/diy/ - Do It Yourself

>>2662600
>or
>and
>not
>not
did i get it right?

>>2662621
They call it SUSTAIN because guitar players don't care about how the control actually works beyond GIMME MOAR DISTORTIONZ and also they didn't want to call it DISTORTION,GAIN or DRIVE — which are more typical labels — for some reason. Might have something to do with how the unit was intended to be used, idk.

>>2662600
Nice OP. Some might complain about not waiting for page 10, but don't worry about their autism.

>>2662455
>I'm not sure I understand the point you're making
The totem pole is inverting, not non-inverting using emitter followers like they usually are. It's powered directly by the +5V source. The BE junctions of Q9 and Q8 behave like diodes, so you basically have two series diodes shorting from 5V to 0V. And this happens regardless of the input voltage, though when it's within 0.7V of a rail at least it's only shorting through one BJT. It's like shoot-through but worse. Pic related.

Replace it with a common-emitter inverting stage or a totem-pole buffer stage, but not both.

>>2662541
Those are just two different file formats of the same files I suspect. The lower ones have gbr extensions, while the upper ones have a selection of extensions that presumably denote the layer in question. gtp = gerber top paste.

>>2662604
>The problem is that this adds more treble even at high distortion settings
Post the schematic of the part that uses potentiometers. If it just uses single-gang pots, it should be possible to replace a single-gang pot with a dual-gang, and use the extra wiper to adjust the tone control as a function of gain (or distortion).

>>2662639
YWNBAW

No go cry to the jannies, faggot. Hope you're pinned between two vehicles and burned alive.

Look what i found in a trash pile
Should i toss or keep?

>>2662753
Let's see: shitty pic that nonetheless took 3 megs and is 4k x 3k, and you are asking if it has any value or use, so I'd say that for you, it has no use and you should toss it.

>>2662753
>toss or keep?

you can never have enough home-made variable power supplies.
bonus if it's safely made, and is regulated.
if you have no use for it, thousands of Ebayers would be willing to take it off your hands for $1, shipping included.

>>2662753
Keep it and electrocute yourself.

>>2662753
Did you try plugging it into your ass?

i'm back at home now, here better pic
the thing is i already have digital power supply so i'm wondering if it's worth to keep this analog one where you have to set the voltage and current by looking at the little hands like some caveman

>>2662790

here are some nudes from the PSUchan

>>2662790
Seems like some power supply which has been made during the training of electronics technician. My father has also been create one of those but honestly those are not worth keeping them. They can't adujusted precise enough for work that is in electronics. More qualified for a short time alternative for a device where you do not have no more the power supply and have to buy a new one.

>>2662792
Keep the toroidal transformer and heatsink.

>>2662799
since it's using a tranny and not a switching design would that mean there will be no ripple or noise at the output?

>>2662800
No, that's a voltage regulator.

>>2662800
The big electrolytic cap is a smoothing cap, the two tantalum caps are filter caps, and the small electrolytic is the output cap.

>>2662790
>wondering if it's worth to keep this analog one

really naive question.
you need two supplies when you have 2 diff loads you wanna vary, which is often if you do a lot of experimenting.
you need two supplies when you need more voltage, like 48V to light LEDs strips from TV or monitors.
you need two supplies when your single one cant deliver enough current.
you need a clean analog supply when dealing with RF or very high gain amps.

>>2662790
Yeah, keep it.
You can always slap one (or a few) DC/DC converters on it if you want something more precise. Good for testing things like audio amplifiers, motor drivers, electroplating, battery charging, etc, etc, ad nauseam.
Won’t have any high frequency bullshit in it for audio (e.g. powering a monoblock)
Picrel is what that autist mathias wendall has been using for for 35 years or so. If it’s good enough for him, it’s certainly good enough for you.
Looks like those are audio meters with a log response.

>>2662792
LM2596?

>>2662604
My solution: Instead of the usual SUSTAIN control that shunts part of the first gain stage's output to ground, there is a control that allows you to choose between full output and a treble boost that compensates for the high end at minimum SUSTAIN setting.

>>2662600
So I have a question: I am considering making my own Welding Shop and mainly I want a small Vulcan Brand Stick and TIG Welder known as a ProTIG 165 and then eventually once I save up a Syncrowave 351 from Miller. How would I go about calculating the power requirements for both of my machines for my ideal Welding Shop? Do note that I want to be able to power both simultaneously and ideally use a reasonable amount of power to do such.

>>2662875
What is that anon? This is my first time on here - so yeah. Just curious.

>>2662885
>What is that anon? This is my first time on here
It's the secret plans for a bitcoin mining rig, each one capable of making 4-7 dollars a day.

>>2662902
C'mon now anon. Are you being real?

>>2662902
back in 2017 i was into eth mining, built a multi-gpu rig and all that. i'd make 6$ a day after electricity costs. i used to sing "another day, another 6 dollars" to the tune of another day another dollar, and it drove my girlfriend fucking insane

>>2662929
Based. Least you're making money. She should love the fact she isn't dating a useless slob who isn't making any money.

>>2662792
>that 5-pin TO-220
Hmm, wonder what chip that is? Since there's no op-amps or other active components, maybe it's doing both voltage AND current regulation?

>>2662600
OK I need some help figuring out how to get this to work the right side of the transformer makes high voltage positive and negative.

The left side has two npn and 1 pnp, yet I tried building it in the simulator and it's working. Diagram says. 24v on left. Clues on photo. I'll post schematic next. I think it's an ossilator? The power inputs which are not shown, it says 24v, I'm assuming DC, any feedback or clues, thank you.

>>2662987
>yet I tried building it in the simulator and it's workin

yet I tried building it in the simulator and it's working

yet I tried building it in the simulator and it is NOT working

>>2662987
Circuit is weird. Either i am dumb, or it is wrong.

>>2662987
can you post video link?

>>2663018
>can you post video link?

Here it is and time stamp. Don't laugh, this is the power supply for anti gravity lifter.

The right side is positive and negative that powers the craft.

https://youtu.be/2JrkOfqm6eY?t=1509

There is also a slyer exciter lower left, I have that working.

Upper right is a sonic wave generator and I don't think that is needed to lift. So I'm ignoring that for now.

>>2662991
https://everycircuit.com/ is the simulation software. I'm not sure if there are better ones, I like how it shows things working.

>>2663013
You helped me already, I didn't know that's the symbol for potentiometer.

>>2663026

>>2662600
>>antigravity and/or overunity?
>Go away

He is driving a CRT TV flyback trasnfromer with it. That thing produces probably 20kV DC with rectifier/voltage multiplier, with pretty significant current. I suggest not to dick around with it, since you don't understand basics.

And yes, I think circuit he provided wont work.

>>2663029
>And yes, I think circuit he provided wont work.
But its already working. Did you. see it flying?
If I die, I die.

I know the dangers of highvoltage and Iknow many of the techniques not to get zapped.

I'm already getting progress. I didn't know about darlington or the Joule thief. Thank you. all for this help.

TO the pessimistic, with. your attitude. nothing is possible, with the right attitude. anything is. possible.

>>2663029
>CRT TV flyback trasnfromer with it.
I had a guess he might be doing that. Thanks for confirming. And thank you. for the danger warning.

I'm looking for an RC oscillator circuit (no inductors) with the following specs:
1. that is fully tunable with the RC component
2. Only 1 capacitor (not 2 like the astable vibrator)
3. preferably only nmos or npn transistors, although I can also settle for pnp transistor but no PMOS.
4. no IC
5. Outputs either a sine or square wave, doesn't matter, if you got a circuit for both even better. Triangle or sawtooth is also fine.
6. Preferably not too complex, should be similar in size to the astable multivibrator, maybe a few extra resistors or transistors here and there, but not more than 1 capacitor that affects the freq/duty cycle


I decided to try to design my own, but i couldn't get the simulation to work, but I do know that for some reason a lot of simulators don't end up showing oscillating circuits (for instance, the astable MV doesn't oscillate on ltspice or kicad) and I was too lazy to try to build it myself for now. I was thinking something along the line of a schmitt inverter oscillator with RC feedback. I wanted to build a schmitt trigger out of transistors but I couldn't figure out how to do an inverting schmitt trigger iwthout adding way more transistors than I wanted. If you think you have something like that post it, worst case I won't use it, don't be shy.

>>2663071
This is the circuit I personally came up, can anyone confirm if this oscillates? According to the circuit simulation, it somtime does sometime doesn't depending on the value of C4, so I assume it does and the simulation parameters are just not unbalanced enough to start the oscillation? In the simulation I used ideal resistor/caps and an nmos with vth of 1v and remaining parameters as default ones for the nmos on kicad.

>>2663028
I never realized how stupid the iec symbol was for a potentiometer was. Should have stomped that iec codswallop out years ago, making all those scam artists on the standard cry themselves to sleep every night.

>>2663142
Resistors are also blocks in IEC. It's just getting used to, but it's pretty annoying that everyone just have their own standard, including teachers at school.

>>2663149
>>2663142
it gets worse, our prof in uni can't decide what he wants and he uses both in the same schematic.

IEC is superior though, much easier to draw by hand than ANSI resistor which is always crooked.

Don't know where else to post,
but I tried replacing my Nintendo Switch battery and I think I somehow fucked up the battery lead part, pic related
My question is, how fixable is this?
Can I solder or something? Or do I have to completely replace the Switch's board?
I will post a picture of what it's supposed to look like for reference

>>2663092
Looks like a variation on the standard multivibrator.
Sometimes the need a “kick” to get started, which is why the standard circuits have the extra components sometimes.

>>2663152
>My question is, how fixable is this?
Fixable
>Can I solder or something?
If you have to ask - nope. You can't. Not enough skill, plus you've fucked up a connector which is a monkey job.
Professional can.
>Or do I have to completely replace the Switch's board?
Not necessary. Depends what would be cheaper for you.

>>2663154
I beg for a spoon feed. How much roughly would it cost for a professional to fix this problem? I doubt more than 100 dollars.
Where would I even go, and yeah I'm noob I don't know shit about electronics honestly.
Unrelated but the Switch is hacked, so I don't know if replacing the board will also "unhack" it.

>>2663156
>How much roughly would it cost for a professional to fix this problem?
I don't know, depends where you live.
> I doubt more than 100 dollars.
Probably
>Where would I even go
Phone repair store, idk, electronics repair. Just make sure they do component lever repair or at least know how to solder ports (if you see soldering station with some sort of microscope and a dude in the corner - yes this is a place to go).
Usually they are small businesses, not chains.
Well, you can try soldering stuff youself, but there is a big chance to ruin it further, so I think it is better to leave it to somebody who knows what they are doing.
>Unrelated but the Switch is hacked, so I don't know if replacing the board will also "unhack" it.
Prob it would

>>2663158
Thank you so much anon, I appreciate you.
I guess I will do a cost analysis, but I know what to do now.

I have an incredibly stupid idea:
I want to build a closed circuit radio station for my HiFi.
Since I don't use the short and medium band on my tuner, I figured i could use it to pipe in my own, curated, radio broadcast in a more interesting way than hooking up my phone to an aux jack.
I think I will start with one station, since the audio streaming seems to be a bigger hurdle than the AM conversion.
Is there any project or product I could take a look at to help me with this?

>>2663168
So you want to make an FM transmitter?

>>2663174
yep

>>2662875
Already tested it (modified a bit) and it sounds great. At the lowest gain setting I get a treble boosted sound. At the highest gain setting the response is practically identical to the base circuit's. Turning the pot just a little bit toward the un boosted path and the treble boost quickly goes away. That's fine, I think. Success!

>>2662885
This is a modified section of a well known guitar distortion circuit. It adds a treble boost when distortion is at low setting.

>>2663153
It is indeed very similar to the standard MV, but this one seems to be unable to oscillate, the transistors end up in the active region. Perhaps it's just because the transistor is a 2n7000 running on 5V and a higher voltage or better nmos could make it oscillate?

>>2663152
I'd need a better picture close up of the board, but from the looks of it, you ripped off a pad. Definitely fixable, quite easily if you have the right tools, but if you're not experienced with SMD soldering, I'd avoid doing it myself and let a professional do it since you might damage it even further and then potentially make it unfixable. Like the other guy said, go to a component level phone repair store, and also as he said, it's usually small stores, NOT big chains like the geek squad at best buy (if that even still exists). It's a simple re soldering of the connector and digging into the solder mask to expose a copper trace that can be soldered on for the lifted pad. A phone store could fix this easily because they won't have to diagnose the board (which would require a circuit schematics). The only potential issue I can see is if your connector is also destroyed, in that case you might have to solder the wires directly to the board or buy a new connector (if available). Those connectors are actually noob magnet trap if you're not familiar on how to disconnect them, they are not as intuitive as other connectors. My best advice for now is don't fuck with it any more. As for price, personally I'd say no more than 20$ if I had to do it (i'm not a repair guy) without feeling like i'm stiffing the guy but since they are a business they might charge more. IMO if they charge more than 50$, they are pushing It since it looks like a 10 minute job if they don't have to disassemble the entire thing. If they do have to disassembly a lot of stuff to remove the board, it can make the repair harder which will probably increase the price. If they ask over 100$ find another store since this is something that even a relatively new person with a bit of soldering experience can do on it's own. If you're interested in seeing this type of stuff getting fixed, you can look up on yt component level phone repair videos, a popular one that I used to watch is louis rossman.

What is a good resource to learn power electronics with?
Specifically I want to learn more about IGBT converters and the inverter side of 6 pulse SCR circuits. Maybe even build my own on the bench, or even just in ltspice.

>>2663071
Try this one. It's even easier if you use a reverse-biased BJT oscillator, but those are a hack.

>>2663168
There are some common ICs for FM like the BA1404, datasheet and appnotes will help. I wouldn't do AM since you don't get a stereo signal.
>closed circuit
Do you mean to hook your modulator directly up to the antenna input socket of your unit, instead of transmitting a wired signal? If so, I'd go FM but with a splitter so you can still connect your normal antenna for the rest of the signals you get. If your receiver can handle digital radio I'd look into that. Though I'm not sure if the quality would be better than a direct analogue input, considering the finite bitrate.

>>2663198
Look for datasheets and application notes by the people who make IGBTs and/or their drivers. IIRC there are some good guys on youtube specifically for SMPS design, some of which should carry over.

