

>>1944668
>I had two 4700uF caps lying around that I'm using for C1 and C2
You can get away with much smaller caps. 35V gives you some headroom (always go a little higher that normal operating voltage on capacitor voltage ratings in your design).
>Do computer fans actually have back EMF issues
Some have internal diodes across their windings, some don't. Better to play it safe and throw a diode in there.

Does anyone know if you have to disable pull up/down resistors on Raspberry Pi GPIO pins if you're using them as outputs? You'd think if it's an output pin both the pull up and pull down resistors would be disconnected, but it's a setting in config.txt. I note in possible config.txt settings for GPIO pins:
ip - Input
op - Output
a0-a5 - Alt0-Alt5
dh - Driving high (for outputs)
dl - Driving low (for outputs)
pu - Pull up
pd - Pull down
pn/np - No pull
dh and dl specify driving high or low "(for outputs)", while pu and pd do not specify "(for inputs)" making me wonder if one's supposed to manually set "pn/np - No pull" when a pin is used as an output.

It's possible that the 50-60k pullup/pulldown resistor wouldn't actually make a practical difference to an output pin... they are at the left hand side of this block diagram

I got a small development board, but it won't fit into my breadboard. Every other component I have fits into my breadboard just fine. I've tried applying a LOT of pressure to the dev board, but the pins won't fit down. Any tips?

>>1944764
I'm wondering if it's an issue of the headers being too thick.

>>1944764
Is the pitch different? Try plugging one row into the power rail.

>>1944764
I've had a lot of issues with cheap breadboards having mis-aligned or slightly misshapen metal clips under the hole, such that you have to insert the wire at an angle to get the wire to fit in-between the clips.

That board in your pic has 40 pins - it only takes a few "bad" holes to prevent it from inserting, so the statistical chance that all of them are aligned is quite low. Theres probably a few pins refusing to insert, the rest are ok.

Ofcourse assuming the obvious >>1944769 isn't the case.

>>1944769
ok, I can plug in one row into the power rail. so I just nee to find a way to get all the pins in at once

>>1944773
Now you know the pitch is the same, so get a paperclip and start poking the breadboard holes to straighten/loosen them up a bit before inserting the dev board.

>>1944774
oh, good idea. okay will do

>>1944764

breadboards arent made for something as thick as terminal pins. if you force it, you'll make the springs permanently loose, so thinner items like 1/4W and 1/8W resistors will make intermittent contact.

>>1944674
Look for the spice model on a manufacturer's website and drag it wherever necessary. There should be a tutorial for installing spice files to multisim.

where can i find a 9,6Volt battery that is at most 62mmX25mmX35mm?

>>1944864
Depending on what current you want you could just hook 3 3.3v cell batteries in series.

>>1944868
unfortunately i've got np clue what current i need i'm trying to replace the batteries in an east german radio that has no good documentation and uses batteries you can't find any info about online

>>1944864
Use a newer battery.

>>1944884
use a 7809, not a fucking zener

>>1944887
7809 output is too close to 12V which is presumably the type of battery he would choose for replacement.

How should I choose a transformer for a 12V power supply? I have found some schematics which use a 12V transformer but wouldn't that be too low due to the dropout voltage of the diodes and the voltage regulator? I'm also thinking I might not need a voltage regulator since I'm just powering a LED strip, is this alright?

Any interesting design projects for a beginner at electronics? I don't think I could the projects in the github since they look a bit too challenging. I was planning on doing a clock with a 12hr-24hr display that also shows the weekday but that sounds a bit boring now that I think about it, is there any suggestions for something roughly around that level? I would have to do it from the ground up with the design schematic etc.

>>1944933
12V transformer = 12Vrms = 17Vpk. This gives you 5V to eat up with the dropout of your regulator (say, 2V min), allow for mains voltage variation (say 1.5V max on the secondary), and take up with the rectifiers (say 1.5V with a bridge). For higher voltages it’s a bit wasteful and lower voltages it’s a bit marginal, but for 12V it’s basically bang-on. Naturally this depends on what regulator you’re using, diodes too.

