/diy/ - Do It Yourself

a textbook I'm reading states that this configuration of npn transistor (where the base is connected "to the connector" (ahead of the resistor from vcc) implements negative feedback.

How is this any different than just hooking it up to vcc (behind the resistor)

>>1682640
>to the connector
to the collector

>>1682640

the more current passes thru the transistor, the lower the collector voltage drops, which cause the base voltage to drop, which acts to shut down the transistor. that's feedback: an action causes an opposite reaction.

if you just connect to Vcc, it's fixed so nothing changes.

>>1682642
alright that makes a lot more sense. I'm familiar with what feedback is, I just didn't see how it applied in this sense. Thanks for explaining.

back again from the last thread with new questions.
I bought a bunch of capacitors thinking some were bulging/eroded, but at the end, made no difference, so now im back to focusing on the transistors. (aka it will be 5 days before digikey gets me the new transistors so I wont be able to do much until then but take values)

here: again TR931 and 924, according to the schematic numbers, the voltage to the solenoid should drop to 0 on playback (P) mode. at TR931 Collector. It is currently stays at 24.5V on playback meaning the solenoid is NOT engaging (bad).

compared to TR912, which correctly shows 24.5V at C on standby and 0 on playback, and this solenoid does work (I can physically see it activate)
is it as simple as just needing a new transistor? I was told in last thread 2S2222A could subsitute for a 2SC1384 (even those the casing is different..)

>>1682659
*2n2222a sorry

>>1682659
N, not S. 2N2222A
>as simple as
it's never that simple, lel
you gotta trace it back further until you find the part that isn't producing the right output for its inputs, so to speak. so check the bases of TR931, TR924, and on back. wherever the collector of TR924 goes south of your pic, look at that too, somewhere there will probably be a transistor that has base drive but is open at the collector

>>1682663
fuck. i'll take some measurements and report back in a bit.

>>1682553
>ready to solder
>with no solder mask
>with that tiny trace-to-trace clearance
Better go a little heavy on the flux to avoid bridging.

>>1682667
you better have a scope

ALL FEAR THE REGULATED WALL CUBE

>>1682671
>flux
can you quantify that on the Rossmann scale?
>scope
y tho, it's discrete RTL. he's not even to the good part yet

>>1682671
a scope would be useful but right now i just want the mechanisms working again. I havent touched any pots. Actuallying getting any kind of video image is the next step if i ever get that far.
anyways here's what i found out:
the C leg starting from TR922 out is bad, TR921 may be the culprit, a 2SC828. which would explain both SL1 and SL3 failing to activate.

>>1682692
so here's what we got. all the transistors give a bad C leg reading except for TR925 which is faulty on all three legs in Play mode.

>>1682680
>y tho
Testing parts in circuit. If the signals are changing faster than a couple of times per second a DMM won't work for shit.

I downloaded all the books but I still don't know how to make electronic noise makers out of garbage. I might read them a little tomorrow and on monday I might post some pics of tvs and alarm clocks and whatever else I could find to possibly make electronic noise makers with. anyone have ideas about this?

Guys it works, thank you so much for all your help. With the spark gap set at .008" like the manual says and the series resistor set to its lowest, the arc at the electrode will start at a gap of about 1/2" and I can pull it out to about 1 1/4" before it goes out which is plenty good enough for my needs. I ran continuous HF with the torch clamped 1/4" above the workpiece for about 5 mins and the capacitors never even got above room temp. Tomorrow, I may open up the spark gap (assuming this would increase the size of the gap the arc will jump while decreasing the frequency and increase the stress on the capacitors) and see where they start warming up and hopefully get someone to take an arc pic. Thanks again and know that you all were by far the most helpful and open minded people I discussed this with.

>>1682759
Great news! Now do some measurements and figure out what's making the biggest impact to your Q factor so you can get that baby to 6"! Probably the transformer though. You could maybe try removing one of the parallel capacitors and seeing if it makes things better or worse. Tuning the spark gap distance may also make a difference.

>>1682759
don't test it to death
you have it working, use it

>>1682767
>do some measurements
>Q factor
I have no idea what to measure or how to measure Q-factor, but from the 10-15 year old welding forum posts I've read, it seems like people open up the spark gap to get a more 'aggressive' start and somewhere I saw that you could get a 2-3' arc presumably only by opening up the spark gap. Not sure if those guys immediately fried their capacitors or not.
>removing one of the parallel capacitors
In one of the manuals I think it says if one of the capacitors fails, you can still use HF, but they recommend decreasing the spark gap to avoid putting excess stress on the remaining cap... I have no idea what the implications of that are.
To be honest, I still don't really understand how the circuit works. I get that the spark gap and the capacitor form some kind of oscillator, but the only 'oscillating circuit' I ever learned about was the one with a voltage source, an inductor and a capacitor all connected in series. What really gives me a mind-fuck is how the spark gap is connected in parallel with the capacitors, R5 and T3. I don't see how it does anything at all. Like if the electrode is too far from anything to start an arc, does the voltage on the right side of T4 just jump sparks across G, that's why I felt the need to actually have an HF arc going for 5 mins to 'stress test' things, but I have no idea if that test was actually valid...

>>1682773
I might go that route... thanks for the reassurance dad.
also sorry for the redundant pic, I get a connection error unless I attach a pic.

>>1682774
The Q factor is basically a measure of the relative impedances of the reactive components compared to the resistive ones. More reactance (capacitive + inductive) means higher Q. More resistance means lower Q. Note that in practice the reactances of the inductor and capacitor are equal and opposite and don't limit current, but the resistance doesn't do that. Now that I look at it, R5 will be the primary limiter of your Q, which is probably a good thing as it's a known resistance that will limit the maximum voltage (and current).
I'd agree that opening up the spark gap seems like the way you'd increase the voltage. What's the total distance (A-B and B-C) of the spark gap at the moment? With that I'd be able to estimate the breakdown voltage of it with the dielectric strength of air, and see how close you are to the capacitor's maximum voltage. I'd want to stay somewhat below (1/2?) of the capacitor's voltage rating, for safety reasons, but if you're only running it at 1/10th the voltage I'd say bump it up a little.

Anyone know why the fuck my eurorack started smoking? The only difference from usual was that I had unknowingly turned on the phantom power switch before turning on the normal one. The smoke was funneling out of either line 5, line 7, or one of the returns above those (can't remember). I had a guitar plugged into line 1 and a pedal loop into send/return with the rack connected to the computer.

>>1682777
it's because it's covered in pubes

>>1682779
it's european, what do you expect?

>>1682776
Thanks for the Q-factor explanation, but my mind turns to sludge every time I try to wrap it around reactance, I think I have a good grasp of DC circuits and straight resistance however.
>What's the total distance (A-B and B-C) of the spark gap at the moment?
Do you mean what's the physical measurement across the two spark gaps? If so, they are both set to .008 inches. While you're at it, do you know why there are two spark gaps connected in series instead of just one? I assume they had a good reason. The original factory capacitors were rated for 5KV 3A at 1 MHZ and the new ones are rated at 10KV 10A at unspecified frequency. Unfortunately I don't know the actual voltage nor frequency the capacitors see out of the right side of T4, all I know is that the input (left) side of T4 is nominally 115VAC (I think).
Pic related is the specs on the new caps, only because I can't post without one.

>>1682779
Didn't even notice them, it was probably there when I got the thing I've only needed line 1

>>1682722
I even did a test by replacing those two (TR925 and TR921) with good transistors from somewhere else in the circuit and the same problems still happened.. weird

>>1682782
>my mind turns to sludge every time I try to wrap it around reactance
You have to be both fairly comfortable with complex numbers and with the calculus behind how inductors and capacitors react to arbitrary conditions. It takes time, but I think there are some other explanations of Q factor that might be a little more universal. Considering the case of a mass bouncing on a spring with some form of damping, its oscillations will die out over time. That's basically a mechanical version of an LCR oscillator, though in your case it's driven by a spark-gap circuit.

>0.008"
Really? 0.2mm each? That would correspond to a total breakdown voltage of about 1.2kV, which sounds pretty low, you should have plenty of room to increase this distance at least three-fold. But note it will increase the thermal load on the rest of the circuit. If nothing is getting toasty as-is (R5, T4, T3, G, C2) then it should be safe to go for. On that note, I suspect heat is the reason they have two spark gaps, such that the thermal load is more spread out. Also consider covering up the new capacitors such that the UV from the spark gap doesn't attack their plastic outsides or the zip ties.

>>1682786
>other explanations of Q factor
>basically a mechanical version of an LCR oscillator, though in your case it's driven by a spark-gap circuit
Thanks, any tips on finding resources/explanations on these? The only thing I found on spark gap oscillators are the old spark gap telegraph transmitters.
>Really? 0.2mm each? That would correspond to a total breakdown voltage of about 1.2kV
Yes, 0.2mm each. Sorry for stupid question, but in practice, how does one actually measure HV capacitative circuit voltages? Alls I have is a 600V multimeter. Seems weird if this is all done theoretically.
The way it is set up now, none of the components listed in the diagram is getting above room temp with R5 set to the minimum resistance. Just based on experience with other stuff, it seems like R5 should be the changeable value that increases or decreases the temps and stress on the caps. Even with R5 set to minimum it seems like the caps hardly register a difference. Also, I kind of skipped the detail earlier where when R5 was turned up to max resistance or even 50% resistance, I'd have to scratch start the electrode and weld for at least a few seconds before it would HF start and jump a gap and then if I stopped and let it sit for a while, it would only start if I tapped the electrode against the workpiece with everything off and no current flowing before it would jump an air gap at the electrode, so it would be nice to be able to turn it up a little bit I think. Brb I will check the min and max resistance of R5 assuming that would be useful to your analysis.
Again pic required, R5 for your viewing pleasure.

>>1682794
Again I can't get a good reading because of the many layers of animal shit embedded onto the contacts, but the manual describes R5 as a "150W, 5 ohm rheostat". The most confident measurements I've gotten off of a couple days ago are 6-7 ohms with the dial set to zero (weakest HF arc) and 3 ohms with the dial set to 100% (strongest HF arc). Speaking of, how should I clean this thing - meaning the whole machine? A lot of guys who refurbish welders say to hose it out with a literal garden hose and wait a week or so for everything to dry before firing it up. My plan was just compressed air and CRC contact cleaner. It seems to be working the way it is but all that corrosive animal urine / shit can't be good to keep in there for the next few decades. WTF gay pic.