>>2663223
It's a mid 70s Tuner, so no luck on DAB.
And yes, by closed circuit I mean coax-cable to coax-cable, no over-air transmission in between.
On FM you're correct, I thought MW had AM standard but I was mistaken.
I'll look into the IC you mentioned, that should be a good starting point, thanks!

>>2663237
>I'll look into the IC you mentioned

get one of these instead.
you often find 'em at the thrift store for $2, or from a friend.
coz a lot of people have moved to bluetooth versions.

>>2663223
> try this one
This looks like a ring oscillator.

If I had a 12v DC to 230v AC converter with a transformer at the end making 230VAC out of 12VAC how would the AC voltage on the secondary side react if the load changes?
I believe the voltage would go down as the current increases
How would you have to compensate for this on the primary side? Increase the current by using a DCDC regulator to control the 12VDC input? ?
What's the relation here?

I plan to send power in the range of 500 kW in AWG 7 wire for a few tens of ms.
If I consider 3 ft to be 2 milliohm it will receive 500W of power and thus receive 10 J of heat.
Am I right ? Will I fry my wires ?

What are the most practical variable capacitor range for experiments with radio receivers? The vintage ones could be very expensive, but there are cheaper ones in the <170pf range so I was wondering if that's enough range for any kind of radio receivers.

>>2663531
>if that's enough range

of course.
frequency is determined by an LC circuit, so you can half the cap if you double the coil.
365pF just became the standard by happenstance.
and the attitude ''if it works, why change it?''

>>2663337
>How would you have to compensate for this on the primary side?

you dont compensate anything.
you simply buy a unit that's rated for the load you have.
plus 15% headroom.
it'll do all the regulation/compensation required.

>>2663531
> not enough
We used to just put a 10 position rotary switch with caps soldered on it for adjusting the bandwidth.
Almost like a “range” switch but it’s not multiplicatative.

>>2663531
I've used 12-180pF varicaps fine, they were about $2 each
>enough range for any kind of radio receivers.
pretty much anything conventionally used, yeah. get a plastic screwdriver.

>>2662880
You can calculate the amperage by using Ohm's Law tracing backwards from the tip to the wall, but that data should be on the data plate on the back of the unit - maybe even on the retail box.

>>2663589
I guess with varactors and a rotary switch and a DAC one could programmatically achieve pretty linear tuning. Or maybe not since high capacitance varactors may require a relatively high bias voltage, certainly higher than 3-5v. But there must be a way. Maybe a purely analog approach and use a log pot to vary the bias voltage.

>>2663609
i meant to say encoder, not a switch

>>2663331
Yes, and? Just one transistor inverter doesn't have enough instability to oscillate with less than 90° of phase shift from the RC circuit, while two transistors cancel out one another's inversions so they'd be bistable. 3 transistors is the minimum for a conventional single-RC oscillator. Though as I said, you can use a single transistor if you use a reverse biased BJT.

>>2663223
>It's even easier if you use a reverse-biased BJT oscillator
I know about that one, but it requires a fairly high voltage to break down the junction, so won't work on 5V. Thanks for your circuit, funny thing is... I came up with this exact circuit and tried to simulate (pic rel). This showed 0 oscillation, so I assume this is a simulation issue? I'm not the best with spice, i know how to design circuits and simulate them, but not familiar at all with the different spice commands/directives. Maybe the circuit didn't oscillate because of ideal components, not giving enough electrical noise to start the oscillation? How would I go and simulate this circuit to oscillate? should I just add .trans 0 100m startup in the command? Or is there other things i need to modify in the simulation. Thanks again.

>>2663727
Add "startup" to your spice command, or otherwise perturb the circuit initially with a single-pulse voltage source. It may also be because the spice model of your FET won't turn on enough at 5V, specifically go for a model of a small-signal transistor if you're not already. Also I'd use higher impedances for your filter.

>>2663764
>Also I'd use higher impedance for your filter.
oh yeah, forget about the values of R and Cs, I just redrew the circuit from scratch quickly, didn't tune the values. The model for the FET i used is the generic one for Kicad with Vth changed from 0V to 1V, so I assume it should be okay for 5V right? As for startup, what if I change the 5V source to a 5V source with a rise time of a few nS, so the source initially starts at 0V in the simulation and after a few nS, it increases to 5V (linearly), would that do it? I just tried to simulate it and I got the following. It looks like it oscillate very little, then stops completely (used R = 1k, and Rf = 1k). If I change Rf to something higher or lower, it stops oscillating completely.

>>2663774
my bad, forgot to post pic.

>>2663651
I always end up making a one transistor oscillator whenever I try and make an amplifier. Sometimes it only oscillates when I’m testing it, or when my hand is near it, but I don’t know the schematic symbol for a hand placed near it. It’s a HCO.

>>2663774
>Vth changed from 0V to 1V, so I assume it should be okay for 5V right?
I'd assume so, but I've never used KiCAD's sim so maybe just idiot check that with a simple ramp on such a FET.
>what if I change the 5V source to a 5V source with a rise time of a few nS
That should work, since it's basically what "startup" does. In my case, it goes from 0V to 5V in 20µs.

The fact that it stops oscillating means the effective round-trip gain is too low. Increase both the resistance and capacitance, the filter resistance should be significantly larger than the pullup resistance, and the filter capacitance should be significantly greater than the effective gate capacitance of the MOSFET, I think. That oscillation frequency is really fast.

Is KiCAD's sim easier to use than LTspice? What about if you already have a schematic full of parts that almost certainly don't have spice models, can you make a simplified simulation that's based off your circuit? Or completely different, but without creating a new project?

>>2663813
I've definitely made accidental oscillators with op-amp circuits, but not with transistors. Is it a pretty high frequency oscillation (e.g. hundreds of kHz or more)? If so it's getting inductance from somewhere and yeah that's pretty doable. If not though it's hard to get oscillation from a single forward-biased transistor without the equivalent of cascaded RC lowpass filters somewhere in the circuit.
With op-amps I quickly learnt that a cap from inverting input to output will reduce high-frequency gain and improve stability. The same should apply to transistor circuits, using a cap from base to collector (or gate to drain if you're using FETs in the linear region for some reason).

>>2663843
>Is KiCAD's sim easier to use than LTspice?
It's easier if you use the built in components with pre-made models, but LTSpice seems significantly more powerful. The Kicad simulator is pretty much the entire screenshot here:>>2663776
You got a run/stop button to run the solver

>the add signal simply allows you to add notes to the plot from a list of nodes
>the probe is just a GUI way to add signals to the plot by opening the schematic and clicking at a specific node
>the tune option allows you to click on a component (such as a resistor) and tune it's value (not it's model though), such as changing a resistor's value from 1k to 100k easily with a slide bar
>the sim command will open another window with simulation parameters like timestep, start time, end time, etc. as well as a small tab for AC freq sweep analysis, DC sweep analysis and a tab where you can also put spice directives manually

They are still developing Kicad though, so I'd expect a lot more features in the next one, it's just more convenient to run on linux than ltspice in bottles, that's why I use it. It also has the advantage of having a nice schematic editor and allows for PCB design all in one, so you don't have to constantly rebuilt your circuit for simulations.

>The fact that it stops oscillating means the effective round-trip gain is too low
that was my thought exactly, but didn't know how to fix it, thanks, i'll give your tip a try and increase capacitance and resistance of the feedback branch.
> What about if you already have a schematic full of parts that almost certainly don't have spice models, can you make a simplified simulation that's based off your circuit? Or completely different, but without creating a new project?
In KiCad all you have to do is right click on a component in your schematic, click properties, then click simulation parameters. You can choose from some basic pre-made stuff or you can import a model from the manufacturer or make your own.

>>2663543
But I want to make my own

>>2663843
hey, i tried to give the circuit another trie, it definitely doesn't oscillate in simulation with a higher rf/c and also i built it on a breadboard, doesn't look like it oscillates either. I think the circuit needs hysteresis

I want to measure the level of a feed silo.
2.5m / 8 feet wide and 4m / 13 feet high
I want an Arduino to evaluate the sensor and illuminate an LED warning light green/yellow/red accordingly.

What is the best sensor to use for this?

I was thinking of a JSN-SR04T ultrasonic sensor, but don't know if that will work well in the silo.
Is there alternatively a proximity sensor, which detects from the outside through the fiberglass silo whether there is still feed behind it or the silo is empty at the point?

>>2663851
>The Kicad simulator is pretty much the entire screenshot
It's about as LTspice on Mac. I do basically everything with keyboard shortcuts because there's no other option.
>buttons
I'd check for other keyboard shortcuts, or otherwise look into writing custom spice commands.

>You can choose from some basic pre-made stuff or you can import a model from the manufacturer or make your own.
Yeah it was more for complex things like gate driver ICs, switching controllers, or even entire MCUs. With LTspice I'd just replace the MCU with a series of voltage sources making the right kinds of pulses, and I'd look for a close-enough built-in model for a switching controller or gate driver IC. Would I have to do the same substitutions in KiCAD? I also only simulate some parts of the circuit at any one time, would there be a way of separating just the current amplifier sub-circuit from a motor driver to test its frequency response?

I guess I should probably get into making models for parts, especially for designing high-power MOSFET circuits like SMPS and motor drivers. Making those models and running simulations seems like a much cleaner method as compared to trying to calculate switching losses manually.

>>2663889
Ultrasonic may work, you can buy ultrasonic water level sensors that have a pretty narrow beam for this kind of purpose, but I'm not sure how well they'd reflect off something like feed. Otherwise you could make an array of the sensors for better directionality, or add insulating foam material around the sensor to prevent it bouncing sideways.
You could also almost certainly use a laser time of flight sensor like you get in handheld laser distance meters.

As for side-based sensors, you might be able to measure the dielectric constant, reflectivity at X wavelength, thermal conductivity, or maybe radio opacity using sensors that "see" through the fibreglass. It's trivial if you make a tiny window for a photointerrupter, watch for feed dust occluding it though

>>2663894
>Would I have to do the same substitutions in KiCAD?
yeah, you're going to have to do the substitute if you're not going to use the manufacturer or a custom model.
>I also only simulate some parts of the circuit at any one time, would there be a way of separating just the current amplifier sub-circuit from a motor driver to test its frequency response?
You can create multiple sheets in kicad within the same project, then copy paste the circuit you want to simulate on a new sheet. Also, when simulating things in Kicad, you need to use a special voltage source since ngspice does not understand +5V flags for instance. So if your circuit is somewhat isolated from the rest of the schematic, you could always reconnect that part to the new vsource and run the simulation and i think it would only simulate that section and ignore the rest since ngspice would ignore it, but i'm not 100% sure. I've started using kicad sim tool only over the past week, before that i was using ltspice. I'd stick to ltspice unless you want a good schem/pcb editor on top of it.

>>2663899
>then copy paste the circuit you want to simulate on a new sheet
Oh so you can simulate single sheets at a time? That's probably what I'll do then.
>I'd stick to ltspice unless you want a good schem/pcb editor on top of it
I already use KiCAD for everything else (well except for what LogiSim is good for) so having it as my simulator also would be nice.

>>2663889
I highly recommend this video
https://youtu.be/2MF1FCjottk

>>2663894
I'll also lok into the laser tof sensors
>>2663930
thats a pretty good video. cant work with a tube, as my silo is not filled with a fluid, but interestinf nevertheless

>>2664015
Maybe you could use a pinball knocker with a piezo disc or MEMS microphone pickup, and DSP to process the feedback. Like knocking on a wall to find a hollow spot.

>>2664035
or maybe a passive (magnetic) floating gauge that rides the top of the grain to the bottom of the silo, with a vertical column of Hall effect sensors that track the position.

>>2663889
Laser rangefinder modules that would cover that range are available pretty cheaply ($15 on Amazon for MengJie MJ-VL53-400). That particular one has I2C and serial, should be pretty easily to communicate with.

Main concern that comes to mind is how much dust the feed generates when filling. Too much dust could give false 'high' readings until it settles.

>>2663889
need?

>>2664068
sneed

>>2663843
> accidentally discover darlington pair transistor oscillators all the time
Yeah, it’s almost always high frequency, like in the MHz as you suspected, you can see it on the scope.
With my luck, maybe I should try making an oscillator and it will turn out to be a stable amplifier!

>>2664035
>>2664039
Interesting ideas, I'll probably try a distance sensor from the top first, but I'll keep that in mind.

>>2664058
That seems like a good alternative, will order that one as well and give it a try, to find out which one works best.
Misreading while filling is off no concern for me.
The problem that I want to tackle is, that right now there's no easy way to check how much is still in there without getting on top and looking inside.
Obviously that isn't done regularly enough and most of the time you roughly know how long the feed will last. But every once in a while your a day late with refilling the silo and the cows only have the regular feed and not the extra feed.

Tldr. The sensor reading only needs to be accurate with the dust settled.

>>2663889
You could use a photo sensitive receiver and a light, LED make the gap one or 2 inches and if there's green in the way it's full and it's vertical with no distance so when the grain is gone it falls away. I don't like those doorway alarm bells when somebody breaks the beam it triggers an alarm.You could use a photo sensitive receiver and a light, LED make the gap one or 2 inches and if there's green in the way it's full and it's vertical with no distance so when the grain is gone it falls away. I don't like those doorway alarm bells when somebody breaks the beam it triggers an alarm.

Picture not related yeah it is unlimited electricity zero point energy we gotta talk about this sometime friends.

>>2663889
a lidar or one of those new stm tof sensors attached to the top facing down. 4m is well with in their range and if youre measuring solid objects and not liquids its quite accurate.

Confused about a loopstick antenna being a bare parallel tank circuit and the ouput of it is fed directly into an amplifier. How can this possibly work if the voltage is the same at L and C, since there is no voltage division unlike in a series LCR circuit?? Compare with #1 which is a capacitive divider and most voltage is dropped at the coupling capacitor Cc when the tank is not at resonance.

>>2664219
Found some links that explain that the even though LC are in parallel (and there also Rlosses there somewhere), the ferrite rod generates voltage in series with the inductance. So I am not sure what the equivalent circuit would look like, perhaps it just Vin->L-R-C all in series and the signal is taken from C? Then it would make sense.

