/diy/ - Do It Yourself

>>1663694
How do you know they're all like that? At the very least they don't have bad sound quality from what people who've bought them say.
Though now I think I might buy a $20 condenser capsule and make my own housing and board for it. That way I won't need to bother with baluns or anything like that. Any reason I should make a circuit like that with discrete transistors instead of just using a JFET op-amp? I don't see any reason to use differential signal lines either.
I also have a space blanket I bought specifically for making my own condenser, which I guess I could attempt again. Some microphones use the condenser as part of an LC oscillator, and use FM demodulation to receive the data as opposed to having a high bias voltage. Which would be a nice first project with PLLs.

>>1663702
>Electic sleep
Was it a meme? Why it was a thing only in USSR? How 50-100 Hz on body would fix anything?
I think Americans had better electric sleep device.

>>1663657
>The audio transformers I see aren't exactly too expensive, but I suspect they'd be more prone to noise
in a metal box, probably not enough to matter. it's a question of how you want your sound colored
>48V rails
just put some nice caps in there. dc level is mostly irrelevant to the output anyway
>diff amp circuit
I probably wouldn't, only because I can't be sure how long a cable I might like to use with it someday
>controller on secondary side
much more effort than it's worth. most of the controller's concerns are primary-side, e.g. switch current, switch temperature, line voltage, and that's before you consider the bootstrapping problem of getting enough energy over to the secondary to bring the MCU up. could indeed be worth it if you're doing some quad-pumped arc welder kind of thing
>switching patterns
the weirder the pattern, the more bandwidth needs to be covered by the input filter needs to be to prevent using your house wiring as a broadcast antenna. a touch of spread spectrum could be welcome
>rewind
hope you have room on the bobbin. neatness counts~
you might find transformers for closer to 50 cents each if you shop a little harder. as long as the operating frequency is about the same or lower, you're probably in good shape to substitute. worst case is reduced output current due to core saturation, switch current limit, and/or thermal shutdown

>>1663694
>no electret bias
that doesn't look quite right

>>1663688
>lower voltage wall supply and boost it
this is probably the correct answer. either LM5001 or LM5002 are probably the correct IC, but expensive from ali, even worse from Digi-Key, and Jaycar sends Pic related to your house to collect the price of one

>>1663724
Have you heard of the magic that is RS

>>1663712
Suggested modification:
Large externally polarised condenser capsule
Omni - Cardioid - Dipole (figure of 8)

>>1663702
What's the translation for this again?

>>1663724
>that doesn't look quite right
Electrets themselves don't require any bias, it's the jfet inside their package that needs it. A normal condenser has an output signal that's proportional to the electric field, i.e. proportional to the DC voltage across it. But the electret produces its own electric field without any voltage, so it should produce a weak signal without needing the bias. Question is, if they're using one of those, why bother with 48V when 12V would more than do the trick? If they do have 48V, couldn't a cheap condenser be cheaper than an electret?
If it is a normal condenser, it's arguable that the DC bias is coming as leakage current through the JFET, but I'm unsure.

>>1663728
What is that trying to do?

>>1663702
Rule 4 is null and void
The name field is there to be used and you can smoke what you want.

>>1663732
It's a deterrent for attention-whores

>>1663731
>What is that trying to do?
improve linearity and sensitivity
allow selection of characteristic

>>1663731
>What's the translation for this again?

>Original schematic of device "Electronic sleep"
>Schematic modified by our reader.
FYI, electronic sleep is some sort of placebo? device that is used in exUSSR. Pic related I guess

>>1663712
>Any reason I should make a circuit like that with discrete transistors instead of just using a JFET op-amp?
JFETs are cheap. JFET op amps suitable for 48V, a bit less so
furthermore, the sound will be colored by every component you place in the audio path and some that lie outside of it. op amps are not very colorful and they don't clip well, which makes pop noises especially unpleasant
>I don't see any reason to use differential signal lines either
avoiding common-mode interference doesn't count? you might regret it later
under general circumstances, not knowing whether the device on the other end is chinked in a like way is a close second
>first project with PLLs
maybe, if you can find some that are fast enough, or you're just miking a kick drum. it's possible you might have to dip into FM radio parts to find a demod that is suitable. I'd probably be cheap and try a discriminator first, who knows, maybe it'll sound good if I get it tuned just so and hold it just right

>>1663731
>the "electrosleep" as designed by our authors
>reader submitted modification
loosely from memory because I'm too lazy to hunt and peck Cyrillic rn
>why bother with 48V
it's the standard for phantom power in pro audio for half a century. it's Norway's fault, informed to a lesser extent by the rotary telephone

>>1663734
engineers, on average, are known to give much more consideration to "can we" than "should we"

>>1663734
Freedom of speech
The real attention-whores are the entitled goose-steppers and their triggered conformism can also be tolerated

Is psychotronics the future?
The Suckerberg thinks so,
because it works both ways.

>>1663746
you know what's better for sleep? not sleeping in your work uniform

>>1663731
it appears a new revision of IEC 61938 recommends 24V for new systems

>>1663752
what catalog is that, Sharper Image or some other hipster shit?
>Hammacher Schlemmer
close enough
>Zuckerberg thinks so
you don't get to be a billionaire by being honest. merchants gonna merchant

>>1663747
In that case, it might be worth buying the BM-800 even if it's got an electret, because it starts with a decent circuit and I can replace the electret if it has one. Not to mention the stand and pop-filter it comes with. But we're still at square 1 with the power supply.
As far as a custom circuit goes, I'd likely run the voltage a bit lower or have a seperate 12V rail for my more sensitive electronics. I could also use coax instead of differential twisted pairs, which I think would help.

>>1663749
I mean, there's nothing stopping you from continuing to post with a name or tripcode or whatever, more power to you. But as there's little point to doing so on this general the rule is more there to let people know that it's frowned upon, not socially acceptable, however you want to put it. Make the thread next time if you want to delete it I guess. At least you read the fucking OP, unlike 90% of the newfriends that turn up here with junk they don't know what to do with.

>>1663762
Holy shit:
http://www.ti.com/lit/ds/symlink/tl783.pdf
>125VDC max adjustable linear regulator
That's pretty damn good.

Can someone help me diagnose this battery charger? The 7 segment display should light up when its receiving power and from there I should be able to press one of the two chink buttons to set how much voltage I'm charging the battery with.

When I touch the positive lead for my multimeter to the positive Input terminal on the charger, I get the exact same full voltage reading no matter where on the charger I touch the negative lead. I got the display to turn on once by shorting the relay but I don't want to destroy the whole thing by doing it again

>>1663762
based on the schematic in the last bread >>1663694 it should work fine on 24V phantom power

>>1663789
intredasting. I did not know linear regs went that high

>>1663807
If it's an electret-based one it should work fine on 12V too. Whether or not it has an electret or not will be pretty easy to figure out just by changing the voltage I feed it. If the audio amplitude varies linearly with voltages from ~8V to 48V than it's a condenser model, if the audio amplitude increases sharply from the standard 1.2V to 8V or however much it takes to get the circuit fully-biased but then remains mostly constant from 8V to 48V then it's an electret model.
Actually I could potentially ask for an anon on /csg/ who already owns one to do this test.

>intredasting
Yeah gonna buy a pack of 5. Be nicer if there was a 400V one I could use on rectified 230V mains, you bet I'd use that to get a >1Vp-p signal amplitude from a condenser cartridge right off the bat. Be useful for testing nixies and the like too.

>>1663811
are you overthinking it or underthinking it? look at R1/C7/D3/Q6
>400V linear
there's a point at which dropping power in a linear regulator makes no sense, and I'd say that TL783 is probably close to it. but there are 600V npn transistors without an insulated gate, if you insist
https://www.youtube.com/watch?v=8j9ejovA7qw

>>1663815
>it's a 9.1V zener with an emitter follower
never mind it's obviously an electret model, though since they're using a resistor instead of a constant current source the voltage will vary slightly from 12V to 48V supplies
Plus it can't be modified into a condenser model, which is kinda a buzzkill. I'll likely still get it though, and just run it off 12V, which will simplify things. Not sure if I can fit a linreg in between ~9V and 12V with room to spare, so running it off 24V or something in between would give me some peace of mind, and let me use that obscure LM317 package I've got lying about. Plenty of room for jfet opamps too, if need be.
The upside being, I get all the hardware (anti-shock mount, articulated boom arm, pop-filter) that I can use with an actual condenser mic, and I can stick with the meme for now.

>there's a point at which dropping power in a linear regulator makes no sense, and I'd say that TL783 is probably close to it
If you're dropping under 100µA I think it's still sensible. Got my packs of 1M and 10M resistors handy. 600V NPNs are all well and good, but getting zeners at that voltage is a bit of a pest. I wonder about just using 1N4000 series diodes in reverse for multiples of 50V, with a 10M resistor or so.
>without an insulated gate
I don't suppose FETs can work as voltage buffers, can they? I've got some IRF840s lying about.

>>1663821
>emitter follower
If you can't stand the noise add R2 and C6 from the modified version.

>>1663824
Oh that's right.

>>1663801
>diagnose this battery charger?
A circuit diagram would help more than megabytes of dark blur.

>Crude sketch incoming
Is this a viable circuit.
Basically I want to be able to press a non latching button that will turn on a relay and time out after a set time.
Time period is about 40 minutes so a 555 timer in one shot won't work.

So I'm thinking 555 in astable running at a period of around 20 seconds.

Pressing button sets a latch that powers the relay and switches a FET allowing the 555 output to input to the clock of a binary counter.

7th stage of the binary counter resets the SR latch, causing the relay to close.

Ignore the XOR gate.

I'm aware I could just buy a delay off relay

I didn't know this thread existed. I've already made my own thread but no one who replied seems to know the answer. So here it goes:

I just discovered I've been trusting my equipment with a fake line filter and surge protector. I barely know any electronics and I can easily tell there are no inductors, no beefy capacitors and no varistors. It's pretty much empty.

I'm working on DIYing this into a true line conditioner and surge protector—I've already bought some capacitors, thermal fuses and varistors according to a few schematics I have.

Before I start doing the mod, I'm doing the following courses:

1- Lynda's Electronics Foundations: Fundamentals
2- Lynda's Electronics Foundations: Basic Circuits

Despite already knowing a tiny bit how a capacitor works and some of its uses in DC circuits, I cannot figure what in the fuck this small-ass capacitor is doing in this AC "line filter" circuit. Can you help me understand why did they put that capacitor in there?

1- Why is it not connected between hot and hot instead? What would happen if that was the case?
2- What is the purpose of this capacitor in that configuration? Does it filter noise, or fills the gap for microsecond interruptions?
3- Why is it not connected at the beginning, before the current hits the outlets? Does it even matter?

Here's how it's wired without illustrations.

Front interface.

>>1664037
>Why is it not connected between hot and hot instead?
because it would do nothing
it would be shorted/bypassed by the red wire marked HOT near the center of your pic.
the WIRE that goes from HOT to HOT
>>1664037
>What is the purpose of this capacitor in that configuration?
electrical noise filter - but not much of one
>>1664037
>Why is it not connected at the beginning
electricity travels at near the speed of light and the line frequency is 50 or 60 Hz
at this frequency, it doesn't matter where it is

>>1664037
>a fake line filter and surge protector
>>1664039
>Front interface.
I see "Line Filter" and what appears to be some language for "Electronic Protector" but I don't see any reference to "Surge Protector"
You got what you paid for.

>>1664037
>case isn't grounded
>that strain relief
thats a big nope from me lol.
>I just discovered I've been trusting my equipment with a fake line filter and surge protector.

to put it another way you discovered that probably every other person in the world uses some shitty knockoff like this and literally nobody ever has had a problem before.
its almost as if electrical surges aren't a big problem. or even exist at all.....

if your electric line is above ground then you need some serious business to take care of lightning strikes. nothing you can fit in that box is going to do the job.

if you have a below ground service cable entry then guess what you don't need it.

>>1664088
>electricity travels at near the speed of light and the line frequency is 50 or 60 Hz
at this frequency, it doesn't matter where it is
>>electricity travels at near the speed of light
lol no

>>1664039
Every decent electronic device nowadays has a MOV. And why do you want to filter your mains?
>br
é nois

>>1664135

>>1664143
That is very wrong, google what Velocity factor means. Lossy medium (copper) and insulators makes EM waves travel way slower than c.
t. comms engineer

>>1664020
The 555 can do 40 minutes. The time is 1.1RC so 10MΩ and 220µF will get you 2400 seconds. The optimal IC for long time delays is the 4541 and another option is the 4060. Both have a built-in clock oscillator and a binary counter.

>>1664138
True, but not all of them have a MOV. When they do, the MOVs are usually small and can't take too many joules during their lifespan. They will explode easier with violent surges, especially if the circuit is simple and doesn't include fast acting thermal fuses beside the MOV. Small MOVs also tend to have a shorter lifespan when used in power delivery for 220V AC, because, like I said, they have a very limited capacity for transient absorption. As a MOV takes each transient hit, it degrades and it's clamping (operating suppression) voltage is reduced. It will eventually overheat at nominal operating voltage, and explode or catch fire if not protected by a thermal fuse. Plus, one MOV is not enough. You need at least three for basic protection. This is because to achieve both common mode surge suppression and differential mode surge suppression, you do need at least three MOVs. See pic related.

>>1664152
Oh, I forgot to mention:
>é nois
Tamo junto, piá! Saudações de Joinville, SC.

>>1664088
>>1664091
Thank you for the clarifications. Here in Brazil, even decent surge protectors with proper varistors (MOVs), thermal fuses and gas discharge tubes (GDT) are mainly referred as "line filters" by the companies that sell them. The higher end ones are referred as line conditioners in Portuguese, just like in English.

Also, thanks to an anon on the other thread, I got pointed to a Reddit page with good info on this: https://www.reddit.com/r/ECE/comments/47lmqy/why_is_there_always_a_capacitor_across_hot_and/
So that capacitor in my fake line filter is serving the purpose of a decoupling capacitor. Its main functions are to prevent arc formation when the on-off switch is set to off and filtering high-frequency noise.