>>1682794
Technically you could use a voltage divider, perhaps 10MΩ-100kΩ to get a readable value with your DMM (though it could require some compensation for your DMM's input impedance), but I'd be pretty hesitant unless you have an old moving-coil meter. If only because the high-frequency garbage from that spark gap will have a bit of a wireless interference range to them. Often with HV circuits even a 10M resistance could load your circuit down, but judging by the girth on that rheostat that isn't an issue you'll have to worry about. Judging by the sound of that poor low-range HF performance, I'd definitely at least double the spark-gap distance if you can. If only because your capacitor voltage is double what the old one was. If you do measure the voltage with a divider to see how safe it would be to increase the spark gap, a few clipping diodes in series to ground could stop the voltage from exceeding the forward voltage of the diodes and more reliably protect the DMM, something with a forward voltage of less than 600V like a 1N4004 would likely work fine. I don't think HF transients would get past the diode.

>>1682812
5Ω sounds reasonable. For cleaning, I'd try compressed air to get rid of anything that isn't stuck down, and some sort of light solvent + scrubbing with a small brush to get what is stuck down. Isopropyl alcohol should be a reasonable choice. Just hosing it all down could work quite well, but if you aren't in a warm dry climate I'd second-guess that thought, plus if there's anything stuck down the hose alone probably won't get that for you. But I have little experience with cleaning anything other than PCBs. IIRC the Post Apocalyptic Inventor has a few videos about refurbishing old welders, he tends to soak things in soapy water but I'm unsure if he does that for things like transformers.

Try posting with the original form instead of the quick-reply.

>>1682774
spark gaps close their circuit above a minimum voltage (tick on V axis) and stay closed until below a minimum current (tick on I axis), conversely to the SCR/triac
the power for the oscillator is being provided by the primary of T4, which is almost certainly mains ac, which is far enough below HF we can pretend it's dc for study, kindergarten show-and-tell related. of course, inductors (and transformer windings, being inductors) resist changes in current
to start, current through T4 primary induces current through its secondary. voltage builds up across C2 at the rate of 1/C = 250V/A·µs (red) until G's breakdown voltage, which other anon estimated ~1.2kV so about 5µS. G strikes an arc and suddenly becomes a resistance on the order of a few ohms. the capacitor discharges very quickly through R5/T3/G (green), taking all the T4 secondary current along with it btw, until the current drops below the holding current and no longer suffices to keep the air in the gap ionized. current then falls off abruptly (blue) and the cycle repeats. now, while all this is going on, that sneaky little shit T3 couples the current going thru C2 into the electrode where we want it

>>1682777
>behringer
found the problem

>>1682780
haven't they gotten over that lately?

>>1682812
fuck yeah, cool that it works
garden hose might be more effective but it's your time you're spending. wrap the trannies and your new caps in some thick stretch wrap first to keep water well off of them

>>1682614
I haven't tried making any double sided pcbs yet, the one in the pic is the first attempt at milling a pcb made
For alignment there are pretty much just two good options in my opinion, either take a piece of wood a mill the exact shape of the pcb into it and then put the pcb in, so when you flip it it is still perfectly aligned, but that takes a lot of time and wood and wood dust, it makes sense if you need to make at least say 10 of those pcbs (but if you need to make more then one prototype, you would just let chinks do it for $5 and send it to you all masked and tinned and all) milling pcbs nowadays is good for fast prototyping, or when you need just one board for your project

The other option is to drill holes into the pcb and then put matcing screws into a wooden board and then align the board by simply screwing it into the holes

i guess you could also make something like a clamp setup that can be rotated on one axis and then after you mill the pcb you rotate it without losing position

If I wanted to turn on a whole bunch of multiple relays at once from a microcontroller pin, is it okay to use a single transistor to drive all of them (assuming the transistor can handle the current needed for all the coils)?
Something like pic related

>>1682913
Common-emitter is the way to go friend, put that transistor on the low-side. Also read the datasheet. Minimum beta / hFE for the transistor and 3*maximum relay current will determine your base resistor, maximum current for the transistor (may need to look at heat dissipation values) will determine if it's feasible at all. It probably is, but your MCU/other circuit might not like driving up-to 30mA too much. A MOSFET would be a better choice to avoid having a high base current, or maybe a darlington if you're one of those kinds of people.

>>1682885
anything thin and hard with an inside right angle should be enough of a jig. screw it onto the table and you'll always know where zero is. two scrap pc boards are good for making an inside right angle, two more can serve as the movable jaws of a vise. should be pretty easy to do a rotation transform on the coordinates in your file, in case you can't be fucked to square the angle with the x/y axes (I mean, you could let your CNC drill for the jig to ensure it's straight... or just mill off the edges of the jig)
you could also place some pins in the table which you can remove when doing other work. you could also use these to secure a square as above

>>1682913
never put loads on the emitter or you'll pull your hair out wondering why it works so poorly. pic highly related

>>1682915
>>1682916
Cheers lads.
I'm actually after finding a bunch of low power DPDT relays in my cupboard so I'll only need to drive 2 coils from each transistor

>>1682915
In sat mode a transistor has no "minimum beta", you set the B yourself.

want to get temperature readings from multiple locations in my house.
I can get 5 cheap digital thermometers from BangGood for $9, but I'm wondering if there's any smarter /diy/ way to go about this.
could I get many cheap sensors and components for about the same price and make my own?

>>1683006
you can't beat china at hobbyist diy quantities. you're looking at $3 s&h min for the circuitboards and $1 each for the displays which already loses to china before you even consider the rest of the components: the uC, battery clip, temperature sensor, and passives

>>1683020
alright, thanks. that's what I thought.

>>1681861
>using an inverter when 90% of household electronics run just fine on 100VDC
it's a little late to answer this BUT I WILL

the inverter has only 1 or 2 functions: running the gas furnace in the winter, and running the fridge in the summer. it doesn't run the fridge yet because the inrush current is too high.
Everything else important in my place runs on a 12 volt bus derived directly from the 16v battery voltage. Then 24v and 5v buses are derived from that 12v bus.

and, running shit on DC for long term will overheat the input rectifier because 2 of its diodes will be working twice as hard as they should be. Not to say it doesn't work though. <--
and besides that, a DC-DC converter to step 16v up to >100v isn't something I could find easy and cheap.

>>1682881
>behringer
>found the problem
Ok but do I need to worry about the thing blowing up or just don't turn on the phantom again?

>>1683072
#yolo

>>1683006
adjust your vents dingus

>>1683047
>that pic
What the fuck. That shit is legit dangerous. DC is MUCH MUCH deadlier than AC, even with weak 0.2A or whatever this shit is capable of producing.

>>1683084
>That shit is legit dangerous.
simply don't introduce yourself or anything combustible into the circuit

>>1683091
over-protected children are different

>>1683121
I hear they improve stability when introduced into the circuit

>>1683084
it'll do more than 200mA for sure.
plenty of things in this world are legit dangerous, but this one is safe as long as you resist the urge to lick it.
why do you think DC is more dangerous?

>>1683144
because Thomas Edison electrocuted a dog to prove it

>>1683047
https://www.youtube.com/watch?v=8hwLHdBTQ7s

>>1683150
>Thomas Edison electrocuted a dog to prove it
He electrocuted an elephant with AC to prove it was more dangerous than DC.

>>1682844
>use a voltage divider
Ok, that makes sense. I think I am going to keep winging it and if I open up the spark gap, keep an eye on the temperature as a proxy for how much stress im putting on the caps. I didn't get a chance to use it today.
>Isopropyl alcohol
Thanks, I saw that recommended somewhere else, can I use denatured alcohol you think or just get the medical isopropyl from the grocery store?
>>1682881
>to start,
Thanks for that explanation and graph, I think it's starting to make more sense. So is the spark gap distance essentially the 'adjustment' for the voltage (and frequency) of the HF arc? Like say the transformer T4 is capable of 5kV, the way the spark gap is set now, the voltage is only going to get up to 1.2kV before it fires a spark and the capacitor has to charge back up again?
> transformer windings, being inductors
There's the inductor I was missing in trying to compare it to an LC circuit.
>fuck yeah, cool that it works
Thanks man, all you guys helped a lot, someone here suggested the doorknobs in the first place, and everyone else convinced me to go for it and why it should be fine.

>>1682622
hey! you made me double check and not only had i wired the speaker between everything and the ground, i had also switched out what i intended to be the normal connection and the one going to the jack. so fuck me i guess.

>>1683204
91% or better IPA from the first aid section is good, 70% has a lot of water in it, which isn't helpful
>spark gap distance essentially the 'adjustment' for the voltage
yes. don't push this too far, air conditions vary
>and frequency
it's complicated. T3+T4 and C2 set the resonant frequency by 1/(2*pi*sqrt(L*C)). G will increase the frequency by pulling back on the LC circuit when the voltage gets too high, similar to how plucking a string very hard increases the pitch noticeably during the attack of the note, which settles down to its normal resonant frequency after dissipating excess energy. so, it will behave differently as the phase of the mains power wave proceeds

I'll be moving in a few months so I can have more space, but this will do for now. Finally happy to have at least a semi-decent work station, plus here's hoping this power supply actually works.

>>1683144
It's not an issue at that voltage, but there are two main risk factors with DC, both related to the fact that it's constant rather than intermittent.

One is that you're more likely to die from touching a live contact; with AC, you can usually pull your hand away, and that's a reflexive response, whereas DC paralyses the muscles preventing you from letting go of the source of the shock.

The other is that arcing is more pronounced. Once the air becomes ionised, with DC it stays ionised; with AC, the arc has to be recreated after each zero crossing, which requires a higher field strength than maintaining an existing arc through already-ionised air. This is why switches and relays rated for 240V AC are typically only rated for 50V DC.

An additional (usually non-safety) factor with DC is galvanic corrosion.

>>1683270
>at that voltage
amperage*?

this is the lm358 textbook VCO diagram. The non-inverting input of the second op amp calls for .5VCC.
I plan on using a voltage regulator for VCC of course. However, going off of Ohm's Law, voltage fluctuates with current. How am I supposed to linearly halve the value of VCC when a plain resistor wouldn't do?