>>2664219
Inductive antennas act like current sources, not voltage sources. The actual voltage they produce is arbitrary, it’s dependent on surrounding circuitry, and the energy accumulates in the tank circuit across multiple cycles anyhow.

>>2664231
>the ferrite rod generates voltage in series with the inductance

that's a weird way of putting it.
what's actually happening is the coil captures (or collects) radio waves coming across the ether and turns it into a voltage.
by the usual means: turning magnetism into current.
(same way rabbit ears catch TV channels.)
the LC tank circuit will tend to reject frequencies outside resonance.
the cap will short out higher frequencies, and the coil will short lower frequencies.

how the FUCK do I properly star ground stuff

>>2664330
It’s the same idea as the pex manifold. See picrel.
Star grounding is a fad that popped up recently.
Just like everyone suddenly needs to test the ESR of their capacitors.
Pex manifold is also a meme.

>>2664330
Wrong question. You have to ask why do I start ground.
Shitty LED blinker? LED light? Motor controller? Even audio amp? Probably you're good by just having one common ground.
RF? Measuring equipment? Are you Marcos Reps? If yes stop fucking shitposting go repair an 8 digit multimeter or something, idk, not your mom.

>try to make thing
>only 2 schematics exist
>one made by a turboautist from 50 years ago
>other made by retard caveman who doesn't put the fucking numbers
>derivation requires a PhD in math
I hate this fucking hobby sometimes

There's a guy called Wintergatan who tries to make these hand-powered marble music boxes etc. And lately he was testing one of his machines and tried to figure out how close you can operate it to the chosen tempo.

Anyways I'm wondering if it is possible to take an arduino, wire a 3.5mm jack to an ADC, connect an LCD display and check the timing between two peaks that come from a piezo microphone on an instrument. ADC should basically spit out a digital signal right? But how does it loon like? I basically need to find a peak of the incoming small voltage generated by the piezo sensor.

>>2664629
I’m not an expert, but a piezo mic might not have enough voltage output, so you might need a preamp.
And, if you make a preamp, you might as well make a zero crossing detector with a comparator instead of an op amp and hook it up to a gpio pin or interrupt.
The incoming signal will look like a sine wave, the adc will turn it into unscaled integer values.
If you use a properly adjusted comparator, you can just check if the gpio goes high or not. If it does, it’s started. Or interrupt, obviously.

>>2664557
> Guy posts pepe meme instead of schematic on an image board and complains that someone didn’t put the numbers
I hate this BBS sometimes.

>>2664296
>>2664300
I've seen an equivalent circuit of a loopstick antenna that looks like this. I guess you could do the same with the current source but a voltage source is easier to understand and simulate.

>>2664648
Well of course the capacitor is a tuning capacitor, not part of the antenna. And R combines the DC losses plus radiation resistance on the order of a 100 ohm.

due personal reasons I need to reduce my electronics kit to the bare minimum, what would you say it's a good portable (ie single small briefcase) electronics hobbyst set?

>ifixit tool kit + extras
>pinecil + soldering accesories
>portalbe o'scope
>usb-c variable power supply
>small assortment of modules
>rpico/arduino + breadboard and jumper cables
>assortment of THT components

what else?

>>2664729
What are the reasons?

>>2664735
lost my job a year ago and now I have lots of debt, I will have to sell my house and rent a small apartment b'cus i can't pay the mortgage anymore

>>2664737
Squat in your own house. If niggers can do it, so can you.

PlanB: turn your whole property into an EPA superfund cleanup site.

>>2664737
RIP bro... Lost my job last year as well but in Germany I just got a local welfare that covers 60% of my previous salary for a year of unemployment. Not ideal but better than nothing.

Was it a job in electronics?

>>2664739
not possible in this country (i'm not american), there's a law that allows landowners to evict tenants by force

>>2664741
I got 3 months of 60% pay, but after that you're in your luck. I've been just selling my stuff and doing uber. My job was software, backend dev, it's a tough market right now.

>>2664729
They should put variable pwer supplies into oscilloscopes, along with function generators. Usually there’s already a cal clock.

Another brilliant idea!

>>2664557
Post the actual schematic.

>>2664629
Yeah an arduino can do that easily, just have a little display or even just indicator LEDs to tell you if you're going too slow or too fast. Even better, you might be able to have some sort of governor mechanism to keep the speed more constant, even with variations in cranking speed.

>>2664729
What kind of projects do you plan on doing? Repair? Making new circuits? Analogue or digital? Do you make PCBs, or just use protoboard? I'd definitely keep a bunch of THT passives, maybe also a book of SMD passives if you plan on doing repair.

>>2664737
Shouldn't you have sold the house a little sooner than 1 year later?

>>2664775
There's a microcontroller-based USB "oscilloscope" that also acts as a function generator. Not sure if it works like a power supply.

>>2664746
That sounds depressing as fuck. I am in software too, also doing backend and random shitty tasks, but it is better than nothing. Why is market so bad? That's surprising. AI is not quite there yet even though it is inevitable. I thought now the sw market should be great since you could also find remote work or worst case relocate. Perhaps that depend on seniority?

>>2664776
>Shouldn't you have sold the house a little sooner than 1 year later?
Not him but it would be logical to try to find a new job which may take a really long time and you never know.

>>2664779
Yeah I was thinking that too, but 6 months should be more than enough to realise the market is dead. Cut losses BEFORE going into debt.

>>2664781
unless i misread and the debt IS the mortgage

>>2664781
True, but the market situation has been so confusing, there are conflicting signals all the time. I pulled out of the stock market a while ago and have been on the sidelines for almost 2 years which turned out to be a bad decision in the hindsight but I don't understand anything about our economics anymore.
>There's a record $5.3 trillion in cash on the sidelines as investors get more bearish on stocks.May 8, 2023
apparently I am not alone

>>2664776
>>2664779
>>Shouldn't you have sold the house a little sooner than 1 year later?

Yeah it's not like I'm not getting interviews, but I keep failing in the last stages. It's a constant rollercoaster of getting a little hope and then disappointment. I put up selling the house for as long I could but now my account is zero and I started eating off the credit card, so it was time to just cut loses and sell.

>What kind of projects do you plan on doing? Repair? Making new circuits? Analogue or digital? Do you make PCBs, or just use protoboard? I'd definitely keep a bunch of THT passives, maybe also a book of SMD passives if you plan on doing repair.

Bit of everything actually, from repair, learning electrical theory to making random kicknacks using microcontrollers/arduino and modules. I just like tinkering with stuff.

>>2664838
What in the actual fuck?

>>2664839
What, you don’t hide your 51.2 v lopo4 batteries in a wall with a secret hardboard hatchway?

>>2664776
>>2664641
>here's your ultra low noise, high sensitivity receiver bro
>I know I only drew two gain stages of 10, but I actually used three gain stages of 30ish lmao idk
>capacitors have values?
>here's my cat btw :3
shielding is apparently essential, but other people have gotten away with unshielded plastic just fine? but then this nerd just routes coax to a plastic box
fucker can't even remember if the stages are inverting or not

>>2664839
don't worry we are almost ready to connect

>>2664851
If you actually want help maybe try posting the proper schematic made by the old (actually qualified) autist engineer?

Anyway assuming NMR stands for nuclear magnetic resonance (why the fuck are you trying to diy this) google says there's decently high frequency RF involved, so that means your normally "good" jellybean opamps (TL0xxs, 5532s) aren't gonna cut it, they don't have enough bandwidth even at unity gain

maybe you'll be able to get away with video opamps but GBW product might still fuck you over (in that case you i guess you could add more stages with less gain per stage but it's gonna depend on the part you end up using)

Oh and if you're dealing with like uV inputs (which im assuming you are since you have all that fucking gain) then you need that first stage opamp to be a precision part with low offset voltage and noise and drift and all the bells and whistles while also having enough bandwith which isn't going to be cheap

Also you're gonna need to shield the fuck out of that thing or the only thing you're gonna be nuclear magnetic resonating is all the local radio stations

Honestly going off this guy's schematic this seems like the kind of thing that's really not worth diying, you're probably just gonna end up with an overcomplicated noise circuit lol

But to be fair I don't know shit about NMR in particular so some of the assumptions I'm making could be wrong, so feel free to correct me in that case

>>266485
>>2664914
Wait wtf i just looked at it better and this whole thing makes no sense why is the output being fed into a crystal?

Also assuming you're running the opamps single supply, Vref is the bias voltage and the feedback resistor is connected to ground the non inverting input will get shorted to ground via the coil (assuming that it's grounded because the fucking schematic is cut off) and the opamp output will clip or do weird phase reversal shit, either way that's not good

Dumbass here, have dabbled in electronics in the past but thinking of getting more into it.
Considering part options for a basic neon project. Never searched for parts on here before, is there any meaningful difference between these two, or should I just go with whatever's cheaper?
https://canada.newark.com/multicomp-pro/mpsa42/transistor-npn-to-92/dp/08N8077
https://canada.newark.com/onsemi/mpsa42/trans-npn-300v-0-5a-150deg-c-0/dp/97K5369

Need a sanity check. If I short the output of an LM317 current driver (inb4 hurr use better parts) to ground, will it still deliver the right current through the resistor?
I want to use my load itself as the "10 ohm" resistor. It has a well defined resistance and I'm a bit short on high power components right now

>>2664924
Most general purpose bjts are interchangeable, the difference is mostly in extreme cases like max current, max hfe, parasitic capacitances etc.
For standard <100mA <20V <100kHz use cases anything's fine

>>2664914
>to be a precision part with low offset voltage and noise and drift
can you elaborate on these a bit? the LC tank should have fairly low impedence, and low gain (10x) per stage should nullify the offset problem.

and what determines if an amplifier is noisy or not? surely the internal schematic doesn't rely on reverse avalanche zeners or anything crazy. does TI have a guide on this sort of stuff?
(earth field resonances are only ~2kHz btw)

I am looking to make a camera flash mechanism for a project, but at a higher voltage than I could find from a normal one. Ideally I'd like to make one with a variable voltage. Where should I begin?

>>2664936
That schematic, lol.
Anyway, so you’re going to shunt lout to ground then? Like that would work, but it’s not the greatest idea.
Adj is a high impedence input.

>>2664990
tested it, it drives 125mA just fine

>>2664923
> Crystal
i assumed that was a piezioelectric transducer

>>2664851
NMR basic principles are undergrad math at best.

>>2664914
The frequency of nuclear magnetic resonance is proportional to the static magnetic field strength. In an MRI machine, you have multi-telsa superconducting coils that result in MHz if not GHz RF. But in an earth-field NMR spectrometer the frequencies end up being in the audio range.

But you're right about the op-amp's DC offset swamping the high gain, even with decent CMOS op-amps that's a problem when you chain so many together. Best bet is to add a DC-cancelling feedback loop from the final op-amp's output to the first op-amp's "signal ground" connection, via a low-pass filter. At least I think that should work, I've never seen such a circuit before.

>>2664936
Yeah it should be fine. The only requirement is that your current source can handle it thermally, in this case your LM317 will be dissipating the most power. But the real question is: why are you trying to use a constant-current source when your load has a fixed resistance? Seems like you just want a 1.25V fixed voltage regulator. Which the LM317 basically is. If you actually do need a constant current because your load has a negative temperature coefficient or other unstable property, then using an LM317 as a 1.25V source won't give you the regulation you need.

>>2664981
Check the maximum voltage of all components, compared to the voltage it actually gets to. Chances are that measurement won't be straightforward though. If you need to get a greater voltage than your parts can handle, you'll need to upgrade those parts. If the boosting oscillator is open-loop, then varying the output voltage is probably best done by varying the supply voltage, though I'd reverse engineer the schematic beforehand to see what the limiting factors on too high a supply voltage (e.g. input capacitor voltage rating) or too low a supply voltage (e.g. mosfet gate threshold voltage) are. If it's closed loop, you can just change the feedback resistor(s).

>>2664851
Wait didn't Applied Science post a video about an NMR spectrometer? He goes pretty in-depth, but if I recall it's not an earth-field one but rather one that produces reasonably high-frequency RF. Because he did some funny stuff with coaxial impedance matching. The basic principles still apply, but you'll need to adapt the circuit for a lower frequency operation, which is easier in a lot of ways but still not actually easy.

>>2664923
>a crystal
I suspect he's using it as a high-Q filter. Still extremely strange though, doubt it works.

>>2665035
It was something else, similar but different.
IIRC he used to work as an MRI tech, so I bet he'd love to make the thing but probably considers it cheating or something
>high-Q filter
it's a piezo buzzer for sound lmao

Why is gold used instead of copper for bond wires?
The whole thing in covered in epoxy so oxidation can't be an issue, can it?

>>2665163
The wires are very fine and gold is among the most malleable of metals so the wires don’t break.
Gold is not that expensive, the wires are very fine.
Bigger chips can use different metals, even aluminum (like power igbt)

>>2665180
>All aluminum, all the time. 96.9 Power LGBT.

>>2665180
>$2000 per t.oz (31g)
>gold flake and bond wires still somehow affordable
mindblowing

>>2665467
>gold flake and bond wires still somehow affordable

The bond wires in most chips total in like the low microgram ranges. Even if gold is dummy expensive, so little is used that it basically doesn't matter.

Bought a madcatz mouse sensor because I want to build my own mouse. I opened it up and I see what looks like 20 connectors. How do I go about figuring out what each of the connections are?

>>2665530
Look up the datasheet.
>there is no datasheet
Sucks to be you. Better luck next time.

>>2665532
There is no datasheet. I was hoping there was a way to figure it out with just a multimeter or testing it

>>2665536
a multimeter is not a magical tool, any ICs or complex sensors requires a datasheet. Even a thing such as a simple capacitor requires a datasheet since you can't measure everything with a multimeter (most don't even have a capacitance option). Your best option is to reverse engineer the circuit with the sensor in it, and see if you can understand what each pin seems to be doing, not an easy task and might very well end up in failure.

>>2665536
There will be a datasheet, whether or not you can find it is the question. It may only possessed by people involved with the manufacturing of the mice, and the sensors aren't intended for anything else. If that's the case, it calls into question the legitimacy of the vendor you purchased it from. Maybe if you email the mouse company and ask nicely they'll send you one, doubt it though.