>>1664124
>to put it another way you discovered that probably every other person in the world uses some shitty knockoff like this and literally nobody ever has had a problem before. its almost as if electrical surges aren't a big problem. or even exist at all.....

Careful with that thought. I lost the onboard Ethernet adapter on my motherboard last month during a thunderstorm. Power went out three times, 5 minutes intervals on and off. Third time power went back: no Internet. No network adapter on Device Manager. Did a whole day of troubleshooting, no success. It's dead. I had to buy an off-board PCIe Ethernet card to get my Internet back and, once the card arrived, I did. The supposed surge probably came through the coaxial data cable, but if that was the case I find it weird that the modem is still working perfectly. Who knows what really happened. Could be related to the thunderstorm or not.

My country is top dog when it comes to lightning and surges:
"In particular, lightning forecasting is very important for Brazil due to the large incidence of lightning in the country. Every year Brazil has experienced about 50 millions of ground lightning strikes (Pinto 2009). This number puts Brazil in the top of the list of countries in terms of lightning incidence in the world, followed by the Republic of Congo in Africa and the United States of America. As a result of the large incidence of flashes, on average, 120 people are killed by lightning annually in Brazil. In addition, lightning causes more than one billion dollars in losses every year in the country (Pinto 2009). Brazil is the seventh country in terms of lightning fatalities." - Pinto Junior, Osmar et al. "A 24-Hour Lightning Forecast System In Brazil." Journal of Aerospace Technology and Management 7.4 (2015): 396-397. Web. 11 Aug. 2019.

>>1664124
Also, keep in mind that not all surges immediately kill your electronics in just one hit, and a lightning doesn't have to directly hit power lines or even the ground to produce power surges. Cloud to cloud lightnings do induce powerful surges in the electrical grid through big electromagnetic fields. Everyday surges are in the hundreds. They are produced internally by the operation of your household appliances and externally by atmospheric discharges, outages and maneuvers from your power distribution company. Give this a watch: https://www.youtube.com/watch?v=iMUuVqB1vI8

According to everything I've read so far, the cumulative long-term effect of noise and numerous everyday smaller transient surges can be just as fatal to electronic devices as a direct lightening hit on power lines.

>>1664149
Yeah, in theory the 555 can do large periods but you need electrolytics for that large capacitance and when you start charging a 220uF aluminium cap through a 10M resistor, the leakage current is large enough to never charge the cap enough

>>1664163
Don't assume what you can measure. Simple experiment: charge the cap with your 10MΩ multimeter as series R.

>>1664020
y tho, this is the sort of task that PIC10 and ATtiny4 live for
but if you must into discrete logic, maybe? did you simulate it? the XOR seems totally unnecessary here, the Q output itself should be enough. also it's usually bad form to gate clocks directly, use an enable or async-reset input on the counter if available

>>1664038
it's not fake. it just uses air discharge to stop excessive surges

>>1664147
it's pretty near on the log scale, by comparison to sneakernet

>>1664179
>measuring a single random copy instead of reading a datasheet for guarantees
3shiggy5me

>>1664179
testing this now with a 680uF 250V capacitor. we'll see if it's anywhere near its RC curve in 2 hours.

>>1664214
>it's not fake. it just uses air discharge to stop excessive surges

WHY ARE SWITCHES SO CLICKY?
Click to find out! -> https://www.youtube.com/watch?v=jrMiqEkSk48

>>1664220

note that you can cascade 555 monostables in a chain of any length. each stage adds its own delay to the total delay.
in the attached pic, the last stage feeds back into the first, so it restarts the sequence. you probably dont want that.

>>1664252
>555
>2011 + 8
cringe/10

>>1664254

wont let any of those filthy micro cuntputers sully my noble workbench.

>>1664260
>being this 13
>on an engineering thread

>>1664220
with a 3v supply the voltage across the cap has increased to 2v after 2 hours. pretty close to ideal RC. good job rubycon.

Can anyone tell me how to get an FFT cutoff-style low-pass waveform out of passives and such? I want to get rid of the sharp edge at the start of the transition as well as at the end. I want a smooth waveform to minimise the bandwidth of amplitude-shift keying.

>>1664252
>dealing with the awful quiescent current, pinout, and inability to vary duty-cycle independently of frequency
>not using schmitt inverters and digital logic for timing purposes

>>1664295
an inductor, obvs
>and such
define
>using discrete logic to slowly charge a capacitor
if you don't care whether your timing is accurate, why don't you just set a mechanical alarm clock

>>1664305
Could you post a circuit? Doing that (in a sim) just seems to limit the gradient at the start of the edge, but still leaves the corners sharp. Also it has a ringing problem.
I tried a circuit where I used an op-amp integrator to turn my waveform into something of a triangle wave, then low-pass filtered it to round over those corners, before differentiating it again, which just gave me a trapezoid-wave. Not awful, but it still has sharpish corners on it and it's not the most stable circuit.

>>1664313
CW radio operators have this problem too. if active elements are fair game, try Pic related

>>1664317
here's another one that might provide the waveforms you're looking for

I'm starting to get more into electronics so I bought a couple nice meters, I'm looking to buy an oscilloscope but I dont know what's a good beginner scope, there's a tektronix 465 for sale locally for 60 dollars, is this a good option?
People seem to be saying the rigol is good but that's been on the market forever, has anything changed?

Ok so now I'm powering my rs232 hpgl reader with the low-current regulator directly off the rs232 line itself, but the voltage has dropped from 5V to 3V due to the damn power LED on the arduino itself. No clue why they make the things eat up a full mA when 100µA should easily be enough. Didn't even know a 5V regulator could do that, the sooner I get this onto a standalone MCU the better. It might be still working however, since the CTS LED is lit. It may or may not be toggling once about every second, which would mean a particularly long transmission time for 45000 bytes (over 12 hours). I'll leave it for 30 minutes and check on the text file then, but chances are I'll have to do the testing after I've dumped it onto an ATMega IC.

>>1664397
It's still going. I was correct, it transmitted 2300 characters, which is about 1.2 characters per second. I guess it discharged the capacitor enough with each write (or read) action that it needed to wait that long for it to recharge before the CTS line was recognised as being high. If this is still a problem without the LEDs I can see about toggling the CTS line via a transistor to the 12V rail (which is currently at only 3.9V).

I want to put together a bunch of 18650 cells to make a battery bank.

My confusion is how am i supposed to charge them? Since they are supposed to be balanced and shit.

A single cell is easy, i just shove like 1C through it with a cc source until it raises to a certain voltage
But how do i balance charge multiple cells? Do i really need to balance them them? Otherwise i could just have them say in series and shove 1C through them collectively with my cc source, if i have 3 cells in series

>>1664397
>100µA should easily be enough
probably not in a classroom setting

>>1664421
you use a protection board with balancing and/or a balancing charger. happy shopping

>>1664423
>balancing charger.
not thats normie shit, i am going to do it myself

>>1664425
I'd start off by looking at what balancing IC datasheets, preferably ones with block diagrams in them that you can recreate with discrete analog and logic ICs. Same thing applies to the charging and discharging ICs, though, the discharge protection IC should be easy enough to figure out even without a block diagram. Sounds like it could be a fun project, if not a practical one.

>>1664425
>collectively with my cc source
Abnormies can do that. Limit to 12V or so, don't charge unattended.

>>1664426
Seems easy to me.
A wire at each battery contact going to an array of mosfets and then i simply open the two mosfets for the battery i want to juice up

>>1664428
Ablic (formerly Seiko Semiconductor, iirc) makes protection ICs for battery packs that can turn the MOSFET on and off automatically, see S-8209B

>>1664428
Don't forget that open=close and close=open. Welcome to Memeland.

>>1664434
it makes sense because pipes are negative space while wires are positive space

>>1664441
interesting. is coax double-neutral or a casus sui generis?

>>1664372
It's good to start out with an analogue scope, since you'll be forced to learn the basics without the aid of magical auto buttons. It makes you more proficient in the long term, but I wouldn't get a CRT scope for productivity.

Get a Rigol 1052E if you want to cheap out, a DS1054Z if you want to cheap out a little less, or a Micsig tablet scope with buttons if you want to be mobile:

https://www.youtube.com/watch?v=C7Ee0wYtAZY

Also yiff in hell

>>1664512
just because you learned on a CRO, doesn't mean everyone should, old timer
>>1664372
just buy a cheap digital scope, you're definitely going to be buying another in the future, anyhow

>>1664582
Everyone should get an analogue because they have a super schway aesthetic d e s u

>>1664582
You can skip them if you want. I think it's useful to get the hands-on understanding of how timebases and triggers work, well worth $60.

Again, I wouldn't recommend one for anything but learning, or special applications. A cheap DSO is going to get you much further than most analogue scopes will.

What the fuck is this shit?
Wuth usb 2 you only have 4 cables and you short the data pins to tell the device it can draw as much current as it wants
How the fuck do i mage a simple charging cable from this shit?

>>1664606
http://www.ti.com/interface/usb/type-c-and-power-delivery/products.html#p1694=PD%20controller

>>1664606
If you just want 5 volts you don't need to do any active communication, just a resistor or two. For a 5V sink put a 5k1 resistor between each of the CC pins and ground, for a 5V source put a 10k from CC1 to 5V to declare up to 1.5A, 22k to declare 3A. And short the D+ D- for legacy devices.

>>1664600
I'm meeting the guy today, gonna offer 40 bucks lol
Itll be a cool piece to have around if anything

>>1664641
Oh no my name fell off

>>1664643
Be sure to get the probes.

I'm taking my Electrical Engineering FE exam in a couple months, do you guys have any good tips and resources for it?

>>1664582
I say exactly the same about people trying to use layers of discrete logic in practice

>>1664589
go eat another avocado toast on your way back to >>>/tv/

>>1664606
did you even read the USB Power Delivery spec

If i wanted to run a 10MS/s ADC for a second or so, where would I send the data to? A memory chip of some sort? Or is there some USB class that I could use without drivers to store it directly on a computer? I'd be using whatever bitrate necessary to get a reasonably priced ADC, so likely no more than 12 bits.

>>1664743
my mechE FE exam was easy. i downloaded an old FE study guide and repeated the section quizzes until i got them all right, and then took the study exam at the back of the book. as a result nothing really took me by surprise on the actual exam. one thing to note is that it seems to include sections of vaguely related topics. the ME exam had sections on economic statistics and humidity, for example.

>>1664746
>avocado toast
You DO know that analogue scopes are cheap as fuck, right?

>>1664752
i've never had to do this so i'm useless, but two approaches would be to glue the adc to a dram using a cpld or to find a microcontroller with suitable DMA. i'd avoid plds like the plague as the software is tantamount to a human rights violation

>>1664746
sneed

>>1664770
I don't suppose it would be possible to get an 8-bit ADC that's pin-compatible with a DRAM and use some sort of high-speed counter to increment data addresses in-time with the clock? It's only MHz range so I figure it wouldn't be too hard to get counter ICs that hasn't any troubles at that speed, though I would need 24 bits of counter for that, plus a little logic to stop the clock at the end.
From what I've seen most ADCs don't have parallel outputs, but perhaps that's different for these high-speed ones since the very first datasheet I look at I see D0-D7 on:
http://www.ti.com/lit/ds/symlink/adc1173.pdf
Looking for DRAM ICs (specifically 8*16Mb) I found this:
http://www.issi.com/WW/pdf/42-45S81600F-16800F.pdf
Which also appears to have D0-D7 parallel I/Os, though the address select pins seem to be multiplexed. I don't blame them for not having 24 address pins on the damn thing, the extra logic wouldn't be too much of a hassle.

So you know, the idea is to make a radio noise analyser. I'd offload the data onto a computer and FFT it to see at what frequencies the noise spikes. Naturally I'd include features for correcting for the antenna itself. Inversely, it would be useful for calibrating antennae.

>>1664787
some of these drams have a lot of features that could help you cut back on glue, like auto address incrementing. you'd have to read the adc and particularly the dram datasheets in detail to determine if you could interface them directly. memory banks and such may need low speed uC intervention, and you'll definitely need the uC for dram configuration and offloading to usb, ideally through a serial usb chip like a ch340g to eliminate miserable usb stack code. good luck.

>>1664752
>A memory chip of some sort
something close to this. a dual-ported RAM would be ideal, but might be hard to find/buy in 16MByte packages
I'd probably break out the FPGAs at that point. among other reasons, I might make some use of its Vccio for integrated logic level translation. as some commercial products do
in capture mode, you sample from the pins into one of two on-chip buffer RAMs, and simultaneously pump the other one's contents out to external SDRAM in burst mode, until the requisite number of samples have been taken. a 20-bit up counter with synchronous clear would be enough of an address counter for the RAM, a synchronous preset would be a useful optimization
>USB class
http://sdpha2.ucsd.edu/Lab_Equip_Manuals/USBTMC_1_00.pdf
>without drivers
CDC-ACM, since this designer is in lazy mode and prefers a smaller, cheaper FPGA. a UART built from logic blocks and connected to the fastest feasible external USB-to-serial chip takes commands from the host. when in readout mode, the SDRAM controller is under the control of the UART with the help of a finite state machine
otoh if you're willing to drop some money on a hefty micro with an external bus interface, SDRAM controller, and USB hi-speed, you might just be able to pull off as close to real-time as USB allows,

>>1664769
that's a good reason. muh hipster consumerism is a bad reason

>>1664796
>3-4 minutes to get your capture to the host is ideal
>not using the CH330N even at that

>>1664799
are there any usb libraries, for pic18s for example, that don't require reading the usb spec? i feel like having to deal with usb directly increases the difficulty of this sort of project by an order of magnitude.

>>1664295
Crystal ladder filter, cavity filter, or like a 10-pole LC filter (good luck with optimizing that)

It's hard enough to approximate brick wall filters with active components let alone passives.

