/diy/ - Do It Yourself

This thread's digits brought to you by: The MAX9427 Lowest Jitter Quad PECL-to-ECL
Differential Translators
http://pdf.datasheetcatalog.com/datasheet/maxim/MAX9424-MAX9427.pdf
Didn't even know what ECL and PECL were until now, at least now I know what to look for when I need a logic IC that works at 100MHz. Maybe.

>>1889059
>Whats the quickest, most compact way of cutting 5v to like 1.5v
>AKA going from High pin to Low pin signal, without using a buck converter board?
By the sounds of things, you want a logic level shifter, not a power supply shifter? If it's a two-way signal (like I2C) that you want to cut down then I'd use a MOSFET-based logic level shifter like >>1889064, but if it's just one-way then you can get away just by using a voltage divider or zener/ to step down (transistor to step up). If it's only 1.5V I'd just slap two silicon diodes in series though, probably can't find a zener rated at 1.5V too easily. Can probably use an NPN in place of that N-FET in a pinch, shouldn't encroach on the turn-on and turn-off thresholds.

Of course, that's all assuming it's a positive correlation (higher voltage = logical 1 for both) and that they're both return-to-zero (~0V = logical 0 for both).

>>1889070
>I should be able to send 5V to it for high signal, then cut it below 3 for low correct?
Probably not, 3V is in limbo region, you'll want to pull lower. The actual thresholds vary by logic family, but a look at a 74HC or 74LS datasheet should give you a good idea. I wouldn't go above 1V for a low.

>>1889559
Wait, this works doesn't it? Regardless if it's step-up or step-down.

Also for the original poster of the question, pretty sure you're doing something wrong if you need a 1.5V logic level signal.

I want to connect a piezo speaker on this on KL.
the voltage, no matter what I do is atmost 8V in those nodes.
the diodes D1 and D2 might need changing.

>>1889708

when you have 9V coming out of pin 3, there's 2 diode drops when it gets to the emitter, so you should see 7.7V at point K, and some unknown voltage across the buzzer. the buzzer will create a voltage divider with one of the 200-ohm resistors, so that's why its unknown. this is one reason you shouldnt use a voltage divider to power things. use a regulator, so the signal doesnt change with the load.

when pin 3 is low, you should get 0.2V at K, so 10V across KL. that'l drop when you actually use a load.

In a battery powered device what's the point of using the enable pin on a voltage regulator vs switching the battery Vin directly?

>>1889801
it's cheaper. you either need a FET, a load switch IC, or god forbid a mechanical switch to do that separately. i think enable pins are annoying because i never use them but they're pretty trivial to stuff into a regulator circuit.

>>1889801
sex without lube

>>1889804
>>1889806
I have to make a decision today and I'm trying to keep things simple. I have a mechanical switch. I might just keep the enable pin HIGH and not bother unless there's something important I'm missing

>>1889810
>I have a mechanical switch.
use it. soft switches or intelligent power control are annoying.

>>1889810
Just switch the enable pin like a normal human.

>>1889813
>>1889811
okok I'll do it and be done with it thanks anons

Where can I get an omnidirectional utrasound emitter? I want to build an ultrasound beacon system but all transducers I can find are designed to provide a narrow beam.

>>1889708
Totem poles are emitter followers. The output voltage can never be greater than the driving voltage. You need an inverter transistor with a pull-up to 20V in order to give your totem pole sufficient voltage.
Also 200R is a bit low for a 20V source imo.

>>1889898
Might have to make an array of transducers. Ultrasonic transducers are basically just piezo discs, vibrating in and out in just one direction. Higher frequencies travel straighter and don’t diffract as much, hence why these ultrasonic piezo discs seem more directional than an audible piezo disc. The only answer that doesn’t mean having a bunch of them would be to fix some sort of a hemispherical shell atop the piezo disc, I’m pretty sure. And that would be a real trial to get working properly. Since they’re so cheap, I’d just make an array. Or maybe cut one into a star-like shape so you can bend it into a hemispherical form, provided you can mount it properly.

Anyone have a good soldering iron they'd recommend?

Similarly, heat gun?

Are piezoelectric buzzers the go-to transducers for generating +120 dB sirens?

>>1889958
i enjoy both my hakko 888d and my hakko 851

>>1889980
I don’t think so, loud piezoelectric discs need somewhat high voltages, so you’d probably find that motorised klaxons or air horns / deliberately loud impellers will give you better bang for your buck. Still, probably cheaper than magnetic speakers, if you don’t mind being restricted to a narrow output frequency bandwidth. A piezo’s high frequencies might also be particularly unpleasant compared to other sirens, for better or worse.

t. not an expert

>>1889958
I’m real fond of my STM32 T12 station, it’s got the grunt and convenience I require and for a pretty alright price. But it’s known to have potential soldering issues out of the box since there are so many different versions and iterations out there. It’s like 70W at 24V, but my one runs a little optimistic at close to 25V so it’s probably a bit more. The T12 tips are also somewhat expensive but pretty good. So I’d recommend it if you already have a usable iron, are fine with 70W and the T12 tips, and don’t need a portable iron. No clue about hot air though, there’s a T12-style hot air station but it looks kinda generic.

do any of you use a PCB clamp?

>>1890135
no. i ordered one of the cheap chink clamps and it was worthless.

>>1890135
No, I try to use my chinky helping hands but their clamping force is insufficient to stop it from sagging when I put force on them. Apparently the chinky ball-joint vices are reasonable if you shim up their ways to counteract the terrible casting tolerances. The supposed gold-standard is the “panavice” brand name, and they’re not too too expensive.

>>1890135

>>1890095
cheers, will look into these other options

>>1890256
Be sure to check eBay or other 2nd hand stores. Things that used to be on vehicles or industrial machinery are probably your best bet for getting them affordably, bonus points for getting broken parts and fixing them yourself.

first for fuck adafruit and their insanely overpriced dropshipped garbage

>>1890291
based. somehow the chinks don't make clones of every single thing on the adafruit store, so i've had to buy from them on occasion.

>>1889427
hey guys, im new to /ohm/. Rolling for a project

>>1890315
remote controlled powerdildo

>>1890315
Original fuzz circuit

>>1890380
>Original fuzz circuit

good choice. fuzzy guitar was a big hit 45 years ago, so it's due for a come-back. watch all the Tik-Tok hoes start to dance to it any day now.

>>1890387
It's sad that guitar became a nerds instrument

>>1890380
>>1890387
Not him, but fuzz is hard-clipping, right? I made a variable hard-clipper with a dual-gang pot and two inverting op-amps, with the pot hooked up such that the total gain is always unity but the amplitude in the middle can hit the rails. Can add antiparallel diodes in the middle too, in order to get a soft-clipper instead. Zener diodes that fit within Vcc give somewhat harder clipping than normal silicon diodes. Might be interesting to add some sort of avalanching part (backwards NPN or DIAC?) or a part with hysteresis in place of the diodes.

But it was just an experimental circuit built with a poor power supply (single rail with janky rail splitting), resulting in a noisy sound if the input gain was too low. The I/O ports were also a mess of 3.5mm sockets. So in the future I want to get/make a split-rail DC power supply that can handle the current, and use DC and signal connectors that let me easy hook up audio and power to whatever combination of audio effects I want. RCA is probably significantly better than 3.5mm, but it connects the signal wire before ground.
Anyone have other suggestions?
I won't be making balanced audio signals because I'm not into hifi and those connectors are expensive, and while I have some RG58 + BNCs lying about they're inflexible, large, and a pain to crimp, so I'd rather not. No clue what connector to use for ± power though, needs to be directional enough not to fry my shit with reverse polarity. For the power supply I'll probably just grab a couple of isolated 2nd-hand power bricks and stack them in series, with a 78XX+79XX plus some caps. Might want the +ve one to be grounded but the -ve one not so I can get that ground connection. 12-15V for the linregs sounds good, as some of my ICs won't run above that. Might need more rails for other voltages too.
IIRC those computer power supplies have ±12V rails and +5V, but I don't know if I can get one of those 2nd-hand.

Get one of these filters for easier soldering anons
Super cheap compared to consoomer gear and refillable easily, can filter an entire room.

>>1890408
What am I looking at? Are those standard parts in some other industry/hobby?

>>1890412
They're commonly used for cannabis growers, available as a kit to absorb odors from indoor "grow rooms"

>>1890408
I've read that carbon filters don't do really do anything for flux fumes. I'd rather just save up for a HEPA like the Hakko

>>1890417
You don't need a filter, just something to push the fumes into an open window.

i have a laser from chinks, but i has no driver board. so i wonder if it would be save to hook it up to bench power supply set the current to 1A which is what it operates at and let it rip? i know using CV power source would kill it instantly, so i am hoping setting the bench supply into CC mode should basically work as a poor mans laser driver board right?

>>1890426
just use a cheap pc fan with a duct on it then

>>1890429
Should work fine, so long as your benchtop PSU output doesn't have big spikes to worry about.

hello guys, I'm learning electronic and there's something I don't understand with the buck converter: why, when the switch open, the tension at the coil doesn't get very high? I think the tension at the load is the same than the capacitor, and since point A is shared, the coil have the same voltage difference, right? but why?

>>1890455
While the switch is closed, current flows through L into C and supplies R. When the switch is opened the current keeps flowing (now through D) and the energy stored in L transfers to the output. After a while the voltage (tension) across L is zero. Then the next cycle starts.

>>1890460
OK for the principe, but what avoid VL to get massive? di/dt become near infinite...

>>1890464
>di/dt become near infinite...

this may be true in simulators using ideal components, but if you muck around with real coils with real-world characteristics, you will get, at most, 62.76V voltage spikes, unless it's the weekend when you get a couple of volts more.

>>1890455
>tension
Are you ESL? Reminds me of chinese sellers talking about 'MOS tubes' and 'high pressure'
>>1890464
Short answer is the diode stops it. Yes, when you shut off the switch the voltage will tend to get very large (negative) but as soon as it gets bigger than the output voltage the diode will turn on and supply enough current to prevent it going any lower.

Can anybody here tell me what im doing wrong? Im trying to rewire this ethernet jack, because the guy who wired it before didn’t use the t568b wiring code, so I wasn’t getting internet on my windows 10 pc, only on windows 7 was I getting something.

Pic related is the result, according to every manual I’ve seen this is the correct way, yet it’s still not working? I have to note that it was pretty hard to get the individual cables into the holes or whatever you call it. Any ideas?

>>1890476
ESL?
so the diode act like a freewheel diode. aight.

>>1890482
>Can anybody here tell me what im doing wrong?
>shared media ethernet
Your mistake is you haven't thrown that shit in the trash and replaced it with something made this century.