>>1682638
im getting more into ham and electronics, and have started shopping around for equipment on ebay to build out my electronics/rf work lab better. thinking of getting back into tek tm500 series again. looking at aston linear PSUs, taking products into consideration that eevblog mentioned in home lab vids.

post your or someone elses electronics lab for inspiration.

also shill test equipment you think i should consider.

>>1683302
Use this version.
>>1680989

>>1683302
I don't understand why a resistor divider won't work here. Your voltage regulator providing your VCC will spit out a constant voltage regardless of how much current it is drawing. That's what it's designed to do. It doesn't behave according to Ohm's law.

The only trap you need to be wary of with this circuit is the load the resistor in the amplifier feedback network create. Use low value resistors for your divider here. 10k. Do not use high value resistors like 1M.

>>1683305
Got it. I'm not the best with op-amp theory yet. Does the varying voltage on the noninverting input actually serve a purpose? Or is it just to bias the op-amp

>>1683312
It's a (1/3) -- (2/3) Schmitt trigger.

>>1683304

>>1683304
You're gonna need a spectrum analyzer. You can get 9kHz to 26.5GHz used spectrum analyzer for around $1-2k. Sometimes a bit under $1k if you're lucky and you're willing to take something a bit beat up or broken.

>>1683318
>to 25.6 ghz
>scan everything into into and including SHF, and 1.2cm band
i dont even know what goes on beyond UHF.
i am trying to build my lab out to handle up to at least UHF so up to about 3ghz frequencies.
thanks for recommendation though, buying equipment for just a bit more that can go into SHF or even EHF is always interesting too.

Hi friends, my subwoofer has a volume knob that is really scratchy, and loses continuity sometimes when you adjust it. how would you go about cleaning a potentiometer? just hose it in contact cleaner?

>>1683302
I got it kind of working I think

I think my shitty Chinese components and a breadboard are contributing a lot of parasitic effects so the waveforms are goofy as fuck

>>1682998
Oh, beta is the ratio of I_b : I_c, while hFE is the datasheet value. The more you know.

>>1683047
Both of those appliances could be replaced with 12V/24V variants, but at that point you'd be balancing significant immediate costs against possibly minimal future savings. As the original poster of the question I don't see a problem. Though 12V may be a bit low and could require somewhat beefy wires, what's the most current an appliance of yours is drawing from it?

>>1683072
take it apart

>>1683355
Try using higher capacitances and lower resistances, bumping down your impedances will result in less noise and interference from stray capacitances. But at 50nF you'd think that it wouldn't be that pronounced. It could be the opposite and your op-amps are struggling to source/sink the current wanted by the circuit, but I doubt it.

>>1683363
The one book I managed to find on synthesizer design said VCOs were used in lieu of other oscillators because of their flexibility and stability

with each resistor and the capacitor influencing the pitch in some way, I'm beginning to wonder how at all they managed to design synthesizers that had some semblance at all of staying in tune regardless of atmospheric conditions, manufacturing aberrations, and what all else

>>1683405
>synthesizers
I can only assume that they use some sort of crystal to keep time, else require tuning and low-drift parts.

>>1683412
I've always associated crystals with digital circuits. How do you use a crystal to synchronize an analog circuit?

I'm looking to make this synth real old school style, like the original Moog synths from the 50's/60's

>>1683420
>How do you use a crystal to synchronize an analog circuit?
PLLs

>>1683425
Got it, that makes sense. Every time it feels like being able to design my own synth is within reach, a whole new rabbit hole opens up that I gotta chase to the bottom. Oh well I'll get there eventually

>>1683405
> I'm beginning to wonder how at all they managed to design synthesizers that had some semblance at all of staying in tune
For the first couple of decades, they didn't. The Clavioline-type synths of the early 60s had a fairly distinctive "warble" which was just part of the sound they made. It wasn't uncommon to have a sound engineer constantly correcting the tuning while they were being played.

By the 80s they'd largely moved to using PLLs which lock the VCO to an integer multiple of a frequency generated by a crystal clock (semi-digital), and by the 90s fully digital synths (samplers, basically) were common.

>>1683443
At this point it seems like designing an analog synthesizer is a lesson in futility, it genuinely seems like it would just be easier to bit bang a lot of this shit

I guess I really do need to integrate a phase locked loop into my synth now, though at least I'm learning this before I've really designed it. It feels like I'm re-discovering a lost art what with the dearth of information there is on the subject.

>>1683452
What output format do you need?

>>1683452
> It feels like I'm re-discovering a lost art what with the dearth of information there is on the subject.
There's plenty of information, but most of it's in magazines which are either decomposing in landfill or decomposing more slowly in a box in someone's attic. Analog synths stopped making sense when you could buy a 32-bit DSP chip for the cost of a couple of JFETs.

Forums for open-source software synths might be worth trying. You have to understand how analog synths worked in order to emulate them.

>>1683363
>Both of those appliances could be replaced with 12V/24V variants
you mean the furnace and the fridge? I don't fuckin fink so mate. The fridge maybe, if it was for a boat or RV or something, but it would certainly not be worth that. And the furnace? not a chance.
My standby loads are <20 watts now. Just the modem, switch, a few tiny computers and one slightly bigger computer. In a power fail I'll also hook a phone charger to it. Nothing else is important except the furnace.
The wiring issue is why I have a 24v bus, it's distributed on 24ga wire in a cat5 cable. for relays and arduinos, mostly.

>>1683270
>DC paralyses the muscles preventing you from letting go of the source of the shock.
are you sure about that? there's a lot of misinformation about these things that most people will never experience directly.
>arcing is more pronounced
true. this is one reason why I gave up on DC distribution. because it's true, most things (except transformers and induction motors) eat DC mostly fine.
I have some antique switches that are DC rated. All of them use a snap-action mechanism to quickly separate the contacts and break an arc. a very expensive way to solve a problem that doesn't need to exist.
AC won the current war for good reasons.

>>1683512
DC fridges for those RVs are probably about as efficient as a household fridge, but they're usually much smaller. No clue what a furnace does to require AC power, is it just a fan?

>AC won the current war for good reasons.
They didn't have SMPSs back then, especially not ones that were 95% efficient. With a properly designed grid that could be more efficient than each transformer required in an AC grid. There's less and less reason to actually have AC transmission these days, though it's still hardly obsolete for more industrial cases where induction motors are still the workhorse. Though I wouldn't be surprised of spindle/headstock motors were being replaced with BLDCs or steppers in CNC machines.

>>1683517
>furnace does to require AC power, is it just a fan?
fuck no theres an HV transformer that generates a spark to light the fuel in most cases. But yes, there is also a blower to provide combustion air.

>>1683524
>an HV transformer
Sure it isn't a flyback that rectifies itself first and has some switching transistors? In any case that would be a relatively easy to replace with a LVDC circuit, like just a car ignition coil (and input filter). Fan too could arguably be replaced with a DC fan that happens to be the right size. If you want to spend time doing so, that is.

Makes me wonder what the standby current of that inverter is.

>>1683517
>With a properly designed grid that could be more efficient than each transformer required in an AC grid.
what would a properly designed DC grid look like?

>>1683540
Just selecting the right voltages for convenience vs losses, and ensuring that any people or industries wanting to move to your idealistic ancapistan uses appliances and machines that are compatible with one DC voltage or another.

Hi, is there a better way to arrange these doides to do what I want?

Hello /ohm/, new into this stuff, I have a very small electromagnet I have, and I have no idea how much current to pass through it, what can I do ?

>>1683363
>beta is the ratio of I_b : I_c
Never seen that. Normally both beta and hfe are used for the forward current gain in a common emitter configuration. I use B=Ic/Ib because I can't easily type a decent ß. The context was using a transistor as a switch, not as a 'linear' amplifier the hfe values in the datasheet are meant for.

Connecting arduino ground to earth pin in the outlet made this super big spark on arduino and killed it, what the fuck?

>>1683292
*current

>>1683405
>he thinks they managed to design synthesizers that had some semblance at all of staying in tune regardless of atmospheric conditions
cute
the analog computer guys controlled the environment by any means necessary. the analog synth guys just let them warm up to thermal equilibrium and tuned them often. it is possible to include compensating components with selected or crafted tempcos and arrange circuits such that tempcos mostly cancel out. look up the Sequential Circuits Prophet-5 to see how one late-1970s synth designer took care of the problem :^)
the analog synth IC guys solved the problem by increasing thermal coupling between all the transistors. integration wasn't just for cost savings and deskilling

>>1683580
well, where did you get it from? it would certainly know
the electromagnet will decide how much current to pass based on the voltage across it. if you can get an accurate measurement of the wire size, you can work out how much current you can pass with 20-30°C temperature rise (should be very safe) using a wire size chart and some Ohm's Law

>>1683590
obviously your system ground wasn't really at ground. look up Y capacitors and imagine what happens when they fail

>>1683557
given what you've got to work with, that's about as good as it gets

>>1683602
>look up Y capacitors and imagine what happens when they fail
It can't be that, since there aren't any in the power source

>>1683589
Well from what I saw the beta refers to the instantaneous (or closed loop?) current gain as limited by saturation or whatever, while hFE is the actual datasheet current gain, if that clarifies what I was getting at. That's what I think https://www.eevblog.com/forum/beginners/bjt-beta-vs-hfe/ was trying to say.

>>1683605
In any case I imagine it's a failure of isolation, possibly through the transformer core if it's one of the conductive sorts, or from winding-to-winding, or even from gunk on the PCB, from live to 0VDC. Check your breaker and/or RCD too. There's probably a skidmark on the thing that will help tell you what's happened.

>>1683606
>if that clarifies what I was getting at
It doesn't, there is no "I_b : I_c". nuf sed.

I've got a busted 144hz monitor sitting my my closet for over a year. Think it's worth the time to open it up and fiddle with it, maybe recover some of it's parts for future use?

How do you guys organize your tools and components? Any pics?

So ordered some buck converters to fiddle with this fan controller idea, but they won't be here for a few days. Didn't have a way to get the display to light up as it needs the 12v source with the 5v source to do it. Completely forgot I had a Uno board until a few minutes ago. Was able to use that 5v source to power the whole thing up. Real fucking neato actually, and let me test my PSV to make sure it at least functioned. Though I don't trust the DSO oscilloscope that I have, it's cool but I'm not sure of it's accuracy so I'll have to get my multi-meter out of my car in a while and compare the readings.