>>1664796
Those serial chips are a bit shitty, never found them terribly reliable across operating systems. I think I'd rather have bloaty code in an MCU than deal with drivers, but preferable to either would be using a dedicated USB memory IC I guess. Makes you wonder how easy it would be to have an IC that would allow the memory to be written by the ADC but also allow reading of the memory as if it were the memory on a flash drive...
The easiest method would just be writing the data to an SD card or flash drive like that oscilloscope anon, which isn't terrible. Using an ESP and transmitting the data via wifi could also be acceptable.

>>1664799
>USBTMC
Ah, makes me wonder if there's an IC out there that will directly interface with an ADC for this purpose. Or an ADC with USB class compliance built-in I guess.
Obtaining full real-time isn't exactly a goal of mine, if I really wanted to I could just use one of those $10 USB logic analysers directly on the ADC outputs. Not to mention I'd want this to be more portable than a computer, being able to use it out in a field and carry it back to a computer to receive the data would be something I'd want to do. Avoiding the computer's own electrical infetterance too is something to consider.

>>1664804
Well there are MCUs with native USB support and those without it, pretty sure most pics won't have it. Common AVRs lack it too, but there is V-USB for bit-banging USB 1.0 from these AVRs. It's not entirely fast, but probably a reasonable way to go for a keyboard or something similar. Probably reasonable for custom joysticks too.

>>1664804
tbf, it kinda does
otoh, with a direct connection, it's worth knowing that the bit rate set by a CDC-ACM host is discretionary and can be ignored, and therefore you can shovel data to the host as fast as it can pick it up. also you don't have to use the T&M class if you just want it to show up as a virtual COM port, for which the microcontroller vendor likely offers some loopback-tier sample code that you can just plug into. or pretend you're a mass storage device, for which some vendors or github also offer sample code with pretty basic callbacks on the level of "pls copy block x to/from buffer y and return status"
chances are you would not need to do capturing and readout in the same session, which is a limitation that imo is better to live with than trying to fight the host's cache (but maybe a removable mass storage device e.g. an SD card reader is a good example for you). also, mainstream flash is typically a bit slower than your sampling requirements so you will probably need to buffer the whole sample in RAM anyway
>PIC18
I'd probably go with an STM32F4. something with a built-in SDRAM controller, some DMA, and hi-speed USB

>>1664813
>off-the-shelf?
probably not. 10Msps ADCs are a bit of a niche market, which you probably wouldn't find for much less cost than the assembled product
>electrical infetterance
*yawns loudly* there is also that. may I suggest a small LCD or OLED display to preview what you're bringing back from the field

>>1664254
>integrated circuit elitism about your faggoty nerd subject
Few things are gayer than this.

>>1664827
i can think of at least five things gayer than that. for example, writing 8 bit code in anything other than asm is pretty gay.

>>1664351
>tfw it just werks
Neat, though I do need room on either side of the 0V and 5V rails to pull it off. With any luck a rail-to-rail op-amp will work fine instead.

Sorry to pester you. I ordered the TS80. I'm having doubts. I wanted a soldering iron with temperature control that's fairly cheap and has no station because I don't solder much and is limited on space.
I just don't want to feel like I'm messing up. It's an alright soldering iron right?

is it real or is it memorex

Retard here, I have a manually controlled 120V AC winch that I'm trying to control via Arduino over WiFi. I popped open the controller and this is what I found:

When:
Red+Green && White+Black = winch pulls
Red+White && Green+Black = winch unwinds

I figure I need some type of relay switch, but I have no idea what kind I need. Any ideas, anons?

>>1665055
>Retard here
Confirmed. Either Rule 0 or more information. Until now your winch has no motor, no mains connection and it never stops. A complete wiring diagram would be useful.

I want to change electrolytic caps to polypropylene caps on the signal patch of an audio amp.

How close does the capacitance need to be?

Can I change for example 10uf to 2.2uf or 22uf to 10uf?

>>1664251
i love clicky switches, but i do have high standards when it comes to the clickyness of a switch

i collect different types of switches, have about 400 to 450 DIFFERED types of switches in my old room at my parents house. i love good switches, and whenever i go to a second hand store / goodwill etc. i always check the appliances for switches i don't think i have already.
my favorite switch came out of a 1956 film projector, its a DPDT switch with a perfectly smooth action and just the right amount of spring pressure. i recently installed it into one of my guitars as a pickup phase switch

in a professional setting when you're designing a circuit, how do you choose what ICs to use?

For example I'm back-of-the-napkin designing a bluetooth speaker. I look up "Bluetooth IC" and I get a million different results. How do I know which ones are actually good for the application I'm designing for? (And then, how do you know how to program it if they're all so disparate from different manufacturers?)

This happens a lot when I'm trying to brainstorm circuits, what do actual professionals do to figure out what IC they should use to implement a design?

How do I get into PCB design? Is EAGLE a good start?

>>1665098
Depends where it's being used. If it's as some kind of filter changing capacitance will change the frequency response.

>>1665118
By designing PCBs. I'd prefer KiCad over EAGLE and it's fully free.

>>1665118
Seconding KiCad

>>1665118
Every start is a good start.

>>1665111
Think business-wise: the cheapest one that works and allows you get to market first

>>1665111
First classify the sensible ones by their specs (like in a spreadsheet), then figure out what chip to select that would give you best value for those specs. Being easily compatible with the rest of your circuit is also important.
In particular, I'd see about ensuring BT 5.0 for this application.

why shouldn't I buy all of my components from aliexpress

>>1664993
ikr? I plugged it into falstad and had the same mfw
>RRIO
is perfect

>>1665055
dual Form C (aka DPDT) with a dual switch of whatever form (A or C) to stop the motor, Pic related

>>1665102
coolest autism I've read all day. have a (You)

>>1665118
meh, EAGLE is kind of on the outs. KiCAD is love. KiCAD is life.
also watch EEVblog's multi-part series on pcb design. it's dated but the principles should be solid

>>1665212
because LCSC is sometimes cheaper and always better

>>1665212

when shit dont work, you're always gonna worry if it's coz you bought fake or out-of-spec components.
worry kills.
life is worth more than money.
unless you're jewish.

>>1665213
>>1665172
>>1665165
>>1665136
Thanks, I installed KiCad and it seems easier than eagle

>>1665118
Sprint Layout.

>>1665055
>>1665213
PIC *clap*
RELATED *clap

>>1665217
I think it would look better with diagonal lines, just like for flip-flops made of NAND/NOR gates.

>>1665102
>1956 film projector, its a DPDT switch with a perfectly smooth action and just the right amount of spring pressure
some engineer designed this perfect switch and then faded into the ether without ever getting any acknowledgement for it

>>1665258
"We don't remember the farmer, we remember the fruit."

>>1665102
based

Shit, I think the atmega328p is using some sort of internal oscillator instead of the 16MHz crystal I stuck on it (via breadboard). The symptoms being, the arduino delay() function is about ~16 times too slow, and nothing changes when I remove the crystal from the breadboard. I'm guessing that's some sort of setting I'll have to change via avrdude?

how to make this or functional equivalent as cheap as possible? how to make it wireless? Where to source the plugs? The one plug I know is a NEMA 250V L2-20P, but they seem to want a lot for them.

>>1665290
>not looking at the datasheet
>>>/diy/arduino

>>1665290
Yep, you'll need to use avrdude to change the fusebits, here's a site that'll generate the command https://eleccelerator.com/fusecalc/fusecalc.php?chip=atmega328p

>>1665306
the little 2p are probably Hubbell midget twist-lock connectors
>how do I make this
if you have to ask, the best thing to do is to source the original parts, or parts of equivalent value and same or better current/power rating, and assemble as per the diagram. nobody knows what is available to you. go shopping
>make it wireless
you've got about 2-4 years of experimenting and reading before you should try this

>>1665111
>choose cheapest, shittiest part that soft of fulfills what people want
>codemonkey gets the barely working software running
>sell the junk for a lot of money because it's "designed and engineered" in the west

>>1665313
>Hubbell midget twist-lock connectors
Thank you, they certainly look it, I'll see if I can find a part number on the female half of the plug that I have. Thanks.
>make it wireless
Isn't there like some thing that I can wire a tiny little switch and a linear potentiometer to like a board with a bluetooth or wifi capabilities (replacing the left side of the diagram) and wire appropriate transistors to handle the same voltage / current ratings as specified in the left side of the diagram to another wireless board where the high-current transistor board would just plug right into the connectors on the right side of the diagram?
Also here's what it looks like inside pic related.

>>1665319
>Isn't there like some thing
at 1/20W, perhaps. at 135W, no. feasibly, you'd be looking at something mechanical like a servo to move the lever and a relay to replace the switch, which could be controlled by snap-together maker microcontrollers, OR redesigning the power stage for digital, which could cost more time/money and perform worse than a Hazard Fraught inverter welder

>>1665319

we need to know what kinda load is used. something like a heater or motor would work well with PWM controller like this or similar. the pot would replace the metal arm.
as for wireless, cant think of anything unless it's just used for on/off.

>>1665323
In theory it's only using 60 W, I thought you could get a nice high rated transistor about the size of a grain of rice that could handle that?
>perform worse than a Hazard Fraught inverter welder
Are you saying if I use any kind of IC / PWM like >>1665331
suggests, I'm going to get like stepping action or something worse than the carbon wiper slideing up the copper resistor? I don't know what a digital output would be like, but based on the brush / resistor combos I've seen, there are always dead / weird spots at a few points in their travel. Also I messed up the first wiring diagram in case anybody is wanting to build one of these. Thanks.

>>1665331
what do you mean by "what kind of load"? Sorry for stupid questions I didn't know about electronics until late last week.

like my board + standoffs?

>>1665342
If it's 30V it would need about to 5A to be 135W. Obtaining such a transistor wouldn't be terribly difficult at all. Assuming 100mΩ on resistance, that's 0.5W of power, which a TO-220 package can more or less handle even without a heat sink. 100mΩis being generous, low-medium voltage MOSFETs are often closer to 20mΩ. As that guy suggested, a simple speed controller box will likely work, the image title says 40A and I believe it. To avoid the chopped up waveform (important if you're powering something that wants straight DC and would be adversely affected by high-frequencies) you can add a large inductor in series with the output followed by a large capacitor to ground. The inductor needs to be able to handle the full DC current without dropping much of the voltage at all, and the "LC time constant" needs to be longer than the inverse of the PWM frequency. Ringing may occur, I don't know I'm not a doctor.

>>1665343
What kind of load means "what will the 30V be powering?" A motor? Heater? Electrolysis bath? Used as a battery charger? More sensitive electronics? A range of those? The "types" could more refer to: inductive or resistive (or capacitive), analog or digital, dumb or smart, sensitive to overvoltage or overcurrent, sensitive to magnetic or electrostatic interference, expensive or disposable, ohmic or non-ohmic, self-regulating or prone to thermal runaway, switched or continuous, using its own power supply or directly using the 30V, soft-start or high inrush current, the list could go on.
If you can say it is or isn't even a couple of those categories that would help define what circuit is right for the job. Knowing context helps a great deal.

What is a good soldering holder? I don't even know what these are called....amazon calls these "helping hands". I read that these cheap ones with the magnifying glass are pretty bad. Anyone have one they like? I suppose it either needs to be heavy enough, or able to clamp to a bench.

>>1665331
the part numbers at the bottom of that schematic suggest that anon seeks a wireless remote control for his welder

>>1665342
there are a ton of caveats that must be considered when applying power semiconductors, in particular, that power transistors deal in dc only (i.e. if this is an ac output welder, you're already fuqd), and do not directly and predictably translate a signal into a resistance. for that to happen, there is almost always a load of supporting circuitry and one or more feedback loops, which your particular welder may not accommodate. context is *everything* in electronics. without a full wiring and component diagram of your welder, we cannot even consider whether there might be e.g. starting chokes that will BTFO any semiconductor in its path
>stepping action
if you PWM, you will probably get an arc that starts and stops at a high frequency, which would likely change the feel of the process and maybe really cramp your technique

>>1665379
helping hands is a generic term for light-duty benchtop work holders. what's your budget?
the PanaVise 300 system is pleasant. the standard base + a pc board holder head will set you back about $70

bad news: Cypress fixed a "bug" with their PSoC SWD port so now only their programmer can acquire the port before the core locks up
good news: their newer PSoC dev boards include a break-away programmer that can program the entire line of 32-bit PSoCs
best news: this one's only $10

>>1665217
>*crap*
AC motor doesn't care about polarity. The controller looks probably more like this. Best solution would be two 1P1T relays for ON-left or ON-right. Alternatively one 1P2T for L/R and one 1P1T for ON/OFF.

>>1665420
The former wouldn't be good if both were turned on at the same time, having one for direction and the other for on/off would be better for this reason.

I wonder if you can have a bidirectional shaded-pole motor by having multiple shorted "windings" around the circumference of the stator, but only half are properly shorted via relays so you can change the direction of the starting bias. Probably easier to just use a starter-winding with a cap at that point though.

>>1665420
>imlpyign
just trying to replicate the switching action observed by the OP at >>1665055

>>1664752
A decent microcontroller with DMA. If you're only worrying about 12 bits, use a 16 or 32 bit controller.
I'm using a 16 bit dsPIC and fancy ADC to record audio at 24bit/44.1.
ADC sends data into the dsPIC via its DCI port, the DCI module sends data to a DMA port, then feeds it into an SD card.

>>1664804
Check out Mikroelectronica, they have an IDE for the PIC series, lots of built in functions.

>>1665098
>What is ESR
You have to keep the values the same
Don't fuck around with snake oil caps, they won't make your audio better

>>1665118
EasyEDA is another option

>>1665422
The shade rings can be pretty significant current, lower power relay + winding + start cap is likely cheaper per starting torque.