>>1890488
So uuhhh... Is there something I can do to make it better? Or I’m just fucked? Sorry I’m illiterate in these things

came her a couple weeks ago for a bike generator phone charger, got some tips and encouragement that kept me going! Some trial-and-error later, made this splash-proof delon voltage doubler + voltage converter with some smoothing caps, tried it out thus morning and it worked! Need the switch, because my phone won't charge after a standstill without a disconnect (phone software problem I think). I'm totally ready for my week-long wildcamping trip, independent of power outlets. Thank you for the input, kind anons!

>>1890491
Imma be honest I don't know much about that ancient stuff. When it was working what speed did it link at? Where do the two cables go? What is the other end capable of? What kind of cabling is it? Looks like it's embedded in concrete so hopefully it's not CAT3. Probably you will want to install two jacks and a switch in between, or you might be able to do the old 2x100base-tx through a single cable trick.

>tfw chinks have a way to do multiple colors of silkscreen/soldermask on PCB
How it is even done? I'm not even sure if it's multi-color soldermask or silkscreen.
I've seen and done various PCB art but I've never seen anything like that.

>>1890482
>>1890491
Use 10/100 BASE-T wiring (2-pairs) instead.

>>1890417
Werks fine on my soldering just recirculating air through the filter

>>1890508
Flux fumes are mostly fine particulates.
A carbon filter won't do much.

>>1890482
>it was pretty hard to get the individual cables into the holes

as with double-penetration, they woulda gone in much easier if you hadnt twisted them together.

the real problem seems to be you have 2 ethernet connections in parallel. this doesnt work. you cant splice ethernet: you need a switch or hub so you wont get collisions at the electrical level.

>>1890482
The two cables go in the same or different conduit?

tough choice .. do i keep fucking around with hot air or do i splurge for a REEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEflow owen?

>>1890514
Are you building single-sided PCBs often? Reflow oven.
Are you mostly doing repairs involving SMD components? Hot air (or IR for BGA rework).

>>1890511
Yeah I didnt set this up, i just opened the hole to look inside, it was already twister like that. So I guess ill have to call someone to take a look at this?

>>1890512
Same conduit

>>1890529
You need to find out where the cables end and eliminate one, or install a switch and plug both cables into the switch.

>>1890408
thanks anon ill look into this

>>1890568
too expensive

>>1890577
Maybe you could build a soldering bong and filter the fumes with water.

We had an LED board with an onboard DCDC converter to supply constant current tested in an EMV lab
It's part of a ca. 1,5m long lampshade with its power supply sitting at the bottom and the LEDs on top
The converter runs on 800kHz and the radiated emissions peaked 3 times between 100MHz and 250MHz or so.
Assuming the 800kHz wave is triangle the amount of frequencies in the 100MHz-250MHz range must be very low.
What are the odds that the ca. 1,5m wires are acting as an antenna? The corresponding wave lengths would be around 3m to 1.2m.
Would adding a common mode choke to the wires be sufficient enough?

>>1890719
I think what I'm looking for would be a differential mode filter

>>1890719
Put a dummy load on the power supply and test for the same spikes without the long wires connected.

>>1890750
We don't have access to the lamp/power supply nor can we conduct EMV tests at our heart's content
Though I too suspect a cheap power supply was used or it is running too close to its rated power

has anyone come up with the magic search term to find these THT mains converter modules on aliexpress? they're useful and i want to buy a handful of them but digikey charges too much. aliexpress just gives me bucks and wall warts.

>>1890776
SMD flyback transformer is as close as I can get.

>>1890426
>he doesn't breathe the fumes while holding the solder spool in his mouth

>>1890797
i've literally done this

>>1890802

>>1889427
/o/tist here with a 6 volt motorcycle, is pic related a good deal?
The stock bulb is on my bike 25 watts (so 6 volts @ 25 watts = 4.16 amps), would it be cheaper/better to throw a step-up converter on a 12 volt bulb?

>>1890803
looks good to me

hey /ohm/ is it best to make your own lab bench PSU or buy one?

>>1890827
how much $$$ u got?

>>1890827
buy one so you have one that works
make a reasonable one with the intention of upgrading it over time.

>>1890827
buy one of those 20-30$ Riden buck modules and hook it up to an old laptop power brick.

>>1890855
That's what I was thinking but my laptop power brick is only 19V not 24V like those things need

>>1890845
As much as I need

>>1890848
(you) I gotta take a shit

>>1890827
If you have a shitty ATX power supply laying around you can get a breakout board for like $5-$15. Only problem is that ATX only has 3, 5, and 12 volts

>>1891009
I would do this but I want a variable power supply

>>1890848
after taking a shit that took not one but two flushes, I have come to the conclusion that I dont have the money to do both

>>1890827
Economies of scale ensure that buying literally anything will always give you a cheaper and better quality product than an equivalent you make yourself at the same price point. Buy a decent dual channel lab bench power supply. Hard to recommend output voltage current without knowing what kind of projects you want to do. Mine's a GW-Instek model that can do 30V 2A per channel and they can be wired together in series or parallel to do 60V 2A or 30V 4A. It's generally fine for me but if you're doing higher power projects that pull more current you may want 5A per channel or something like that. You usually get a dedicated 5V 3A output as a bonus with these supplies too for digital stuff.

But I'd say also build one anyway. It's a good experience and you can never have enough power supplies really. If you only have the budget for one right now though then I'd buy first. You'll have a lot more flexibility in what you can do.

>>1891013
thanks anon, I'll do both desu, my real question should have been is it safe to make one yourself

>>1891034
I didnt mean to say desu im not a weeb

>>1891013
>Economies of scale ensure that buying literally anything will always give you a cheaper and better quality product than an equivalent you make yourself at the same price point.
Eh, it really depends, mostly on what the markup is for the thing in question.
Economies of scale absolutely means that you can't make something in single quantity for cheaper than what it would cost per mass produced unit, but if the markup is like 500% you can probably still make it for less than MSRP

>>1890719
Those won’t be harmonics of the main wave at the very least, be it a triangle or a square. They could be switching transients, but even 100MHz is pretty high for a MOSFET made to run under 1MHz. The lack of lower frequency harmonics would lead me to suspect that it’s radio as you guessed, should be trivial to use a cheap SDR to search for nearby broadcasts on the same frequencies. Maybe it’s the local hams deliberately doing it as retaliation for the shitty LED power supply.

>>1890827
Power brick plus a cheap class-D amplifier, with a pot on each input. Add a switch so you can have each rail be the opposite of one another with a double-gang pot. Instant high-efficiency push-pull split-rail power supply, that you can also use to buffer the output of your 50 ohm function generator.

>>1891082
>>1891085
Do you have a boner for RF?

>>1890827
Buy a decent one, not the cheapest junk you can find. You don't want to spend 3 hours trying to debug your circuit to find that your power supply is injecting switching noise into your circuit or falsely triggering current limiting. If you're poor get an old one or use batteries.
Get decent leads too for similar reasons.

>>1890827
Hook your dingleberry pi up to a potato and make a LED blink for $75.

>>1891091
Yes, but I don’t see how my recommendation of a class-D amp implies that. They can also go arbitrarily close to 0V as well as sinking and sourcing current to some extent. I wouldn’t use one for low-noise applications though, unless there’s some trivial way of driving a linear stage with one that I’m missing. It would have to be a totem pole too if you want push-pull capability, and the feedback loop to account for that dropout would be difficult to incorporate into an existing class-D.

>>1891082
arent emv tests done with shielding on the walls of the lab to prevent something radio signals from other sources to skew results?

>>1891249
Yes, but by turning the appliance off and just listening for RF signals with an SDR, you can determine if there are any radio signals that could possibly be the culprit. With any luck, your SDR is more sensitive to those signals than your LED circuit. So it's like a counterpart to an emv test, instead of shutting out the radio signals and listening to the appliance itself, you shut off the appliance itself and listen to the radio signals.

>>1890802
Several of my EE professors do that.

>>1889427
I have a CCTV module that can be programmed using an included "ADKEY" cable. From what I'm seeing, it looks like it sends commands to a processor based on different voltages. What is this control system called in English? Most of the documentation on the "ADKey" are in Chinese.

>>1891266
It's a pretty common method of "multiplexing" buttons, often seen in automotive steering-wheel consoles IIRC. "Analog button circuit" and "button resistor ladder" both show similar circuits. Though the actual resultant voltages will vary depending on the system being used. An R2R ladder is an easy and efficient way of getting a lot of digital lines on one single analog line, but I'm not sure if they work with buttons.

>>1891257
Sadly I don't have access to that lab so there's no way to be sure

I'm changing my alternator from a 60 amp'er to a 250 amp'er. As I'm upgrading my starter and wires anyway, I'd figure I'd add an ammeter. Independent of any specific brand of ammeter, in general would it be safe to put a -60/+60 range ammeter on a 250 amp system?

In terms of power usage, the stock starter only draws about 30-50 amps and the stock battery is a standard group 65 with 850 cca. The new starter only draws around 10-20 amps. I'm doing this whole job as a way of future-proofing the vehicle in case I want to use it to charge a lot of (4+ batteries worth per shift) of power tools. Also, the bone stock vehicle does in fact use an internally-regulated alternator and not an externally-regulated alternator or a generator, in case that information is useful. Any other advice would be appreciated.

First time /ohm/ OP here.
So if /diy/ has a new thread approximately every hour (4stats.io doesn't give any better number), and has 150 threads at any one time, then it will take about 6 days for this thread to go from initial bump limit to the archive. Dividing that into 10 pages, that's about 1.5 pages per day. So if I'm available to refresh the general once per day, then I should do so at page 9 or 10 when I see it, any sooner will be jumping the gun and any later I'll be potentially unable to post the link in the old thread.

This post is also aimed at the people who make a new general as soon as we hit bump-limit.

>>1891360
>I should do so at page 9 or 10 when I see it,
page 3-5 sounds more reasonable.
You can guarantee the new thread will always be at least 3-5 pages higher than an autosaged thread (due to sage), and as soon as the old thread dies, we'll have 2-3 days to bump new thread back up

>>1891432
>page 3-5 sounds more reasonable.

You underestimate how anal the guys who think they run this thread are.

>>1891440
there is a right and a wrong way to address autistics
don't get offended, don't give them an excuse to go off topic, and frame everything as dry and logically as possible and everything will work itself out

>>1891360
Why does it matter?
>>1891442
>autistics
Is this the novel smear du jour?

Can I get some advice on how to find an equivalent IC for something that's been end-of-lifed? On the board is a HA1127, which is a general purpose transistor array. I'd prefer something currently in production.

I also need to find something equivalent to a MC3740A, but I'm guessing nobody's making that anymore and I'll just have to get NOS chips at the usual suspects.

>>1891797
NTE912
ECG912
MC3346P

>>1891797
>MC3740
getting zilch on that search, might be an in-house exclusive or a rebadge, unless you've got better search terms than I. What does it do?