Now I just need to pick a project and start working on it.

>>1683628
klipsch speakers?
nice setup you got there

The whole reason I'm picking this up is to learn the conversions math and circuitry better. So quick question.

Right now, this fan draws 3.36W on the high end, 12v, 280mA. If I wanted to use a step up from a v5 to 12v that would be around 672mA (740mA after efficiency )draw on the 5v source?

>>1683634
Dali Lektor 1, And thanks, though I'm not putting it to good enough use yet.

>>1683628
>e 12v source with the 5v source to do it
breh wot
every computer power supply provides separate +5v +3v and +12v sources

>>1683638
I don't have a spare power supply is why I did this. Only have one source output on the tester I have.

>>1683639
you can also use lm317 they cost pennies
or even just string diodes

>>1683619
no

>>1683622
no pics but:
>sample books for SMD small R/L/C
>LCSC boxes for SMD power L (sorted inside by shielded and unshielded), microcontrollers (sorted by family), and misc (including 0-ohm resistors, hundreds per size, and unsorted back stock)
>polypro bait boxes for THT electrolytics and film caps, THT trimmers, SMA connectors and plumbing, and Dupont housings/terminals
>SMD discrete semiconductors in anti-static zipper bags in a 2-column storage bin. zener diodes in a couple of spare pages of above sample book. rectifier/small-signal diodes organized by max I(f) then max V(r). transistors organized by polarity with a divider, then type number. varactors in an anti-static zipper bag off to the side
>connector housings/terminals in one box, with sub-boxes for housings/terminals by family and a bunch of miscellaneous items each in their own bag loose in the connectors box. 2.54mm pin headers/sockets in zipper bags by type and color, all in an outer zipper bag which goes in the connector box
>everything else SMD in individual zipper bags per type/value, inside 1L zipper bags by category
>other THT discretes bagged by value, then by decade, then sub-boxed by type in small chink shipping boxes, in a big misc box. also a sub-box for magnetic cores here
>aaaand one last box for oddballs and backstock

>>1683637
math is pretty good
>90% efficient
a bit optimistic for a boost, especially for chinkshit. many variables in the step-up converter have to be just right. estimate 80% and see also
http://www.ti.com/lit/an/slva372c/slva372c.pdf

>>1683641
>>>/g/

>>1683641
>lm317
Already thought about this, but I intend to fuck around with this a lot, and eventually I plan to build it to it's max, which is 12Wx4 loads. So going to try a low power version from a 5v source for just a few fans, and then putting the system to it's max, and the lm317 won't work for the amount of power that's going to be.

I have 6 MP1584EN I just ordered(spares) to run it all from a 12v source and see if I can get it to work like that. Which from what I understand should work, having it as a 12v,1A supply, and then sending 12v to the fan controller, and the down stepped 5v to the display.

>>1683643
Appreciate it, and yea 90% is pretty high, I just pulled it near what the down step says I'll get (doubt it'll be close to the 92% efficiency) I'm more just glad that I was able to understand the logic behind it and get the math close.

>>1683644
lm338 can do up to 5A

Can someone doublecheck my noobmath?
I have a fan that draws .1A at 12V
I have 24V power source and i want to run the fan at 4V which draws 0.035A

So
20V = 0.035 * R
R = 571Ω

So at 24V source i need 571Ω resistor to run the fan at 4V 0.035A

24V * 0.035A = 0.7W the resistor has to dissipate
those small blue shits can do 0.5W at most, so if i put 2x240Ω resistors in series i should be golden
Does my math check out?

>>1683647
>24V * 0.035A
oops was supposed to be 20V * 0.035A

>>1683648
Now find the fan current at 480Ω.

>>1683652
it's just 0.05, the fan can handle up to 0.1A no problem
i just want to run it silently and it's not like they make 285.5 resistors
It doesn't have to be exactly 0.035

if anyone wants to put together a brainlet's guide to niggering PC fans on pastebin, I'll link it in the next OP if I get to it in time

>>1683656
Distraction.
Calculate the power the two resistors have to dissipate.

>>1683661
eh fuck that is 0.1w extra but thats fine i'll just place the resistors in front of the fan

>>1683664
it's generally a bad idea to run components at full rated current, leave yourself some headroom. use three 180 ohms?

I tried making this phase fired control circuit for inductive motor loads and I have an issue - when I connect it to the fan it is supposed to regulate (120W fan) it emits a low frequency, quite audible rumbling noise. Is this to be expected from something like this? Is there some way to get rid of it?
From what I understand, this is basically a typical dimmer circuit with an add on (on the left of C1) of a circuit helping to keep up the symmetry of the phase on positive and negative wave-halves (also discharging leftover C1 charge). The presence of the add-on doesn't influence the noise.

>>1683695
>and of course I forgot the picrel

>>1683696
Is the rumbling in any way affected by an adjustment of R6?

>>1683524
The furnace has 3 induction motors. One to pull combustion air in from outside, one to force air through the combustion chamber, and a big circulating fan.
No HV transformer, it has glow plugs like a diesel engine.
But I'm not about to heavily mod the furnace for an extra < 10% efficiency, the Xantrex will break 85%. The 1800 watt model will hit 90%.
Inverters are efficient, enough so to make DC not worth the effort.
>>1683544
You just listed some reasons WHY AC won the war.

>>1682785
>>1682722
bump for help/advice

>>1683703
Yes, but only slightly. Unlike omitting the whole controller

>>1683644
in the case of the buck topology, substantially all inductor current goes into the load. 90%+ efficiency is definitely possible, especially with a synchronous rectifier and due attention to the details. not so for the boost topology
(synchronous rectifier: replaces the buck converter's diode and its voltage drop with a second switch turned on and off out-of-phase with the pass switch, thus improving both power and board space efficiency. but can be destroyed if power is back-driven onto the output (I'm only a little bit still sore about it))

>>1683695
>inductive motor loads
very clever. almost all motor loads are inductive
inductION motors are designed to depend on the fixed frequency sine waves from direct connection to the mains. they do not react well to phase angle dimming. they require a VFD for efficient, effective speed control
only shaded-pole inductION motors (the flat ones with the C core and the shading rings) and motors with a permanent start capacitor react sort of well to phase angle dimming. iow, yes, that's to be expected

>>1683712
ok. so for TR925-B to be at 10.5V (yellow), TR921 must not be turning on at playback. TR922 could be stuck on, but more likely you have to keep working back (blue). those two caps could be short. D918 could be open-circuit. and so on

>>1683725
>yes, that's to be expected
Thanks. The press article I got it from claimed this triac solution could work with inductiON motors (which is why I picked it in the first place) instead of expensive inverter based ones. I checked it with a smaller 30W motor and there is the same noise, but it's almost unnoticeable.
>inductiON motors
I don't know how much of it was a factual error, and how much is my english to blame, thanks for the correction, I'm just a humble Saturday amateur.

>>1683739
it's all good. to be fair, it does work for *some* induction motors, just not most of them. I blame the original author

I want a good soldering station.
I think I want direct heat, but I don't want paying thousands of dollars on brand names. And I don't want chink shit either.
So, I basically want to get a good soldering iron handle and heater, yet find transformer somewhere in local dumpster and make control electronics myself.
Who makes best out of best?

>>1683787
Also, does red part get hot on those direct heater soldering irons? How hot?
I want to solder wires directly to heater and thermal couple, and call it a day

>>1683771
From what I read all triac-based fan controllers seem to have their hum at lower speed and the authors recommend to use switchable series capacitors instead. All of the hysteresis-free circuits I came across (mostly for AC motors) use much smaller values for R1 and R2 but none had R6 for adjustment. Just fyi.

I remember that last time you posted that circuit it didn't work at all, what was it?

Considering picking apart an old 37" LCD TV from ~2004. It still works, but it's old/heavy/slow/energy-inefficient. Any particular components that could be worth salvaging? Will possibly cut out some of the polarizing filter for glasses/screens.

>>1683817
nothing you'd know what to do with, since you're asking instead of looking inside

>>1683829
I'm using the screen right now to look at your snarky comment, which is why I haven't looked inside this one specifically :^)

I'm asking for particular components that /could be worth salvaging/ in other people's experience, given a LCD TV around 15 years of age and use. since you're not open to sharing your extensive plethora of knowledge you're very free to not reply.

>>1683796
No, really, how hot it gets?

>>1683800
>I remember that last time you posted that circuit it didn't work at all, what was it?
It's embarrassing. Somehow it was hardcoded into my brain that MT1, MT2 leads of the triac are interchangeable. I was so sure of it, I didn't even question it. I was staring at the diagram trying to figure it out, looking at the triac symbol, and a question popped into my mind - "hey, why is the triac symbol asymmetric?"
That's what you get if you read just tutorials with no hands on experience - you are left with a bunch of half-assed impressions.

>>1683879
hot enough you won't want to hold it there.

>>1683725
Thank you friend, I will take a look when I get home from work and let you know

>>1683933
Hm. Shit, that means I have to get a proper handle.

I've been trying to design my synthesizer around op amp voltage controlled components, but I'm now remembering I bought a shitload of lm13700 OTAs to make all of the VC(O,A,F)s out of

good thing I hardly know how the fuck an op amp works, redesigning everything to work with an OTA ought to be a cake walk

>>1683941
>Hm. Shit, that means I have to get a proper handle.
You just don't have what it takes

>>1683948
Thing is, some fag on the internet told me that TS100 cartridge doesn't get hot, only tip is heating.
T12 has similar structure.

>>1683956
Here.
Makes me think, that I can use it without any handle

Isn’t he a kike?

>>1683956
well that fag is a fag.
I have a TS100 and I just tried it out. the part of the element where the logo is gets hotter than 100 degrees. I can't find my thermocouple but water sizzles on it.
the narrow part above that doesn't get as hot but still too hot to hold.

>>1683975
>the narrow part above that doesn't get as hot but still too hot to hold.
Sheit. I guess I will have to buy a handle.
Or take wooden dowel, drill a hole and make a soldering iron with wooden handle

Why is the online world of electronics so inundated with people from India
Like I understand their prominence in all of STEM, but it seems to be that they're especially prevalent in EE/CompE. Youtube videos, quora, websites, they're everywhere

>>1683998
Because India is a DIY shithole.