For an LED strip, you’ll probably still want a regulator even with 12Vpk, due to potential mains fluctuations.

>>1944985
>Any interesting design projects for a beginner at electronics?

try this. good for young and old men alike.

>>1945003
Haha... I wish...
It has to be a digital logic design...
I'm actually pretty stressed about it haha......
Thanks for the suggestion though...

Bros, is there a "beep" display or something that makes a noise in circuit design? Is there a name for it? I want to incorporate it into a design but I have no idea if that exists.

Why are resistors incorporated in switching systems like gates and stuff? I thought that resistors are used in circuits while switching systems is more "assume ideal". I'm completely new though, so I can be completely and utterly wrong.

>>1945056
>is there a "beep" display or something that makes a noise in circuit design?

the thing you find most often are piezzo buzzers. note that some require just a DC voltage, whereas others require square waves and coils to boost the signal. you probably want the DC ones.

>>1945066
>Why are resistors incorporated in switching systems like gates and stuff?

coz transistors are wild things that will easily pass too much current and burn up without resistors to limit current.

>>1945078
>the thing you find most often are piezzo buzzers. note that some require just a DC voltage, whereas others require square waves and coils to boost the signal. you probably want the DC ones.
Thanks for the quick reply.
I was just planning to build a simple design to get myself started where a timer counts down and then at 00, it makes a beep. Hopefully it's not too complicated but from everything I've seen, it looks very complicated.

>>1945043
How about a seven segment hexadecimal counter? Or a hardware CRC or some other finite-field algebraic gizmo? Multiplication ALU? Something with ASCII and/or RS232?

>>1945084
>How about a seven segment hexadecimal counter?
That sounds good! Could you elaborate on it? Out of the stuff you listed, the only thing I know is seven segment displays and counters...

>>1944890
He wants 9.6V, 8.3 is either enough (great, use an LM317) or it isn't (use an LDO). Then again, the poorer regulation from using a zener doesn't really matter since it's replacing a battery. I guess he should just use whatever he has the parts for.

>>1945115
So, a normal 7-segment decoder will turn BCD (binary coded decimal, 4 bits from 0/0000 to 9/1001) into the 7 signals needed to feed a display, but often circuits don't have BCD at all. If you want to display a multi-digit (more than 4 bits long) number, then you'd need to do some (frankly stupidly complicated) digital logic maths to split the number into seperate BCD streams. On the other hand, you could just take each set of 4 bits and treat them as a hexadecimal number (base 16; 4 bits from 0/0000 to 15/1111), and display them like that. Not so useful for a human to count, but fine enough for some purposes. Using "A, b, c or C, d, E" you can display 16 different digits as commonly done in hexadecimal.

In order to make a hexadecimal 7-segment decoder, you'd need to write out (or automate via based logisim) the 4-input karnaugh maps for each of the 7 segments, and implement this in either discrete logic, some sort of monolithic logic array, or an EEPROM (see a Ben Eater video where he uses an EEPROM exactly for this). Note that it's quite a lot of discrete logic to do, especially only using 2-input gates.

Then the "counter" part is just some peripheral to make use of the decoder, like two buttons to increment or decrement the count. Could also put it alongside a parallel ADC to measure an analog voltage or sensor. Or any other parallel binary value for that matter.

>>1945127
>In order to make a hexadecimal 7-segment decoder, you'd need to write out (or automate via based logisim) the 4-input karnaugh maps for each of the 7 segments, and implement this in either discrete logic, some sort of monolithic logic array, or an EEPROM (see a Ben Eater video where he uses an EEPROM exactly for this). Note that it's quite a lot of discrete logic to do, especially only using 2-input gates.
Ooooh, I did this before, it was an insanely nasty design at the very end...
>Then the "counter" part is just some peripheral to make use of the decoder, like two buttons to increment or decrement the count. Could also put it alongside a parallel ADC to measure an analog voltage or sensor. Or any other parallel binary value for that matter.
I don't really understand this part, I thought counters was in essence just a timer... I need to study this more desu, thank you anon...
I have a month to plan all of this out!