Your best option will be to reverse engineer an existing product with that IC. Be that by owning a mouse, or using internal photos from someone else's mouse. Chances are you won't be able to fully reverse engineer it just from images of an intact mouse, considering traces will be going beneath components, your best luck will be looking at people doing their own replacement jobs. It may be that the pinout and package look identical to that of another mouse sensor (preferably one with a datasheet) which would suggest the mouse company just write their own name on someone else's sensor.

You will likely be able to find the power pins with just a multimeter, by using the diode mode of the tool in order to find where the I/O protection diodes are. I'm not sure if this can possibly damage an IC though, look that up yourself. This alone would hopefully allow you to power it up and try looking for signals on the pins, and maybe reverse engineer it that way, though I imagine there will be some ambiguity from multiple power pins or other unusual pins that don't have diodes. Also the mouse only works when it gets polled via the USB link (assuming it's a USB mouse sensor) so you'd somehow need to find the USB pins first. Maybe they're giving 3.3V while everything else is 5V, who knows. Having images of from a teardown will likely be a requirement even with this method. If you don't mind burning some chips, you could likely get through it. But I'd cut my losses and buy a sensor that DOES have a datasheet.

>>2665029
>though I'd reverse engineer the schematic
If I had a schematic I'd just make one completely from scratch.

>>2665542
>>2665554
Thanks anons, that's kinda the direction I was going. I did find the data sheet for the sensor pin out but not for the board it's mounted to. I'm hoping by following the wires on the board I can figure out what most things are. But the sensor is 16 pins, then seems to go through a micro controller, then out to 20 pins. I probably need to find a data sheet for the MC now. So I'm going to try my best to not fuck this up

>>2665565
>If I had a schematic
No I mean you reverse engineer your existing PCB to draw out a schematic.

>>2665614
>I did find the data sheet for the sensor pin out but not for the board it's mounted to
Huh? I thought you just had the sensor, do you have the entire board instead?
>microcontroller
That makes things tough. A lot of mice these days make the sensor and controller all in one package, but if you need a microcontroller it might be quite difficult to get it working. If you have the datasheet that makes it doable to write custom code for a MCU that has native USB.

What kind of flip flop do I add to a cordless drill to make it change direction between successive pulls of the trigger. Another option would be to pick the rotation direction based on the most recent direction (in/out) of the trigger. Counterclockwise could be full speed regardless of trigger position. Is /mcg/ more suitable especially if I want to choose between the two or four different variations?

>>2663029
>CRT TV flyback trasnfromer with it.
Is the flyback transformer OUTPUT alternating current?

>>2665972
>OUTPUT alternating current?
My guess is yes, a/c output.

>>2665972
Most flyback transformers have a diode on the output.

>>2665988
>Most flyback transformers have a diode on the output
OK I think what you're saying is there's a built-in diode.

wtf is this Roel Van de Paar account on youtube.

>>2665972
AC, but some flybacks have multiplier and coils cast in same casing

>>2665622
>your existing PCB to draw out a schematic.
I don't have one.

>>2665962
Some sort of T flip flop, you can also use a JK flip-flop, maybe a D-flip-flop but they're more prone to noise. Debouncing might be a requirement, so I'd put a schmitt buffer/inverter and an RC circuit before the flip-flop, a time constant of 10-100ms or so should be good. You may need some trial and error to find a node that's on whenever the trigger is depressed, if not you may need to just rectify and filter an output of the motor PWM/inverter stage.
Then the output of the flip-flop needs to drive some sort of reversing switch. There may already be a reversing switch inside the drill, especially if it has a brushed motor, but this will just be a hand-operated switch, which easily can't be electrically controlled. If it's a brushless drill it's more likely that the thing runs with a software switch controlling the inverter. To invert that you'd need to flip two phase wires. The simplest way of doing either would be a DPDT relay, but those are bulky for the kinds of currents drills handle. You might be able to downsize the relay since it won't be switching while current is flowing. I'm thinking that if you have a brushless drill with a software reversal switch, you could use an XOR or XNOR gate to combine the outputs of the software reversing switch and the flip-flop before feeding them to the motor driver. That will be much smaller, lower quiescent current too. Bonus points for using a schmitt trigger XOR gate as your debouncer, though by the looks of things those don't exist.

>>2666156
Oh I thought you already had one and just wanted to change its voltage range, my bad. Should be pretty easy to find a schematic of a standard fixed-voltage one, I'm guessing they're a flyback converter of some sort. A more popular and likely easier circuit to make would be a ZVS transformer switcher. They're hard to change the output voltage of, but you can still just change the input voltage. Within a range that is, for gate driving reasons.

>>2666211
Thanks. Since I also want to intercept/override the trigger I ordered optocouplers and relays to do it with esp32. At the same time I discovered pic related which doesn't need reversing like the thread die I was using.

I got a corsair PSU with a dead 5vSB supply (when my PC shuts down on a working PSU, the RGB on my mouse stays on, but on the corsair it quickly fades off, also the PSU won't boot up after sleeping which is a problem).
no warranty, it was a used for a quarter of the price of new.
I've played around with ebike motors and have a basic understanding of electronics, but basically not much beyond that.
do you guys think it's worth opening it up and trying to diagnose it?

EE question. Sometimes in datasheets an impedance is given as R+jX, for example 50+j40, and what am I supposed to do with it? 40 reactive ohms at what frequency? For example if I need to build a matching network, how do I convert 50+j40 into a series resistor and a series inductor? L depends on frequency. Am I supposed to pick a frequency and then calculate L from Xl? Why do they do this? For example an untrimmed half dipole Z= 73 + j42.5 Ohms. For what frequency?? Is the reactance always 42.5 ohm regardless of the frequency, and if f goes down, L will go up so that Xl=42.5 is constant? Doesn't make any sense

>>2666402
one of the reviewers referenced a component for the 5vSB and I found a zener (I think?) diode that seems to have a broken trace.
I couldn't get the PCB off the shell without cutting anti shock globs.
I have never repaired traces before, I have a 15watt soldering iron and a multimeter. how should I approach this?

>>2666429
choose the frequency in the centre of the band youre matching for

>>2666464
buy a new power supply

>>2666429
Usually the 73+j42.5Ω will be tested at a certain frequency. 100MHz is a pretty common frequency for things to be measured at, which is useful when you're designing FM radio hardware and literally nothing else. For some parts (e.g. ferrite beads) there will be a frequency/impedance graph showing both the real and imaginary components. What component are you looking at the datasheet of? If it has a shitty (chinese) datasheet, try looking for equivalent parts with datasheets from different companies.

>>2666464
Can't see much from that image. Don't feel afraid to cut through hot snot or silastic to unplug connectors, but modifying the board is something I'd be reluctant to do unless I thoroughly understood what was going on. Something that looks like a broken trace might just be a deliberately cut jumper wire from the factory, or even a fusable link that is just a symptom of some other problem.

>>2666540
I don't remember a specific datasheet, but I am just wondering in general. For example a datasheet for an opamp specifies an input resistance and capacitance in pf. So you know two specific values R || C. But an impedance of an antenna is given as R+jX. And I also remember some specs in datasheets could also be given in this form. So I was just wondering why that is.

>Usually the 73+j42.5Ω will be tested at a certain frequency.
But isn't supposed to be a function of the physical length, not the frequency? This value tells me that the real resistance is always 73 ohm and it also has an inductive reactance with a phase angle that lags by about 30 degrees (atan2(42.5,73). But it doesn't tell me anything about the frequency.
>100MHz is a pretty common frequency
So at 100MHz the inductance will be about 67nH so the equivalent circuit would have a 67nH inductance in series with a 73Ohm resistor.
But if f=1Mhz, then L=7uH etc. But somehow magically 73+j42 ohm still holds.

>>2666429
>40 reactive ohms at what frequency?
it should clearly say. if not, look for chinese markings along the edges

>>2662600
Is spice simulation even worth the trouble with free solutions like LTSpice? There's no way of knowing whether or not a model is easily obtainable for a given component. I hate the idea of paying hundreds for Proteus just to have access to the models. I could buy all kinds of ICs, an oscilloscope, a large bread board, a power supply, maybe a signal generator for the same money.

Is spice in general for the birds and should I just calculate everything by hand and build everything on the bench?

>>2666605
yes
>no way of knowing whether or not a model is easily obtainable
annoying, but easily solved with substitutes (or git gud with custom models, check the community)
Spice doesn't do very well with ASICs, so you're mostly using it to test RLC networks and basic stuff like trannies/op-amps

>>2666558
>But an impedance of an antenna is given as R+jX
I'm assuming that's an antenna designed for a particular and relatively narrow band of frequencies, in which case that nominal frequency is told for you. If it's an electrically short capacitive antenna, or an electrically long magnetic antenna, they should give you the effective capacitance or inductance instead as that impedance is going to change as a function of frequency.

The thing is, it's basically impossible to get a constant phase shift as a function of frequency. The given complex impedance will only hold for a certain frequency, it will be vastly different at a significantly different frequency. If you plan on using such an antenna at its nominal frequency, you don't have to worry about converting the impedance to anything else. Only if you want to use it wider-band for things outside of its intended range will you have to convert that impedance into effective R, L, and/or C, then use that to backwardsly calculate the effective impedance at your new desired frequency. The inductance and capacitance do not change. The resistance kinda can change since it's a function of LC networks like a coaxial cable.

Actually calculating the inductance and capacitance is more complex than just assuming a series or parallel RLC circuit though. Your picture would be a function of like 6 ideal components minimum, you'd have to research what kind of equation results from each type of antenna.

See the ham radio general for people who actually know what they're talking about.

>>2666558
as others have said. it depends on the component in question. you cant generalise this like the way youre thinking. for your picrel example, impedance is determined by the length of the antenna relative to the wavelength. which is proportional to the frequency of the rf waves the antenna is designed for. designing a matching circuit isnt trivial either

silly anon, im not nearly smart enough to understand this.

>>2663365
Can somebody please shed some light on this topic ?

>>2666664
>baby me!!!
divide by the heat capacity of the wires and there's your heat jump in kelvin. any weak spots in the circuit could thermally run away (hot = more resistance = more heat = more resistance =...)
try it and find out

>>2666665
I'm just not sure about my calculations; it seems so low. But at the end it's only a few ms.
How much do you think should I theoretically reduce the wire section to account for weak spots in my dimensioning ?

>>2666664
> 500 kW
So, you’re at 1,000 V at 500 A.
Just move to 10,000 V at 50 A. You can almost do that continuously.

If you keep following that progression, you’ll get a Tesla-like epiphany and realize with enough volts you can remove the wire entirely.

>>2666676
Yep resistivity checks out, total power checks out, energy checks out. 16J for 180g being 0.24K feels about right, though I haven't done the math on that. It's also only like 1.68V drop relative to your 1000V source, so it won't result in much power loss as far as the load is concerned.
> theoretically reduce the wire section to account for weak spots in my dimensioning ?
A non-issue, even if you assume the wire has 100 times the resistance it will still work, though barely.

All I'd watch out for is inductance of your system causing a spike after the 500A turns off.

>>2666686
>A non-issue, even if you assume the wire has 100 times the resistance it will still work, though barely.
Ideally I'd like to repeat the operation after a very short period (every second is what I'm aiming at) so I need to conduct a thermal study too. I guess I'll over-dimension it by 50% for the prototype.

>All I'd watch out for is inductance of your system causing a spike after the 500A turns off.
I'm dimensioning for the max current of a capacitor discharge, power input will be much lower at t>15ms. I remember fuck all about how inductances behave in the temporal realm. Do you have an idea about what behavior I should expect ?

>>2666694
>every second
Ah, that's when you need to look into power dissipation of the wire. There's likely a good way to calculate this, from intuition I expect your wire to be able to dissipate ~100W before getting too hot. From the american wire gauge article on wikipedia it suggests ~47A as a reasonable steady-state current for 7awg wire. From experience I know that rating is really conservative, but we'll run with it. For your 2m long 3.36mΩ wire that's P=R*I^2 = 7.4W total. If you're dissipating 840W for 20ms every second, that's an average of 16.8W. Which is double the rating from wikipedia, but again their values are really conservative for free air. For example, a 10W power resistor is about 10cm long, 1cm in diameter, and gets up to probably no more than 80°C in free air at that power output. 2m of that would be able to handle 200W. Maybe your wire has significantly less surface area than a 1cm*1cm square resistor, but not more than a factor of 6 or so. I expect it will get up to maybe 50°C, no more.

Do a test run or find the thermal specifications of your wire if you need more precise data.

>I remember fuck all about how inductances behave in the temporal realm
They store energy proportional to the square of the current. Suddenly shutting off current in any inductive system will cause voltage to rise on its output until either the energy is stored capacitively, or it reaches the breakdown of something. In practice, this means arcing across switch contacts when you break the current, or spiking and killing your semiconductor switch. At 500A you may have quite a bit of energy there, even with a short wire. At 1000V, you risk it breaking down insulation or air gaps. Keep your loop sizes small by having the supply and return lines as close to one another as possible, maybe even twisted together, though that's worse for thermal reasons. Using an RC snubber network, TVS diodes, MOVs, etc. would be what I'd do, if possible.

>>2666621
Most of what I can find online is IBIS models and there do not appear to be any free simulators that can even use the models. The IBIS models are basically a black box that tells you nothing about how the IC actually works or behaves and the simulators are all proprietary and don't even list the price of the product on the web page because they know you would lose your lunch if you knew what the price was. Electronics is Jewish as hell compared to software.

>>2666753
well duh, ASIC's are black boxes too. wtf are you even trying to use?

>>2666700
>Heat topic
Thing is I'm supposed to operate in a low pressure environment so I'm not sure about those figures. Maybe this highlights the fact that some noble gas should get involved in the cooling of the system. I'll for sure test that out.

>Inductance topic
I can't really get the supply and return twisted together as they are moving. I'll definitely look into these solutions.

Thanks for the valuable pointers.

I'm making something in KiCad. A large part of the circuit is more-or-less a single template subcircuit copy/pasted 16 times, where each copy-paste has slightly different values or layouts.
Having to hand-edit things is becoming a pain in the ass and I can feel that it's quite dangerous and error prone. So, what's the move here? What the smart way of doing this?

>>2666864
>So, what's the move here? What the smart way of doing this?
You need to look into kicad python scripting. But depending on how complex your problem is, doing it manually could still be faster in the end.