>>1683973
He sounds and behaves like kike.
And it is difficult to tell apart southern european from kike without precision measuring tools

>>1683973
>>1684001
>>>/pol/

>>1684001
My jewdar was calibrated in August to sub microkvetching precision.i assure you it is not wrong

>>1683998
because they have a billion people and are rushing to advance industrially

>>1683998
world population ~7 billion
population of India ~1.4 billion
1 in 5 people of the world are pajeet

>>1684022
How about we rush them to forced labor camps

>>1684025
superpower in 3 months
god help us all

>>1684001
>>1684015

>>1684035
E. Coli will help

>>1683948
kek, another one?

>>1683998
four words: air conditioning in Bangalore
also easiest to get other people to do your work for you which is all Brahmin are good for

>>1684042
I was looking for that extra "o" in superpower

>>1684026
>>>/pol/
mein gott, the whole thread is /pol/ now
what the fuck happened

How bad was Differential Equations and Calculus III for you guys, compared to calc II?

>>1684063
too lazy to read the post you're replying to but almost all of the engineers i've worked with have been conservative. i think it's related to the object-oriented mindset. since 4chan leans right already, the combination naturally means that everyone itt is a literal neo nazi.

>>1684067
i don't even know what a differential equation is and i passed difeq. calculus in my experience was entirely dependent on the professor so do your research.

>>1684063
/diy/ is rightful /pol/ clay

this is what happens when people start mentioning /diy/ on /g/ or other boards
go back to /pol/ or wherever you came people, there's nothing here for you

>>1684086
/diy/ is a jewish board, because only jew would repair or create his shit, so he can outjew other jews.
It also happens that jews are into /pol/ too.

>>1684067
Calc 3 was easier than calc 2 but diffeq, fuck that shit.
>>1684069
>dependent on the professor
Yeah mine was a fat Ghanaian or whatever woman. Fuck her.
>>1684086
go back to /g/ tranny.
>>1684091
actually, /diy/ is for Aryan men who exist / wish to exist outside the jewish system of usury and consumerism.

>>1684118
Speaking of /g/, how hard is it to make a CPU using FPGA?
What sort of performance can I expect?
How do I learn FPGA?

>>1682638
why are smc to bnc adapters $50 a piece? holy cow mang.

please spoonfeed me decent places to get electronics components and things like adapters and connector heads. I want to get some parts organizers or some parts drawer thingies with all those parts&components so i can work on things without having to go shopping every time.

>>1684139
And can I implement OS directly in FPGA, so it requires user knowing how to program FPGA in order to modify it?

>>1684139

Read this and do exercises:
http://www1.idc.ac.il/tecs/plan.html
https://www.nand2tetris.org/

The knowledge should transfer over to FPGA.

>>1684140
https://www.amazon.com/SMA-Male-BNC-Female-Adapter/dp/B00QPN7YYQ

>>1684140
>decent places
OP

>>1684139
not too bad, performance scales sublinearly with effort (cf. Spectre vulnerability)
mediocre, on the order of tens of megahertz (but you can have some bangin' peripherals)
start by learning digital logic and Boolean algebra. then learn Verilog. then read the two Hennessy & Patterson books. then learn at least one assembly language of at least one 16- or 32-bit CPU
now you're ready to come up with a theme and start exploring it. for example, what if a 16-bit load-store machine has just one prefix instruction that provides 12 constant bits to the next instruction? the ramifications of these sorts of hypotheticals propagate throughout the instruction set design and architecture, ultimately affecting clock speed and code density. that machine was about on par with the SuperH SH-1 architecture and ran at 66MHz
there are more exercise-tier themes such as implementing various kinds of program flow control such as conditional execution (a la ARM) or zero-overhead loops (some DSPs), or executing some other ISA like Brainfuck or Forth
anyway you have at least several months of study ahead of you. good night and good luck

>>1684069
I think it's because marketing fucks mess up our lives on the daily with their fanciful make-believe.
>i think it's related to the object-oriented mindset
I think it's more a byproduct of whole-system thinking, fluency in thermodynamics, autism (of course), and the Cold War as state religion. I probably skew the /ohm/ curve way to the left on economics, but fuck trannies *and* jannies

>>1684140
good heuristic: if you're using one now, you'll use one later, so buy five or ten or a hundred. resistors and such are best bought in kits covering a broad range of values. git gud at SMD, components can cost less than the time it took you to desolder them

Haven't been to /diy/ for a while. Has the /pol/ cancer spread even here? iir normally they'd be told off to gtfo

>>1684278
EE has always had little time for fanciful shit or bourgeois pretension

>search for "conductive grounding mat"
>get new age hippy-dippy pseudoscience results
tiresome

>>1684287
idc fag, go back to /pol/

>>1684278
also a bit surprised to see it in /ohm/. the fuckers should keep politics out

>a fanatic is someone who won't change the subject and won't change their opinion

>>1683725
>>1683939
Didn't get a chance to work on it last night unfortunately, however here is the complete sheet if that helps you at all

>>1684303
thanks
>trigger pulse from "FV" panel F8-3
what's all this then?
current to turn on TR921 is dependent on TR920 being on. for the purposes of testing the solenoid drivers on the right side, you can temporarily tie the trigger input to +12V, but once you have verified all those solenoids and logic work correctly, next stop is the FV board

>>1684289
now that's some last-level logposting. grow up

>>1684312
FV panel is the Audio Rec/Play Servo Panel. Here's the Voltage circuit for the unit.
The whole schematic book is here, pg 39 is the full sheet. too big to post here without reduction in file size making it unreadable. FV panel is on the center left size, split in two for some reason
http://www.mediafire.com/file/xu6ly6f0qldma63/VR1000-set.pdf/file

>>1684315
sorry just the schematic sheet is this
http://www.mediafire.com/view/vmixgd1dir9aqj9/Full%20Schematic.jpg
dont know why the voltage sheet is such low dimensions, its bigger in the PDF...

>>1684295
based

>>1684317
alright, this is the last of the babby food, just 3 more transistors to walk thru until it's out of the realm of the electronic and probably requires a replacement part almost as old as I am

>>1684325
>pulse generator
oh man. i dont have a scope. you do think a bad pulse gen could be the issue do you?
here's a quick rundown of the history of this unit
>bought in 1976 or so
>around the late 80s the machine no longer rewound tapes
>stopped working for good sometime in the 90s
>when I press play, the head and motor spin at normal speed.
>when I press FFD, the heads and motors still spin as if in play mode
>when I press RWD, the heads and motors still spin as if in play mode.
>note: the unit only activates the drive motor when I bypass the light sensors with tape over it

How good are those silver conductive ink pens? I know that they are meant to repair traces, but could you use them to draw a full circuit board by hand using them?

>>1684329
could be. there's a small possibility it could be a dead servo controller or those transistors, but the only way to know for sure is to use a scope. you can rule out the transistors by testing dc voltages and maybe you'll get lucky and it was just a spike from the head that blew TR711. otherwise, might be time to refer servicing to qualified personnel
is there some precious programming you're trying to pull off some old tape?

>>1684353
not particularly, its more of a sentimental thing. it was my dad's first vtr, he has about 35 tapes full of stuff from tv, movies, live events, etc from the 70s and 80s. Before he passed away we used to joke about how I was going to get it fixed for him so he could see those videos again. I'd just hate to toss it away if it's something remotely fixable. There are units that pop up on ebay from time to time, but most of them are broken too, and the one that does work is going for over $700

>>1684369
it might be fixable but you don't appear to have the required instruments
one more thing you can check before buttoning it back up: read the voltage on F8-3 and F11-3 (right side of previous excerpt) to see if anything's going on. or, it could be as simple as loading a tape

>>1684372
are scopes hard to come by?
would any one of these work?
https://chicago.craigslist.org/search/sss?query=oscilloscope&sort=rel

What's the simplest way to make a latching circuit with a single momentary push button.
i.e. Press once, circuit goes high, press again, circuit goes low.
Preferably with discrete components, I've a load of BJTs, FETs, diodes and resistors to play with.
Logic level will be 3.3V

>>1684431
do you know about switch bounce?

>>1684155
>>1684263
Cool.
Should get FPGA.
Not sure if I will make a CPU, but I'm 100% sure I can copy existing projects like "Amiga on FPGA" and such, which will feature motorolla core inside.
Okay, once you made everything work on FPGA, how do you move to ASIC?

>>1684432
Yeah, I assume I could overcome it with an RC arrangement

>>1684434
Hell, people had even made X86 CPU in FPGA...

>>1684434
nah. dev boards are social animals and like to pile up on your desk
>how do you move to ASIC
if you have to ask, you can't afford it. some FPGA vendors offer mask-programmable drop-in equivalents to their standard devices (MOQ applies). or you could route the masks yourself (on your old bitcoin mining rig), verify through simulation (on a big-ass EC2 cluster), and have a multi-project wafer service make you a handful of chips for $thousands each

>>1684431
74LVC1G80. connect ~Q to D and circuit out, button to clock
>preferably with discrete
the only reason to make logic-type functions out of discretes is education. in the next 30 years there will never be any reason at all, ever, to reenact the mechanical control board on other anon's antique VCR. there are plenty of designs for a D flipflop from discretes on the interwebs. find one and use it

>>1684440
So, ASICs not worth it, unless you need million units, right?

I want to control some 5V fairy lights with my phone. I've been looking into the Shelly 1.
Should I just get a 230 to 5V adapter and plug it into the Shelly and then the fairy lights?

Any other ideas or setups you might have at home?

Guys, does something like this exist in a fairly rugged, self contained, and most importantly off the shelf package with the potentiometer part actuated by a good size dial knob? It would be used to drive a coil on a hydraulic valve that needs 1/2 of the DC input voltage (nominally 12V) for the "neutral" (valve closed) position, and from 1/2 to 3/4 of the input voltage to vary the position of the valve from 0-100% open. Cost is no object.

>>1684441
correct, plus or minus an order of magnitude depending on the chip size

>>1684443
why not

>>1682638
best mutlimeter sub 50 dollars? Is it still 8008 or is the 8009 better.

>>1684457
> does something like this exist in a fairly rugged, self contained, and most importantly off the shelf package
No.