>>1665374
>If it's 30V it would need about to 5A to be 135W. I'm sorry I don't have more info on this, but I believe they are saying the resistor is rated for 135W and at 30V and 15 ohm, it would be dissipating 60W, well within its rating. The thing that doesn't make sense is that as you bring the brush up the resistor to the +30V end, the resistance drops as low as 2 ohms which seems like it should be frying the resistor if it's only rated for 135W.
>What kind of load
I guess I'm not sure, somehow the voltage and current coming out of the variable resistor are going into one set of windings in a transformer and varying the output current of the other side of the transformer. Pic related if it means anything to you. I know this says 24V on the supply to the variable resistor but I'm pretty sure mine is reading high 29's, I will check later.
>>1665381
Yes it is for a welder. 220V 1PH 72A AC in, 300A 32V DC or AC out.
Full wiring diagram attached
> transistors deal in dc only
> if you PWM
This would only be replacing the DC variable resistor that's controlling the output of the transformer that's actually putting out the welding current. I'm not sure yet how it's doing this, but it's a definitely a DC only control circuit on the welder that I'm trying to emulate.
That said your point about semiconductor caveats is concerning, this thing has an AC spark gap high frequency generator to start the arc, and beings it's a >300A transformer, they recommend to have any electrical cables, including lights, shielded with a ground rod every 10' if they are within 50' of the welder.
Thank you both for your knowledgeable and encouraging replies.

Why is vE 0.7V and not 1.2V?

>>1665792
Vce != Vbe

>>1665841
nvm for a minute i thought this was a legit question and not some dumb homework trick question. it's because Q2's base is tied to ground so it's pulling vE lower than Q1 would. Q1 is in cutoff as a result and you can ignore its +0.5V bias. simulate it yourself.

>>1665649
>it's current limiting for the secondary side of a welder
oh fug, getting into sketchy territory here
>over 300A
yeah not going to get that with any common semiconductor, though maybe a semi-cheap ebay vacuum tube might be able to handle that
feel like buying one of these?
http://pdf.datasheetcatalog.com/datasheet2/d/0jk4co198jg44fcgqh6r1cp9s7cy.pdf
Or perhaps try your face at one of these:
https://www.aliexpress.com/item/300A-Industrial-SSR-Solid-State-Relay-300A-Input-4-32VDC-Output-24-680VAC/32746634126.html
lmao at that pic tho

They all seem to be made with HVDC in mind, not ~30VDC, so I guess they're being limited by the resistance of the transistors themselves. Is there no way to simply put a solid state relay before the step-down transformer? Where is the variable resistor on that circuit? R5 is connected to the HF part of the circuit, so I don't think it would do anything in DC operation. Also show a pic of diodes D1-D4, if they're operating at the full 300A they'd have to be pretty beefy.

I am new to using Eagle and designing PCBs so I was wondering if anyone could tell me if this is possible. I am trying to design a board that is extremely thin (every 0.1 mm counts) that uses an esp32 module. The module is designed as a SMD component and should be soldered on a side of the main PCB, however that is not what I intend to do. I would like to design the rest of the board around the module and connect them together side by side (castellated holes to castellated holes). This would save me 0.8mm of wasted thickness on the module being on top of the main board. How could I do this in Eagle? I shaped my board to have the space missing for the module and placed vias along the edge where they line up with the modules vias but just can't seem to figure out how to get them to actually connect so the the autoroute works. Anyone know how to point me in the right direction?

Pic related in case my explanation wasn't clear. Sorry for poor paint doodle.

>>1665868
>>over 300A
>yeah not going to get that with any common semiconductor

the footswitch doesnt handle the 300A. it seems to act more like a valve, where a smaller current controls (modulates) the main power flow. i dont even wanna look at that purposely ugly schematic, so i'll take a guess and say the footswitch runs current thru a winding which causes variable core saturation, which ends up controlling the main power running thru the other windings.

>>1665875

i dont see the problem. you just name your pads (which you are calling ''vias'' for some reason), and run connections to them, as if they were an external connector. dont use the ESP32 footprint at all.

this was fun to route. i wanted to make a hardware version of those arithmetic alarm clock apps. this has a 12 char 7seg display and an encoder for input. i generated a few thousand division problems that can fit in the character limit, eg 5684/196 = XX.

>>1665887
i absolutely could not fit it on the main board, so i also designed a daughterboard with a low Iq alkaline boost converter and reverse protection

>>1665884
>the footswitch doesnt handle the 300A
No I was talking about replacing the rhoestat with a PWM'd transistor

>>1665884
Im calling them vias because, to my knowledge, vias go through the board while pads are just on a face. I would need a conductor that goes through the board so that when they cut it I am left with a conductive surface on the edge that can have connections on the top and bottom of the board and also connect them to the module on the side. How could I make an external connector that digitally and physically lines up with the module? Could I keep the module in the schematic without it being on the board?

>>1665894
>>1665884
Edit: Sorry, just realized I'm swapping the terms via and plated through hole. Am I more on track or still off?

>>1665897
so you want to mate the two boards together via castellations? solder works by forming intermetallic compounds between two metals. it is not useful as a load-bearing connection. milling tolerances in manufacturing could leave you with too much or not enough space between the module and the main board for a strong, reliable connection, especially if the finished device experiences significant handling or shock. also, castellation is an advanced process and often entails a significant cost adder
however, standard processes would allow for you to mount the module inverted on a lip, connected to the usual SMD footprint with a cut-out for the shield can, which would then provide mechanical support to the module. you're still gonna have to ground the module properly, now that the ground pad on the back is just flapping in the breeze. copper strap or braid might do

>>1665894
>Could I keep the module in the schematic without it being on the board?

sure. just place it down but make no connections to it, just to the adjoining pads. dumb software will allow that, ''smarter'' software might complain.
(one diff between pads and vias, is that pads are named, vias are anonymous nobodies)

>>1665894
>vias go through boards
specifically through-vias, but good
>while pads are just on a face
a layer, but you're otherwise correct. and a pad stack is an arrangement of (usually) through-connected pads and/or antipads assigned to particular layers on a board, which in usual parlance is treated as a unit and just called a pad

I didn't expect to get any replies, so I really appreciate everyone who has. You're all really helping this make sense to me.

>>1665902
>it is not useful as a load-bearing connection
It wouldn't have to bear much weight or force as the main board and module would then have a side that is flush on the same flat material and would also be attached to it with double sided adhesive.
>castellation is an advanced process and often entails a significant cost adder
I didn't know about it affecting cost. Checking that option with JLCPCB jumps the price from $2 for 5 to $38. Could I get around it by not checking that box, changing the board shape to go back to a perfect rectangle (without module cutout in shape), making pads and plated through holes on the board, and then adding a milling path through them on the milling layer?
>the ground pad on the back is just flapping in the breeze
Also do I need to ground the the GND pad on the back of the module? I assumed that all pads with the same name were electrically connected and the module already has a few GND pins/pads on the outside edge.

>>1665903
Thanks I will give it a try. I wonder where Eagle falls on that scale.

>>1665909
I forgot multi-layer boards are a thing. Thanks turning gears for potential (albeit I'd assume expensive to fab) future ideas.

>>1665920
>Could I get around it
maybe, but they might reject the job, those through-holes on your castellations might not get plated, and they certainly won't guarantee that pads will extend to the edge of the board much less plated over the side. you should watch some videos on prototype pcb board manufacture to see why (tl;dw: standard boards are aggregated onto a standard-size panel and milled out when complete and that's why they only cost 40 cents each)
Pic related could still be fairly thin. one drawback is that you'll need to mirror the pads on the footprint, or renumber the schematic symbol. you might also have to deal with the new z-offset of the board in your mechanical design. some board houses offer a range of substrate thicknesses for free or a nominal charge. while 1.6mm substrates are the default for prototype boards (mainboard in pic), 0.8mm substrates (module board in pic) are far from exotic and are probably enough to gain back the offset on the shield-can side. on the solder side, you could mount the ESP32 module with double-sided tape and leave the mainboard to float (which will be more stable due to the multiple, flush welds from contact face to contact face. if you have doubts, get some blank 0.8mm copper clad and use as spacers or washers to add support and improve shielding
>Also do I need to ground the the GND pad on the back of the module
signal quality and therefore range will improve greatly if you do. but you could probably get away with strapping the shield can to a cutout in the solder mask layer on the can side using copper tape, foil, or braid, if you have a good iron and hefty tip
>multi-layer
costs a bit more and takes some thinking/calculation, but sometimes it's priceless
>eagle
it's a bit more mature than KiCAD. but really, other anon's idea was lame. it's better off to git gud at designing and modifying your own footprints/symbols, because you *will* eventually encounter a reason to do so once you've done hardware long enough

>>1664254
>needs a simple delay
>has to write code to make it work
this is how we know you know nothing about electronics, Arduinofag.

If you want more accuracy than a linear IC like a 555, then try a 4536. If an RC oscillator driving the thing still isn't accurate enough for you, drive a crystal oscillator into it instead. Simple. Easy. Cheap. No coding required *if* all you need is a time delay.

anyone /FPAA/ here?

>>1665986
but I am using an RC oscillator. an extremely low-power one, on the chip, that's already been trimmed to ±1% at room temperature. reproducibility is a nice touch. so too is smol. so also is non-calibration especially when the time to be tested is particularly long
unless I missed a post, anon never put a number to "about" 40 minutes nor specified what sort of process he was enabling (heating, cooling, blowing, lighting, stirring, opening, closing)
>cheap
Digi-Key prices as of a few minutes ago: about $0.49 for the 4536 alone in singles, not counting support components. about $0.43 for the PIC10F200 in singles. other micros with internal calibrated oscillators can be cheaper still. chinkbay and/or quantities can vary the balance. 100n bypass caps are assumed free
>*if*
agreed. all that said, chained 555s is next-level sheepposting and thanks for the (You)

>>1666009
do Cypress PSoCs count? waiting for some checks to clear so I can pick up the PSOC 5LP programming tool/dev board

If I'm powering a circuit off mains by stepping down voltage with a transformer then rectifying, smoothing and passing through a voltage regulator; should I connect the ground of the regulator side to the mains earth?

>>1666040
You can but you need not. I keep local ground separated from PE so I can stack two supplies to get a +/- dual source.

>>1666040
Depends on what it's for. If it's meant to be a general purpose power supply then what >>1666056 suggested to leave it floating isn't a bad idea (though I'd add a switch to toggle the connection), but if it's a dedicated supply for powering a circuit, especially a more sensitive analog or high-speed digital circuit, having a grounded ground plane can be quite beneficial.
Also you can use two diodes to turn an AC source into a ± dual source (pic related), in which case you're less likely to stack the supply with another. Having to use twice the capacitance for the same ripple rejection is a small price to pay. Throw an LM317 on the top and an LM337 on the bottom and you've got yourself a half-decent variable PSU for pretty cheap.

>mfw you can actually use discrete gilbert cells and current mirrors and some other structures that are usually implemented in integrated form for low frequency stuff with reasonable precision
I'm a noob and had to mix two audio signals, after making a gilbert cell I just realised the people who parrot over and over that "current mirrors should not be made with discrete components" probably have never soldered anything

Why does it seems like there is no literature on this type of circuit?? Look in any good (pre-digital, at least) power supply and you'll see this or-gate type configuration. Maybe because it's not well defined and better to determine expermentally?? Either way, I know that it's pretty common yet fairly undocumented, unless someone can prove me wrong ;)

>>1666062
I'm inclined to agree with you (that the mantra is parroted), and for low-budget (i strongly relate) then for some extra time the trade-off is worth it. But you have to consider the statistical worst-case nature of a production product, and for how cheap bulk lm324s could be used to create multiple semi-precision sources, wrt to calibration labor costs, then it makes sense from the manufacturer's pov. Especially since the writing is primarily industry-to-industry shills

Are there "general rules" for cap replacements? I've done them before on vintage receivers but I've replaced everything at once so it wasnt really an issue. I have another old piece of tech (1978) with way more capacitors so I want to only replace the ones that would need it most, should I start with the power caps?

>>1666092
Apply common sense.

>>1666101
well it does work at the moment, but ive had receivers that also worked fine but sounded better after getting new caps

>use power transistor to allow LEDs to draw as much current as they need
>use resistors to prevent LEDs from drawing that much current

>>1666117
LEDs do not know how much current they need. That's why their current must be limited.

>>1666153
This is me >>1657019
basically want to cut off the bottom, remove the speaker and pedal and turn the volume pedal (swell pedal) into a volume knob.
I have the schematic, is there any way to tell what kid of potentiometer I'd need? Do I have go farther back in circuit?

This is me >>1657019
basically want to cut off the bottom, remove the speaker and pedal and turn the volume pedal (swell pedal) into a volume knob.
I have the schematic, is there any way to tell what kid of potentiometer I'd need? Do I have go farther back in circuit?

>>1665868
Here is a more accurate wiring diagram for the machine, this shows the correct 30v control voltage for the foot pedal, I circled the 'variable resistor' which is essentially a carbon brush wiping up and down a big (1" sq. x 2 1/2" long stack of copper plates - see pic in >>1665319
) I don't know why they didn't just draw it as a variable resistor. From vid i've seen this goes from 15 ohms to 2 ohms depending on how far the pedal is depressed. Still not sure how it can see 30VDC at 2 ohms when it's only rated for 135w....but
>>1665884
Is right, it's not seeing the 300A welding current, or anywhere near it, its a separate circuit controlling core saturation in the welding output transformer. I don't know what that means or its implications for the components I can use trying to emulate the big copper resistor, that's what I need to figure out.

>>1666159
Measure the range of the LDR for clue.

>>1666221
>LDR
oh shit i mis-read that symbol. thank you, it gives me a good starting point.
https://organforum.com/forums/forum/electronic-organs-midi/hammond-organs/31449-hammond-l-143-expression-pedal
looks like a complex issue. i just want a volume pot.