Slightly off topics, but doesn anyone know what the name of this formula is, or what the context is? I'm fairly certain its electronics related, but i have no clue where to find it.

>>1892294
Multiplying gains together makes it looks like some sort of cascaded amplifier array, but I haven't a clue what the N numbers could stand for. Looks somewhat similar to something I might have seen in an equation for modelling a transistor or discrete amplifier circuit though, so have a look at that. Also note that since there aren't any time or frequency components in the equation, it's likely low frequency (or DC), with a corner frequency significantly higher than the circuit cares about. If those were dynamic gain values, I'd expect them to be represented by "A"s, not "G"s.

Gonna give us any more context? That photo of a screen has to come from somewhere.

common base, collector, and emitter amplifiers all use either the emitter or base as the input terminal. is there a transistor topology that uses the collector as the input terminal? with the base tied high or emitter tied low, and a resistor on the opposite terminal, i guess

>>1892301
simming it, i get common base with collector input being a voltage follower of some sort, probably similar to an ordinary common-base amplifier. not very good at common-base amplifiers, but the input impedance is probably near identical here. common emitter with collector input gives me an attenuator, which could be useful if it's lower-noise than using resistors or otherwise easy to tune. looks like i can alter its attenuation by changing the emitter voltage, but i guess that applies to all transistor amplifiers

>>1892294
>>1892296
G's are commonly used when looking at gain stages of Op Amps (as an example). Like the internal gain stages.

>>1892310
Maybe the addition of more and more terms is the accumulation of the noise from each stage? Though the addition of the reciprocals of these gains is entirely foreign to me. But why you'd stack three simple op-amp gain stages in series is anyone's guess. If it were a cumulative phase-difference oscillator I could accept it, but it doesn't look likely. I think it's more likely that it's a bunch of near-beta-limited transistor amplifiers cascaded together to get a few orders of magnitude worth of gain, since cascading multiple op-amps worth of gain would be almost entirely useless.

>>1892341
>since cascading multiple op-amps worth of gain would be almost entirely useless.

wrong. op-amps have a gain-bandwidth limitation. if it's 1 million, say, then you can get a gain of 1 million but only up to 1Hz, or a gain of 1 up to 1 Mhz.

moral: if you want a lot of gain with a lot of bandwidth, you need to cascade some op-amps.

>>1892301
What do you mean by input? You might consider an emitter follower with a current limiting resistor on the collector to use the collector as an input. In general the base-emitter current is modeled as independent of collector-emitter current, so you can't change the base by adjusting the collector current. Of course you can connect both the collector and the base to the "input" and get a diode connected transistor.

>>1892346
>op-amps have a gain-bandwidth limitation
I know, but the equation doesn't mention frequency or time at all, so I assumed that it was a low-frequency one. If I was anywhere near that cutoff frequency I'd replace G_n with the actual transfer function A_n, since it will matter for noise too. But I suppose it's conceivable.

I know that it's an example, but 1Meg GBP is low for an op-amp, right? Can't imagine being limited to a gain factor of 50 with audio signal bandwidth up to 20kHz. Pretty sure I saw 100Meg being more reasonable. At the very least, I don't think I've ever run into the GBP in my experiments, at least aside from testing the bode response of an op-amp.

>>1892350
Maybe he means something like this? Both transfer the signal from one end to the other, at least to some extent.

>>1892351
What does it do? Does out1 get affected by 1/hfe?

>>1892351
it's the gain stage my dude. not the amplification stage. cascaded gain stage

Working on a little pet project for my old beater BMW, it's an adapter board that'll fit an MS3Pro Module and holds the necessary components for the high-current ignition coils, a CAN gateway, A/C control, plus some small power stuff for Fuel, oil and manifold absolute pressure sensors.

>>1890803
>nigger_rigging.jpg

>>1892263
Crap I typed it wrong. The ‘lexia fired up yesterday sorry. MC3470A. It’s a floppy disk read amplifier.

For me, it's the CD40106BE. The best integrated circuit. I even ask for extra feedback potentiometers and the chip is so friendly and more than willing to oblige.

One time I asked for feedback potentiometers and it gave me hysteresis. I said, "Wow, hysteresis!" and the nice friendly integrated circuit buzzed and said, "EEEEeEEEEeeeEEEEEEE".

Now the IC greets me with "EEEEEEeeeeEEEEEeeEEEE" and ALWAYS give me hysteresis. It's such a fun and cool atmosphere at my local lab, I go there at least 3 times a week for hysteresis feedback and a schmitt triggered inverted logic level, 1-2 times for digital waveshaping on the analog input of my circuit, and maybe once for monostable vibrator triggering when I press a button but want a short-lived high logic level that doesn't fall back to 0V immediately, buffers the input, and can be adjusted for digital decay values.

I even DIP my chip in breadboard sauce, it's delicious! What a great integrated circuit.

Suppose you’ve got a circuit schematic of an IC’s internals from the datasheet. Can you just... build the same functionality out of discrete components and it’ll behave the same way?

(I assume there’ll be differences in nuances of the specs especially if components need to be matched. My question is what differences you should expect, how to compensate for them, etc.)

>>1892568
Yes, it's often possible, but the practicality of it will vary widely. So it's highly dependent on the situation and circuitry. Like pic related, super simple, though variencies between discrete transistors will be greater than if they were all on the same die IIRC.

>>1892568
Yeah, pretty much.
The circuit schematic isn't always 100% accurate to what's inside, usually it's simplified a bit, but it's functionally equivalent.
With analog ICs you might have trouble getting discrete transistors which are closely enough matched to meet spec though.

>>1892623
>usually it's simplified a bit, but it's functionally equivalent.
Sometimes they'll give you the count of transistors and capacitors and maybe resistors underneath the simplified internal diagram, so you can count how many transistors and such have been left out.

>>1892590
Funny you should post that pic, the device I'm rebuilding now has a pair of those.

>>1892590
>protection diodes on the outputs
huh? is this to allow blanking by cutting power to the ic?

>>1892680
>blanking
What? Do you really not know what the diodes are there for?

>>1892680
>PROTECTION diodes

they are there to protect the semiconductors from esd and transient voltages jumping around from inductors and capacitors

>>1892568
basically if in the datasheet there is a whole schematic then yes you can almost always make it from discrete parts, it just wont have the same electrical characteristics

>>1892761
>it just wont have the same electrical characteristics
I guess I’m trying to figure out how to determine what the critical characteristics are and how you compensate for that.

But I guess that’s part of the magic you learn to achieve with an actual EE education haha

>>1892783
It's really, really application specific. Because of how IC's are made, the internal component characteristics will usually "match up" better than a collection of discrete semiconductors. Look back at the ULN2003 for example. Just a bunch of darlington transistor pairs (and some protection diodes). Because of how the IC is created, those transistors are pretty much all going to match in terms of, for example, gain, breakdown voltage, forward voltage etc.
Slap together the same thing with discrete components, and those individual components will most likely have have a wider range of values compared to eachother. Transistors with different gain, different forward voltage etc. Sometimes it matters, in which case you would have to pay extra for tighter tolerance components or test a bunch until you collect enough with proper matching values. Sometimes it doesn't matter. Sometimes it only matters years into service; a few weeks ago I traced a serial communication fault in a $250k machine down to a 1N4007 diode (in series with 2 others of the same) that had a slightly different forward voltage which eventually led to thermal related failiure.

>>1892802
God what a mess. So I'm trying to unfuck the analog board on an Apple II disk drive that worked great for 35 years (well, about the last 20 it spent in a basement but it worked great for the last 6 months). The Apple fanatics online told me to start reseating chips and the damn thing went from "something off" to "destroying data"... and then not functioning. And of course the only schematics out there are for a massively different board revision.

Anyway I'm just rechipping the fucker and hoping to god that works. All but two of the chips were available on Mouser for pennies. The two that weren't I can probably get on eBay if it's still not working. Folks online were quick to suggest the 74LS125 did it. We'll fucking see.

>>1892810
>Folks online were quick to suggest the 74LS125 did it.
Very likely, but you might as well replace the electrolytic caps while you're at it. Check your chip sockets and the traces around their pads too.

>>1892761
Yeah, but on the output and not on the input? Helps when driving an inductive load I guess, but for them to neglect them on the input feels like a mistake.

>>1892823
Actually, even then I can only see the diode to GND ever seeing use. Plus with internal diodes for it you couldn’t use the array for a bunch of series boost converters, in case you’d ever conceivably want to do so.

>>1892824
Wait, I meant only the diode to Vcc will ever see use. Unless you’ve got an inductive load from the collector to GND, which doesn’t really make sense.

>>1892819
Yeah I ordered caps and all new sockets as well. I suspect I busted at least one contact while trying to clean them in desperation to make it work again.

>>1892296
>>1892310
Its from some plebbit post about how the variables look like a naughty word. Thats great background though, thanks for your thoughts.

>>1892848
No biggie. Easy fix. You itchin to play some Oregon Trail?

>>1892901
Oregon Trail Deluxe > original apple versions
Its so much better its not funny

>>1892901
Actually wanted to boot up Merlin on actual hardware and develop some shit. I’m in love with the simplicity of the 6502 architecture and like some of the weird unique challenges of the Apple II platform (like doing sound by toggling the speaker driver on and off directly to generate frequencies).

>>1892907
It looks better, sure. But does it tickle your ass with a nostalgia feather?
>>1892911
http://www.mirkosoft.sk/65832.html
There are pin compatible 16/32 bit options.

>>1892915
>There are pin compatible 16/32 bit options.
Oh momma.

>>1892915
>A 32-bit 6502
Huh, never heard of that before.
Sounds like it never materialized though, rumored to have been in development starting 2005.

>>1892918
Bill Mensch is getting old. I hope to god he has an adequate successor that follows his design philosophies.

>>1892918
You're right, my fault. Seems it wouldn't have been 100% seamlessly compatible anyway. Oh well.

>>1892915
>It looks better, sure. But does it tickle your ass with a nostalgia feather?

It was the only one I ever played, I played the windows version not the DOS version too so I have a hard on for the music in that game.

>>1892932
>I played the windows version
When I was a kid, a 386SX-25 IBM clone averaged about $3000 with monitor, keyboard and mouse. My pants weren't fancy enough for PCs back then.

>>1892932
t. zoomer
>>1892942
Nice. My first exposure was in elementary school on this weird IBM PS/2 classroom LAN system. That was in the early 90s.

We moved a couple years later and the area schools were all running much older Apple II-based systems.

>>1893013
If you were in elementary school in the early 90s, you arent any older than I am.