>>1684457
>Cost is no object.
Hire an electrical engineer and an industrial designer. Have them produce it for you.

>>1684467
The 8009 is better because it's a greater number.

>>1684457
Remove 15ohm Resistor and use the coil resistance as part of the divider

Add a rheostat, knob, resistor, and a box
Search:power rheostat or industrial rheostat

>>1684467
AN870

>>1684457
are you sure you've interpreted the proportioning valve wiring correctly? it seems dangerous and stupid that it must be powered-up to not flow.

>>1684497
Thanks do you mean like this? The only issue with is that the hydraulic valve I want to send the 6-9V into has 3 wires going to it, nominally 12V, the variable 6-9V and ground. I don't know what's going on inside the coil, but I think the 6-9V is just used as a reference and the coil is actually driven to the desired position by the power supplied to the constant 12V pin if that makes sense. The coil data only shows gives a current rating on the 12V pin (0.5A) and normally this would be wired straight to a 12V battery with a 2A fuse. Also the 6-9V signal I'm trying to create with the voltage divider is normally supplied to the coil by a microcontroller. The purpose of this whole exercise is basically to replace a microcontroller with a dial.
>>1684489
Sorry I shouldnt have said 'no' object, more like $500 budget when I was assuming there was an off the shelf version.
>>1684511
I'm positive that 6V is neutral, 9V is fully open one way and 3V is fully open the other way (unused in this case). I think there is circuitry inside the coil where say a wire gets cut and you send 0V it will go to a fully closed position, not sure how it does this though, and I will double check this on one of the microcontroller units by shorting this control pin to ground.

>>1684511
Also if the constant nominal 12V that I believe is driving the coil isn't there, it goes to the neutral position. Any safety devices used with this are supposed to cut power to the 12V pin and the valve will go to the neutral position.

>>1684463
Ok.
Obviously, I don't know a fuck about FPGA, which board should I get to blink LED and what not...

>>1684541
start cheap, get a $25 Lattice iCEstick. there is an open source toolchain for it, and a vendor toolchain that isn't completely monstrous like Xilinx's Vivado is. you get 1000 logic cells on the chip that's on it. a very elementary 8-bit "processor" can be built using about 100-150 logic cells, leaving some 85% free for blinkies or I/O ports or whatever
but if you're new to electronics, maybe do some arduino first and learn how to treat I/O pins with due care

>>1684519
>do you mean like this
Yea, but since it's a micro-controller signal rather than a coil load you probably need a divider like you originally drew.
I'd up the resistance values to use less power (the 15 ohm uses 6+ watts = a big resistor)
I'd think 1/4 watt would be good enough (assuming the control signal draws very little current)
Power = voltage*current
1/4 = 12 * current
current = 20.8mA
resistance = voltage / current
R = 12 / .0208 = 577 ohms total
so your potientometer needs to be about 1/3 of that or about 192 ohms
I'd look for a basic pontientomerer 150-300 ohms and decide your other resisters from that. 1/4 watt resistors are cheap.
If it's an electrically noisy environment, you may need a capacitor from the control signal to ground.

>>1684549
>it's a micro-controller signal rather than a coil load
anon is correct, I've worked with these valves too. it's a pretty high impedance input. The valve datasheet might tell you exactly. Read it.
So why use such low resistance? there's no need to max out 1/4W resistors and shorten the life of your potentiometer.
You can probably multiply all these values by 10.

Also, instead of using fixed resistors for the top and bottom (or in addition to them) you can put trimpots to fine tune the "open" and "closed" voltages.

>>1683979
>Or take wooden dowel, drill a hole and make a soldering iron with wooden handle

this was used to put Soyuz in space. nobody really needs any stinkin' temp-controlled baby tools when they have the skillz of Russkiy master craftsman.

What semiconductor solution can i use to switch 2kA DC currents?

>>1684621
SCRs come first to mind. several options are available at and above 2kA, some of which are fast-turn-off type

>>1684605
Who said I will have temperature control?
I will modify laptop power brick, so it outputs 12-ish V in idle, and when I press a button, it raises voltage to 19-24V (or whatever highest stable).
Feedback circuit on cheap power supplies is really simple. Optocupler, zener. Czech guy explained SMPS better than my uni.
Since brick is rated at 65W, transformer won't be overloaded.

I just don't like temperature control on any station. They barely work, and you end up cranking it to the sky anyway.

Also, how should I ground the tip? Directly by connecting mains earth to it, or via resistor?

>>1684626
Well, not zener, but TL431

Thanks guys, all makes sense.
>>1684549
>If it's an electrically noisy environment, you may need a capacitor from the control signal to ground.
Appreciate that tip, does that just resist changes in the control signal voltage due to interference?
>>1684594
>Also, instead of using fixed resistors for the top and bottom (or in addition to them) you can put trimpots to fine tune the "open" and "closed" voltages.
Now this is a good idea. Do you think once it's adjusted right I could measure the values and switch to fixed resistors, or do you mean it might need adjusted differently every single installation given the variation in wire resistance / tolerance on a fixed resistor? Also with the microcontroller version, there is a deadband around the neutral value that you can adjust out in the software that I have been setting to around 7% of the 3V range or else the valve won't open until you set it to like 15-20% open. Not sure why these percentages aren't the same, but it works. I'm assuming these trim pots would allow me to do the same thing?
>pretty high impedance input. The valve datasheet might tell you exactly.
It's all gay proprietary shit where they are expecting (forcing) you to use their microcontroller so it's scant on those kind of details. Can I measure resistance from signal pin to ground on the coil? Can't say I've ever tried.

>>1684334
Yes, but don't expect to put more than a few mA through them. You can probably find demonstrations in youtube, and probably people measuring their resistance too. But in a pinch a ≥4B pencil can draw conductive traces on paper. Maybe with a 2B if you're heavy-handed.

>>1684378
You'll basically want a 2-channel, preferably ≥20MHz, AC/DC coupling, doesn't really matter if it's analog or digital for the most part, but digital ones have their advantages. That ES Tech Pro for $20 would probably do, if it works.

>>1684457
just use an LM317 lmao

>>1684624
Cool, thanks

>>1684629
>trimpots
are there to adjust for the varying resistance of each copy of the main pot. the higher the resistor values you use, the less current flows, and the less compensation you would need for long wire runs etc.
>switch to fixed resistors
I don't recommend it. usually manufacturers just put a dot of red Loctite to fasten the shaft in relation to the body once initially adjusted, in case you need to readjust later due to aging, wear, etc.

>>1684629
>once it's adjusted right I could measure the values and switch to fixed resistors
no reason to. Trimpots are cheap and convenient. it frees you from almost all precision because you can just measure the valve position and trim them until you get the valve where you want it. That takes care of your deadband adjustment and any little variations in your system.
>Can I measure resistance from signal pin to ground on the coil? Can't say I've ever tried.
me either. but what I'd do is put a high value resistor (~100k) in series with the signal line and measure the voltage dropped across it. that will tell you.

just tried out using my function generator and scope in X-Y mode to plot the I/V curve of a red LED. I'm easily able to see a forward voltage of ~1.8V or so, but the curve after that is just a straight line. I think the reason is because of my circuit; to measure the current through the diode I'm measuring across my resistor (100kΩ because I forget how high a voltage my function generator can manage), and to measure the voltage I'm measuring the total voltage across both the LED and the resistor. This is because I don't have isolated/differential probes. Swapping the function generator's wires so I can measure the voltage directly across the LED and representing the current as the voltage across both of them also results in a bit of a shitty waveform, with a straight line but the voltage offset in a different position because I swapped where I'm measuring.

Is there some technique I'm missing out on, aside from whipping up a little USB or battery-powered differential probe with a quad op-amp inside it to make an instrumentation amplifier? I'd dead-bug the circuit and cover it in hot snot.

I also have the fact that my scope's and function generator's grounds are linked to worry about, but it isn't a pressing issue since a common ground doesn't pose much of an issue. Interestingly enough, when I turn the frequency up to 100kHz (as opposed to the 1kHz I'm using for these measurements) I get a different voltage path, in normal time/voltage view I can see this as a significant phase difference, and/or capacitance causing asymmetric smoothing of the forward/reverse bias transition. Because I'm on a breadboard.

>>1684647
On the left is the circuit I was trying in that image. Just tried the right in an attempt to turn the function generator into something closer to a current source, but I get a very similar slope.
I'll try some other types of diodes to see if it looks any curvier.

>>1684653
Oh jesus these 1N4004s are slow, only 10kHz and they're dipping badly. But they still have the same constant slope.
Just checked, and yes that slope is basically identical to that of just a single resistor.

>>1684069
fuck you i hate all the damn indian videos. I would 1000x prefer if they just spoke in their own language instead of clickbaiting in english

>>1684656
>english
please to be needing citation thank you

>>1684659
clean your armpits ravi

>>1684636
>usually manufacturers just put a dot of red Loctite
>>1684637
>what I'd do is put a high value resistor (~100k) in series with the signal line and measure the voltage dropped across it
Thanks men.
>>1684660
open bob

>>1684653
>>1684655
Ok looks like the problem with circuit 2 is that I was using a 1kΩ as a current sense resistor, which was too high (with the lights on what I thought was a 100kΩturned out to be a 10kΩ so that makes sense I guess). Trying a 10Ω was too noisy and I had to have the thing cranked up to maximum gain, but a 150Ω works ok (picrel). The resulting curve still isn't good enough. Using a 100kΩand 1kΩ I get a very similar signal:noise ratio. I won't go down to a 1kΩ current limiting resistor because that could blow a diode.

I guess I'll see about making a differential probe after all, but I don't really have a good dual-rail power supply for that sort of thing yet. Be nice if I could directly access the power rails of my scope, I bet it has some nice handy ones like ±12V and the like. I have a multi-tap transformer lying around, but it needs a case and plenty of circuitry to be properly usable. With the option to use higher voltage rails, I might go for a jfet op-amp this time, like some of these LF347s I have easy access to. Actually since it's an instrumentation amp, I shouldn't need a split-rail, I can instead use a cheapy boost converter, like this adjustable one that goes up to 28V. Using the 4th op-amp to buffer a voltage divider-based ground rail.

Why do differential probes cost megabux again?

>>1684660

Holy shit, these are a lot smaller than I thought they would be.

>>1684674
The miracles of SMT.