>>2666628
>The given complex impedance will only hold for a certain frequency, it will be vastly different at a significantly different frequency
>The inductance and capacitance do not change
If C and L do not change, why would reactance change with frequency?

Anyway I think I know where I am wrong in my assumptions regarding antenna impedance. I assumed that R+jX of a an antenna depends on just its length, but I forgot that it is a normalized length relative to the wavelength so of course one can say that R+jX depends on frequency. Then it all makes sense. C / L can be calculated from the wavelength/frequency. Duh.

>more complex than just assuming a series or parallel RLC circuit though.
Yeah I've seen some insane equations to calculate the radiation resistance and I don't want to go there. I was just curious about this R+jX thing but I think I get it now.

>>2666864
>>2666879
obligatory

mom, im 1930 capacitor factory

>>2666919
>If C and L do not change, why would reactance change with frequency?
That's like the primary function of inductors and capacitors. They act like filters that attenuate signals above or below cutoff frequencies depending on whether they are wired in series or parallel with a given circuit.

>>2667009
Oh never mind, I don't know why I read that as "C and L change with frequency". Of course Xl and Xc change with frequency. I confused myself.

I'm reusing a bluetooth headphone board for a speaker, but the on switch is pressed and held for power on/off.

Am I correct in assuming a momentary switch would be the correct replacement switch to move it from PCB to a panel?

>>2662797
you're a dolt. linear, regulated supplies have plenty of uses.
or maybe he should add a touchscreen and twitter huh asshole?

>>2666864
Are you using subsheets? If so I think you can duplicate the sub-sheet and edit each one without it influencing the others. How many values do you have to edit per schematic? What sort of differences in layout are there? It may be easier to just have a flexible layout for every one, and just not populate some parts.

>>2666985
Now get yourself a big square-bottomed storage bin and make consecutive pours of paraffin wax with aluminium foil between them.

>>2667026
Yes a momentary switch will be what I'd use for a direct replacement, the toggling and such will all be done in software. That said, if you want more advanced features you could use a microcontroller to control the buttons, for turning it on/off remotely or with some sort of timer, or turning it back on again if the thing automatically timeouts. Timesout?

>>2667030
Eh, those adjustable ones are kinda shitty for everything. They can't handle high power draws due to the linear nature and lack of switched windings. So even with ones that have current regulation, they'll burn a bunch of heat regardless of output voltage when they're self-regulating. Furthermore, it's a single rail output, so for the main situation where you need a nice smooth linear supply (audio) you don't have the twin rail output, and you don't need that to be continually adjustable anyhow. They're still usable for sure, but just aren't flexible enough to waste bench space on. He'd be better off tearing it down and making a different type of power supply from its nice toroid. Be that a low-noise split-rail audio supply, or a relay-switched multi-winding linear supply. Or just a fixed AC supply to power something reliably.

These days you can get away for 99% of stuff with a cheap CC/CV switching supply, and a cheap fixed split-rail linear audio supply. Having multiple fixed rails for the linear supply is a nice addition for little extra cost. For anything better you should make or buy a relay-switched multi-winding linear supply.

>>2667041
>Now get yourself a big square-bottomed storage bin and make consecutive pours of paraffin wax with aluminium foil between them.
Eh, Id rather order ceramic SMDs from china

But yeah, FM transmitter works :D
Shits far away

>>2667049
>not wanting to store thousands of joules at tens of thousands of volts
ngmi

also nice work lmao, be glad you're not working with microwaves

Here is my fm transmitter

>>2667041
So I should be good getting something like this? https://www.amazon.com/uxcell-Momentary-Button-Switch-Mounting/dp/B07Z2TDDTZ/

>>2667053
la creatura

>>2667053
>soldered onto dupont jumpers
yameroo

>>2667054
Sure. There are definitely cheaper and simpler panel-mount switches out there, but if that matches your aesthetic go for it. That one has an LED in it which you have to wire up separately. Ideally there's some pin on the headphone board that goes high or low only when the thing is active, or else maybe there's a built-in LED on the headphone board you can solder small wires from. You can find buttons like that with multicoloured LEDs in them, so if you had multiple signals (standby, bluetooth, aux mode, low battery, etc.) you could feed them into different colour LEDs inside the same switch.
You could just wire the LED up to the button itself so it only glows when you push the button, or maybe make it glow all the time, or just ignore it.

>>2667049
Actually, quality of sound is surprisingly good, but entire circuit reacts badly to thermals it seems (no fucking wonder, capacitors are paper)

>>2667071
Are they oil-soaked paper?

>>2667075
Masking tape and alu foil.
Like heating kicks in, frequency moves somewhere (unpredictable direction too).
But yeah, I think you can make pirate FM transmitted rather compact and disposably cheap (phone-> $2 MP3 player and obv id add some sort of amp there on the output, and possibly filters so it shits less outside of designated freq.) and put it in neutral spot and dab on FCC/their alternative.

Whew, spent the last few weeks cleaning my bedroom/workspace, now it's looking pretty good. I still want to add some USB and 240V outlets on the plywood under the storage drawers, along with an RS232 port or two, one of which goes to the oscilloscope. Hooking up some of those USB ports to a USH hub that goes to my computer would be nice, so I can program microcontrollers and such without having to stretch a cable over each time. I still have to make some wire reels, some sort of tool racks, racks for storing IEC, USB, banana/alligator, and BNC cables too, and another rack of 30 drawers to add, in the near term. Long term it would be nice to integrate a good variable power supply into the setup, for now I've got the Heathkit breadboard unit with its linear ±12V and +5V rails, and a CC/CV LM2596 module crammed in a box with external pots and panel meter sitting on the floor. I'll put some magnetic tape on it for more convenient use for now. I have an ATX PSU to use to create a high-power buck-boost CC/CV power supply with linear postregulator in the long term. Not sure where it will fit, maybe I'll hide it behind the plywood and just have the outputs coming out somewhere convenient.

>>2667053
twist your leads together and presolder them into 3-5 component modules to reduce the amount of space you take

>have 700V EV battery that can deliver 500A
>apply that across a 100u ohm resistor
>have a power dissipation of either 490kW or 25W depending on which formula you use to calculate the power dissipation
Someone explain please

>>2667236
>have 700V EV battery that can deliver 500A
It can deliver 500A, it doesn't mean it will deliver 500A. How much current will flow is determined by the load and internal resistance of the battery (+ cables if you want to be accurate). In ideal case for your 100 uOhm load current is U/R = 700V / 100uOhm = 7MA which is obviously way more than 500A. Battery itself has R_bat = U/I = 700V / 500A = 1.4 Ohm of internal resistance. If you connect 100 uOhm resistor across it current will be I = U / (R_bat + R_load) = 700V / (1.4 Ohm + 100 uOhm) = 499.96 A. This is the current through load resistor so power dissipated on it is equal to P = I^2 * R_load = 499.96^2 * 100uOhm = 24.996 W.

for me it's circuitmaker2000

>>2667245
Thanks!

where the FUCK do I get cheap single sided copper PCBs? I see mauser selling 6x6" for $8+ fucking WHAT
Even the chinese are selling for like $3 each, fuuuck this garbage should only be like a dollar it's bakelite ffs

>>2667334
I just get the kilo of copper PCB cutoffs for 3€ from Pollin.

>>2667338
thinking of just putting copper tape on cardboard/plastic, the fuck to I need shitty bakelite for anyways

>>2667334
>I see mauser selling 6x6" for $8+ fucking WHAT
You can get a custom 2 layer PCB from chinks for that kind of money, shipping included. The truth is, since covid money printer madness economy is totally fucked. I suggest you start learning chemistry instead, because we'll soon be making everything at home from scrap. All it takes is a few more upsets, natural disasters or chinks invading taiwan and you can say goodbye to your hobby. Might as well throw everything into the fucking bin and start weaving baskets instead. It's only going to get worse from here on.

>>2667351
>I suggest you start learning chemistry instead, because we'll soon be making everything at home from scrap.
How plausible is home IC fabrication? Is everything just going to revert to analog and vacuum tubes and shit?

>>2667334
https://www.mpja.com/Copper-Clad-Boards/products/393/

>>2667353
Back in the old days people would steal car radios to feed their crack addiction. In the new shit world, people will steal entire cars for the $2 microchips and scrap the rest. lmao

>>2667362
Car manufacturers are jerks though and always lock their sweet sweet micros and epoxy everything.

>>2667353
>How plausible is home IC fabrication?
Pretty much impossible unless you can find old fabrication equipment and can source raw materials. Even making a vacuum tube at home is a big challenge. Electronics is a very modern field and requires society and its supply chains to function perfectly. One small upset and you get 24 month lead time for new ICs. It's actually quite scary that there are material suppliers that are almost solely responsible for supply of critical materials for manufacturing. Anything from resins to ultrapure water. If something gets fucked you can add another 12 months to lead times. If a big upset to system were to happen it would be very difficult to restart everything. It's better to not think about any of this and just accept reality as is and solve problems as they appear. Future doesn't exist and past is already gone, what matters is now.

>>2667353
You’re not going to need it.
Most of the shit we need and want can be done in /ohm/ rather than /mcg/

Filters for example. Now it’s common to implement a filter using ad converters, billions of transistors for cpu/dsp and ram, lots of code, then a dac.
This is an epic waste of resources, and it’s artificial.

In reality, even before transistors, we would implement things like active filters using a few tubes, and later bjts, fets. The whole filter theory was already well established in the times of tubes, and not much has been “invented” … the last breakthrough was sallen-key in the mid 1950s.

Remember, even television was originally analog!

It’s kind of like going to the moon. We used to have the capability, but we lost it along the way. We probably couldn’t even manage such a project of that complexity today, let alone pay for it.

So, when you consider your first led flasher project, please consider doing it with a single transistor before breaking out an arduino.

>>2667377
From fabrication point of view there is little difference between making a single transistor and IC with millions of them. Today analog is a complete waste of time and resources. Creating components with tight tolerances is expensive and even in best cases things move around with temperature and age. Then there is a question of physical size, number of components, flexibility, repeatability, assembly cost...
>Remember, even television was originally analog!
And those boxes weighed 100 kg, whined all the time and image quality was shit. Now you can have 8k displays that are as thin as a painting and it's all done with few ICs.
>It’s kind of like going to the moon. We used to have the capability, but we lost it along the way. We probably couldn’t even manage such a project of that complexity today, let alone pay for it.
Going to the moon was just a PR stunt to "own" the soviets. The real reason why space tech developed so fast was military applications (ICBMs). Now that tech is sufficient for it's purpose nobody wants spend money on another expensive PR stunt. Once China or Russia start (probably not anytime soon) to do various things in space things will resume. US literally can't do anything unless there is some boogeyman to defeat. Its default state is to just decay and rot in decadent lifestyle powered by entertainment industry (and drugs from south).

>>2667377
Based as fuck.

>>2667245
I doubt the ESR is 700V/500A. It's likely significantly lower, but the battery simply isn't safe to operate over 500A.

>>2667396
I think his point is that you don't need to make your own transistors in the first place. Buy a million of them for a paltry sum, you're set for life. Special purpose ICs (SPICs) are something you need to buy specifically for each project, so in a collapsed supply-chain situation you can't rely on them. That said, cheap microcontrollers are also general purpose enough you can feasibly stock up on them, so I might be misreading him. Same for jellybeans like op-amps and TL494s. These are the kinds of ICs people design projects around anyhow, because it's expensive to make a new digi-key order for every little project.

stupid question, i want to drive an led display module but i cant find a schematic/pinout of it, only its spec sheet and ratings i only have 1 so reverse engineering is out of the question
>https://buzzer.co.kr/goods/read.php?M2_IDX=19462&SC_ALL=N&SC_SC1_IDX=488&SC_SC2_IDX=1582&SC_SF_IDX=Array&SC_SM_IDX=&SC_TYPE=&SC_WORD=&SC_TI_IDXS=&SC_PA_LEN=&SC_LIST_TYPE=&SC_BOOKMARK=N&SP_CODE=20030YYF

>>2667610
That has over 100 different addressable LEDs being controlled by 8 pins, so it's not like one pin is going to connect to an individual LED. From the picture of the back of the device on the website, there are three reasonably large surface mount chips behind the epoxy that could be basically anything. Might be a microcontroller and two pin drivers, three shift registers, or something weirder. If you don't have a working device that it connects with so you can at least observe how it communicates you will probably have a bad time. Unless you like doing this kind of thing, in which case you wouldn't be posting this question here.

>>2667610
8 pins can at most give you a 4x4=16 LED matrix, even with charlieplexing you can get at most 56, and that board has 106 LEDs on it. So it will be some form of digital comms protocol like SPI, and that connector is hardly a standard for such a protocol. I'd look to see what kind of device it's intended for looks maybe like a replacement part?) and see if you can get info on the pinout from that side. Otherwise, email the manufacturer or one of their agents.

>>2667526
>I doubt the ESR is 700V/500A. It's likely significantly lower, but the battery simply isn't safe to operate over 500A.
You are correct, but detailed battery/cell data is most likely not available. BMS would most likely trigger protection for currents above 500A so that would also be the limit for the load. But the original question said
>battery can deliver 500A
nothing about safety so only thing you can assume from given data is that this is short circuit current. It's probably not a practical example anyways but a question for students studying the basics.

>>2667526
> don’t need to make your own transistors
Well said, that’s exactly what I mean.

> Stock up on common ICs, too.
There’s not really a lot of transistors in some ICs, like inverters I think there is like 2 mosfets in there for each inverter.

You could sand off most of the black epoxy of some 2N7ks, solder the leads together and encase it in epoxy again and there’s your IC. I assumed OP was talking about something more complicated than that. We call these things “hybrid” ICs. Sort of.

One thing about using transistors directly is that the specs are higher, like usually around 50 V and 100 mA at least. And generally more robust.

> lcd television is high rez
Wasn’t talking about the CRT per se, but the signal processing. Anyway who’s to say we wouldn’t have miniaturized everything from the 50s until now?When we took apart a downed MIG, we found it had miniature tubes in it. Now we’d use VFD tubes like the Korg nutube. Maybe display technology would be based on tiny gas discharge tubes. Oh right, it was—plasma TVs were a thing.

And in both cases, these make things more resilient to EMPs so you should brush up on your tube technology before it’s too late.