>>1666183
>30V transformer secondary
>feeds bridge rectifier
>feeds into variable capacitors and potentiometers in series
>whole thing feeds into another transformer coil
What the fuck is that circuit even meant to do? Is it supposed to be filtering into 120Hz like some sort of awful f*2 guitar pedal?

>>1666183
>>1666238
>core saturation
Oh that makes some sense. Still no clue what the varicaps are for. Are you able to measure the DC resistance of that transformer coil? That way we could figure out if it's feasible to use a dimmer unit instead, because you basically just need to feed it a stable variable DC voltage with current limiting. A variable constant-current supply of some kind may be ideal.

>>1666239
>measure the DC resistance
I measured across the three sockets in >>1666183
C-B, C-GR, B-GR and got open circuit on all 3. Maybe it should read open because of the capacitors? If it should read a resistance, I'll try to measure tomorrow with a different meter or maybe wait til I get the mating plug so I can be sure I'm making contact. I'll post some pics when i get it wheeled out. Thanks for your interest.

>>1665359
yes. i use the same boards/similar style. but i mostly use SMD parts. i don't post pics because nobody gives a shit usually. so i thought i wouldn't ignore your post. show the wiring on the other side.

>>1666239
>>1666291
Nevermind im an idiot. I measured with another meter and in pic>>1666183
with switch S2 in RMT (remote) I got 211 ohm across what I believe to be C-B, 4.7 ohm on C-GR, 215 ohm on B-GR. Unless Im reading the plug schematic in >>1665306 backwards. With S2 in STD (standard) I get 4.7 ohms across C-B, open circuit on C-GR, open circuit on B-GR. If I'm reading the plug backwards I think you swap C and GR.

Is the ETK3699 a pin for pin replacement for the PT2399?

>>1665379
It is like communism. Great idea but I've never seen one that actually works well. I don't know if that's because they are cheaply made or the idea is flawed. I mean it should work in theory but the clamps are always loose and I could never adjust them the way I wanted.

>>1665381
> PanaVise 300
I looked it up and it looks more like a vise (well hence the name i guess duh) not a set of flexible/adjustable clips/clamps. It may work but it seems a different concept.

>>1666062
>luck is a reflection of my skill as an engineer and my value as a human being
>>>/pol/

>>1666064
it's not a gate because it's not digital. it's common and trivial to use such arrangements to select the highest/lowest voltage of many. low-power battery-powered devices with dc jacks use this all the time to avoid back-feeding the battery or the external power supply

>>1666102
you can get good return-on-effort by replacing the filter caps on each supply rail and anything that looks swollen

>>1666183
electrical schematic symbology varies slightly by sector, generally simplifying that which the drafter would use most often. therefore, industrial equipment uses two parallel plates to refer to switch contacts rather than the comparatively ornate hinge/arrow/etc, while small signal work uses the same two parallel plates symbol to represent capacitors. for local color, IC fags use something very much like two stacked hydraulic valves to represent an electrical transmission gate
also, yuropoors use the empty box for resistors

>>1666239
see above. the / signifies normally closed

>>1666305
personally I love pics, even if I don't give quite enough of a shit to reply every time. post some tits fåm

Is the ETK3699 a pin for pin replacement for the PT2399? From what I saw in the spec sheet, they look exactly the same & it appears that the same circuits are shown, but I can't tell if I'm missing something important.

>>1665359
>SHCS
based senpai

>>1666305
>show the wiring on the other side
kinda wish I had ptfe-insulated wires

>>1666314
kek, sweet i recognize some of my own creations. i usually re-do the ones i fuck up particularly bad. that wire is crap btw. melts/burns way too easily. i found some sort of a silicon wire (see pic) that doesn't melt if you accidentally touch it and it doesn't shrink at all when you tin it. it is stranded but it works well for me.
>lifted pads E13 F13
again, funny as that happened to me too. these boards are pretty good quality and pads are pretty tolerant to overheating. so it usually takes like 10 resoldering attempts with a really hot soldering iron to fuck them up like that :) but i've done that more than once. i need to find some and post them lol.

>>1666322
The mixers and long tailed pairs I see online all have diferential inputs, I'm not going to build one but how the heck is someone suposed to make a floating AC current/voltage source to put betwen the RF+ and the RF- terminals? (the base of the main transistors in the pair)

>>1666332
Also why they only work when supplied with +-vcc? For single supply operation do I need to bias the bases and AC couple the signals?

>>1666332
>>1666334
I don't know what schematics/devices you have in mind but in case of amps only one pair is used as input, the other one is feedback. Here is an example of a single supply amp. You can google op amp biasing for single supply operartion. One of the inputs is biased at 1/2VCC. And yes you can then AC couple the input signal.

>>1666332
If it's RF you're dealing with, a balun is often used if just to convert from unbalanced coax to balanced dipole. Depending on the circuit, you can often just ground (or tie to DC)one of the differential inputs, which you would typically do with a ring diode mixer. Assuming you don't need to augment the open-loop characteristics of your circuit.

>>1666332
>floating AC
Inductive or capacitive coupling, balanced or single ended.

>>1666439
Case of a fully balanced mixer (except LO)

>>1666439
>>1666440
>SA612 has an oscillator port
Oh neat. I guess it makes a sine wave with that crystal in some sort of linear oscillator circuit, right? Can it be used as an unbalanced mixer too, and with an external oscillator signal? If so, I'm wondering why I bought that pack of SA602s.
I recently learned how PLLs work so realised it would be far easier to just use one of those to generate an RF sine wave (assuming I'm correct that they make good sine waves) and feed that into the SA602 instead of using a square-wave oscillator and using cascaded low-pass filters.

>>1666442

>>1666443
oh thanks
Can't really see if that transistor circuit by 6 and 7 would result in a sine or not, in fact I think that specific circuit would just reach a steady state, but since they didn't show a schmitt inverter across the crystal I'm guessing it is a sine. The "equivalent circuits" are just representations after all. So it will work with unbalanced inputs, and will work with an external clock. Neato.

There's not much price difference, but they're both semi-pricy compared to other ICs. For that reason I bought the SMT 602s, but now that I look again I can get THT 612s at 0.68USD a piece, which is pretty good.
Apart from making me a subpar boardmeister, are there RF/high freq reasons I should steer away from using THTs?

Where's the most ideal place for me to start off in this hobby if I'm one of the monkeys who do commercial electrical work?

>>1666442
>square-wave oscillator
There is no such. The 602/612 oscillator is typically a tuned Colpitts type circuit. Suggest you study the data sheet, it's all in there. And look for examples using that chip. Of course you can inject any externally generated signal into the oscillator port. Be aware of the low dynamic range, <100mVpp or so. I can't fully decode the text in the VLF/LF up-converter circuit but it seems to hint at that.
>>1666443
Do you see the famous error in the drawing?

>>1666447
>typically a tuned Colpitts type circuit
Wait, the 602 has an oscillator-driving circuit too? What the heck is the difference between it and the 612 then?
>famous error
I don't suppose it's the missing pin 8, is it? Both the Phillips datasheets I'm looking at have the pin 8 shown. The missing resistance value from the input resistors to ground however is the same in those datasheets. I'm not too familiar with gilbert-cell or differential-transistor circuit topologies, so I couldn't say if there's an improper connection in them.

>>1666445
for the SA6x2's own purposes it doesn't need to be a sine. but pretty much any RF generator that outputs a sine has a high-order filter to clean it up enough to be transmitted without shitting harmonics everywhere
>reasons I should steer away from THT
no real electronic reason, at moderate frequencies. it's notable that 1/8W resistors/inductors were popular in radio circuits above the broadcast FM band near the end of the THT era, to keep parasitic effects low and reduce delay from the velocity of current flow inside the component. compactness is a virtue, so keep your leads short, but beware unintentional capacitors and inductors created by wire dressing and component layout. also, don't use solderless breadboards for RF, it only leads to tears

>>1666446
you probably already know about Ohm's Law and have a working understanding of Kirchhoff's current law, and we won't have to tell you that a power source's current rating is a maximum and it's perfectly okay to connect loads requiring lower current. that's a good start
what did you see or hear about that made you think "cool, I wanna do this stuff"? also, how's your higher maths? high school algebra is mandatory. parts of trig and calculus are important too but you might be able to pick them up piecemeal as you need them

>>1666447
kek, I sure did. let's see if the young players get it

>>1666460
iirc the 612 is a low-cost 602 with some relaxed figures of merit

>>1666466
>without shitting harmonics everywhere
Not as important for a receiver circuit (personally I was considering a 25MHz upconverter to listen to SW with a cheap SDR).
>1/8W resistors
Do they somehow have better high-freq characteristics stemming from a smaller main element? Or is it just due to them being shorter in total?
>don't use solderless breadboards for RF, it only leads to tears
I wasn't going to, don't worry. Though being able to use sockets to conserve ICs and maybe transistors when prototyping would be reason enough to use THTs. Makes me wonder if it's worth using extra-thin copper for RF boards to reduce capacitance between traces, especially for veroboard.

>>1666466
>iirc the 612 is a low-cost 602 with some relaxed figures of merit
I see, I may order some for prototyping then, and use my SMT 602s for polished projects.
>doesn't answer famous error question
damn I'm gonna be scratching my head all night

>>1666472
>25MHz
maybe if you're only upconverting one band at a time and using a variable LO to get there. I'd still filter the output to keep the LO from desensitizing the SDR, and at least stop the AM broadcast band on the input to keep nearby stations from swamping everything
>1/8W
the resistance of a volume of a material from end to end is inversely proportional to its cross-section and directly proportional to its length. typical film resistors have a helix laser-cut into the resistive material, varying the angle and length of the cut to vary the width and length of the conductor, and thus the resistance. inductance is proportional to cross-sectional area, i.e. diameter of the resistor core, and inversely proportional to length. all else equal, a smaller resistor of the same proportions will have equal resistance, but a smaller inductance proportional to the smaller cross-section, which is only partially compensated by the shorter length
there are low-inductance film resistors that are trimmed with a laser-cut L shape, a coarse trim laterally, then a fine trim longitudinally. afaik chip resistors are trimmed similarly, having no cylinder
>thin copper
capacitance is proportional to area, and inversely proportional to the square of distance between the plates. probably not worth it when increasing separation by 41.4% yields the same improvement as half-thickness copper. one advantage of stripboard is that you could peel up part or all of a whole row to massively reduce stray coupling. harder to do that with dotboard, especially if they're plated-through. but you might want to look into "Manhattan-style" construction, it's pretty quiet
>famous error
I can only tell you it's part of the output section

>>1666480
The SDR has a minimum frequency of 23MHz or so and a maximum of 1GHz, so just a single constant upconverter that I can turn on or off would handle all my needs. Your points about swamping from the LO or AM stations are valid though, maybe a PLL-based variable oscillator would be worth using instead. No clue how I'd do a variable filter though.

I think I see the error in the diagram, and I feel silly for not noticing it earlier. And both my Phillips datasheets have the same error too. I wonder if the older of these datasheet circuit diagrams were designed in some sort of ancient wysinwyg cad. They look like the sort of thing that would be easy enough to write with some sort of plotter language, but then you'd have to do the rest of the datasheet in a plotter too, and I don't think they look right for that.

>>1666482
There were/are actually two versions of the famous error602|612.
The 1997 error was already present in the early Philips datasheet NE612 of 1990, corrected a year later in the Philips Application Note AN1982 but not removed from later datasheets. A 'revised' version of the error was then introduced which is still published by NXP and maybe others.

>>1666346
>>1666439
>>1666402
Thanks, so just like an OPAMP then

>>1666482
There's something on /ham/ about SDR and short wave overload problems.

>>1666510
>>1666482
>>1666466
Listening to shortwave/HF on an 8bit sdr is kind of liking running windows 10 on a single core atom netbook. It can be done, but it's so terrible you likely won't get the enjoyment out of it that you could.

I know, I'm the guy from /ham/ that built a 160 meter dipole for his rtlsdr. Upgrading to a 14bit sdr made literally all the difference in the world. Even with only a 20 meter dipole I was having severe overload from an AM station 22 miles away and a 130,000 watt shortwave station 1200 miles away. You simply need more than 8 bits for this kind of thing.

>>1666228
>>1666159
>>1666221
two questions, how do i measure resistance in a LDR, does it have to be totally dark?

second, is this "pedal slider" any relation to the swell pedal? i cant follow the schematic properly so its a stretch. there is no pedal slider on the organ itself unlike the melody slider next to it..

>>1665387
>$10
>not getting your dev boards for free on an trade fair

>>1665387
>$10
>not getting your dev boards for free on a trade fair

>>1666550
If you look at this crop you see all you need to estimate the value of the variable resistor (VR, pot) you need to replace the LDR in the swell pedal. The LDR acts as part of a voltage divider along with R101 (270K). A 1MΩ VR for instance would give you a max volume of 1M/(1M+270K) or nearly 80% of what is possible. The main level adjustment R131 allows you to correct that. The minimum is almost zero. Pots larger than 1MΩ may not so easy to come by and are not really needed even if the totally dark LDR may show a few MΩ after a while.

>>1666616
thank you for the insight. I can say that pushing the pedal all the way down produces sound so loud it might as well blow the speaker. I dont really need the full range if that's what it comes down to. Ideally id install a pot next to the power switch because the variable vol swell pedal will never get used. (99% of the times ive used the organ i just set it to one volume and play)
Worst can I can just cut the wires and make a plastic clip connector on each end so you can unplug it when moving the unit. (like a molex or something)

>>1666625
also, would this be what i would eventually need?https://www.taydaelectronics.com/1m-ohm-logarithmic-taper-potentiometer.html
a log pot, not linear right?

>>1666633
Yes, log is ok, makes it easier to set lower volumes.

>>1666644
im going to order the pot now because it will take a few days to arrive and i wont be able to start working until sunday.

1) what if i order 1M and its too much? will the volume just max out before i get to the end of the rotation?
2) what if I order a 500k and its too little, will i not hit full volume by the end of the rotation? will it burn out from too much power going through it?