>>1684626
>Also, how should I ground the tip? Directly by connecting mains earth to it, or via resistor?

i've had occasions when i had to solder to live wires (coz couldnt get the key to turn off the breaker) so was glad my tip wasn't grounded. you can use a 2.2meg resistor, i guess, but i wouldnt bother: your hands are injecting just as much stray voltage onto the circuits as the iron itself.

he guys... any tips for becoming an electrician? trying to quit my crappy wagecuck job.

or should i say fuckit and do something like welding?

>>1684680
I had burned MOSFETs in the past with ESD, so idk.

>>1684674
one of the most used pieces of equipment on my desk: a pcb ruler

>>1684681
don't care about sparkies here. >>>/adv/

I'm working on a design that will have several QFN chips which have a large ground pad on the bottom as they sink a lot of current.

I'm looking for suggestions on how to solder that. Preferably without a hot air gun. The most popular suggestion online is to create a through hole in the board going to the middle of that pad, pre-tin the pad, and solder it from the other side via through hole however I do not want to put holes on the opposite side of the board as it's the display area and the whole board is visible. It'll ruin the cosmetics of the design.

Other options were to make the pad slightly larger than the chip itself and pre-tin it and melt it from outside the bounds. I think this would work for one chip I'm considering that isn't QFN but for all the QFN chips this is no good as they have pins on all sides. I might be able to do a thick trace coming out from the chip diagonally but it'll be hard to put enough heat through that to melt the whole pad, especially since it'll all be via stitched to the ground plane.

The final option I was considering was applying no solder at all and instead using thermal paste between the chip and the pad on the board. That's not ideal since thermal paste isn't nearly as conductive as solder but it's better than nothing I guess.

>>1684694
You can use solder paste and reflow it by putting the board on a regular hot plate. I've done this with success.
if you're doing a lot of boards you can order a solder stencil with them to apply the paste.
Hot-plate reflow is pretty janky but it does work.
if you're interested I'll describe my setup

>>1684490
idk they seem the same but 8009 has easier to get fuses.

>>1684680
>>1684685
Why not have a toggle switch, with an NC, direct-to-ground, and 1MΩ to ground modes? A little kapton tape around the ground pin of your plug can work in a pinch.

>>1684697
The 8009 is the updated model, it has a few more features.

>>1684372
Thanks for all the help. I’m at the machine now, f8-3 is showing .1v and f11-3 is showing .6v
TR711 is showing C .6v when it should be 11.6 now in the schematic there is 12v coming in at r748 but only .6 on the other side going to TR711

>>1684733
>only .6 on the other side
sounds like a collector-to-base short. first verify TR711 base is no higher than collector. if so, replace TR711. as used here, it's kind of a wear part

>>1684736
Base should read 0 on standby and 0 on playback. Base DOES read 0 on both.
Collector should read 12 on S and 11.6 on P, but DOES read 0.6 on both.
Tr711 is 2sc828, seems like they still make it so I can replace with the same part.

>>1684745
well actually it looks like its a 2sc828Q or 2sc828R according the schematic

>>1684726
Good idea.

So, I have following idea.
1) In idle laptop power brick outputs minimal stable voltage (sources say 12-14V is about as minimal as it gets)
2) Once pushbutton is pressed, SMPS enters full retard mode with maximum possible voltage (sources say 23-24V is aboutas maximal as laptop PSU can get without loosing power).
At 12-14V soldering iron barely gets energy (sources say T12/T15 tips have 8 ohms, thus I can expect 18W which I think won't make it go red hot and overheat, but I'm not sure). At 23-24V I can expect full 65W, because it is the limit of PSU I want to sodomize for this purpose.
In case it overheats at 12V, I may add simple ass T12 regulator (pic https://cxem.net/master/images/87-8.jpg)), because I know where can I steal P-fets and where I can get OP amps

>>1684750
Nah, it will overheat. It can reach 350C from 12V no problem...
So I would do this

>>1684745
Also side question, there’s abour 4-6 capacitors with greenish patina/stuff on the negative lead only (one has it on both). Does this mean the capacitor is shot or is it okay?

>>1684745
any common general purpose npn transistor in TO-92 package with a hFE of 100 or better will work just fine to test that the pulse gen head is still functional after all that time spent half broken. check pinouts first ofc

>>1684757
It’s the white thing there, any way to test it in place? Not quite understanding on the schematic what the voltages should be to know if it’s working

>>1684760
the two ways are replacing that transistor and using a scope. not very testable with just a meter

>>1684763
gotcha. i'll go ahead and order the transistors tonight, maybe even contact the guy about that ES pro tech https://chicago.craigslist.org/nch/ele/d/lake-forest-oscilloscope/6973178068.html

If you spin up a blank hard drive at a low speed, could you write analog data to it with a solenoid and read it with another solenoid to have an analog delay? Add an array of permanent magnets/electromagnets to erase the data before it gets written over.

>>1684803
In principle, but not in reality. They used tape loops for that. The "white thing there" mentioned above looks like an erasing head. Do you know how a tape recorder works?

https://www.banggood.com/5PCS-DANIU-Electrical-Cutting-Plier-Wire-Cable-Cutter-Side-Snips-Flush-Pliers-Tool-p-1387787.html

11 bucks for 5.
Anyone used this DANIU brand before? Starting to get into electronics and it seems these are good to have laying about here and there. Has the "buy one high quality tool" concept served you better? Don't know if there's a huge difference between chinesium cutters and "not chinesium" cutters.

>>1684837
just shop till you drop

>>1684837
The difference is how well aligned the blades are, how sharp, how hard, the quality of the grips, how worn out the tooling was, etc

>>1684842
I'll just might. just picked up 180 ft of paracord for $6.

>>1684844
thanks for the reply.
but are those differences great enough to justify the price difference in your opinion?
at $2.2 I'll just toss 'em when they break.

>>1684824
Roughly. I've heard of tapes being used for delays (when watching a video about an electrostatic method of audio delay where a rotating disc had charges deposited by conductive wipers), but heard they're a pain to erase afterwards, which isn't surprising since they aren't just a single reusable loop. So I figured a scrap hard drive would be able to do this job for me more easily.

Also if HDD disc material is meant to be magnetised, I wonder if you could machine them into ferrite memory cores? Those cores are kinda elusive.

>>1684824
>>1684851
Well i found a video where a guy used a HDD as a turntable-scratcher-style instrument, and one where a guy literally carved a spiral groove into the surface of a HDD to play on a record player, and a few stackexchange posts that look semi-relevant. One of which mentioned a specialised hard drive used to record freeze-frames of video for instant replay playback, which is one of the rare occurrences where the convenience of a hard disc was more important than the low price per length of a magnetic tape.

One semi-analog way would be to write raw class-D/PWM binary onto the disc, assuming you could get it spinning at approximately exactly 1Mb/s. The output of the head could be buffered, totem-poled, and fed into a low-pass LC filter and that would get you audio. If you did it correctly and could somehow use some of the existing control circuitry, it may even be possible for it to automatically read or write a stepped-spiral groove into the disc. But this implementation would certainly mean it would be impossible to read at the same time as writing.

For a proper analog solution you'd have a variable speed somewhere between ~1 and ~1000 rpm, with diametrically opposed heads for reading and writing so you don't get cross-talk between the heads. Not sure what you'd do with the multiple discs, maybe have read-heads at different positions on them, or have one where the data is erased after reading and one where it isn't for echoing, or all sorts of other thoughts.

>>1684847
quality's not that crucial for those little cutters.
but protip, if you do any amount of multipin connectors, buy the genuine crimping tools. it's worth paying $200 for molex even when the chinese "OEM quality" shit is $2, because the chink tools will fuck up every single connection and pretty soon you'll have burnt through the equivalent value in wasted time and trashed pins.

>>1684844
Blade alignment and sharpness can be fixed with a grinder, but shitty steel cannot.
some of these are so soft they wear out at nothing.

>>1684837
they look like knockoffs of the Plato 170 shears. should be bretty gud but not particularly amenable to rough handling or repeated dropping. do note the 1mm Cu restriction. I have some slightly smaller, tougher, misc chinese nippers for stuff that's larger or stronger
but if I could, I would have my Weller snips from my youth. those were tough enough for a semi-stupid teen

>accidentally use the PCB-cleaning toothbrush to brush my teeth
and that's why I use lead-free

>>1684903
How'd you mix those up, do you do electronics in your bathroom?

>>1684862
cheers. I'll look into these quality brands some more.

>>1684886
>Plato 170
oh nice. I found the 170S for up to 1.5 mm here:
https://www.digikey.se/product-detail/sv/techspray/170S/170S-ND/8593021
will probably go with those instead.

>>1684756
This is what I’m talking about

>>1684920

>>1684913
do you do wet chemistry in your bedroom? don't answer that

>>1684920
looks like the same shit that pops up on car batteries. mix a medium-strength amount of baking soda in water and gently rub it off with a cotton swab

>>1684920
corrosion caused by the electrolyte that's supposed to stay sealed inside the capacitors. they're certainly due for replacement, but don't expect a miracle cure.

>>1684974
>>1684924
for 25c for a 47uf 16v cap it wouldnt really hurt, just get killed on that $10 digikey shipping fee to mail a 3oz package

How are you cleaning your tip?
I used to use cellulose pad with water, but it always left nasty limescale on tip. Now I use just toilet paper, works great, temperature doesn't drop, slighly abrasive, cheap.

>>1684977
I use a jugganut bitch

>>1684981
I tried that, I didn't like it at all, doesn't clean a fuck.

>>1684987
it gets my tip shiny and raw with just a few plunges and pulls no problem
this soldering instructor https://www.youtube.com/watch?v=wPqf2RhRINk recommends brass wool and a Kimwipe. TP is too linty

>>1684977
I find paper towels work better, they don't leave any fuzz behind like some TP. Pads cut from old jeans work best imo

>>1684837
>>1684886
>knockoffs of the Plato 170
Get the real thing for $1.91
https://www.amazon.com/Pliers-Electrical-Cutters-Cutting-Diagonal/dp/B0188DHO40/

>>1684920
I would replace any cap that looked like that.
I'd also replace any other caps of the same brand in the same device.