>>2667688
>And in both cases, these make things more resilient to EMPs so you should brush up on your tube technology before it’s too late.
If we have to revert to tubes we're already fucked beyond repair and messing around with electronics would be the least of your worries. Learning chemistry and biology would be much more useful if you want to have an advantage in such situation.

>>2667685
Yeah but the point is, electrically it's different. When you connect a 2Ω load, the resultant voltage is different. So using the "maximum safe current", which is what the 500A value almost certainly is, to calculate ESR will get you a wrong value.

>>2667688
>One thing about using transistors directly is that the specs are higher
Yes but they're also worse for quiescent current and balance. Try getting as low an offset voltage as a halfway modern op-amp using seperate transistors. The common CMOS arrangement of transistors in an inverter is also kinda hard to replicate without getting shoot-through current, not sure how they avoid that in 4000 series CMOS, considering their maximum supply voltages are more than twice the gate threshold voltage of their FETs. You also miss out on a bunch of stuff found in ICs like 4-terminal MOSFETs, multi-gate FETs, multi-collector BJTs, floating-gate FETs and ferroelectric transistors for memory. P-channel JFETs too.

If you care about EMPs then store your components in an iron container of some sort. Actually a server rack enclosure might work pretty well. Could you make a discrete BJT circuit EMP-proof by adding TVS diodes at every junction? I live in a rural town in a small pacific island nation and think I'm pretty low down the nuking list, but that won't apply to most people here.

>>2667693
>So using the "maximum safe current", which is what the 500A value almost certainly is, to calculate ESR will get you a wrong value.
I still think that original question was about understanding of thevenin equivalent circuits and power dissipation and not actual physical battery behavior. If we were dealing with actual physical battery the way data would be presented would be completely different.

>>2667707
Well yeah even then, the Thevenin calculation is different from the ESR = 700V/500A case. There isn't enough information to calculate the ESR because 500A is just a guideline.

Voltage spikes can purely be determined by V/L = dI/dt, right? So if you abruptly stop 100mA going through a 1mH coil in 20ms, you'd only get a 5mV spike?

>>2667887
Voltage is determined by rate of change of current what you need to write out is actually

v(t) = L * di(t)/dt

There is one simple case when current "shape" is equal to

i(t) = k * t

where k = amps/second, a constant. This means that every second current rises by some amps (linear ramp). Because derivative of i(t) is just k then voltage across inductor will simply be

v(t) = L * k = 1mH * 100mA/20ms = 5mV

which is not a spike but a constant value. Voltage across inductor will be 5mV as long as current through it is changing by 100mA every 20ms. As a first order approximation this will work well enough, but once you move into real world an inductor is not ideal and has parasitic capacitance and resistance which will change how it behaves. This is a good method for quick approximations and might not always work because either inductor is not ideal or your current is not a ramp but some weird waveform that moves all over the place.

>>2667693
> an offset voltage as a halfway modern op-amp using seperate transistors
I’m sure you can, there is ways to compensate them—especially if you make your own—depends on how much time you want to spend with your 12 digit meter and dropping blobs of solder on the nichrome/copper wires like we do with μΩ-accuracy current shunts.
Even the commercial discrete op-amp clones in picrel are comparable to, say, the NE5534AN.

> i said modern
Well, like I said, it’s certainly possible, but the average audiophile doesn’t need it to be that accurate so the don’t laser trim the offset compensation network out of the factory on these things. Just like I don’t really need a 10k television set so I can see the individual pores of the BBC newscasters.

In many applications, an op-amp is overkill for a lot of specific tasks. Sure, it’s incredibly useful and versatile, but we didn’t see a lot of tube op-amps back in the day, they were more of a re-usable piece of lab equipment. Imagine needing to shout out “okay, who has the op-amp” in the EE lab. It’s like that.

I have a brilliant idea, liquid ass on a chip. Put liquid into a container and burn open a hole in it using an electrical fuse. Think of the possibilities.

>>2667639
> like SPI
Yeah but what’s the cable length on SPI, like 1" or something like that?

>>2667951
I was thinking more OP07-tier. And even if you trim them right initially, there’s thermal drift to worry about as they get to different temperatures.

Though you’re right that you don’t need that high performance for most uses of op-amps. But I still think op-amps and TL494s will be workhorses of the collapsed supply chain. Because they’re cheaper and more reliable than discrete transistors. If a single long-tailed pair is enough then that’s fine, but the higher gain and more disparate impedances really are useful if just to make design easier.

>>2668072
10m officially, but with good wire you could push 30 I’m sure. Heck they put I2C lines in HDMI cables. If that’s not good enough go for RS485. Or Ethernet.

>>2668152
>SPI
>10m
Maybe on very low clock frequencies nobody is seriously using?

Is there a decent forum or somewhere I can go to discuss a project other than 4chan?

>>2667887
Yes, as the other guy says. If you're using a capacitive snubber, you can calculate the peak dI/dt based off the resistance and inductance of the snubber circuit, which will tell your the voltage spike size. I'd also calculate the energy storage of the inductor, then figure out what the voltage of the capacitor will be if it's full of that energy (in reality it will be something lower due to the resistor), I think what applies will be the higher of the two voltage values.

>>2668153
Try twisting your wires together for once, Gomez.

>>2668241
The EEVblog forum is pretty cool. Making your own blog or youtube channel is what I'd recommend if you were wanting to make a more permanent project log. Maybe ScienceMadness does stuff other than chemistry, idk.

>>2668152
> spi goes 10m, use ethernet
Roflmao, there’s no way most spi goes that far, it’s practically an on-board interconnect.

Ethernet is a millard times more complicated than SPI, there’s no way i’m putting an ethernet controller and pulse transformer on there thats way bigger/pricier than the rest of everything.

Imagine your calculator using ethernet to run the 7 segment display controlled by the calculator blob chip or z80 if you have one of those deluxe jobbies.

>>2668153
Wait is that a chio-chan?

>>2668253
>Roflmao, there’s no way most spi goes that far, it’s practically an on-board interconnect
You know you can just decrease your clock speed, right? Not like your calculator blob chip needs more than 10 updates per second, which for 106 elements is like 1kb/s.
>Ethernet
I'm not telling you what standard to use, I'm suggesting what standard it might come baked in with. Considering it has an 8-pin header it's not completely out of the question. They might use such a protocol if it's a standalone industrial display not built into a piece of equipment, though not by the look of the front. More likely than 8 pins being used just for SPI though.

That said, looking again maybe it is more than a display, and is itself some sort of controller. In which case, some of those pins would be for inputs like buttons or sensors, or to control outputs like relays.

>>2668241
>Is there a decent forum or somewhere I can go to discuss a project other than 4chan?
Did that long-winded Dutch faggot Arsenio ban you again? He should go back to hackadong and stay there.

>>2668329
>long-winded
why does nobody seem to get this?
100 words or less motherfucker, nobody wants to read your walls of text

>>2667981
I do believe you've invented the Stinkzistor. However it's already been accomplished: https://www.youtube.com/watch?v=OOA1NaoKV6I

What kind of tools I need (to make) for being a ghetto HAM faggot?
After making FM transmitter, I want to make an amplifier, but... I want to know a lot of shit, like.
- is frequency what i want?
- SWR
- harmonics and other waste of power
and this is a big problem, because i've no money to buy oscope, spectrum analyzer, even SWR meter is too fucking expensive for me, but I do want to shit in the air, and shit i want far, so I'd rather waste money on piece of precious BeO ceramic transistor and copper pipe for antenna.

>>2667199
Just use SMD like I did. And prison capacitors, of course.
>>2667334
Your local supplier. Idk how, but they manage to get a better deal than chinks, and unlike chinks they know the difference between FR4 and phenolic.
Alternatively, you can just order PCB from usual suspects. Quality is high, comparable to laptop motherboard in terms of resistance to lift pad and solder mask. You can't get that quality at home.
>>2668241
I'm making one, but I've another question, how do I advertise it?

>>2668552
I bought a shitton of FR4 copper boards and now I hate it cuz i'm a dumb notoolz in my bedroom and I can't break it or cut it and it's shitting splinters everywhere reeeeeeeee

noobs, if you're reading this get copper-clad FR-2 for prototyping. Easy to break, low noise, piss easy grounding, no fucking LUNG CANCER all over the place or dulled sawblade nonsense

>>2668562
>I bought a shitton of FR4 copper boards and now I hate it cuz i'm a dumb notoolz in my bedroom and I can't break it or cut it and it's shitting splinters everywhere reeeeeeeee
Get good, use water and diamond wheel to cut it, and sand.
>noobs, if you're reading this get copper-clad FR-2 for prototyping. Easy to break, low noise, piss easy grounding, no fucking LUNG CANCER all over the place or dulled sawblade nonsense
FR2 sucks, it delaminates instantly and asborbs moisture like crazy, and about just as hard to cut.

>>2668597
>FR2

>>2668562
Cut using shears instead of a saw

>>2668635
Shears get fucked quick. I think best way is just cut it outside or with dremel and pool of water so fiberglass doesnt get thrown everywhere

>>2666666
just checking

>>2668635
tried that, fucking tin snips leave a serrated edge, cut inward at a weird angle, and warp the board

>>2668655
>>2668597
Sadly, I don't have the space or ventilation for a fucking diamond saw in my apartment, otherwise I'd be doing way cooler hobbies. Considering how many hobbiests are in similar situations, I think it's naive to say power tools are a solution.
I can see FR-4 being used for final revisions of projects, clearly it's the better material, but sometimes you just need the rf/low noise equivalent of a breadboard - cheap and to the point.

>>2668562
When I was in your situation I used a multitool’s knife to repeatedly score and cut through the copper. Once that was done I just bent the thing in half using the edge of my windowsill. Finished it up with a piece of sandpaper I found on the ground by a construction site.
Now I just clamp it in a vice in my garage and go ham with a hacksaw, dust be damned.

>>2668767
>I used a multitool’s knife to repeatedly score and cut through the copper.

cavemen tech for the win!
shoot, at least get an acrylic scoring tool.
used to get 'em at Home Depot for $2.
with covid inflation, Ukelele war, and Trump's Chyna sanctions, they're now closer to $20.
only place to get a deal now is on the slow boat from Chyna.

>>2662600
Do you guys think a 20 MHz oscilloscope and a 24 MHz, 8-channel logic analyzer is good enough for tinkering with old, 8-bit microcontrollers?

>>2668779
>good enough

they're adequate for the job.
but you DESERVE nothing less than 1Gz with 8 channels.
coz you're THAT special.

>>2668788
>you DESERVE nothing less than 1Gz with 8 channels.
coz you're THAT special
y-you too

So uh the tutorial is telling me to plug a .1uF capacitor between VDD and ground, as you can see in picrel.

I've figured out that this is a bypass capacitor with the function to filter out noise, so only a clear 5V or 0V is registered (?)

Is my reasoning below correct:
So basically this capacitor conducts charge up to a max of let's say 2V or so, and this is a function of its size (.1uF - a larger capacitor would e.g. filter even 5V). This makes it ideal for 5V-high-style circuits, as low-voltage noise will be sent to ground but any higher charge close to 5V will be insulated/resisted against and thus registered.

>>2668668
>fucking diamond saw in my apartment
Like, no space for dremel and cardboard box with vacuum cleaner attached to it?
>>2668826
bypass capacitors are there in order to overcome line inductance.
In other words, if you have switching load and no capacitor locally, voltage would first dip, and then raise above 5V level, possibly burning shit, with capacitor nearby, sudden load would discharge capacitor first, which will reduce the voltage drop, but also it will absorb the inductive spike produced by power wires.

>>2668329
>Did that long-winded Dutch faggot Arsenio ban you again?
What?

>>2668826
>Is my reasoning below correct:

You're overthinking it. By their nature, microchips tend to draw huge (relatively) spikes of current in operation. This can cause voltage at the power terminals of the chip to sag, if the impedance (particularly the inductance) of the conductors feeding them is too high. If that voltage goes too low, the chip can start glitching and you get unexpected and unwanted behavior out of it.

The solution is to simply put a capacitor very close to the power terminals of the chip. The capacitor can provide the necessary charge to the chip during the brief periods when it needs it. Bigger isn't necessarily better. As long as the capacitor has enough capacity to keep the chip happy, adding more doesn't do anything. In the extreme, it can actually make things worse, as larger capacitors tend to have higher inductance, and may be less capable of covering the spikes in current demand from the chip because of it.

The exact requirements aren't usually too demanding. Unless otherwise specified, a 100nF ceramic capacitor is the go-to size for anything from a 7400 logic chip to an 8-bit microcontroller. Larger 32-bit chips might ask for a combo of 1µF and 100nF capacitors. The larger one handles larger, lower-frequency spikes in current, the smaller handles smaller, higher-frequency ones. (Similar is often seen in power supply filter capacitor arrangements, just with larger values.)


The arrangement pictured isn't great, DESU. It's better than nothing, but bypass capacitors should be as close to the chip as possible. Is that a dedicated breakout board for that chip? Those usually have the bypass caps on them.

>>2668779
You want the sampling frequency to be twice the maximum expected frequency (Nyquist sampling theorem). If your MCUs can handle no more than 12MHz then you're fine, otherwise you may run into situations where it isn't adequate. Well not like you'll be shitting out data from DMA at clock speed with an 8-bit micro anyhow, you'll be fine. As for the scope, you'll almost certainly not be seeing analogue stuff up in the mid MHz anyhow. Better be getting a chad analogue scope instead of a shitty Hantek or whatever.

>>2668826
Yeah, whatever the other guys said. I = C*dV/dt and all that, run an ltspice or falstad sim if you're curious. More importantly, your breadboarding looks wack. Did you stack two together? Are there power rails underneath the breakout board? Also I'd put the bypass capacitor as close as possible to your microcontroller, meaning plugged in right next to your board. Or even soldered atop it.

Have you considered buying a dedicated breakout board for your microcontroller of choice? They come with all the bypass caps and crystals and other supporting components you need. A 32 bit 100+MHz microcontroller on a breadboard is pushing it.

>>2668843
Why would a sudden load discharge the capacitor first? Simply bc of geographical proximity?