>>1892823
fug, meant to reply to >>1892760

>>1893025
cope

Just found this next to the road near my house, alongside some other drivers, which were basically all full of water. These look intact though, and less prone to water damage even if there was water in them.
Anyone have any idea why they're connected together like this? Reminds me of acoustic delays, maybe with a pulse forming network I could store a couple bits of memory in here. But since it's only 8cm (0.08m) of separation, and the speed of sound in air is about 350m/s, the total delay is 230µs, the inverse of which is 4400Hz, and going a few times that would likely be difficult for a driver of this size.

It says "454246" and "8H412j" on the back of each driver, with no other markings. If anyone happens to know what that means (better than just googling those) that would be neat.

>>1893039
isobaric configuration

>>1893041
Never heard of that before, interdasting. But from what I can tell by reading wikipedia, they rely on the air being sealed between the two speakers, which looks difficult in this instance. The distance between the cones is also larger than it could be, which too contributes to a less ideal system.

>>1893045
Yeah that's the best I can think of with though. Maybe it was an experimental rig or the guy tore them out of an isobaric enclosure to replace them with better drivers?

Or maybe it's an entirely experimental setup. Wouldn't happen to be a college town would it? If so it might just be someone into acoustics who was cleaning house.

>>1893054
>college town
It is, but it's a suburb away from where most student houses are. I suspect it was left here since it's right outside the junk shop that I live next to.

Speaking of which, the junk shop also had this. Since the case isn't screwed together properly, I had a look inside and saw a vacuum tube, probably a tetrode or pentode judging by the number of pins. I wonder if it would be useful, considering I have a scope rated to 400Vpk anyhow?

Additive synth anon here, I was wondering how to go about making a PLL with a fast enough response time but a really low ripple at the VCO input. Standard procedure for non-sinusoidal waves is to use an XOR gate with a low-pass filter after it. So I realised I could improve on the result of this by adding a peak detector to the output of this filter, but then it would only have fast response in one direction and be as slow as just using an RC filter in the other direction, provided I didn't use a transistor to pull the cap low. But pulling the cap low to drain the hold caps would require variable timing, as the frequency of the VCO will vary by at least 1 order of magnitude, hopefully 2-3.

So instead I could use two peak detecting sample+hold circuits, with enable inputs being fed off an f/4 square wave's rising edges, and the draining of the hold caps also be triggered by an f/4 wave that's 90 degrees out of phase. Then I just use op-amp ideal diodes to give an output that's the greater of the two. The result being, the output will update to the phase comparator+lpf waveform's voltage peak every cycle or two. Might be able to get that f/4 to an f/2 if I use a schmitt trigger on the VCO's triangle wave itself to get a usable phase shift. So if I'm down at 20Hz, I'll get really low ripple but it will still update fast enough for any lag to not be audible, and the PLL itself shouldn't desync. I think that's about 3-6 op-amps, hopefully I can fit it all in a single quad package. But it might still be too complicated to have 8 of in an additive synth, alongside the planned 1 op-amp + 2 comparator + 2 NOR gate variable phase shifter.

I can draw a circuit if anyone's interested, it probably sounds like a bunch of mumbo jumbo.

>>1893093
I have no idea how to design a PLL, but all I know is that the Earthquaker Data Corrupter is supposedly a good-tracking PLL. Not sure you could reverse engineer that, though

>>1893093
Unless I could just trigger a sample+hold circuit (one without diodes for peak detection) off the XOR's output directly.
Pic related, then a second one triggering on the opposite waveform. Though maybe to get the timing right it would require the same amount of op-amps and TFFs and such, plus requiring some sample-timing transistor wizardry.

>>1893098
>Earthquaker Data Corrupter
I'll look into that, thanks.

>>1893098
>>1893104
Never mind, it just uses a CD4046. While its second phase comparator does intrigue me, I don't think it's worth using an entire CD4046 for each stage, as I'll need my VCO to output a proper triangle wave with a circuit I've already designed.

Also just realised that I'll get a gradual frequency dependant phase shift to all my harmonics, since the phase comparator's phase difference needs to change as a function of VCO voltage. Might need a seperate fundamental frequency oscillator to avoid this.

Hi Frens,

Bought some stupid idiot Arduino Leonardo pro micros from China, 4 months later they've showed up in DFU mode. Just trying to get the proper Arduino bootloader on it, hooked all the SPI pins up to an Arduino Uno setup as an ISP, however it keeps hitting me with the avrdude: Device signature = 0x000000

Thoughts and concerns lads? I can't seem to find any good documentation on this process. Surely I'm not the first dickhead to order one of these shite Chinese pro micros and want to use it like an Arduino.

>>1893112
>I can't seem to find any good documentation on this process. Surely I'm not the first dickhead to order one of these shite Chinese pro micros and want to use it like an Arduino
Programming any arduino via ICSP is identical, it shouldn't be any different with a pro-micro compared to a nano or uno or whatever, as they all use an Atmel standard.
>Thoughts?
Firstly, ensure everything is wired up properly. Breadboards have been known to have shitty connections that make this kind of work very annoying. Try swapping MISO and MOSI and back again after each troubleshooting step. Do you have a capacitor on the reset pin of the programmer? That's also somewhat important (at least it is for some programmers), but I don't think it matches your particular symptoms.
I spent hours on end trying to fix a very similar problem (trying to program a nano via the ICSP header), turns out the solder joints to the header that I'd done years prior were cracked, reflowing them fixed the issue.
Personally I use a USBasp clone from ali for this stuff, it works well with little hassle, from either the arduino environment or the command line via AVRdude. Except for when I had to flash new firmware on it by using an arduino nano as a programmer. I also made a little PCB with an IC socket on it that plugs right into the USBasp with its own ICSP header, in order to program DIP ATMega328p ICs to use on a breadboard or to solder into a project.

Also note that there's an arduino + MCU thread already on /diy/, so it's possible that you'll get better suggestions there. I still doubt you'll get anything more helpful than "check your wiring" though.

>thot gif
Please leave that kind of content for the less autistic boards. Either post a pic of your problem, or don't post any pic at all.

>>1893115
>leave that kind of content for the less autistic boards

please be less gay, kthx.
cutie asian tits should be included in every fifth post, coz they brighten up any life situation, and make life worth living,

>>1893135
This wahman taught me Ohm's law.

I want to convert an old flashlight to led and have it use a differnt style of battery. Its a 3 C cell body and i believe the reflector is like 40mm

What do?

Anon that likes doing repairs and mods on headphones here.

Did a simple Portapro driver to MMCX conversion a while ago, along with slapping a few chink drivers into older Sennheiser sets i had myself and got from friends like pic related with some weird Onkyo drivers that fit and screwed in right into the shell. Sorry for the terrible cold solder, as i'm using a 10 year old 10$ iron and have a T12 clone and accessories stuck in the mail.

One thing i'm wondering if there are any simple inner PCB boards for headphones you can buy separately? Meaning the board that split the R/L/GND to R+/R- and L+/L- instead if just wiring the ground together.

>>1893204
>board that split the R/L/GND to R+/R- and L+/L- instead if just wiring the ground together.

you could use a couple of audio transformers, tho i can think of any logical reason to do so. of course, you'd lose some fidelity and bandwidth.

>>1893202
>What do?

simplest answer is to find the model of the lightbulb and google for an LED exact replacement. they do exist.

>>1893221
That would be a overkill on complexion as the board has to be rather small and short, and would require quite a bit of power (essentially making a headphone amplifier). Which isn't going to work since i'm basing it on working on any basic 3.5mm output and or headphone amplifier, and i don't need to make a headphone amplifier inside a headphone.

I know that it's much easier just using either 3.5mm jacks or if i want to run it balanced, a MMCX or 4pin miniXLR connectors but not all headphones i get to fiddle with it's fully possible to put a jack on or use a drill to make room for one. (Or have room for a miniXLR plug).

I think the best option is just taking a standard universal PCB, cutting into the maximum size that fits inside like a 1.5cm on 1.5cm (or 2x2)board , soldering some pins, shorting said pins with solder or wires, and soldering wires to said pins and call it a day.

But i'm still hoping there's a board you can buy separately.

>>1893112
>>1893135
>>1893190
>posting giant lumps of fat
yikes

>>1893232
>would require quite a bit of power (essentially making a headphone amplifier)

it's just 2 transformers, dude, they're very efficient, so power out = 95% of power in. no amplifier needed. but i still dunno what this is all about, so i'll shut up now and forever.

>>1893224
Its super old so i dont know if it takes a normal c battery though. I wanted to make a battery bank/ housing for it

>>1893233
those are bags of sand

>>1889427
I need to be able to pull BOOT1 high with a button press. But I also need that button to function as a regular button when in the main loop. Im a little confused because internal pullups wont work here. All in all Ill probably have near 30 buttons connected to this. My thoughts were that 29 of them would be internal pullups with the buttons connected to ground, and 1 would just be input and the button connected to 3.3v and BOOT1.
Im worried because Ive read somethings about the voltage regulator being crappy so would having a button connect 3.3v to the pin be a bad idea?

Just made this little LED flasher with 74AC14. This IC manages to work with just 1,5 V supply voltage, it's a great IC.

>>1893724
neat, though i'm more fond of the 4000 series as they can run up to 15V or whatever

>>1892819
Mind my asking why you say "very likely" about the 74LS125? Is it because that's just a common problem with the Apple drives? Or are they a likely weak point in circuits integrating them?

Are stupid questions allowed here? If they are, I have one regarding speakers. I'm trying to connect a speaker system to pic related, but the wires are open.

I was thinking of just buying these:
https://www.amazon.com/CERRXIAN-Terminal-Headphone-Converter-Adapter/dp/B06Y5YJRPD/ref=sr_1_3?dchild=1&keywords=speaker+wire+to+3.5mm&qid=1597702184&sr=8-3
but is it really that simple?

>>1893786
a speaker *system* like an amplifier, or just bare speakers?

Replacing all PDIPs on an old board with CDIPs: Based or cringe?

>>1893797
por que?

>>1893797
Are CDIPs more thermally stable or less noisy than PDIPs?

Is 8.77 volts enough to get a 12v dc fan to start spinning? Trying to fix this stupid dehumidifier

>>1893892
yes in my experience.

>>1893204
you dont need any PCB board for that, you gonna have to rewire the drivers so each will have their own ground to "the outdside world"

theres no way to separate the grounds (no gadgets) headphone ground is generally soldered together (the 2 dirvers shorted) this short is generally located in one of the drivers

>>1893797
based, but expensive

>>1893774
>Is it because that's just a common problem with the Apple drives?
I think it's because of people misaligning the IDC connector on the analog board header pins which blows the chip.

>>1893115
Yeah thanks man, I found out I had my SPI wired up wrong, it's always something dumb, ended up following a generic tutorial thanks for the tip.

Sorry I thought the tits were a bit risque but was keen for a quick reply

>>1893233
>you must be 18 years or older

How difficult is it to replace through hole mounted electrolytic capacitors on a computer PSU?