>>1666664

1M is fine. 500K is fine. no possible resistor you choose would cause anything to burn out.

>>1666678
thanks for the guidance. i didnt anticipate this to be this complicated but with the sparse info on the schematic, its what its become. ill keep any good updates back in my old thread >>1657019

>>1666685

just FIY, when you use a log pot, it matters which of the 2 outside terminals you connect to: one gives you log, the other anti-log. there is no right or wrong, but just know that if you're unhappy with it one way, you can always change one wire and get a diff action.

It's so weird that you can talk about male plugs and female sockets all day at work, essentially thinking about sex non-stop and nobody bats an eye.

>>1666712
When all female sockets are plugged then everything is right in the world

>>1666711
If so, there would be no need for antilog pots. But you want to maintain the common cw to max and ccw to min volume characteristic. In this VR context you connect to the ccw end and to the slider. If you chose the cw end and the slider you reverse the characteristic and turning cw reduces the volume. Or is it.

>>1666738

the pot in question is only gonna use 2 wires, not 3 like a normal volume control. so what i said stands.

>>1666744
>what i said stands
Offer it a seat.

>>1666543
>You simply need more than 8 bits for this kind of thing
Right. If I just wanted to measure the noise amplitude at a variety of wavelengths, perhaps I'd want some sort of logarithmic ADC or logarithmic amplifier stage before the ADC in order to make the most out of a lower-resolution ADC.

why aren't there any SMD male hdmi connectors? i found a discontinued drawing for one but that's it. why don't the chinks share their stock?

>>1666773
I don't think there are non-linear ADCs (or easy to find ones), usually you do the compressing/expanding before the adc.

>>1666789
>smd male
why though

>>1666804
Not surprising, I guess I'll consider using a 12-16bit ADC instead.

>>1666811
i need a 270 degree low profile hdmi adapter but none exist that are actually low profile. also if i did my own board i'd give it a mini hdmi connector on the female end since i need that anyway.

>>1666498
>thesis
>antithesis
>synthesis
absolutely Hegelian

>>1666609
>having an employer to pay for trips to trade fairs
>being this cosseted

>>1666819
have you considered flex circuits? profile becomes irrelevant then

>>1666819
Why not use a socket and have a standard male-to-male cable going to it? There are HDMI sockets with two rows of pins that you may be able to squeeze on either side of a PCB to solder as an end-on socket, otherwise just using a self-wired HDMI plug that leads to your circuit. should work fine. At the very least I wouldn't like an HDMI dongle relying on my computer's HDMI port to stay vertical.
Not quite sure what you mean by 270° though, pointing upright from the PCB or pointing along it's plane?

>>1666819
Oh and there are also mini and micro HDMI ports last I checked.

>>1666835
here's an illustration. i'm wall mounting a spare monitor and the dvi port blocks the standard orientation low profile right angle hdmi adapter. if i could just find a solderable hdmi plug I could design and oshpark an adapter board and be done with it. if i can't find one i'll pull the monitor board out and desolder the dvi receptacle.

>>1666848
nevermind, page 14 of aliexpress had precisely what i needed. the /g/ solution beats /diy/ this time.

>>1666482
not only is that a terrible demand to make of a draftsman let alone a designer, ECAD has been around for decades, even if it did require a bangin' washing-machine-sized graphics workstation to run. personally I think it got DESIGNATED

>>1666852
my rule of thumb: if the world needs 100k of some adapter over the next five years, the chinks have probably already made and warehoused it

>>1666852
There are more elegant right-angle converters than that (i.e. the rigid ones), but yeah that's the right way to go.

>>1666311
>>1666064
I was specifically referring to their use in feedback loops (those are amps, not comparators), wherein it is definitely not as trivial lol

>>1666900
okay, not exactly trivial, but once you get it you might facepalm
first, when a non-linear gain element is supplied with negative feedback, the circuit gain is reduced by some factor and the gain element's error is reduced by approximately the same factor
second, an opamp's function is to amplify the difference between its two inputs by a large factor, usually about 100k. thus, the opamp will drive its output Vmid to whatever voltage the ??? box requires to drive Vout in turn to a voltage that makes Vsum ~= Vin. in fact, there is a tiny error voltage between the inputs, about (Vout-Vmid)/100k times the closed-loop gain iirc, which is required in order to push Vmid to that point. therefore, opamps can compensate out a lot of crap in post-output circuitry, as long as said circuitry's transfer function is monotonic and non-inverting, Pic related. likewise for feedback circuitry, used to great effect in filter and log amplifier circuits among others
now, with no opamps (and the pullup resistor you didn't show), the transistor will be fully on. with both opamps and diodes in place, one of the opamp outputs will be lower than the other, and its diode will be forward-biased, while the other is reverse-biased and effectively out of circuit. under normal, voltage-limited operation, the voltage opamp output will be the lower of the two, allowing just enough current to pass to maintain the set voltage (but no less than zero), while the current opamp output will call for more power, output near the + rail and effectively disconnected by its diode. in current-limited operation, the current opamp's output goes more negative than the voltage opamp's output and limits the amount of transistor current to approximately the set current, regardless of the voltage opamp's calls for more power. the ultimate result is that the transistor which is normally fully on is throttled back when *either* output measurement exceeds its set point

>>1666918
For that sort of current and voltage regulated power supply, what sort of voltage will end up being across those diodes? Assuming they feed into a FET or another op-amp buffer. The current through the diodes will be piss-all, so can I assume that the voltage across the diodes will be insignificant?

Also I'm not sure if he implied that it was a cc/cv circuit, but yeah I know I've used similar circuits for that purpose before.

>>1666936
>current will be piss-all
depends greatly on context. the pull-up resistor will have to be at least enough to convince the pass element to pass maximum current. its gain and cascading were left unspecified, so who knows
>insignificant
yes and no. assuming around 0.65V-ish as usual but, once divided by the opamp open-loop gain(s), will induce an output offset on the order of microvolts. it would take some profound, aggravated autism to be bothered by a bit of roundness at the Vmax-Imax corner
>cc/cv
he mentioned pre-digital power supplies as an example, so it seemed like a safe guess

>>1666939
>pull-up resistor
In that case if it's only powering a MOSFET without a pull-up, or an op-amp buffer. I've seen op-amp ideal rectifier circuits that use a single diode in their feedback loop that seem to have an effective forward voltage of very low. Actually that wouldn't be too hard to simply calculate the closed-loop gain function for.
>pre-digital power supplies
Oh as part of the output stage. For some reason I figured he was referring to something further in.
Also the NPN-based circuit didn't look like it would work at all since there's no way for current to get from the collector to the base and through those diodes.

>>1666944
I figured it was just a very rough pseudo-circuit fragment and filled in the blanks assuming the n-type devices drew current to activate a p-type device at the + rail (the pass device itself, Darlington pair, or Sziklai pair), as is typical of such linear regulators

Best resources to learn radio frequency?

Best resources to learn audio frequency?

BIOS modding.
Where can I read about it?
I have an Acer Laptop (A315-41, mobo Compal DH5JV LA-G021P rev unknown), and CPU clock speed is artificially limited.
I updated BIOS, CPU clock speed became higher. BUT. Fan is not as aggressive, so laptop throttles down to same 2.0 GHz rather quickly. I think even quicker than on previous BIOS, where it was keeping steady 2.4 GHz for couple minutes. And there is a head room in thermal design, fan rarely spins at 100%.

It doesn't have legacy mode (i.e. only UEFI), and apparently in last 20 years electronics got so smart, that I can't even downgrade a BIOS (it won't let me, Windows-UEFI flasher, or boot block flasher, same "Fuck you"), despite me trying to edit out version.

Also, I have seen people on other earlier Acer laptops having "Advanced" secret menu (press Fn-Tab 3 times), which looked really advanced to me, since they can edit so much there, including fan and shit. So I tried playing with efi variables, but without luck.

>inb4: I'm banned from /g/ for whatever reason, and /g/ doesn't care about such low level stuff anyway.

>>1667278
And yeah, stupid question. Can I use Arduino as flasher, or better get CH314A with SOIC-8 clip?

stumbled on what seems like pretty decent, free content. Haven't checked it out yet so I can't say for sure
https://training.ti.com/
https://training.ti.com/ti-precision-labs-overview

what kind of textbook or branch of EE would I be looking to double down on if I wanted to learn the principles behind and how to design VCOs, VCAs, and VCFs? I know it's signal processing but I want a textbook on the subject matter

>>1667326
if anyone has any recommendations on specific texts too, I'd love to hear them

>>1667279
don't care. you have to >>>/g/ o back. try >>>/g/ oogle

>>1667326
audio synths are just 1950s analog computers with a speaker on the output. git gud at network analysis and calculus, and much will be clear. sometimes there's also a /synthdiy/ general which might have more specific recommendations

>>1667336
I want to learn the theory behind them from an EE perspective, not just how to mash them together to form a synthesizer

>>1667336
I'm banned from /g/, and in /g/oogle every serviceman is a fucking jew that wants $3-$20 or some retarded invites, or solving electronics problem for a fucking bios dump, or want to steal your Windows activation key by asking dump from YOUR eeprom.
I obviously went for "solve electronics test to access download" and got BIOS and schematics this way in order to find out that BIOS update contains this firmware unencrypted and apparently 128k of firmware for multicontroller (idk how this thing is called, in MacBooks it is called SMC).

In any case, thanks for really helpful answer you fuck...

>>1667278
>>1667349
don't flatter yourself, that's nowhere near as low level as you think it is
even if we did give you the answer you're obviously not equipped with the requisite knowledge or skill to perform that which is required anyway

this is not the right place for you
why don't you go back to breaking xorg while trying to rice i3 (you couldn't hack it with dwm) on the arch install you performed by copy/pasting from the wiki handbook

>>1667355
>don't flatter yourself, that's nowhere near as low level as you think it is
It is pretty low to me, since in controls hardware. It tells fan what to do. It tells touchpad in which mode it should work (USB or I2C), etc. Sure, it is not SMC, but still low enough level.
>even if we did give you the answer
So give the answer. Don't know the answer? Then stop telling others to fuck off.
> you're obviously not equipped with the requisite knowledge or skill to perform that which is required anyway
What is difficult here? Connect clip to EEPROM or solder couple wires and click magic button? This is easy.
Disassemble BIOS? Sure, it is hard, requires knowings assembler and disassembly software like Ada. Perfect opportunity to learn how to do it.
Edit BIOS with a tool? Easy as clicking a button. Hardest part is to steal the right one.
Edit correct efivar?
>why don't you go back to breaking xorg while trying to rice i3 (you couldn't hack it with dwm) on the arch install you performed by copy/pasting from the wiki handbook
Because breaking xorg and ricing Arch is boring and gives zero (null) benefit.
Changed fan curve would give result 100%, since cooling system is decent sized, and can dissipate more. (not sure about VRM, but whatever)
Programmer that I would buy will allow me to get money repairing laptops after /g/ which tried to play with BIOS editing.

welder repair anon from above here. How do I recreate a 5-10kV .002 uF mica capacitor using cheaper shit (the original is mica for high-frequency tolerance and about $150) I read about some goys wiring like 5 1kV capacitors in parallel. Say I wire 5 capacitors in parallel, I'd want 5x .0004 uF to get a total capacitance of .002 uF? Also what key descriptors to look for to match the performance of the original mica cap? Also also, these are AC oil filled caps, someone somewhere said that if they leak like mine did you can maybe just refill them with mineral oil and be good to go? If so - what do you seal them with? JB weld? Thanks!
PS. I found the original wiring diagram for my actual machine in case anyone is interested pic related.

>>1667364
Oh thank God russkies for stealing Insyde H20 dev tools.
They have like complete guide and shit. All I needed is to google this in Russian. Fuck. I'm really stupid. RIP

>>1667367
Well for starters it probably doesn't need to be mica. All you'll need to match is the capacitance, voltage rating, tolerance, and possibly the ESR and ESL too. All but the capacitance could benefit from being improved values, but the capacitance itself should be equal to what it's listed as. Have you a datasheet or part ID for that $150 replacement?
>refill with mineral oil
I wouldn't, unless you can find a comprehensive guide to doing so and know the dielectric strength of your oil.

>>1667392
Thanks, here is a link to a site that has? surplus capacitor that I need: https://www.radwell.com/en-US/Buy/MILLER%20ELECTRIC/31602.. One of the many things I don't understand is there are capacitors listed with the greater V ratings and same capacitance ratings that are $3 and look tiny: https://www.rfparts.com/capacitors/caps-highvoltage-plate/002uf-10kvhda.html pic related. Is there a current rating I need to be shooting for as well? Because the capacitor in pic looks a lot more wimpier than what's on there unless it's an illusion and that thing is like 4" diameter with 4 ga wires coming out of it.

>>1667345
there is nothing mystical in analog synths. all is taught as part of basic analog design. if you understand the behavior of diodes and transistors (e.g. basic transistor amplifier topologies, Shockley equation, Miller effect) and basic analog building blocks (e.g. long-tailed pairs, current mirrors, op amps, OTAs, log and antilog amps, analog multipliers, oscillators in general, passive and active filters in general, etc.), and you can translate fluently between the time and frequency domains (Laplace transform and foundational math), their application in analog synthesis should be pretty obvious. I mean, if you weren't so proud, you could learn a lot of that by example, by looking at and reverse-engineering proven circuits. once you can describe the operation of the Moog ladder filter in your own words, you are probably also able to design such circuits from scratch
if you are interested in something more fundamentally oriented, you should look for books from the 1960s and early 1970s where such fundamentals were critical to understanding but not as widely known. note that some components from back then are not available today due to material cost and hazards, but the general idea of an oscillator hasn't changed much or at all since the pendulum
where are you even starting from tho? (/ohm/ RULE 2.5)

>>1667349
>nobody wants to reverse-engineer my 128k binary blob and fix my expensive consumer computershit for free REEEEEEEEEEEEE fucking jews
/ohm/ RULE 0
fine, google "flashing bios with arduino" and watch or read any of dozens of walkthroughs. don't they teach zoomers basic research skills anymore?