>>1945136
>I thought counters was in essence just a timer
I'm referring to binary counters (ripple or synchronous), where each clock pulse results in the outputs advancing by binary 1. See the wikipedia page for "counter (digital)" For an up/down counter (74HC169 for example) the circuit needs to be more complicated, and synchronous. Have a look at its timing diagram.

>>1944065
Wait, I'm not understading how a (current+voltage)/time would give you an (energy+voltage)/time. I know I=dQ/dt, integrating that would give you Q. Also the way I'm understanding a current+volatge)/time graph is literally 1 graph that represents current/time and overlaying that is a voltage/time.
I don't know why this shit isn't making sense to me.
>>1944082
Very interesting, I might have to look into jfets if maintaining a constant current draw is something they're capable of.
Not that I really have much else input on them to tell you the truth. Again, just something I have to read up on.

Apologies for not replying sooner.

>>1945175
divide part of the I/t + V/t graph into an arbitrarily small chunk of width dt
this effectively has a flat top with length
I(t)*V(t) = P(t)
dE(t) = E(t+dt) - E(t) = P(t)*dt
in other words:
E(t) = int[ I(t)*V(t) ]dt, where t = t_start to t_fin
I think so at least. Add some boundary condition because theres a +k in there maybe.

And charge*voltage = energy too, so you could integrate to find the graph for charge Q/t then integrate Q/t with V/t, but that's one extra step.

>>1945003
You just KNOW someone has seriously crippled themselves trying to make a DIY sex toy.
>Local man found dead in bizarre suicide pact. Doctors say he applied mains voltage directly to his prostate.

>>1945127
>karnaugh maps
Never heard of them before but they look neat.

>>1945179
I had to stare at your math a little bit there. It's been a while since I messed with calc and physics equations by hand.
I do understand;
>dE(t) = P(t)*dt
Since E=VQ & P=VI, you can work it out to the above.
I also understand;
>E(t) = int[ I(t)*V(t) ]dt, where t = t_start to t_fin
Is just the resulting integral.

But I'm not sure how you got;
>dE(t) = E(t+dt) - E(t)
I take it this is just a calc expression?

>>1945201
dE is the added energy over that tiny timeframe dt. 10V and 1A makes 10W, 10W over 1µs is 10µJ. The energy after that 1µs (say N=151), E(t+dt) = E(151µs) is going to be 10µJ higher than it was before (N=150), E(t) = E(150µs). The accumulated energy is the sum of each step.

>>1944862
Ok thanks

On a scale of asperger's to downs how retarded would I be for trying to make my own headphones?

The drivers are the most important part right? So if I were to go on aliexpress and buy some replacement drivers for good models I could bypass the brand markup.

>>1945242
Wouldn't the most important part be the headphone case? If your headphones aren't ergonomic/comfortable you'd be better with speakers.

>>1945245
I plan on 3d printing the enclosure so I'm not too worried about ergonomics.

>>1945144
Timing diagrams are my worst subject but I think I can handle it, thank you anon!

Is there a reason why one would use an active-low decoder over an active-high decoder? My book explains an active high decoder and then starts showing a bunch of examples for active lows such as the 74138.

>>1945293
low = 0 = off
high = 1 = on
Active low encoder pulls outputs low/0/off.
Active high encoder pulls outputs high/1/on.

>>1945299
Oh shit, I can't believe I went full retard on that. In the real world, we would use both of them at the same time to make sure we cover all of the options available or something of the sort?

>>1945304
What is the decoder controlling/communicating with? Does it expect high or low signals?That's the real question.