>>2668851
In both your responses you used "inductance", the "tendency to oppose a change in current flowing through it". Not too sure how an object can oppose a change in current, is there some specific metal/element in a capacitor that does this?
Furthermore, I can see how a material can oppose a change, but only in terms of delaying. i.e. I can see a spark in voltage being delayed by some substance, but not sure about this capacitor-as-battery property.

About your other concern, there are other power sources connected to the microcontroller which have capacitors right at the source.

Here is the current setup so you don't get concerned, with capacitors at source. The tutorial is very thorough

>>2668878
>Why would a sudden load discharge the capacitor first?
>Simply bc of geographical proximity?
All wires are resistors. There's a lot less resistance between the MCU and the capacitor than between the MCU and the power supply, so current can much more easily flow from the capacitor. Also Q = C*V and Q = I*t, so if the MCU draws 500mA for 1µs, that's 500nC of charge. If you had a 10µF capacitor, its voltage would drop by no more than 0.05V, which is fine. If you had a 1µF capacitor, its voltage would drop by up to 0.5V, which is too high for comfort.

>Not too sure how an object can oppose a change in current, is there some specific metal/element in a capacitor that does this
It's a fundamental property of electromagnetism, and it's most significant in the wires in this example, not the capacitor. The wires exhibit inductance, which is the storage and release of energy inside the magnetic field surrounding the wires. The magnetic field strength is proportional to current. To increase inductance, you wrap wires in coils to make the magnetic fields of each wire stack, these are called solenoids or inductors. But more turns means more resistance, so you often want to fill the core of a solenoid with a material that stores more energy in its magnetic field, to get a higher inductance with lower resistance.

Capacitors exhibit capacitance, which is the storage and release of energy inside the electric field between two conductors. Electric field strength is proportional to voltage. To increase capacitance, you make the distance between the conductors as low as possible, and the mutual surface area as high as possible, in practice this means having two large plates close together, often wrapped up or stacked atop one another. But thinner gaps means lower voltage tolerance, so you need to put a material between them that is a good insulator. Different materials also store more or less energy in their electric fields.

>>2668878
>Simply bc of geographical proximity?

yeah.
if the chip creates a current spike, it's gonna pull that current from the closest source it can find (the cap) then from the supply if the cap has exhausted its reserve.
a cap is, among other things, a lot like a short-term rechargeable battery in that it stores a charge.

>>2668882

breadboard still looks weird, tho not necessarily wrong.
normally the pins would be inserted in columns D and G.

two of my transfer terminals accidentally got shorted together and the transformer got really hot. is it still okay to use?

>>2668878
>Why would a sudden load discharge the capacitor first?
>Simply bc of geographical proximity?
Yeah, but you have to understand why. Lemme explain it with picture.
Don't worry about this, just put capacitor in there as close as practical, and don't think about it, until you figure out Laplace, Fourier, diff. equations for various shit, Gibbs phenomenon, frequency domain, control theory, and other bullshit they teach you in uni that makes you want to jump of the bridge on railroad tracks or crash into WTC.
Hell, Im not even sure if im 100% correct with my explanation.

Lemme answer couple other questions you might have later:
>but why 100 nF instead of 10 000 uF
Well, that has to do with fact that capacitors that go to 10 000uF usually have pretty big ESR or/and pretty big inductance too. 100 nF ceramic ones act more like a real capacitor, which is why all of a sudden tiny 100nF capacitor makes bigger difference than big electrolytic.
>>2668898
eh probably, idk, check if it gets hot if you plug it without load

>>2668874
> You want the sampling frequency to be twice the maximum expected frequency.

At *least* twice
And that’s for a periodic waveform
That’s also a sine wave.

Note that you can use a bunch of tricks to deal with things, such as those found in a logic probe; and you can count pulses with bcd counters, lock out the other signal with a flip flop to see which came first, use comparators to ensure you’re at the right threshold voltages for fan-out problems, etc. These are tiny circuits you can build on a breadboard, or put them in a little enclosure if you use them enough.

Usually you can just downclock things though, then turn up the clock until it breaks.

>replace my mouse encoder for the first time
>feel bretty gud after it works flawlessly
>dies two days later
Fuckkkking chinkshit components. It took me an entire afternoon to fix it oh well I guess I can use more practice.

>>2668901
realistically you need to have 5x the sampling rate, or more.

I have this food vacuum sealer but sadly milk got sucked into it and now it doesn't seal as well as it used to, i opened it to clean it but looks pretty much clean on the inside which means the milk damage must be on the inside of the pump parts so i guess i'm fucked
i was thinking about pouring some isopropil and letting it suck it in to was out the milk, would that work?

>>2668953
Water. Use water.

>>2668901
>>2668934
>sine wave
>5x
For a logic analyser, on an MCU circuit where all time periods will be integer divisions of the main clock, it doesn't matter if it's just 2x. For an oscilloscope I'd agree, though the frequency rating in MHz for a scope would be the passband bandwidth anyhow, not the sample frequency. Judging from digital 100MHz scopes with 1GS/s sample rates, they use a full decade worth of anti-alias filtration.

>>2668929
Is that scopejak OC? I gotta make me a scope XY DAC array.

>>2668953
That looks like a membrane pump, so there's not really any risk of fluid ingress damaging the motor. Less so for the pressure sensor though. I'd detach the tubing from the sensor (flat black round thing on the board) and try to wash it out with washings of some 98%+ IPA or distilled water. Might want to look up what is safest to wash out a pressure sensor with. Maybe it's just a pressure switch of some sort, idk, hopefully there's a part number to find a datasheet with.
Putting distilled or even regular water though the rest of the system should be fine. I'd probably avoid soap or alcohol since there may be a sealing grease that it could dissolve. But chances are it's not a big problem, if the lethargy persists I'd give IPA a shot.

>>2668959
>. Less so for the pressure sensor though
thats exactly the problem
when i hit the button to seal food, it starts pumping the air but it never stops i have to stop it manually where before it could automatically detect when the air was sucked out so it's 100% consistent with the sensor being damaged, i will try squrting some technical alcohol into it, it's not like i can break it even more since it already don't work

>>2668962
Fair enough. I imagine any washing step will take multiple goes to flush out any contaminants. You may also be able to buy a replacement sensor for a few dollars.

>>2668767
you aren't worried about pets or small children?

>>2668966
how does the sensing part of pressure sensor work? like if the milk got in there how would it make it not work? clogg up something? could i stick wire in there maybe to poke through?

>>2668972
Really depends on how it's constructed. Worst case it's a bare MEMs silicon die, in which case it's fucked. Could be a rubber membrane with a strain gauge on it, which is probably pretty cleanable. Might even just be a rubber membrane pushing against a switch when it's distorted enough. Considering the shape I think it's likely to have a rubber membrane, so you should be in luck.

tempting as fuck
should i?

>>2668997
I think Miniware (the same people as the TS100) make a good electric screwdriver, not sure if I'd trust a $17 one. Look for reviews, either on youtube or r*ddit, or on /g/csg.

>>2668959
>scopejak
back in my day it was called "feels guy"

>>2667396
>Once China or Russia start (probably not anytime soon) to do various things in space
Russia just launched an epic mission to the Moon. It went as expected.

>>2669083
>Russia says Luna-25 spacecraft crashed into the moon's surface
I'm surprised they made it out of our gravity well.

>>2668997
No that one is weak AF and you can't manually turn anything tough.because it backerives the gears

This at least has a speed and lock on the tip and comes with.the Gucci metal case.

>>2668929
Yeah, in my mouse eventually both switches went after about 10 years.
Bought 8 identical omron replacements.
Didn’t think I needed 8 but oh well….
Now I need replace the switches every year.
IBM PS/2 ball mouse still going strong.

I need help guys
I got Yamaha CLP 535 piano and I want to fix touch sensors detecting maximum strenght sometimes
I need to remove keyboard first but I got that one connector I just can't unplug
It won't move no matter if I try to push from the sides or just grab and pull approach pic very much related

>>2669138
>It won't move

it looks like A is the hook keeping it in.
so slide a small screwdriver behind B, pull towards you, and pull off the wires.
prob easier if you do both sides at once, so grow an extra arm first.

>>2669157
it should be possible to just unplug it without detaching wires from what I'm seeing on yt
>https://youtu.be/fmB8D8k8e80?t=295
here he seems to just pull it and it works but hindis are electro warlocks so I should try a mortal way

>>2669161
>he seems to just pull it and it works

sure, after you do that a few times, the hook becomes rounded, and the tabs are permanently semi-bent, so you can just pull it off as you twist towards you.

>>2669165
I will try doing it your way tommorow
there is too dark now and I don't have good lighting
nevertheless thanks for all help

>>2669161
>without detaching wires

just realized i phrased it wrong.
i meant *pinch and tug* on the wires to remove the connector, not pull the wires off.
loosen one side, then finish it off on the other.

>>2669138
Looks like JST to me, you just pull those.

Need a solid state relay. A small one.

Problem: Cheap electromagnetic relay (Chinese) welds contacts occasionally and appliance stays on. House will burn down one day.

A better (but more expensive) solution would be a triac with a zero crossing opto isolated driver. 5V control from processor can then switch 120Vac, 10Amps. Problem: the space I have to work with is small. Attached pic is the electromagnetic relay form factor I have to work with. I know I can fit a triac and driver chip in there. But add a suitable heat sink and it's going to be tight.

Does anyone make small footprint SSRs to be pin replacable for such a old tech relay? Or am I asking too much?

Typical SSR finds with Google lead to 3x3 inch block units. Too big.

>>2669400
Maybe try an electromechanical relay with a big snubber network? Otherwise you could use a MOSFET or SCR inside a bridge rectifier, though you’ll need to find an isolated driver circuit for those.

Or just mount a big SSR somewhere else.

>>2669413
>Maybe try an electromechanical relay with a big snubber network?
Looked at that. But most of the snubber application notes deal with inductive kick on contact break. This is a resistive load (heater).

MOSFET or SCR will probably need a heat sink as well. Smaller? Larger? I'll have to work the numbers. TRIAC is sort of the go-to solution for AC load control (heart of most SSRs).

>using desoldering braid to remove solder
>removes the tinning from the iron tip instead

>>2669418
I see. Maybe putting a high-C low-R snubber in parallel with the switch contacts will work here? It's worse for the switch when it closes, but it should drop the effective voltage across the switch momentarily when it opens, possibly mitigating arcing. This method relies on the ESR and ESL of the wiring and load to drop the voltage. Simulate it I guess, see if the voltage across the switch contacts rises faster than the distance between the switch contacts.

>>2669400
I've exactly the same problem with AC. I have to give it a whack from time to time so it turns off.
SSR sucks as it has like 2 volts drop across the triac so it will produce 20W at 10A which is fucking a lot.
What you really want is triac that turns on the compressor, then relay that shunts the triac, so it doesn't weld itself, because main problem is not turn off, but turn on, which involves insanely high currents. For this you need to modify the circuit extensively which aint worth it.

If I were to simulate a homogenous solid of conductor (say copper) as a mesh of wires (section correlated to length to conserve volume) could I get a good finite-element simulation of the Joule heating of this solid ?

>>2669510
I guess that's the best way of doing it, assuming you don't have to worry about the skin effect.

>>2669445
just Whack It®

>>2662600
I want to make a seismograph, my first idea was using a raspberry but these days are pretty much expensive and (for me) overvaluated. I was thinking to use the rasp with something from the shake but I think they are astroturfed, (I read overthere they are shilled by the usgs).
I think it's very expensive and somehow caped, there is a version measuring acceleration in one axis only (DIY version of $354.99 in offer $229.99), that's pretty fucking and the response of the device and the response in amplitude/frequency it's ok, but for the price you would expect something with more precision.
Now, the really valuable component is the RGI-20DX geophone, but using mems accelerometers and arduinos could achieve similar results into early warning or quakes registers.
Also
How can filter/conditionate the device to avoid false positive (like people running or cars/trucks).
2nd also
I'm not geologist, or seismologist so could you suggest books in this field so I can develop something better.
Thanks anons, very appreciated.

bros the first stage of my coilgun burnt through after my mosfet board failed.

>>2669666
glad the new IO board survived but the mosfet board is ded.

>>2667163
This is so clean.
What do you do for work?

What are some good techniques to build cases and enclosures for your stuff? I'm inclined to use plywood, because it's easy to cut and cheap. But it got bad properties for electronic stuff, isolates heat, burns easily, and no EMC isolation.

>>2669718
> enclosure
Acrylic, polystyrene is pretty good.
There’s a bunch of people making enclosures from Schedule 80 pvc pipe, which is a good material. You could also use black ABS pipe which is pretty good, but I haven’t tried it. Most black electronic devices (e.g. TVs and stereos) are made of ABS.
PVC is more fire resistant, so use that if you’re worried.
John heisz just made an amplifier and made the back panel out of what looks like PVC.
Other good stuff for pros only in there too.
https://www.youtube.com/watch?v=7FRhn0OVD28

>>2669706
Professional poker player

>>2669706
I work retail at a Jaycar. The bastards keep on calling me in when I only nominally work 15 hours a week.

>>2669718
>>2669754
Sheet metal is ok to work with. Though I’d rather have antistatic plastic. A mildly conductive PLA filament would be pretty nice for that, but none of the conductive filaments are cheap.

Personally I’m planning a rack of drawers to store large amounts of THT ICs in, 3D printed but with 3mm acrylic walls. Not really sure the best way to connect the acrylic edges together, glue feels like the only sensible option if i can’t screw them into solid corner pieces of wood or whatever, which I guess is a worthwhile design change.

>>2667163
>got the ksger with the nice tight handle for upclose grip
>it doesn't fit the stand

My brother.

My truck has variable-color accent lighting that I want to tie into to control led accent lighting in my subwoofer box.
I bought some LED strip amplifiers that I tied into the RGB wires in the truck, but they only stay on for a minute before my trucks shuts the lights off and they will only work intermittently for a few days afterwards before they come back on normally.
I assume the truck is detecting that the current consumption is out of range and shuts them off for protection.
The amplifiers are an optocoupler with a 7.5k ohm resistor and the output of the optocoupler goes to a mosfet.
The factory LED has a 5.6k ohm resistor for each color.

What circuit would likely give me the most success at successfully getting a signal from the factory LEDS?
I was thinking a transistor would not draw much current. Or a mosfet, which is already controlled by voltage.
Should I tap into the factory LED AFTER the resistors? Right now they are tapped in before the resistors.