>>1667397
It would greatly help if you show the part of the circuit the capacitor is from.

>>1667397
Yeah I can't find a datasheet of it. Best bet is to infer what current and frequency will be going through it from the circuit diagram and plan accordingly.
Cheap modern components will be using more advanced dielectrics than old mica capacitors, it's no wonder they're only $3 a pop. At 2nF I doubt it has much current going through it, but better safe than sorry.
Also check out other sites, pretty sure tubes-store.com has some old HV soviet doorknob caps for pretty cheap.

>>1667398
here is the section of the wiring diagrams that shows the capacitor in question. I don't know the voltage or freqency, but there is a spark gap HF generator that will send a transformed-to-high-voltage 115v AC frequency over 1/2" distance to initiate the arc. Full wiring diagram (of the similar but easier to read machine) is in >>1666183

>>1667403
>doorknob caps
I heard this was the way to go... thanks for the confirmation.
Just out of curiosity, what does 'doorknob cap' mean and how is it different from the little pancake with a nipple on each side like in >>1667397?

>>1667399
>128k binary blob
8MiB, grandpa. Modern confusors have 64 mbit flash EEPROMS.
>fine, google "flashing bios with arduino"
Technically that is flashing with FTDI/whatever USB-UART it uses. Problem is, those are 5V, and I think that tiny flash is 1.8V (according to schematics), which means it is not that easy. And I'm not a parts store, so it would take ages to get this IC if it is not 5V or 3.3V tolerable.
And nobody mentions.

>>1667403
>tubes-store.com has some old HV soviet doorknob caps for pretty cheap
dude, thank you so much. I read somewhere that the doorknob caps have similar HF tolerance to the old mica cap I need replaced, would these really work the same way? We're looking at like 1/15th the price here...

>>1667096
www.antenna-theory.com

>>1667364

if all you need is a faster fan, simply clipping the fan's PWM control wire should send it to 100% speed 100% of the time.

>>1667417
This is African american way of doing this.
I want laptop to stay quiet when CPU die (not heat-sink or surroundings) is colder than 50 C, and make it go full retard when it is higher.
And "Overclocking performance menu" is tempting...

>>1667345
so, having said all that, I can outline the general problem that analog engineers try to solve and tie it together with your design needs: given a transfer function of some circuit (frequency domain for VCF/VCO, time domain for VCA), introduce the desired variables (here, a control voltage) and factor the transfer function back into a circuit

>>1667410
I thought all but 128k was uninteresting to you, but you're welcome to root through a bunch of compressed/encrypted shit if it satisfies your autism
>Technically that is flashing with FTDI/whatever USB-UART it uses
I'm going to take a wild guess that you were b& from /g/ for shitposting aka having no fucking idea what you're talking about
*technically*, you are feeding data to a microcontroller and having the microcontroller talk SPI to the flash and send it commands and data, which is a far different process from directly controlling a serial flash via bit-banged IO on a USB-to-serial bridge. as you pajeets should well know, having your name in a group project doesn't imply you actually did the work

>>1667413
they look compatible, except mechanically. you'll need to work out a mounting system and possibly add some insulation to prevent arcing

>>1667413
To be honest I think those cheap, new, tiny caps will work just fine. On octopart I've found a dozen or so reasonably priced capacitors using a ceramic dielectric that fit the bill (https://octopart.com/615r100gad33-vishay-39696724)), but if what you've read shows that a doorknob cap is more likely to work, go for the doorknob. At that price, could get both.
I think doorknob refers to both the shape and the mounting style, the doorknob caps I've seen have that screw-on terminal on either end like a doorknob. Then again, the only place I remember seeing doorknob caps is on that site. They've got nixies too, even ones with Ω and V and °C symbols in case anyone wants to make a benchtop multimeter or other practical readout that isn't some hippie clock.

>>1667408
Ah, it's got some wacky spark-gap oscillation happening. Be pretty tough to say what sort of frequency it will be working at, but I guess it won't be above 100kHz, so ESR and ESL shouldn't be much of an issue. What value resistance is R5?

>>1667429
>they look compatible, except mechanically.
What about the current ratings? Again the one that's in there just looks a whole lot beefier than the $6 version... also I have no idea what the current is, but it has to be pretty low if the HF is running at 2-6 kV of a 240 / 60 / 1 drop.

>>1667433
>What value resistance is R5
Idk, but I will check tomorrow, I think it is adjustable along with about 10 other resistors on this machine...

>>1667429
>as you pajeets should well know, having your name in a group project doesn't imply you actually did the work
heheh
Now I really feel stupid. I though CH341A is USB-UART converter (because all my chink arduinos have exactly this chip), but apparently it also does SPI and I2C...
And I all this time though that SPI is a serial port...
One more reason to get a proper programmer with 1.8V, because I would probably blow up IC if I'm gonna use Arduino.

>>1667441
Just use a logic level shifter bro. If it's only a few lines use a few MOSFETs.

>>1667433
>>1667434
peak current ≠ average current. if that's a starting cap, peak current is going to be at maximum when the arc starts and, ideally, should be delivered in relatively little time
as a general rule I wouldn't consider a replacement with terminals/leads smaller than those going to the cap, and would be careful about terminals/leads closer together than the ones on the original cap
>ESR doesn't matter in an LC circuit
well, consider dI/dt, then consider (I^2)R divided by volume, then consider wearing a face shield. those itty bitty thangs would probably melt their plates if the dielectric didn't crack first. if not on the first start, then certainly by the hundredth. which is why doorknob caps are even a thing and kilowatt radio transmitters use them

>>1667449
I think I can buy 1.8V converter adapter (from CH341a programmer kit). It has some sort of special IC, which would be faster and safer than anything I would make.
Or even better, buy entire kit, and couple spare eeproms, in case I release magic smoke.

>>1667459
these look pretty easy to use, should be able to figure out connections just from the labels. but you might have to turn your data clock rate down a little
https://www.aliexpress.com/item/33049748815.html

>>1667462
https://aliexpress.com/item/32583237273.html
I think this would be better, since it uses IC that was specifically made for this purpose.
I'm really concerned about the voltage, because I don't want to desolder IC (because I have no hot air) from motherboard. And it connects straight to CPU...

>>1667466
Btw, is it OK for SPI_ROM to connect directly to CPU without any protection (resistors, idk, special isolating ICs)?

>>1667468
Well, now I have a legit reason to fuck with BIOS. Sleep mode got fucked with update

I'm dumbfounded.
I need a circuit that translates a signal from 0-5V DC to 4-8V DC and when the input is hi-z the default output should be 6V. I can't seem to make it work.

It's for a motor controller so the default value is important.

>>1667457
>starting cap
What do you mean by starting cap? I don't think that term applies to this situation.
Looking at the circuit, it's part of a spark-gap circuit intended to create a high-voltage high-frequency signal for HF starting. So like a shitty tesla coil, the actual current wouldn't be much at all. It's a 2nF cap remember, so any energy going into it should be minimal. Though there is a continuous mode to the HF operation.
As far as ESR goes, the damn thing is in series with a variable resistor, and I doubt the variable resistor has a value below 1Ω, likely closer to 1kΩ or higher. It's a ≥5kV circuit, ESR almost certainly won't matter. But that's why I asked him to post the value of that resistor in the first place. If it isn't in a power package like the first one, the small cap should be fine.

>>1667468
Compare the input and output impedances for them and see if you can get away with a resistor in series or. Better safe than sorry.

>>1667480
What rails do you have available? Anything higher than an 8V rail? Can a 7V or 7.5V signal substitute for an 8V one? If no for either I think you'll need to use comparators and maybe even logic gates. If yes for either it's possible with just resistors and probably a single op-amp too.

>>1667441
>And I all this time though that SPI is a serial port...
but it is. it's just synchronous instead of asynchronous (SCK), and message-oriented instead of stream-oriented (~SS)

>>1667468
it's far from unusual for modern digital ICs of significant complexity to have a separate power domain and voltage standard for I/O banks, possibly several

>>1667480
Pic related, the zener in the feedback loop provides the offset, which unfortunately varies with the 3.3k resistor and the output voltage. for more stability, replace the zener with a TL431 voltage reference and appropriate set resistors, sketched in grey (you'll probably want one of them to be adjustable). an LM321-type opamp, or any other whose common mode input range includes ground, is fine

>>1667486
He can do this with a single diff amp with the reference input set to 4V to provide the zero offset. Correct choice of the gain resistors will limit the span to 4V giving you the 4-8V. The 6V out for when input goes hi-z is a bit of a curve all though. Perhaps a weak pullup to a 6V rail on the input (just Zener regulate one if one not avail)

>>1667486
This is the old circuit (which I've lost). So I have a 12V rail. However when I simulate this circuit it just locks itself to a constant value. (Sim in next post)

>>1667488
I was just going to have a pullup and a pulldown on the output such that high-z = 2.5V, before buffering that. The output of this would be in series with one resistor, which would have another pullup and pulldown, this time to 0V and 8V. It would be impossible to hit 8V with this method, but I think you'd get close enough with a 1k on one side an a 100k on another. Calculating the resistances though is a pain in the ass. As I said, use millman's theorem.

>>1667489
12V? Then it's definitely possible with my pic.

>>1667486
>>1667489
Or am I being stupid and doing the simulation incorrectly?

>>1667479
I've managed to roll back a BIOS via soft flash. Man, that BIOS was a complete pile of junk.
>>1667484
>Compare the input and output impedances for them and see if you can get away with a resistor in series or. Better safe than sorry.
Better get a hot air gun and desolder it. EEPROM is like 1$/piece. CPU is like $500, since I have doubts you can get anything that modern yet.
>>1667487
>it's far from unusual for modern digital ICs of significant complexity to have a separate power domain and voltage standard for I/O banks, possibly several
I guess every resistor and every mm of trace matters at high speeds.

>>1667488
>>1667490
>>1667486
Holy shit I'm stupid. Just keep it simple and tweak the resistors a bit to hit the 6V default value.

>>1667490
>tfw my calculations tell me I need -30kΩ resistor to ground
oops, gotta redo the numbers

>>1667494
nice one

>>1667480
>>1667487
explaining my design process a bit more: I arbitrarily chose a nice round 10k on the input as within the output drive capability of even the wimpiest voltage sources. to reduce the 5V range to 4V, I needed a voltage divider with a 40k resistor to "ground". so I chose my fail-safe offset resistors to set a voltage of 5V/2 (the center of the input range) and an effective resistance of 40k to the supply (treating the voltage source as a short). I made up the difference between 2.5V and 6V in the zener/precision reference. 3.3k keeps a current flowing through the voltage source, which is an active component and needs some quiescent current to maintain regulation

>>1667492
>I guess every resistor and every mm of trace matters at high speeds.
0.1mm for USB superspeed, and USB hi-speed is pretty tight too iirc

I want to replace my shitty old power tools with fancy new ones and i want them to be battery powered because i had enough of gay cables and extension cords.

So anyway, i only use my power tools couple times a year so what i want to know is what is the approximate life of those batteries?The power tools use those standard liion battery packs that are usually 20v 4ah
I don't want to buy them and two years later find out they are now worthless trash because all of the batteries died

>>1667511
They should last a while, but 5 years is likely past the warranty date, let alone 10. They're definitely not the sort of thing you'd expect to pass down to your children, especially if they change the battery voltage or mounting type. For your situation I'd primarily buy corded tools and just hook up outlets in more convenient locations about your shop. Specifically ones hanging from the ceiling above workbenches. Problem is corded drills have shit-all features (no variable clutch, variable gearbox, etc.), and none of them have the advantages that you get from a brushless motor (unless you want a 1500W hikoki rotary hammer). 5-10 years ago I'd have said to stick to the corded ones regardless, but these days brushless motors are pretty nice. For a lot of higher-power tools, being cordless is a needless gimmick that increases their price and decreases their effectiveness compared to corded tools, and I think angle grinders and circ-saws fall under this category.

So I'd say it highly depends on where you'll be working and with what tools. In general only using them a few times a year isn't worth ditching the extension cords even for semi-remote work, but if it's just a drill you need I guess there isn't any reason not to. Remember that buying a battery-powered tool locks you to that line of batteries.

>>1667516
So say the batteries die, will i be able to take a 20v wall wart and solder the wires to where the battery was connected and keep using the tool?

>>1667518
Provided the wall adapter can handle the current, almost certainly. If any battery management is done off the battery on the tool itself you'd need to spoof that, but from what I've seen they all have their BMS inside the battery casing.

>>1667520
Well then thats fine. In fact i just realized that i can probably crack those packs open and replace the 18850 with new ones, since pretty much anything liion uses 18850s

>>1667523
*18650

>>1667523
>i can probably crack those packs open and replace the 18850 with new ones
Definitely possible. If you're not afraid to do a bit of intrusive maintenance and part replacement, they should last you ages. The next most likely part to go on those is either the switch or the switching MOSFET. The switch itself is replaceable but not entirely cheap, while the MOSFET is cheap but half-covered in epoxy so very difficult to replace. Hence I'd do some research into what has the most robust electronics

>>1667491
You goofed R3. 105K belongs to the middle position of the 10K offset adjustment (6V) but you omitted that entirely and instead connected R3 to 12V. Now you have about 7V at the (+) input which sends the output of the opamp to the upper rail, even without input signal. Would be interesting to see what it does if you fix that.

Analog mixer guy here, aparently a full bridge rectfier can be used as a very bad mixer, but it does make beat frequencies and after that is just a matter of filtering and amplifying.

>>1667598
thanks to the translinear principle god bless

>>1667598
Found this simple mixer using 602 or 612.

>>1667626
I can only find those from china ;_; it'll take a while to get here

Has anyone ever used one of those all in one resistor, capacitor, inductor, and transistor tester things to test inductance? Are they any accurate? I just realized that I have been putting a lot of faith into this thing without questioning how well it works and am wondering would it be worth testing my inductors with another method.