>>1894096
Probably pretty easy, like a 2-3/10.

>>1893786
>is it really that simple?

that amazon thing you found is for mono only, but you need stereo. another thing is headphone jacks cant put out enough power to get any kind of volume from a speaker. if you have a set of powered computer speakers, they would work fine, and they come with the proper audio cable to plug into your box.

>>1894096
>>1894097
>Probably pretty easy, like a 2-3/10.

nonsense. the negative lead is gonna be connected to the backplane, so trying to get those holes clean of solder is a nightmare unless you have some high-end equipment, or know some tricks, like the needle trick.

>>1893918
That would be the case and that's what i'm going to do, as what you basically get is a cable that gives a L/R and ground singal via an aux cord.

However the headphone drivers themselves have the R+/R- signals (Or L+/L-) by design. Now sure to make a standard single ended headphone that would basically require to solder both negative connectors from the driver itself to the ground and call it a day while if you want to make it a balanced output (aka separate grounds) that's a different story and not what i'm trying to make. I'll see if i can find an image of one.

>>1894133
Not like the drivers will sound any different if you feed them proper differential/balanced signals, the only reason to use differential/balanced wiring is for noise immunity in the cable itself.

does anyone have experience with the ebay seller survy2014?

>>1894307
might be worth asking on /g/csg too, but i doubt you'll find anyone who has bought from them either way unless they have 10s of thousands of purchases on their listings

>>1894474
didnt think of that thnx but I got a response on another site and they seem good

FUCK I think I fucking broke a trace and might've torn out some through-hole plating.

>>1894749
damn, son

Yep. I fucked a trace and at least three hole platings trying to get a 14-pin DIP socket out.

I should be able to use a wire for the trace. No idea what to do about the holes. God fucking dammit.

So uh. What do I do about this board guys? Am I fucked? Do I need to buy one of those board repair kits with the eyelets and punches? Or can I just flood the thing with solder and use wires to substitute for where these things are fucked?

>>1894749
Fixable with a bodge so long as the pitch isn't too tiny, post pics.

>>1894815
the top. this is where most of the damage happened. the fucked trace is in middle of the pic. there's probably another one because of that destroyed pad a couple pads up from it.

>>1894831
and here's the bottom

The shit thing is to finish what I fucking started on this board I need to remove 11 more sockets, most of which have that many pins.

>>1894831
Yeah that's fixable. Won't be pretty, but fixable. You're Apple II anon, right? Mind if you ask what the fuck you're using to solder with?
>inb4 PTC soldering iron with no extra flux

>>1894854
Removing sockets without chopping them out is an exercise in futility, at least if you don't have desoldering needles. SMDs are arguably easier since you can just incrementally push a shim under all the pads. Just go at the sockets with your side-cutters (or melt through them) until you can desolder them pin-by-pin, not like you'll need them afterwards.

>>1894874
Yeah I'm the guy fixing the Apple disk drive.
>Mind if you ask what the fuck you're using to solder with?
I have one of those shitty soldering irons with a suction bulb attached to it. Also a Hakko FX-888D iron with a bunch of tips, a thing of rosin paste flux, and some desoldering wick that my brother loaned me.

>>1894874
That's what I ended up doing in the end, using side cutters. I reckon I'd already done a good bit of damage to that one socket. Ah well at least I didn't start on the others before fucking that one up.

>>1894877
I'd maybe drop your temp a little lower, and save the solder removal for after the socket is out. Trying to remove enough solder that the pin just drops out is a damn pain without desoldering needles, and will usually result in your iron being on the PCB for too long and getting it too hot. Thats why your flux is all burned away and leaving oxidised solder behind, and that's why your solder mask is starting to peel off. A lower temperature will slow the time it takes to get the pad itself reflowed, but also slow the time it takes to overheat anything.

As for repairing, you'd better be using those machined-pin sockets, as opposed to the flat pin ones, as IIRC they have a little gap beneath them that you'll be able to solder to the pad to excuse the lack of a plated hole. The delaminated trace should be replaced with a small length of magnet-wire after cleaning the surroundings and gently scratching off the solder mask on either side of the break.

And clean that shit with isopropyl alcohol.

>>1894160
Well yes, and better power delivery overall.

>>1894902
>power delivery
Not really a big issue for headphones, or at least not high-z headphones.

Just double checking to see if I got this right.
Assuming the Vin to be +5V since I'm working with an arduino, these LEDs that ask for a 2-2,4V at 20mA will need a 180Ohm resistor?

>>1894946

calculation is correct but 20mA is a maximum safe value. generally speaking, you dont run anything at max. it depends on actual LED, but typically, it will be adequately bright at 15, or 10, or even 5mA.

also, some arduinos put out 3.3V, not 5V.

>>1894977
>>1894977
Thanks! I'll take that into account when finishing the PCB.
Also when the calculator asks "how many LEDs connected" I'm a bit confused. At most 2 LEDs will be lit at the same time, so I told the calculator but it just puts them all in paralel to the power source, can I share one resistor to light up all the LEDs if they are the same color and rating?

>>1895007
>can I share one resistor to light up all the LEDs if they are the same color and rating?

you can do it, but it wont look good, coz the current will divide between lamps and so 1 lamp alone will be brighter than when 2 are lit, and even dimmer when 3 are lit, etc

I have an idiot-level question that requires non-idiot answers, anyone here knows shit about light bulbs?

I have this high-pressure sodium bulb that broke down and I wanted to make it into a decorative piece, like I have many times before with regular incandescent bulbs. Not thinking it through I broke the ceramic plate at the end and carefully punctured the glass from below to extract the contents (normally the filament etc.) so it can be cleaned; only when the then did I think to inquire if it holds toxic gasses and/or poses a danger when touched with bare hands. From what I can tell on sodium-vapor bulbs it contained xenon and possible mercury.

Yes, I was a dumb cunt. Did I expose myself to respiratory danger and/or should go to the hospital?

>>1895064
>Did I expose myself to respiratory danger
no
>should go to hospital?
lmao

>>1895064

i have one in original packaging and it doesnt have any warnings.

they now sell similar items at the dollar store, all mounted on a base and ready, so you dont have to risk life and limb.

>>1895069
>no
Expand.
>lmao
Explain.
I mean, yeah the last thing I want to do in the Coronatimes is go to an ER but sure.

>>1895071
They also sell rat poison and bugspray under the assumption you don't consume them. That doesn't quite tell me it's safe to break open high-pressure sodium-vapor bulbs in an enclosed environment.

>>1895073

>>1895079
Thank you, this is exactly the kind of information I was looking for. I'm relieved.

Now to a more pressing matter: when the lamp broke down the inner tube broke and sodium condensed on the glass, which I need to remove. I tried the usual (soapy water, rubbing with a rag on a stick, shaking some salt inside to grind it off) but it doesn't come off the glass. It looks half-transparent in the image but is actually very dark and reflects a lot of light. Any suggestions?

>>1895037
Thanks! I know it was a kinda basic question but I'm just starting

Is it possible to have two different hex files flashed to a microcontroller and have it decide which one to use on a button press?

>>1895122
Probably.

>>1895122
Yes.
Can you switch it with a button during normal operation? Maybe
Depends heavily on microcontroller

Most will require reset, almost all will require some fuckery to get working properly because of read only areas, interrupt and lookup tables etc

Recommendation - use dual bank microcontrollers, easier to implement
Another option would be some rtos on mcu where you can easily switch tasks

>>1895103
I have no idea. Is it just baked on or is the glass etched?

>>1894883
Thanks for the advice man. I was feeling pretty sick about it last night. Was my dad’s computer and the machine I learned to code on as a little one. (Of course the comp itself is still solid, just the disk drive is fucked.)

I’ll order one of those sockets for that one I fucked up. I think I can get one with extra long legs so I can reach better with my iron.

>>1894831
The traces under the chip socket should be repaired with copper foil or magnet wire. The bottom traces can be bodged.

>>1895141
I don't know that there's a practical difference. What chemical method could I use to remove that grime?

>>1895190
Maybe someone with a better understanding of chemistry can chime in here. This old article says to boil toluene to clean metallic sodium lmao.
Don't do that or you'll end up like Richard Pryor.

Hi,
I got some water on my breadboard, about 3mL. I removed all the components, 5 transistors and an opamp. No power was running through it. I wiped off the water that was on the surface, could it have caused damage to the board or the component? How do I ensure it dries completely?

>>1895204
The components should be fine.

Squirt board down with your high purity alcohol.

Alcohol displaces water and dries fairly clean.

>>1895211
Thanks, not really sure how to get the alcohol inside the board though. For now I've just put it in front of a fan hope that's enough.

>>1895246
I meant to just give it a big squirt and let it run out of the breadboard.

If you want to be careful, then
You can probably pop the breadboard apart too and remove the metal strips so you can get the alcohol in there directly and watch them dry off

>>1895173
I think I can bodge the really bad top trace actually since it's just from pin to pin on the same chip and neither hole goes anywhere else on the same side. (so in theory the plating isn't even necessary). The second top trace might be fixable with a bodge wire as well provided I can get the other socket's pins out without killing the plating.

>>1895103
Vinegar or some other acid might work, as sodium oxides are basic. Might need an oxidiser too (hydrogen peroxide is relatively cheap) if there's too much metallic sodium present, or just heat it up until it all oxidises. If there's mercury though you'd probably need nitric acid, and mercury nitrate isn't something I'd enjoy making in a lab, let alone in a frat-pad's kitchen.

>>1895157
Is the disk drive circuitry on a seperate board? If so you might want to replace that board in total, though personally I'd aim to refurbish the existing one as neatly as possible.

>>1895306
>Is the disk drive circuitry on a seperate board? If so you might want to replace that board in total, though personally I'd aim to refurbish the existing one as neatly as possible.
Yeah it's a separate board. It's quite modular. The problem is that right now everyone's gouging like fuck for "vintage apple" shit online so it's tough to find anything for a reasonable price. I've got an eBay search saved in case anything comes up. I'm gonna do what I can with what I've got for now. Worst case I'll just get a new copy of one of the more common drives to read my old disks and drop a compact flash board into the computer to have fun with otherwise.

Really only thing I'm worried about with fixing this is if I do nonstandard shit whether it'll change the electrical characteristics of the circuit enough to break copy protection. But I guess I'll cross that bridge when I come to it.

>>1895290
It'll work fine but the DIP socket might not sit flush with the board if wires are in the way.

I haven't seen transformer symbol like this.
What kind of transformer does it denote?

>>1895322
center tapped

>>1895322
It's a matching audio transformer. One tap is for 8-ohm speaker, the other 4-ohm.

>>1895322
Very strange, I guess the core is capacitively coupling to the windings and hence somewhat useful to use as a terminal.