>>1945303
>In the real world, we would use both of them at the same time

no. if you used a decoder to select memory chips, for example, you need to find the one that matches your chips, some of which are low-enabled, and some are high-enabled. it's just catering to everyone's needs: black, white, yellow or normal. LGBTQ+ or normal. republican or normal.

>>1945306
That's a really good point anon, thank you. I forget that this isn't just a closed system by itself.
>>1945320
Thank you too! It's based on the customer's preference as well as what the other surrounding gates/ICs use.

Okay quick question, why is the least significant bit chosen as the "master" in a decoder controlling multiple other decoders and not the most significant bits? Like with inputs ABCD, CD would be used in the decoder controlling the other decoders that have AB inputs.

>>1945293
The decoder's outputs needs to match the inputs of whatever it's driving. Conventionally, "chip select" signals tend to be active-low, so active-low decoders are more common.

The main reason for preferring one convention over another is related to open-collector outputs, which either actively pull the output low or passively allow it to float high. You can connect multiple open-collector outputs together, with the result that the voltage will be high if no output is pulled low and low if at least one output is pulled low. So you essentially get a free gate; whether it's AND or OR depends upon whether you consider the high or low level to be "true": A∧B=¬((¬A)∨(¬B)). The concept is usually termed "wire-OR", due to the fact that the low voltage is the "active" state.

I just spent a lot of time trying to track down the source of a random low-frequency noise in an amplifier.
Turned out the noise was coming from the mains wiring in my house. It's a random noise in the range of 4-20 Hz, low enough to make it through the filter capacitors. And I do mean random, it's non-periodic random very low frequency noise.
I didn't notice it until I started working on tube gear, where stepping it up to the 300v range and rectifying without a voltage regulator made it obvious.
Is this normal? I can't imagine it's normal. I can't seem to find much info on this because searching "low frequency noise" in power lines just brings up discussions about 60 or 120Hz noise. What's going on here is quite different.

>>1945364
How did you determine that it was coming from your mains wiring? Have you tried flipping all the breakers off at the panel and going circuit-to-circuit trying to narrow down the source, or do you think it's coming from outside your house?

What type of switch do I need for this? I want FILT1 to be engaged in one position and FILT2 in the other.
Thanks

>>1945426
SPDT
single pole, double throw

>>1945420
>How did you determine that it was coming from your mains wiring?
I tore the amp down to everything but the power supply and it was still there. I even removed the rectifier tube and jumped the transformer out to an external diode rectifier and capacitor and still saw it.
Then I plugged in a scrap 48v transformer, again rectified and filtered it, and still measured the same noise. Smaller, because the voltage of the tranny was lower, but it was still there.
This isn't the first time I've seen it, either. It was never an issue before because it was easily rejected by voltage regulators so I just ignored the mystery until now.

I'm going to try flipping breakers tonight, after the rest of the building empties out. I suspect it's coming from outside the house, since the noise is so random I suspect it's being caused by load fluctuations elsewhere on the block.

>>1945436
is that a digital CRT scope? My one is less digital looking than yours, I mainly use it in analog mode since it updates quicker, but digital is handy for single-shot capturing and such

>>1945474
Agilent 54622D. I love this scope. From the early 2000s when CRTs still had superior resolution.
100MHz, 2 analog channels + 16 digital. With a very slick UI.
One downside is that schematics are not available, so when it breaks it'll be a challenge to repair.

When I want "analog" performance I keep the time reference at left so it doesn't waste as much time capturing pre-trigger data.

>>1945497
>16 digital
That would be handy, though considering it's 100MHz I'd probably want to get a dedicated logic analyser that's fast enough to scope USB.

I've got a Hitachi VC6545 from the 90s(maybe 80s), which is a proper hybrid scope with a vector display. In analog mode it pipes the signal right to the CRT plates, in digital mode it goes through an 8-bit ADC then back into a DAC, for storage and single-capture. No math functions, let alone an FFT, but it's got two cursors and an RS232 on the back that I can get the data from a digital capture out of if I really want. It's in HPGL plotter format though.