>>1667737
>extra heavy battery
i believe it

>>1667737
I don't have one but if you can zero it and you have a few known low L inductors to check I would trust it. I have a cheap RLC meter that shows -18µH with shorted input probes (10cm) so I have to subtract that from the measured value (--=+) the same way you correct your ohm meter when you can't zero it and you use long probe leads.

>>1667457
>>1667437
>>1667484
Sorry, getting readings off of this thing is a PITA because all the terminals are covered with 1/4" of bat shit, but R5 is definitely a variable resistor controlling "HF intensity", With the dial set to zero and not accounting for any bat shit in the circuit, it reads like 6-7 ohms, and with the dial set to 100 (max HF), it reads 3 ohms? From the technical manual from the similar welder the resistor R5 is described as "Rheostat WW 150 w 5 ohm". The attached pic is the Rheostat/resistor in question.
>>1667457
>>1667484
>starting cap
The way it's supposed to work is that in DC mode, the HF is used just to initiate the arc, once arc voltage is sensed, it initiates the "HF timer" which would turn HF off in a matter of seconds after the arc is established, so yes, HF would just have to run for maybe 5-10 seconds at most and then be off for a minute or so until you start the next weld. In AC mode, I the selector switch would be in the "continuous" position because it's using HF to re-establish the arc every time the AC voltage changes polarity.
>shitty tesla coil
Yes, that's exactly how it works / looks like in operation. I don't know what implications the resistance discussed above has for the current running through these caps, but this type of transformer welder can do an HF arc like 3-4 feet long through the air, whereas the newer inverter ones can do maybe 2-3 inches at most presumably from running less current / smaller caps in the HF circuit?

>>1667893
>>1667457
>as a general rule I wouldn't consider a replacement with terminals/leads smaller than those going to the cap, and would be careful about terminals/leads closer together than the ones on the original cap
Is there a rating on current or something similar on capacitors that I'm missing to see if I'm in the ballpark of the capabilities of the originals or do you seriously just go by physical size / beefiness? The existing, leaking caps have like 1/16" x 1/4" copper bars connecting them, so is the biggest doorknob cap with the same ratings the best I'm going to be able to find? What do you think about smaller, lower capacitance capacitors wired in parallel. Pic related is the current caps.

>>1667635
If you are mixing line level audio signals I do not recommend the SA602/612. It's internally biased very close to the positive rail. The input should not be greater than -25dBm into the chip's nominal 1.5kΩ. You'd have to seriously attenuate line level audio to get to that level and in the process degrade your SNR considerably.
Diode ring mixers are not suitable to audio, only RF. Ring mixers need to be transformer coupled. Not doing so seriously degrades your port to port isolation. For this you'd need large audio transformers with center tap. They are bulky and expensive. Not ideal. With RF it's tiny toroidal transformers with a few turns. Much easier.

In your case you're probably best off with dual-gate MOSFET mixers (hard to find dual-gate FETs) or a JFET cascode mixers which is essentially the same idea. A discrete Gilbert Cell make work as you can bias them how you want though I've had issues getting them to work well without at least attenuating the LO a bit. Or if you wanna go old school, pentagrid converters which is just a JFET cascode mixer with a pilot light.

Could anyone instruct me on what I would need to build a drone jamming gun like pic related? Im new to electronics building so maybe its to hard of a project for me to start with, but I wouldnt know. I guess I need an antenna of some sort, a power supply, and something to broadcast radio signals. Ive got a 3d printer so I could print parts to cover it and make it look like a gun.

>>1667921
Radio jamming is illegal. Fuck off.

>>1667921
>so I could print parts to cover it and make it look like a gun.
this is the least of your hurdles. RF is complicated and you're not going to learn it from a 4chan(nel) post. go get an electrical engineering degree, or pony up and buy something on ebay.

>>1667737
I have one. I don't have anything to test it against, and it's half useless because I broke off the to corners of the display while transporting it, but it's never given bad results. I'd trust it to within 10% or so.

>>1667893
Ok I give up, definitely don't use a small capacitor.

>>1667898
It's basically size determined, the larger it is the higher the surface area and the more readily it can conduct heat to the surrounding air. The caps already in there don't look to be optimised for high-current, so I'd say just that one doorknob cap should be fine, though I can't really get that good of an idea of a size comparison between them.

>>1667919
just tell him to use a transconductance amplifier instead

>>1667939
>the larger it is the higher the surface area and the more readily it can conduct heat to the surrounding air.
Ok thanks, since this thread is practically toast I'll ask a retarted question: Is this something I could make at least for testing purposes out of a couple copper plates separated by a piece of saran wrap, or mica / glass for long-er term durability? I've already sunk 250 into this thing and I hate the idea of buying $30-150 parts without knowing if it's going to work in the end. Consider my labor/time free.
Pic is of the innards, what you see there is about 2' tall... It would just really suck to put a couple hundred into this thing only to find out the most important parts are toast.

>>1667942
At 5-10kV it's not implausible, but I wouldn't recommend it. You'd need to find a good dielectric at the right thickness and somehow get it flat enough without air bubbles or moisture inside to roll it all up. Then you'd need something to contain it all and prevent high-voltage leakage around the outside. I'd say $6 for that cap is a reasonable deal.

>>1667949
>$6 for that cap
The doorknob cap? How can I be sure it is of sufficient "size"? If it's too small, do you think it would work fine for a little and then fail? I'm afraid if I get one that's not quite right and it fails I'd attribute it to a too-small cap but there could be some issue further back that's causing it to fail. Thanks for your insights on this. Here's another pic of the insides, I think the bottom right cap is also swollen up but I'm not sure which one it is / what it does. This looks like a common motor capacitor though.

>>1667957
*top right capacitor looks swollen, sorry picture is rotated, the right side is the floor.

>>1667939
>just tell him to use a transconductance amplifier instead
I forgot about that. That's probably the best/easiest option for low frequency mixing.

>>1667939
>transconductance amplifier
meme vs reality

>>1667919
>bulky and expensive

>>1667975
2x LT44 1:1 ct 4x BAT48
alibay

>>1667975
>can't find datasheet with graph of frequency response
yikes... red flag

>>1667981
experiment, find out.

>>1667986
I don't have any on hand but it looks like other small shitty telco transformers. Designed to drive crappy 4 ohm speakers over a limited frequency range from like 500Hz to 4kHz. These things are garbage and the fact that it came from alibaba makes it even more so. Who knows what the hell you're getting. It could be anything.

What iron are people recommending these days? My aoyou 986a+ has a fucked up heating element and I broke it further trying to replace it. I'm waiting on a replacement handle from AliExpress but need something sooner than later and wouldn't mind a smaller device anyways.

Are the t12's decent?
Amazon.ca has a couple that will ship in 24h. But not sure which one I should get. Ksger has two models a silver one with 2.01 and a black one for $10 less. No reviews on Amazon. There is a "shinenow" brand that has lots of reviews and is $70. Apparently I need to check earth ground and solder that on. YouTube has good reviews on the ksger.

Price is basically the same from Amazon and AliExpress (give or take $10)

Is there another iron I should be looking at? Mostly just want temperature controlled and the possibility of buying high-quality tips down the road.

Hi /diyohm/
I have an old classic ipod-era portable mp3 player I'd like to salvage. It is an olympus m:robe 100 (picrelated) with a dead battery. I don't want to use it as a portable, instead I want to use it as a fixed player in my car.

This thing is like 15 years old and I don't want to buy a new battery for it. Is there a way for me to bypass the battery and just run it off power from the car (like a USB 5V adapter plugged into the cigarette lighter)? Would it be safe?

>>1668027
To be honest a new battery wouldn't be difficult at all, assuming it's just an internal lipo. Running it directly off DC should easily be possible, though you'd have to have a squiz at the circuit to see where you can inject power and at what voltage. Ideally you can bypass the battery circuit and use the USB input directly as a power supply (assuming you can't already do that). Worst case scenario it's just 5V feeding a 3.3V linear regulator that feeds where the battery used to be. Possibly with an isolated converter or something in there, since car audio can be a pain when it comes to ground loops. You'll want to make the power supply as noiseless as possible, for what should be obvious reasons.
pic of inside

>>1668027
>>1668029
Oh you'd better hope that there's no battery-induced corrosion though.

>>1668027

it should be simple and straightforward to power it thru the USB cable, without a battery, but the dumbass developers likely made it so it wont play while it's charging - for no good reason other than hating their customers. if that's the case, you gotta provide clean regulated 4V power across the battery contacts.
alternately, get a LiPo from someplace like Adafruit and scotch-tape it to the back of the unit, running wires thru a hole. i did this on my Sansa, and it was easy except for drilling thru titanium.

>>1668032
Could arguably attach it to a (big?) capacitor and a TP4056 and power it via USB if it has that "won't play while charging" feature.

>>1668029
>To be honest a new battery wouldn't be difficult at all, assuming it's just an internal lipo
>get a LiPo from someplace like Adafruit and scotch-tape it to the back of the unit

I don't want to spend the $40+ its going to cost for that because its not really worth it wrt the player itself. The player has a 5GB HDD (yes you read that right) so once there's mechanical failure its toast anyway.

Anyways I guess I'll open it up to see if there's any corrosion, I'm just gonna toss it if the inside is nasty. There's an ifixit, looks pretty straightforward. Aesthetically its a very very cool device.

>>1668041
>$40
A small lipo would set you back $5 or less.

>>1668043
Ok, I'll look into it. I live in Canada though (in Ottawa) and I doubt there's anything here that's remotely reasonable in price. I pretty much buy electronics shit whenever I have a chance to head down to Toronto

>>1668059
If you don't mind waiting 3-6 weeks, buy from aliexpress. Free shipping worldwide. Buyer beware, ensure you only buy from sellers with a decent rating, and try to buy only from listings that have plenty of other orders and decent reviews.

for those not following the 3 cent micro saga, those absolute madmen at eevblog reversed the shit out of the programmer and there's now SDCC support and a FOSS chip writer
https://github.com/free-pdk/sdcc-pdk-code-examples

>>1668059
if you don't care about form factor, 18650s are cheap too

>>1668067
>shipping Li+ batteries overseas
I wish I could b& you

>>1668071
Ok the shipping isn't necessarily free, but it's still hardly expensive. Not sure to which countries shipping is available.
https://www.aliexpress.com/item/32964049857.html
If you specifically want a 3-wire lipo instead of just spoofing the thermistor I think there are some ones for ~$8 shipping included.

>those absolute madmen at eevblog reversed the shit out of the programmer and there's now SDCC support and a FOSS chip writer
50-pack of micros with a programmer included for $5 from alibay when

>>1668074
for that price it is almost sure not to be registered as UN 38 and might be transported as air cargo in violation of IATA and/or national regulations
>my broken mp3 player indirectly took down an international aircraft and I survived
#goals

>>1668103
I guess there might be ones shipping from within the US? eBay may be better for sellers outside of China.

>>1668103
>UN 38
#yolo

Do there exists standalone modules for Cortex M microcontrollers (e.g. the STM32 line) like there exist for ESP8266/ESP32/SPBTL1-2/etc..?
In a way that to design a PCB around a micro you would need absolutely the least amount of external circuitry to make it reliably work?

>>1668138
somehow:
https://www.aliexpress.com/wholesale?catId=400103&SearchText=stm32+module&switch_new_app=y
but you can very easily get a stm32 working with just a few decoupling caps

>>1668138
I think the only reason you get boards like that for the ESPs is because they were originally designed as wifi chips that are easy to include within a product to give it wifi without having to design the antenna and shielding. If I recall correctly its use as a standalone MCU is more of a hobbyist thing. But I could be talking out of my sphincter.

>>1668146
yeah I considered using one of those dev board to snap on a PCB like a ""shield"" but imho it's not such a good solution, especially for a low profile product..

>>1668149
then what is stopping you from just designing your own mcu pcb layout?

>>1668149
Yeah if you want low profile and are doing a PCB anyway adding the STM32 chip + few supporting components + SWD header is easy.

>>1668138
the reason for those is because of this
most non-RF lower-speed 32-bit MCUs are easily used on 2 layers without the need for such a board. just do like >>1668146 suggests

>>1668160
how many layers does an ESP8266 have?

>>1668178
https://www.espressif.com/en/support/download/documents?keys=Reference+Design

>>1668179
Looks like 3-4 layers on the most recent file. The second from the top looks like it just has vias so I can't really tell if it's a copper layer or just a drill template. But since it still has a ground plane to suit the outline of the board I assume it's a copper layer.
Interesting layout, they arranged the resistor footprints to select whether using the internal antenna or an MMCX connector by having two resistor footprints at right angles to one another share a pad. There are some design decisions that I may have done differently, but it's hard to say without actually opening it up in kicad and messing about with the nets.

>>1668182
>Looks like 3-4 layers on the most recent file
hmm?
you said ESP8266, I see two layers...

>>1668183
I clicked on the ESP32 one because I assumed they'd be similar and those were the only ones listed, guess I'm wrong

>>1668182
>having two resistor footprints at right angles to one another share a pad
so that stub length approaches 0, in either configuration

Is the constant current for lm318 also I = 1.25 / R like for LM317? I can't find it anywhere

>>1668243
>lm318
*LM 338

>>1668243
have you considered reading the datasheet, doing a little math, then killing yourself

>>1668257
5 x 0.24 = ..erm..

>>1668257
>doing a little math
How to calculate B?

If i want to use photosensitive pcbs i need two chemicals right? the developer one to show the circuit after you uv expose the board and standard fecl3 to etch it, is there anything else?

>>1668293
you'll want a stripper for after you etch. acetone or hot NaOH will work. photoresist has an unpleasant nose when soldered

>>1668300
>NaOH
That is the chemical i will use to expose the board, you are saying that after i expose it i need to heat it up and swipe it again...?