>>1895322
What puzzles me is that there are 3 connections on both coils. Shouldn't there be 1 less on the primary coil?

>>1895326
>>1895333

meant to reply to

>>1895328
It's not the output transformer connected to speakers. It is installed directly on the ballanced signal input.

I'm trying to verify is that it is soldered correctly to support unballanced input atm

>>1895333
feedback winding
>>1895336
So it's a balun then? Post more info and pics.

>>1895310
I mean I’d put a wire on the underside of the board between the two points. Unless there some weird capacitive coupling reason not to do that with traces running perpendicular to it (in which case I’d probably just attach a bodge wire directly to the involved chip legs).

What am I doing wrong that makes my solder connections come out burnt and yellowish? Is the heat too high? Too much flux? I only started recently

>>1895383
You might be dwelling to long on each solder joint.

>>1895383
It'll look nicer when you clean the flux off.

>>1895306
>Vinegar or some other acid might work
Yup - that did the trick, vinegar broke it down in minutes. Thanks, chemAnon!

>>1895456
I haven’t done any chemistry in years, that’s just what I’ve picked up from high school. But it’s a handy thing to know if ever you’re etching PCBs or designing something to be in a corrosive environment or whatever.

Looks like it's time to start buying from Aliexpress. Most of what I bought from ebay was chinkshit anyway.
Imagine telling our grandfathers that red-blooded Americans boycott American businesses because they're un-American

>>1895507
most american businesses aren't "american" in the way you'd think they are anyway. i'd rather pay $2 on aliexpress than pay $50 to the jews any day, even if the latter option might marginally help keep some western engineer employed.

Is KiCAD good enough?

What level of projects are EE firms/emloyers looking for when hiring entry level? I'm always told that projects will make me stand out but, say, building a functional FM radio feels too advanced for where I am right now

>>1895525
>want to be an EE
>FM radio is too advanced
surely you did this in your undergrad coursework anon

>>1895530
Sorry, should've said that I'm going into junior year. Circuits is the most advanced class I've taken. I like radios and I've tried reading about how they work, but a lot of it goes over my head right now

>>1895534
If you've got 2 years of EE coursework under your belt you should at least be able to fuck around with radio stuff. If you think it's neat take signal processing.

>>1895525
Read up about how PLLs work. Make one of them that can function up at those frequencies, and you’ve got an FM demodulator. Getting stereo will be harder though, pretty sure you need to mix the FM output signal with a 40kHz sine from somewhere (another PLL?) to get a difference wave, and just low-pass it to get the sum.

can anyone give me some tips on how to charge this battery? my charger broke and new one is over a month out (thanks covid). i have a variable bench power supply that i've used to charge li-ion cells before, but no clue how i can safely hook this up.

>>1895554
Take apart the broken one for hints on what the connections are.

>>1895555
i can't, i sent the broken one in for warranty repair, which is why it's going to take over a month to get a new one

>>1895554
can't charge your vape?

>>1895333
>there are 3 connections on both coils
Not in that picture. On the left, you have two ends and a centre tap. On the right, you have two ends (no centre tap); the centre connection is for grounding the core.

>>1895566
In that case, it might be a bunch of pads that are all commoned together to work with multiple devices or to carry more current. Plus a balancing lead and maybe also a thermistor. So do some form of continuity test with a high impedance or low voltage. But it might also be a bunch of comm lines for some deliberately opaque proprietary battery charging standard made to stop people using unlicensed chargers.

>>1895578
>for grounding the core
Not him, but is this for noise purposes?

Is a water squirting taser possible? I don't know much about electricity but I do know that 1 million volt stun guns exist, and that voltage can create a pretty long spark gap. So connect that power supply to the water stream and as it breaks up in the air the spark should jump across the gaps until it reaches the target, right?

>>1895591
The electric current will need a closed path to flow. So you could make one path with a water stream to the person, and a grounding rod from the taser as the other path, but if he's wearing decent dry shoes that won't work, and any water dripping down the target could short the shock without it going through the target's body. The other option is to use two parallel water streams, but the distance between them would need to be large enough not to arc, basically necessitating the water stream to have a strictly laminar flow (no breaking apart). You may need to tweak the water's conductivity with salts.

Now I'm thinking about some sort of vapour weapon, where you release two gases into a room. One gas is heavy and insulating (like SF6 but not really bad for the ozone layer), the other lighter (but still heavier than N2 and O2) gas or suspension of particles is a much better conductor / it breaks down under a much lower electric field. That way you send a shock through the gas, it goes through their torsos (or lungs), though their legs, and to ground.

>>1895591
Mythbusters did an episode on this.

>>1895591
You’d probably be just as likely to electrocute yourself as your target. Ever used a water gun? It’s kind of hard to avoid getting yourself wet.

>>1895591
>1 million volt
well, let's see. Air has a breakdown voltage of 3Kv/mm. 1 million / 3000 = 333 mm. Wow, almost a whole foot.

That means if it gets closer than a foot to the ground path, it will take that opportunity to arc rather than sparking at your target.
So if you sprayed two streams, they'd need to start more than a foot apart and stay a foot apart. Like how tasers fire two darts.
> the spark should jump across the gaps until it reaches the target, right?
No, it's jumps once. Stun guns fire multiple times to create multiple arcs. It doesn't keep arc'ing when its unpowered. Once it arcs, the electricity has gone to ground.
There's a lot of misconception in what you're asking, to the point where i'd stop thinking about making magic weapons and start figuring out how it actually works.

>>1895612
>they'd need to start more than a foot apart and stay a foot apart
You could reduce that though. Make the nozzles, say, two inches apart, and angled slightly away from eachother. As long as you can keep the drops less than that distance apart it wouldn't be an issue.
>Once it arcs, the electricity has gone to ground.
But the ground in this case is just a drop of water. Where does the electricity go from there?
>i'd stop thinking about making magic weapons and start figuring out how it actually works.
That's why I'm here. Never said I'd build anything, just find the idea intriguing.

>>1895619
How do you think electricity arcs?
> You could reduce that though.
If the charged and and ground EVER are closer than 333mm, it will arc. You cannot start small, you have to start big and end small. If you start less than a foot apart, it will arc immediately at the nozzle, like how stun guns work. If you send it out into the world, it could arc in really weird places where you don't want.
> But the ground in this case is just a drop of water.
You cannot just charge a blob of water to 1million volts.
First, It's 1 million volts in reference to something else. You have to give the 1 million volts a path to ground for there to be an arc. This is so you get it to arc where you do want.
Like other said, people will most likely be insulated from ground.
Let's put it this way: when you are shocked on a door knob, you are charge to something like 3kV. Does it bother you when you walk around? Nope, only when you get close enough to ground.
Also, at 1 million volts, it's very very little current. trying to "charge" an appreciable amount of water to 1 million volts won't work since water will dissipate the charge. The arc is created from a point like tip because it has a such a small surface area. It concentrates the charge. Water will disperse it.
Besides, it won't even be possible to move the water fast enough. Stun guns work on AC, at like 1kHz or something like that. Which means you couldn't move the water fast enough to actually change it.
A DC charge that high is too hard to create, look into tesla coils.
Like think of how short a time a millisecond is. To actually charge the water, you'd have to clear the water in a fraction of that time, at the peak of the wave, so much closer to a micro second. Otherwise the water wont charge.
Basically it fails at every stage.

honestly might as well use 5G waves to cook someone's brain and nuts
it'd work better than this water gun business

Does causing a spark always mean damage was done? Was probing my vintage stereo receiver to check bias, required me to insert long probes into a heatsink block to reach the emitter resistors. I powered down before moving the probe but didnt wait long for it to discharge. While unclipping the neg probe I caused a spark. Unit still powers on and runs but didnt do any further voltage readings yet. Thought i'd look around online to try and find out if spark always equals damage.
Pic related is pic I took with heatsink disassembled to better show the situation. The pincers sparked during unclipping as seen, maybe hit that lug or the screw, not sure. I have an imgur with more pics/schematic if it helps https://imgur.com/a/nt5weOi

>>1895666
>devil trips

That aside, as long as the unit still works and there's no visible physical damage there's shouldn't be an issue. As i remember accidentally shorting a small metal cased chink transistor once, it sparked when i accidentally shorted two legs and a pin on a 15VDC chip (one of these 3 legged chips that output 15VDC constant), and at the base of one of the transistor legs it was slightly black and burned.

Probably did short one of the legs on the screw.

>>1895666
>While unclipping the neg probe I caused a spark
That means a reasonable amount of current was flowing through that lead, or at least through the tip of the probe. If you were probing for current then that's perfectly normal, but if not then you shorted two adjacent contacts with the probe. Happens far too often, especially when probing directly on ICs.

Looks like the circuit is a pair of full-bridges (with all NPNs) and you'd have shorted either B+ to B-, or one driver output to B+ or B-. Probably something to B- as you were working with the low-side transistor. This could have damaged either the drive transistor Q2 or the power supply output (especially if it has a linear output stage), so I'd check both. It could also have damaged Q1. So just measure the voltage on the supply rails (AC as well to ensure the ripple is low) and the output voltage of the half-bridge in question, should be near halfway between the rails at idle, and similar to the other half-bridges.

>>1895688
Checked & thanked. Will be more careful in the future. Surely I must be a brainlet, figured japs would've had accessible test points on a flagship kenwood unit.
>mfw moments after checking your digit(z) I pick up service manual and read it's instructions for bias test

>>1895697
Thanks fren, was reading DC voltage for bias, spark occured after turning unit off, quite a few large capacitors mustve still had charge. Aye the unit (KR-9940) is "quasi complementaty" wih all NPN. Has linear power supply, will double check the voltqge outputs there for sure. Unit is partially disassembled, was replacing the bias pots. Will report back within a week or so when I hopefully have time. Screenshotted this answer so I can refer to it. Thanks again gentlemen.

Why do rc aircrafts use 2S lipo batteries instead of a single alkaline 9v battery?
Both are 9v right?
Is it because of the discharge rate?

>>1895700
>spark occured after turning unit off, quite a few large capacitors mustve still had charge
Ah, rookie mistake. Though usually that's in the context of not shocking yourself, as opposed to not shorting something to shit. Since it would just be a momentary short as the caps discharged, you're probably fine. The PSU especially.

Personally I like to use my oscilloscope for this kind of work, because A: it lets me see both DC level and ripple voltage (and frequency) quite easily, and the parrot-style clip works much better than an alligator clip for getting into narrow spaces and not causing shorts. Can probably get parrot-style clips for DMMs too. I clip my negative lead somewhere convenient and just focus on where the positive probe is, one less thing to worry about, though not as convenient for making a differential measurement like that bias reading. But that's a good way of making a bias reading measurable, as opposed to making some sort of measurement on the base-side of things.