I have a service manual for it, sort-of. It's for a slightly different model, but it seems to be accurate.
>https://wiki.london.hackspace.org.uk/view/Equipment/Hitachi_VC-6545
It always boots in 10x probe mode after it's been off for a while, instead of remembering it like it should, I suspect there's a mostly dead coin cell in there.

>>1937846
Is there a good resource about camera modules (for a home security system)?
The main question I want answered is why is footage captured with a 2MP CCD so low quality when that's enough pixels for HD footage.

>>1945601

megapixels arent the whole story. the other part of the story is the size of the sensor. a half-inch sensor will collect more light than a 1/3-inch, and the diff can be huge when light conditions arent ideal.

besides sensor size, there's optics. if you want good quality, you got make sure those are good too. if the thing is cheap, you're gonna get 640x480 quality at whatever number of pixels they brag about.

>>1945601
>>1945611
This anon is right. Assuming the optics aren't absolute dogshit, camera ISO will make most of the difference. While ISO is a term left over from the days of film photography, it generally refers to the sensitivity of the CCD. High ISO means less light required to make an exposure, but it also means more noise, making a lower quality picture. Most digital cameras will adjust their ISO depending on light conditions, hence why your cellphone camera pictures taken at night all look awful. If you want to use a shitty webcam or whatever as a camera, the easiest way to improve the picture is to increase the lighting on the target. This is usually done with visible light, but in the case of a security camera, they're often sensitive also to IR, meaning you can illuminate the foreground with IR LEDs. Decreasing the framerate to prolong each exposure might also be an option.

If you're still choosing your CCD, higher quality CCDs and ones with greater area will be able to take better pictures under the same lighting conditions and framerates. But I've never tried to buy a camera by its CCD before, so I've no clue how easy it is to specify for such things.
Technically aperture width is also a factor, but I've no idea if those small monolithic camera modules even have variable apertures.

>>1945420
>>1945436
Are you able to use a voltage regulator at those voltages? Like something using a TL431 and a pass transistor, or maybe just a capacitance multiplier or two.

I've got no clue what could be causing it, besides maybe schumann resonance (had any lightning near you lately?). The ELF band is really not something that's very loud or easy to pick up as a radio signal, but neither is it often produced by electronics on the same grid. And if it's random then it probably isn't some form of power line comms. Could maybe be demodulation of a radio signal, or mixing of two radio signals, or something like that within imperfect wiring within your house, which I've heard about but never really understood.

Try building yourself a 50/60Hz twin-T notch filter to scope the mains before rectification, maybe you'll be able to hear some lightning strikes while you're at it.

Im trying to get a phone game to play itself for cheap and used a double aa two paperclips and a penny to sit on a touch screen button. its detected but i need it to pulse in order to work.
-would slapping a capacitor on one paperclip create a pulse?
-is there a way to make it not heat up whenever i touch the penny?

>>1945226
I think I understand you now given the graph you whipped up.
You're subtracting the area if E(t) from E(t+dt) in order to get your segment of the graph.
I thank you for explaining things. Will definitely apply all this when trying to make a percentage lookup table.

>>1945683
It detects touch via change in capacitance. I assume it scans the row and column electrodes by feeding them AC and measuring the reactance. Passively speaking, if you keep the contacts electrostatically isolated from the phone's ground and increase the capacitance, you should be able to spoof a touch. But electrostatic isolation sounds tough, I'm not sure the kind of capacitance you need to avoid. If it senses a touch on the order of 1nF or higher, then it should be pretty easy, but if it's 100pF or lower then it will be pretty tough. Check the capacitances of MOSFETs, JFETs, and other switches. Might be able to spoof a low capacitance with feedback from an op-amp circuit, which sounds fun.

Active methods (trying to feed it AC of the right frequency and/or amplitude) are probably not worth the effort.

where should I put a trace that uses vout when I have capacitors connected to vout like this? does it even matter?