>>2669887
I just considered that maybe I should have used the factory LED's ground as well instead of grounding to chassis. What are the odds that the BCM is looking at the return current and comparing it to the output current as a method of fault detection? Like a tiny GFCI

>>2669887
>My truck has variable-color accent lighting

kudos for representing the Pride, bae.

>>2669881
No it does fit the stand. The diameter is maybe 3mm too small for a snug fit, which is a tad looser than I’d like, but it works fine. The crab holder was just for the photo-op.

>>2669887
It’s tripping from the LED signal and not the LEDs themselves? MOSFETs or BJTs should work but they’ll invert the signal. An emitter follower should work, give it a go if inversion is a problem. But first try swapping the ground wires.

This is the turbo panel of some game controller, but...

The tiny square LED indicators on it, what would this part be called? I want to know how to search for it or similar tiny LEDs.

>>2670024
Hard to say without seeing what's actually connected to the circuit board. Those might just be light-pipes or even just clear windows for conventional LEDs. Be they common 3mm through-hole LEDs, or 1206, 0805, etc. surface-mount LEDs. I have seen some rectangular package LEDs (I think they're often called "2mm LEDs" but never square ones like that, so I suspect what you see is not the top of the LED itself.

I have this detectors at work used to enable and disable the AC
how can I make it to stay always on like there's someone detected
only first 4 wires are connected
there's always continuity between pin 3 and 4 so I don't know how exactly the alarm is triggered

>>2670044
Lol, been there fren.
Just rig up a piece of cardboard attached to a rope near the sensor. Just pull the rope every now and then.

>>2670045
>Just rig up a piece of cardboard attached to a rope near the sensor. Just pull the rope every now and then.
it's too far away and I have to go there every 3 minutes or so
I can't do any work if I'm constantly moving around

>>2670044
I think the "alarm" pins (3) and (4) are being used to turn on the AC unit. In which case, I'm pretty sure you need to short them together. I don't think doing this while it's connected to the sensor will damage it, but I can't say that for sure. The safer method would be to undo the screw-terminal or whatever for those two wires and just tape or wire-nut them together.

If this doesn't work, you may need to send a positive signal to one of the wires instead, or maybe even leave them disconnected. Post a pic of the PCB in there, reading the main IC and interpreting any transistors or relays on the board will tell us what that output pinout is going to be like. I had one lying about somewhere but I can't find it for now. I'll be able to check the internals of one at work tomorrow if you'd like. From memory they're just an open-collector output, but I can't say that for certain.

>>2670059
>If this doesn't work
it doesn't sadly
>or maybe even leave them disconnected
will try

>>2670061
Any transistors on the board? D6 is a diode, it must be something with a Q (or maybe T) designator. That metal can might be a relay that could be shorting the pins together when there ISN'T any motion detected, which would be failsafe. If you have a multimeter, you could measure voltage between the pins both when there is and isn't motion detected. You can also measure voltage across the screw terminals and across the wires when they aren't plugged in, to figure out what the default behaviour is.
If it's a relay, I'd expect no less than 4 THT pins coming from beneath it, maybe up to 8. 10+ is a faint possibility. I'd also expect a reference designator like RL, which I think I might be able to see up near that microswitch. If you can pull the sticker off we'd know for sure, but maybe leaving evidence of tampering isn't the best.

The diagram shows something in series with the switch, maybe it's that resistor above the metal can, and the other end detects that it's been shorted with a certain series resistance. Though that wouldn't be failsafe. If you can read a value on the plastic cover, or otherwise read the colour bands on the component on the board, you should be able to find what resistance it should be and add one between the wires.

There's still the possibility that it switches some voltage to the output wires, maybe that's the input +12V, or maybe it's whatever is coming from the onboard regulator that I assume is IC1. Maybe that rectangular symbol on the diagram is a voltage source.

Also see if you can read that IC number, it's too faded from your image but if you can shine a light on it from an oblique edge it should become pretty visible.

>>2670068
thank you very much Anon, I'm saving this post for later
I disconnected the 4º pin and it seems to work
I'll give it some time just to be sure

>>2662753
Keep it, looks pretty nice and you could use it for a variable power supply or just gift it to some nephew into electronics.

>>2662792
Keep it but replace the power supply with a modern one and connect up the voltmeter and ampmeter. The old PSU can be donated to someone or keep it for spare parts, ESPECIALLY THE TOROIDAL TRANSFORMER!!!

>>2662800
>no ripple or noise at the output
There will always be, but very small compared to your average switch-mode PSU. The big tradeoff is that you're gonna waste a lot of power in that regulator unless you use a design with a tapped transformer and thyristor to switch the windings.

>>2669860
>Not really sure the best way to connect the acrylic edges together,
UV cured epoxy

I will never understand transistors.

>>2670089
Nobody does. You only need to understand them approximately. Engineering is not about understanding things perfectly, it's about making and tweaking things until they work. Don't overdose on math/physics, it will only make you understand things even less.

>>2670089
Ic = (1+beta) Ib
Vbc=0.7v

>>2670149
Ie*

>>2670025
I see. Thank you anon.

I want to make a similar panel (3D printed) but I'm unsure what to get. Ill look for tiny 2mm LED. I still don't know thr most elegant way to mount then to a panel and lead wire to the PCB. I guess just gluing then into little holes works if nothing else.

>>2670024
>The tiny square LED indicators on it, what would this part be called?

first guess would be ''square LEDs''
second guess wold be square light diffusers.

>>2670145
>>2670149
What is the Vce saturation voltage? And why would I care about it?
I tried reading up some tutorial on analog. com but it talked about the ce junction going from forward biased to reverse biased or something...

>>2670218
>What is the Vce saturation voltage?
The voltage between the collector and emitter terminals under conditions of base current or base-emitter voltage beyond which the collector current remains essentially constant as the base current or voltage is increased. (Ref. IEC 747‑7.)
>And why would I care about it?
This limits usable output voltage range if you're building an amplifier.

>>2670227
So if Vce saturation voltage is 20v and the voltage between collector and emitter is 24V, base and collector current dependencies are no longer linear?

>>2670242
No in this case 24V of c-e voltage is greater than saturation of 20V (real transistors have 0.1-0.2V). This means you're in active region and i_c = beta * i_b holds.

Base-collector dependence breaks when change of base current doesn't produce change of collector current i.e. base is saturated and no matter how much current you push into it collector current will remain same. It's best to visualize using pictures. Notice how when not in saturation you put in 50uA and get 14mA increase. If you keep adding base current you get less and less increase. Saturation voltage is defined as voltage when collector-base PN junction gets forward biased. In case of NPN this means that collector must be below base voltage.

>>2670261
>In case of NPN this means that collector must be below base voltage.
Ah yes I can see that on the diagram but why does the collector base pn junction change polarity...?

>>2670285
Base voltage is held at reasonably constant value of ~0.7V. If transistor wants to pull some current through resistor via collector it can only do that by changing collector voltage (base is ~0.7V, emitter is fixed at 0V GND). For small currents it only needs to lower collector a little bit, voltage on collector is much greater than on base so you have V_bc = (0.7V - some large value) = some negative value < 0. Now if you keep increasing current, collector will have to go lower and lower and at some point it will go below base voltage and base-collector voltage polarity flips V_bc = (0.7V - 0.1V) = 0.6V > 0. This is now saturation, both BC and BE junctions are turned on.

CS studenet here
>4th out of 5 year of college.
>Liked college really much in the beggining when there were generic electronics/electricity classes, even if they were harder than coding classes.
>feel bored by coding
>try to do software projects in free time
>get bored and do other things
>also read/watch shit about electricity and electronics
>for college do hardware arduino projects instead of software when I can
>still cope about being able to get an ez coding job
>ai arrive and will soon axe half of software dev jobs
>oh no
How viable would it be for me to pivot my free time into studying electronics and soldering appliances first, then fixing computers and complex stuff?
In my final year plan on halfassing pure software classes and focus everything on classes that also teaches about hardware
What are some good Master thesis where i can best hone my soldering and electronics skills? Something like fixing computers and installing Linux OS or some other project for masters

>>2670298
>Something like fixing computers and installing Linux OS or some other project for masters
That's more of a high school class project tier level of difficulty. Master's is generally geared toward solving advanced problems. It would sure be convenient if you could just install gentoo for your master's, maybe do some ricing and publish your dwm config. But in real world you're looking at something like writing an init system, driver for some PCIe FPGA board that processes 32 channels of data from your physics department's particle accelerator or even solve a practical problem for some company during your internship.

new thread
>>2670315

>>2670313
Fuck i overrestimsted myself about masters
I am stupid
How about rest of my post

>>2670319
It's perfectly viable to focus on embedded and real-time systems as those things have a lot of overlap with CS. If you want to do straight analog electronics then you would have to switch to EE or finish CS and then do EE so you have 2 degrees. For now it's better to focus on CS and see if you can do something in embedded space or in signal processing.

But be aware that grass always seems greener on the other side... Dealing with actual hardware for your master's is not as fun as one would imagine. You have to deal with parts shortages, workarounds because parts are not in stock, waiting for chink PCB manufacturers, you have to solder and debug circuits... Meanwhile in CS you just coode and all you need is a computer + compile times are measured in seconds. "Compiling" hardware takes 14 days and causes neck pain because you have to manually painstakingly place resistors that are supplied in bags with retarded labeling schemes (LCSC) that don't even list resistance values so you have to guess from part numbers. Putting actual effort into your thesis or any college work is a waste of time. Your goal should be finishing as soon as possible so you get more time for your hobbies.

>>2670329
Fuck. I guess i needed some sense slapped into me.
Thanks.
I should've taken a gap year after high school to rethink my life
Even in high school i would read news about lack of trade in my economy and my dad would be like
"you're dissonorabu my family" whenever i brought up going into trade college.
I also didn't focus on coding enough and prefered maths, doing manually labour, exercising and cooming (my bad)
I guess with AI i am cut off from entry level jobs in CS so I would probably have to do some were house wagie job or learn electronics/electric installation since i have most knowledge about it and fix shit to become a tradie

>>2670338
AI is a meme. If anything it will create even more CS jobs because everything will get even more broken than it already is. It's just too unpredictable for engineering purposes and can't even do basic addition properly. What AI is good at now is generating bullshit. Only jobs that will get replaced is jobs that generate/build/sell bullshit as now they'll be able to do it more efficiently.

>>2670346
I guess
How about a job where i am field technician for people's internet/phone lines?
That seems fine to me.
I don't mind heat (+35 C is my limit but would survive higher if needed) or cold, that is my greatest strength

>>2670358
Why not be a lineman for a power utility company?

>>2670363
I need to think about my future more. I will hang arround /g/ and /diy/ to see it

>>2670294
That makes more sense
Thank you

>>2670338
Learn about AI, dummy. Start fucking around with tensorflow and pytorch.

>>2669963
>It’s tripping from the LED signal and not the LEDs themselves?
It trips when I attach the LED strip amplifier, which has higher-value resistors than the factory LED. It does it with or without the factory LED connected.
I can only assume that the LED amplifier is putting too low of a load when it is connected by itself, and too much of a load when connected in tandem.
That is if it isn't the ground issue, which will be easy enough to try.

The emitter follower looks like it would be an ideal solution

>>2669954
Don't break out the anal-ease just yet, homo, it is only one color at a time

>>2670316
way too early dude

>too late on /mcg/
fug

How many days until this thread gets archived?

new /mcg/ thread for those who aren't refreshing the catalogue for it:
>>2670671

>>2670673
2-5 i think

hello, i have a little issue with a 2n3906 pnp transistor bought on aliexpress (i bought 100 + 100 2n3904), it appears that the emitter and collector are flipped, why ? it's not a huge issue I just have to flip it but it feels wrong to have the collector at pin 1 and the emitter at pin 3

>>2670791
It's probably fake. Try to measure its hFE to see if it's within spec. Hard-mode: measure its gain bandwidth product.

How do you guys look at discrete electronics and not puke at how ugly circuitry is?

Autism?
The big picture?

>>2670338
>I guess i needed some sense slapped into me.
Yes you do. If you really think AI is going to take the place of programmers then you need to rethink your whole mindset.

>>2671157
99% of Programmers are already.useless sacks of shit. It's just like how all the writers went to shit and no one even noticed they were replaced by Indians, and chatgpt is actually an improvement on that anyway.

>>2671215
Writing fiction is easy to replicate by AI because there is always a formula to stories. Programming is not because there is an infinite depth of logic and variables to consider.
Sure, AI may be able to create simple functions and some logic here and there, but it's impossible to create large programs that just work. It's akin to illustrations; you may get close to what you want but there is no way (or a much harder way) to specify what exactly you want.
Plus, what if a new logic must be implemented that the AI has not been trained on? That ends everything.
Does nobody on this site know how to think for themselves? Like holy fuck, you're just going to give up because of some new technology based on some reason that has vaguely been implemented? It's like the scare that robots would take over manufacturing jobs. "OhHHhHHh, the robots are going to replace everything!!!" Well, have they?! No! Some jobs that would have theoretically been replaced have not because in testing they were found not to be feasible. Others have indeed been replaced but the amount is really overblown by fears. Additionally new jobs will be created to maintain those robots.
It's the same with AI, and the thousands of other new, efficient technology that appeared through history. I'm writing this for myself, but honestly I'm tired of this site. Every time I find something cool half the thread is filled with retards like you with braindead takes. Fuck you. Go to fucking college, get a fucking job where you don't cuck yourself, and stop swimming in this lake of piss with all these other retards.

>>2671255
stfu windbag. if this post isn't AI-generated i'd feel sorry for you
and programming is formulaic as fuck

>>2670919
yeah i might try the gbwp, i still really don't understand why it works

>>2671276
>i still really don't understand why it works

because if you flip and NPN, it stays an NPN.
The E and C junctions are doped differently, but not so differently that they stop doing the job.
just not as well.
same way when you flip your GF over, an orifice still presents itself, not as good, but usable.

Ok page 10 now, move to the new thread:
>>2670315
>>2670315
>>2670315
>>2670315
>>2670315

>>2671511
>"new" thread is 2 days old
It's retarded.

>>2671517
Idk I didn't make the thread. I would have made it at page 10.

still poastin

>>2671773
nigger