>>1895701
Common RC lipos can discharge at 100C, so a 1000mAh battery could discharge at 100A, and that's continuously. Alkaline batteries of similar volume and voltage would be lucky to handle 10A, let alone a single dinky 9V battery. Plus alkaline is not rechargeable. The next best rechargeable tech, NiMH, has worse energy and power density, and worse specific energy and specific power, compared to lithium ion, and the price isn't that different either. For an aircraft, the specific power (power per unit mass) of lipo cells are just unbeatable. Though conceivably, lithium primary cells (like in 3V coin cells) could see some use when they don't need to be recharged.

>>1895701
A single alkaline 9v battery is made of 6 cells.

>>1895704
Actually built a "dummy load" and re-capped/service bulletin mods on a tektronix 2213A, been too much of a pussy to use it but may try it out. Believe I just clip each probe to points "B" and "D" and connect one ground lead from scope to chasis of receiver? Based on all ive read about grounding dangers and scope use I believe thats the proper way to view outputs when supplying a waveform through the the receiver's aux input? Looks like I can also set the bias using that method too based on the excerpt in the pic, maybe not as accurate since the scope hasn't been calibrated.

why use phototransistors over photodiodes or vice versa?
it's the same thing essentially

>>1890505
I'm just making shit up but it looks to me like they printed this onto some sort of plastic film and which then fused with the PCB in the reflow oven.

>>1895770
No, it was hand painted by a 6-year-old Uighur slave laborer.

>>1895554

dont use your bench supply unless you have a smart one that you can program to go from constant voltage to constant current at some given voltage. measure between + and - terminals, if it's around 3.7V you can use any old charger, like from a camera battery that puts out 4V. just connect the plus and minus pins and ignore all the rest. you can get these for a buck at the thrift store.

if it's 7.2V or higher, you need a more sophisticated Lipo charger with buttons and a screen to set things up. often these will detect the voltage and set themselves up accordingly.

I'm trying to upgrade a cheap LED flashlight to run on an 18650 cell and be dimmable.
The LED COB drops 2.6-3 volts when driving up to 1A, so I'm looking at an LM3405 buck-mode driver.
I'm surprised there aren't a huge selection of LED driver chips meant to run at li-ion cell voltage. Most want at least 5v.
Some of the fancier chips have analog dimming, but the 3405 needs a PWM signal, so I'm trying to figure out how to generate a good PWM from a potentiometer with a minimum of components. I really don't want to have to stick an attiny microcontroller in here.
I was hoping there would be an off-the-shelf tiny chip that would take an analog voltage and generate a PWM signal, but I didn't find anything.
Any suggestions?
And is the LM3405 really the best chip for this application? Datasheet says it should do the job, but I feel like I'm pushing its limits.

>>1892823
you dont want to use diodes on the input because you get voltage drops and diodes are notorious for temperature drifting so your input dignal gets distorted , also semiconductors ususally need protection only on the output because they are sensitive to reverse voltage bbreakdowns and are harder to destroy when used while properly polarised...and on the output is where you have inductors and capacitors connected to usually which cause transients

>>1890464
>but what avoid VL to get massive? di/dt become near infinite...
small capacitor after the switch

>>1895383
it looks like your joints are cold and board is burnt ...that would indicate your temp is too low and you dwell too long on the joint ...you might have bad solder though

set your soldering iron to about 330(if you can) clean the part of the board you want to solder with a brush and alchocol , wet the tip of the iron with solder and clean it off the tip should be nice and shiny .....add flux (never too much flux) and put the solder wire FIRST then the iron on top ( solder is your thermal transfer medium) and when you see it starts melting then add a bit more....soldering things like resistors, diodes and other small parts usually lasts around 2-3 seconds max if everything is correct

things that will change the look of a solder joint is if there is oxidation on the solder mask , different types of solder , type of flux , temperature of the tips and the thermal conduction

/ohm/, I have a stupid question.
Can you "over-amp" an electronic component? I have a 4.5-12 volt, 3 watt LED bulb that I want to use in place of a 6 volt, 25 watt incandescent bulb, but I'm not sure if the extra amps would fry it.

>>1895843
>Can you "over-amp" an electronic component?
yes but not the way you think. your bulbs will draw whatever current they like as long as the voltage is right. so your 3W bulb will work off the 6V supply, it just won't draw the same current as your 25W bulb.

also 4.5-12V is kind of a wide range for a LED module, that seems weird. i might be thinking of a different sort of module than you're using though.

>>1895843
if you are talking about a general light socket that has no controller or constant current source then no , electronics does not work that way ...the incadesent lamp drains 25watts because it made like that by design ....3 watt led lbulb will drain only 3 watts because it only needs 3 watts to work (if its a led bulb it has a controller circuit inside the neck of the bulb that converts the alternating current from the wall into direct current the diode uses)

>>1895853
>>1895849
Thank you!
>also 4.5-12V is kind of a wide range for a LED module, that seems weird. i might be thinking of a different sort of module than you're using though.
It's a no-name maglite LED conversion bulb.

>>1895789
>that would indicate your temp is too low
I have the iron at the highest setting. It's this https://www.amazon.com/gp/product/B000AS28UC/ref=ppx_od_dt_b_asin_title_s00?ie=UTF8&psc=1
>put the solder wire FIRST then the iron on top
Really? Every guide online says to heat up the spot with the iron, then touch the solder without making contact with the iron

>>1895888
how funky is your iron's tip?

>>1895890
What do you mean by funky? It's a screwdriver shape head like in the link

>>1895891
as in, is it clean and tinned properly? or is it covered in corrosion?

>>1892296
N is the noise factor of the amplfier, when cascading amplifiers and filters etc together in a Transceiver for example, you would use an equation such as this to get the overall system noise. See pic related from RF Circuit Design Theory and Application - 2nd Edition - Ludwig & Bogdanov

>>1892294
>>1895897
I posted a thumbnail sorry

>>1895892
I clean it on the sponge and tin it before using it but that's it

>>1895888
>Really? Every guide online
i know thats why i made it a point of it , for through hole components your tip touches a bit of pcb and a bit of wire which is not thermally joined (there is air around the hole) which means your component and pcb will heat at different rates since they are not thermally coupled ...you use that solder in just a dab to couple the pieces and when you see the small piece of solder moving that means you got it up to temperature and then add a bit more solder to fill the joint (at the connection not the tip as you said)
with smd parts its a bit different since the components are so small they heat up quickly enough and at smd soldering you focus on heating the PCB NOT the component

>i have been soldering professionally for 15 years

>>1895888
also that soldering station is kinda shitty ...chink soldering stations are actually quite decent and have better temp controll and can be bought with soldering and desoldering tips (the hot sucking pump thing) whic for newbies can be a lifesaver

>>1895902
that looks ok ...also how are the tips changed ? is it just the tip part that joins with the shaft or do you change the whole insert? ....i ha d problems with weller solder irons because they had just the tips exchangable and made contact with the heater with a bit of flat part which would corrode and would need to be polished every now and then to make a good thermal contact

>>1895782
I'm talking about protection diodes; parallel diodes to the rails, not series diodes. The kind of diodes that all MCUs I've ever used have on their I/Os. Saw some guy feeding a 12V signal directly into the MCU and it kept on working. It got really hot, but it kept on working.

>>1895934
most mcu-s have protection circuits builtin including pullup and pulldown resistors ...it has been like that for a while already and most modern mcu-s are quite hard to kill accidentally they dont show that in the schematics because its baked into the crystal ...just like almost all modern mosfets come with flyback protection diodes out of the box

>>1895973
>just like almost all modern mosfets come with flyback protection diodes out of the box
No that's a result of the construction of the MOSFET itself. It's impossible to make an enhancement-mode MOSFET without such a diode, by connecting the body to the source, you create a diode from source to drain. And the pullup/pulldown resistors aren't really for protection. They might help a little against ESD, but the diodes will do most real work. I think 7400 ICs have a diode to ground on their input, not sure about a diode to Vcc though.

have you seen something more pajeet than this?

https://www.youtube.com/watch?v=WUjxtrK1xTM

>>1896012
>drops glob of liquid solder on chip
i wouldn't want to do that even if the chip was ceramic
>uses screwdriver to push liquid solder off smd pads
jesus fuck do i have to watch any more of this?

>tfw i'm the guy that fucked up his apple ii disk drive board desoldering shit and even i know better

>>1896024
oh wait that's mercury holy shit

what do you want to bet that's his mom's toothbrush?

>>1896012
interestingly this was patented at one point: https://patents.google.com/patent/US3187423A/en

>>1896012
Looks like poor`s man chipquik

>>1896026
>mercury
It's indium (Chipquik).

>>1896028
yes but the alloy in question was premade with about 60% mercury alloyed with tin and lead which is also very questionable and is nowhere said its safe ... not some pajeet burning mercury in flux and then fingering it around like its still not highly toxic

>>1896031
it isnt ...in both of his videos he claims its mercury

Honestly mercury OUGHT to work. Ever seen NileRed's video on doing mercury vs aluminum? I'm reasonably sure it'll amalgamate with tin as well.

Elemental mercury's toxicity is somewhat overblown. Pure mercury handled properly (yes, he's not handling it properly and I have doubts as to its purity) it's not that bad. It's bad for the environment and it accumulates in wildlife, and if you start inhaling the fumes it's really bad (he's certainly not under a fume hood), but generally it's the organic compounds of mercury that you've really got to be careful of. Shit, elemental mercury used to be available as a laxative.

The two bigger problems are (1) proper cleanup is a bitch, and (2) mercury isn't fucking cheap. Considering 100 grams of lab-grade (this is about 1.5 tsp) is upwards of $40, it's not really a suitable substitute for chipquik.

>>1896043
Pretty sure old stock dental Mercury is pure enough to be handled, that’s where Cody got his from.
I bet indium prices went up due to ITER using it in their superconducting magnets. I wonder if there’s any of that indium-tin superconductor material being salvaged from old equipment like MRIs? Be pretty interesting to inherit the SnIn of ones father.

According to Wikipedia:
>Most recent 7400 series parts are fabricated in CMOS or BiCMOS technology rather than TTL.
Are there any serious implications to this? Or put differently, what implications could you foresee coming from this? I figure that since they're reusing the old part numbers for currently-manufactured parts they're drop-in replacements for TTL 7400-series chips.

>>1896096
They’re still labelled as 74HC___ as opposed to 74LS___ or whatever, so it’s not like you can’t tell the difference. They should be drop-in replacements for any conventional logical circuit, only when you try to drive them with unconventional voltage levels or high output impedances, or try to drive high currents, will you notice a difference. For the better in most cases, since CMOS outputs should be push-pull and require very little input current, but they are more vulnerable to ESD. There’s probably a comparison out there about different 7400 logic families.

>>1896096
The only real difference is that the input threshold voltages aren't all the same.
For instance: 74HCxx isn't compatible with TTL families (like 74LSxx), but there's 74HCTxx which is a special version of 74HCxx which has TTL compatible inputs.

>>1896096
>>1896103
https://electronicsclub.info/74series.htm
Here’s a good article.

Time for some light reading.

>>1896043
The real problem in the video is him heating the mercury with his iron, imagine how much vapor its shitting out.

>>1896012
>look up boiling point of mercury
>356°C
>mfw he's boiling mercury with a nail without any fume extraction

So I was just thinking... this oddball "disk amplifier" IC that isn't made anymore, what are the odds it's just a "single-chip" form of something people could just make out of a few other components from the manufacturer's catalog? Even this diagram is separated into three discrete "stages".

...and if so, how would one go about finding what those components are?

>>1896163
There's lots of matched transistor pairs, so that could be a bit difficult, but you could probably do it.

>>1896182
Well what I'm thinking is that when this was designed it was probably a rough combination of a few preexisting IC somewhat general-purpose designs, perhaps with some additional parts. Like the preamp/gain section is described as being a differential amplifier in the documentation. Surely Motorola didn't design each of those blocks from the ground up when putting this chip out to market, except perhaps the digital logic section.

Of course knowing 70s/80s Apple, if they could've more cheaply done it in-house they would have. Earlier revisions of the same board use this one's sister chip for controlling writes, but at some point they dropped in a different IC with a bunch of other discrete stuff.

>>1896163
I’m not familiar with differential output amplifiers, but I’m assuming they’re similar to single-output op-amps. So I’d expect it to be quite feasible to recreate the circuit in such a manner. Not going down to the individual transistor level unless you really need to. Could possibly use two normal op amps in place of a diff-amp, or just ignore the differential signalling and use op-amps for everything, shouldn’t make much difference.

>>1896163
What is that logic stage supposed to do? I get that the rest is just a zero-crossing detector after a differentiator to make a peak detector, but then it’s g a bunch of matches and such. I’m assuming it’s just for ensuring the data comes through without any noise, but surely a low-pass and a Schmitt trigger would do that well enough. So maybe it’s also decoding Manchester or some other simple encoding scheme used on the floppy?

>>1896235
Yeah, I think that's for decoding the data.
Pins 6/7,8/9 connect to an external RC timing network, and the D-flip-flop is a part of that feedback loop to the timing capacitor.
I don't feel like mapping out that logic right now, but I have a hunch that it's a self synchronizing clock circuit.

>>1896244
Well it can’t be decoding it completely, because magnetic storage like this looks to use a somewhat complex encoding, both with a final bit mapping to ensure no more than 2 consecutive 0s, and a prior polynomial error correction code + parity check. I think. But it also states that they avoided consecutive 0s because a 0 meant no state change, which makes it something similar to Manchester encoding, which this circuit could well be decoding.
So I think 10011010 would be turned into (0)11101100, 11111111 to (0)10101010, and 00000000 to (0)00000000. Don’t click on those phone numbers. This is probably relatively simple to decode if you have a reasonable estimate as to the clock speed, though the end result could be inverted. This has the advantage over Manchester that it doesn’t require twice as many transitions per second, which matters when you’re cramming as much into an inch of magnetic tape as possible and your tape can only have a definite transition after a certain distance.

>>1896235
It's explained in Motorola application note AN-917. Here's the description:
>After the input signal is amplified, and filtered, it is then differentiated (to change the peaks to zero-crossings) and applied to the comparator. The comparator output toggles each time its inputs sense a zero crossing. The comparator output is then applied to the time-domain filter, which has a dual purpose. The primary purpose is to generate an output pulse (at Pin 10) for each transition of the comparator output. These pulses correspond to valid data read off the disk. The second purpose is to filter out false zero crossings which can be caused by noise during the "shouldering" period as shown in the second trace of Figure 22. Shouldering is a result of media characteristics and recording format, and can be aggravated by media defect and contamination. For example, formats which permit excessive distances between bits can generate shouldering situations, by allowing the read signal to droop and possibly create a false zero crossing. Filtering out false information involves the use of an "ignore window", during which time any zero crossings are ignored.
I understand this amplifier had a variety of noise issues that were actually exploited for copy protection purposes.

>>1896249
>figures

>>1896249
>this amplifier had a variety of noise issues that were actually exploited for copy protection purposes.
Sounds rough.

>>1896253
I think I understand it. I wonder if there's some documentation somewhere on how to design discrete digital filters like that? I'd like to see how to convert a circuit into an equation and vice versa.

>>1896268
>I think I understand it.
Wish I did, hah. I should've taken more EE coursework in college.
>I wonder if there's some documentation somewhere on how to design discrete digital filters like that? I'd like to see how to convert a circuit into an equation and vice versa.
https://en.wikipedia.org/wiki/Filter_design#References might be a good starting point. Particularly Andreas Antoniou, Digital Filters: Analysis, Design, and Signal Processing Applications. There's an epub of it on libgen.

>>1890135
I use a stickvise and actually highly recommend it for small board. For big board, just prop up with whatever random shit is on my desk (harder for SM stuff, would recommend a larger stick vise probably). Plus, they're easy to make.

Any advice on how likely my solution here is to fail? I have a PAPR blower that you wear on your waist, like what Silver Surfer Pajeet in vid above needed. I wanted to run it on a usb power bank and opted to make "permanent" adapter instead of splice. I bought quality parts online (qualtek q4-5x adhesive lined/fluid res polyolefin shrink, te connect barrel connector, qualtek usb a 20awg)
The only part of the job I felt like a nigger about is not crimping the crimp style connectors. I could only find 20AWG USB A cable which was too small to crimp so I:
>tinned wire, put slight hook in it
>fed it through the connector as pictured and shaped it to travel underside of connector represented by dotted line
>Pinched it tight then soldered both sides
>1-2 layers of heatshrink for rigidity
Is this going to fail while im walking around wearing the blower? Should I re-do differently? Am I overthinking this and encroaching autistic levels?

oh shit he said the gamer word

>>1896301
I only took one digital logic paper and a DSP paper, neither of which really mean anything here.

The pulse generator turns the square wave into a series of small rising edges, one for each transition of the comparator (both rising and falling), basically an edge-triggering circuit like you'd put at the input of a gated latch to make a flip-flop. The One-Shot blocks are basically SR-latch-based monostable circuits that turn off with a rising edge and turn back on again a small time later. The pulse generator is used such that the SR latch is able to reset itself immediately, instead of having to wait for the comparator output to go back low again.
By using the output of T1 to clock a D-FF, the circuit rejects any cases where the comparator outputs a signal that lasts shorter than t1, be it a high signal or a low signal. The second One-Shot monostable T2 is used to form the resultant pulses into pulses with fixed widths.

I'll check out those links thanks.

>>1896323
What causes crimps to be stronger than solder joints is that they usually also grip onto the insulation, making a form of strain-relief. If your heat-shrink is more rigid than the insulation it goes over then you'll be fine. Just be wary that the hot glue inside those glue-lined heat-shrink tubes has a bit of a tendency to lose stiction after a life of frequent motion.

I have a test coming up in a week for internship (final sem in bachelors electronics) at Texas instruments. The description simply says there'll be questions on digital, analog and aptitude. Can someone please tell me what topics i should focus on?
inb4
>digital, analog and aptitude

>taking a test to get a job
reason number 2 why i'm glad i went into law instead of tech

>>1889427
I have an LCD screen unit that is connected to its control PCB through a flat ribbon cable like pic related. Obviously, it's fine provided the screen and the PCB are parallel, but my design requires me to "torsion" the cable since the screen is perpendicular to the board. Is this safe? If it's not, is there some kind of adapter that would allow me this kind of flexibility?

>>1896374
It will be fine so long as there's enough room for a nice long twist, and some strain relief on either end to stop the torsional force from torquing the connectors on each end. If you're 3D printing stuff or otherwise making your own enclosure, would be pretty trivial to make little brackets to hold the cables in place.

Hey /ohm/bres, I posted two weeks ago in in the last thread asking for advice on fixing my surgical headlight (>>1884285), and I fucking fixed it!!

Inside the cord there were two small wires covered in enamel, and insulation that I burned away with a lighter. I soldered the wires back together in a similar fashion of fixing earphones. It didn't work the first try, so I desoldered everything and tried again, and it turned on the second time.

Feels fucking great. It's all put back together now, but here's how it looked before I covered the wires with electric tape and shrink tubing. Cheers lads :-))))

>>1896425
Nice. Heatshrink should make a good and tight seal if you need to.

hey guys, I want to make an usb monitoring system form my dumb ups. So far I've designed the signal sensing and conditioning for voltage (left) and current (right), please let me know I there are mistakes. As mcu I want to use an arduino, is it worth to place an atmega and its ancilliaries on the pcb or should I just place a pro micro clone and call it a day?

hi /ohm/, I'm kinda retarded.

I have this 12V transformer with center tap. I want to get both 6V and 12V out of it, both should be independent (floating) but I also want the ability to tie both to a single ground.
I think it's possible with half-wave rectification, but that's not what I want. if I do full-wave and tie both negative ends together, I also get 12V from the 6V side.
how do I fix this?

>>1896472
>is it worth to place an atmega and its ancilliaries on the pcb or should I just place a pro micro clone and call it a day?
Use a dev board if it's just a one-off. Otherwise, design a PCB and use the raw chips.

Just rectify the 12v tap and buck it to 6v.

>>1896497
>buck
linear regulator (LDO)

>>1896497
>>1896498
oh yea I forgot to mention. I do need linear regulators and the 6V end needs to source about 3A. so if I did that then the regulator would have to burn away 30+ watt.

NEW THREAD >>1896523
>>1896523 NEW THREAD
NEW THREAD >>1896523

>>1896501
30W isn't bad once you've resigned yourself to using a large heatsink, or a medium heatsink and a small fan. iirc 3A+ regulators are kind of pricey though so you may prefer to go discrete. NPNs that can handle 30W are everywhere.

>>1895916
I took your advice and practiced on some spare resistors. I took this short video, would you mind telling me if this looks correct?

>>1895916
Sorry only half the video here >>1896651 got converted. This is correct

>>1896668
>>1896526

>>1896668
Put the iron down first and feed solder to the joint. You're using larger gauge solder so it's more difficult to get the right amount of solder on each joint. Keep practicing, you're improving.

Can anyone ID what the fuck type of socket this is? It's DC, in the 12v leisure battery circuit (narrowboat) and is located in the kitchen. I've been living on boats for fifteen years and I've seen some pretty creative thrown-together electrics but I've never seen a socket like this on anything before.

>>1897280
ask in the new thread, this one is already at page 10:
>>1896523