>>1684991
>TP is too linty
I use professional Torx single ply paper. Because my shitter can't handle multi-ply consumer grade TP, it clogs. It is fine.
>>1684992
Fingers work good too, except it hurts a little bit.

>>1685000
I’ve already started on the big ones. Using all Nichicon caps

>>1685002
Tork*

>>1685002
>shitter can't handle
add more flux

>>1685006
Speaking of flux, what flux should I use? I deal with corroded shit all the time, and rosin doesn't do it, and even Amtech (chink) doesn't do it too, nor FluxPlus.
Only plumbing zinc-chloride flux works, but shit is hard to wash, because it is vaseline based, and is not that water soluble.

>>1682638
can someone help a noob out? Is it possible to use a tl431 to output always 1/2 of the supply voltage? I'm working with something that may be supplied with 3-9V. I'm cracking my skull here but I can't think of anything

>>1684999
yep that's mine. I say it's a knockoff but maybe it's just a blemished piece
>the guy who owns the Washington Post is totally honest guise
kek, at least dealing with ali I have no expectations

>>1685007
to become an educated flux consumer see https://www.ipc.org/toc/ipc-j-std-004b.pdf (the first content page has the table you want, no need to buy the whole standard). in particular, note that fluxes are categorized by activity (which is your main interest), halide content (which is mainly only interesting to large-scale manufacturers and yurofags), and chemical composition (which determines some of the material's characteristics)
I use this stuff. it's kinda weak stuff but works pretty well on clean boards, holds shit in place nicely, is nominally no-clean (which should really be called "don't clean", leaves nasty white/brown sediment depending on soldering temperature if you try) https://www.sra-solder.com/sra-tf5000-no-clean-tacky-flux-2cc-squeeze-tubes-6-pcs I'ma try to get some banana oil and see if a dash into the alcohol wash helps pull that extra crap off without melting my components
for harder jobs I use an RMA liquid flux pen from a non-chinese distributor. comes right off in a two step alcohol/water wash
>Amtech
>chink
lel, you got a tube of vaseline with a bit of flux in it
>vaseline doesn't work
>rosin doesn't work
just rosin? not activated?
>FluxPlus
>posting a brand without a formula
hoo boy
>water soluble
this stuff should take care of some next-level shit while still being very safe for electronics when thoroughly washed off https://www.sra-solder.com/sra-80-water-soluble-flux-pen-refillable

>>1685016
no, the whole point of the TL431 is to provide an absolute voltage, not a relative voltage. how about a resistive divider and (as needed) a voltage buffer?

>>1685020
>no, the whole point of the TL431 is to provide an absolute voltage, not a relative voltage. how about a resistive divider and (as needed) a voltage buffer?
Eh. I wanted to avoid using a opamp but that seems to be the way. Thanks.

>>1684999
>the real thing for $1.91
>Pliers Electrical Wire Cable Cutters Cutting Side Flush Cutting Pliers Diagonal Cutting Pliers
>by PekH

That's not it, Anon
https://www.techspray.com/5-ways-to-spot-a-counterfeit-tool-before-you-buy-it
>Amazon, eBay, Alibaba, and other web retailers, is they are rife with counterfeit products. Scrolling through Amazon listings for “Plato 170”, I spotted 13 fake Plato cutters on the first page
>You can verify if a distributor is authorized by going to our distributor page as well.
The store on your link is not on https://www.techspray.com/distributors

>The cost is much lower than certified Plato distributors.
Check

>The reseller is not clearly stated in the product listing.
Check

>There are misspellings or odd descriptions in the product listing
>As you may have noticed, we call our tool a “Plato 170 Shear Cutter"
>descriptions can get pretty bizarre. “Pliers” (that’s a different tool entirely),
Check

When the product also is at a quarter of the price at Digikey or Mouser ... it's obviously not "the real deal"

>>1685020
>lel, you got a tube of vaseline with a bit of flux in it
Maybe. I don't know.
>just rosin? not activated?
Idk, to be honest.
>posting a brand without a formula
I believe it was RMA.
Thing is, I got it a tray of it from a dumpster near my local electronics shop, because shit had expired. I will never buy a flux which is like $20 for a tube. It worked nicely with relatively clean components, smelled sweet, but was pain in the ass to clean (water won't do it, and IPA is expensive), and won't do oxidized shit.
I also tried using aspirin as flux, but shit is horrible, it terms of smell. And acidic.
>https://www.sra-solder.com/sra-tf5000-no-clean-tacky-flux-2cc-squeeze-tubes-6-pcs
Expensive as fuck.
Shit.
I guess I will end up buying zinc chloride in powder, and mixing it with glycerine or something.
>>1685016
Yes and no.
Most SMPS have TL431 in feedback circuit, you may adjust output voltage by changing couple resistors.
But, if you have to ask such stupid questions, I don't recommend tinkering with SMPS yet, unless you have steel balls.
Also, not all power supplies will work at 1/2 of voltage without removing couple turns from transformer.
Easiest way to do it, is get LM317 or chink DC-DC board.

>>1685016
What the hell are you trying to do? If you just need 6 volts from a 12v supply there are better ways. 317 regulator or a generic buck converter.
If you really need to divide a supply precisely in half, use the voltage divider and buffer.

>>1685007
>Only plumbing zinc-chloride flux works
jesus christ

>>1685034
I have to deal with shit pic related quite often.
In the past I tried removing shit with brush, but quite often I would just rip traces.
Plumbing shit removes green stuff real easy, but I can't remove plumbing shit, because it is vaseline based, and using hot water is not kosher at all.

I guess I will try making my own flux, glycerol and zinc chloride. Maybe phosphoric acid and glycerol...

>>1685007
Bakers no.3 is zinc chloride in a water base. I keep some in the shop for dip-soldering brass barrels onto galvanised steel bicycle cables. This is not stuff you should be using on circuit boards or stranded wires, it's closer to an etchant than a flux; it will run into through-holes and wick up wire insulation, and then it'll fuck your shit up.

>>1685040
>hot water is not kosher at all
well ... neither is salty plumber's flux ?

>>1685042
It is kosher, because flux itself is dirt cheap, and deals with nasty green stuff for me.
>>1685041
Hm.
>This is not stuff you should be using on circuit boards or stranded wires, it's closer to an etchant than a flux; it will run into through-holes and wick up wire insulation, and then it'll fuck your shit up.
I have never seen this though.
What sucks, shit is conductive, so you must wash it of really well. I mean, I shitted my pants when PWM IC exploded on SMPS, because I thought I can get away without washing it off.

>>1685040
I use undiluted phosphoric acid for soldering stainless steel cables. The fumes are fucking brutal.
I have no experience using it as an electronics flux, but a quick patent search suggests you could mix it with high weight polyethylene glycol to make it less runny, but you'd still need to wash the residue off with water. Why is hot water not kosher? Distilled water never hurt anything.

>>1685022
you could use a common-collector buffer if offset from ground is no problem, or two in series if it is. opamps are 2x2mm these days, I fail to see the problem with using them

>>1685040
dear god what am I looking at here, water damage?
throw a dash of halogen acid in your rosin flux. congratulations, you've converted R to RMA (or RA if you went heavy on the acid)
>hot water is not kosher at all
y tho, it's what manufacturers use to get flux residues off

>>1684977
I buy cheap chink 900 series tips in bulk, clean them in the most abrasive ways and throw them away whenever I feel like it

>>1685043
>It is kosher
no, it isn't.

>shit is conductive
that's exactly why it's a problem. you're introducing salt crystals into the electronics. every single PCB you've used plumber's flux on is a fire hazard and you should be held liable for any damages.
I shouldn't have to have this conversation with you:

https://superiorflux.com/plumbing-flux/
>Plumbing fluxes should not be used for soldering electronic or electrical components as their ingredients, particularly Zinc Chloride, may be harmful to such components.

https://dccwiki.com/Soldering
>The flux can never be completely removed nor neutralized by cleaning.
>If the ingredients include zinc chloride, ammonium chloride, [list of compounds] and is not recommended for electrical work.
>A major decoder manufacturer stated that the most common reason for decoder failures is the use of acid based flux

https://www.lifewire.com/types-of-solder-flux-818849
>Inorganic acid flux should not be used for electronic assembly work or electrical work.

https://www.reddit.com/r/AskElectronics/comments/6f2d7w/can_i_use_this_oatey_flux_and_solder_on/
>Nope, plumbing solder doesn't fly well with electronics
>Zinc-based flux aka plumbing flux is a huge no-no with pcb boards. The zinc bakes into the board converting it from highly insulating to slightly to moderately conducting. Once this happens, there is NOTHING you can do to reverse it

https://www.head-fi.org/threads/is-zinc-chloride-flux-safe-on-circut-boards.131509/
>Zinc Chloride is a corrosive, "acid" type of flux. In very small quantities it can aide in soldering on corroded parts, but will also eat copper when used in higher, pure flux concentration.

https://www.vintage-radio.net/forum/showthread.php?t=85062
>about the last thing you would want near a PCB

https://forum.arduino.cc/index.php?topic=331932.0
>You have been using acid flux, which will eat through your circuits over time if you fail to remove all traces [which you will]

How can you not know this

here's a fun read about flux residue from back in the leaded days, those of you still using Pb solder and having powdery crap all over the board should read this https://www.kester.com/Portals/0/Documents/Knowledge%20Base/White-Residue-15Apr1988.pdf

>>1685072
he's obviously Ukrainian, they don't do laws

>>1685076
>they don't do laws
ah yes, unlike Russians

>>1684913
I store my toothbrush in my bedroom. Whenever you flush the toilet it puts a fine mist of toilet water into the air, and even though I always flush with the lid down, I can't say the same for my other flatmates.

At least I didn't brush my teeth with solder paste.

>>1685053
True, phosphoric acid makes horrible fumes. But I think it depends on concentration too.
>Why is hot water not kosher?
It is more expensive than cold water.
>Distilled water never hurt anything.
Oy wei, no distilled please, it is expensive.
>>1685058
>dear god what am I looking at here, water damage?
Water damage.
>>1685072
>no, it isn't.
It is kosher, because it is 10 times cheaper as something as effective electronic related.
>The flux can never be completely removed nor neutralized by cleaning.
This is not true at all. You can indeed wash it of completely, even under BGAs. If it wasn't the case, I would receive all my shit back with same green stuff.
>The zinc bakes into the board converting it from highly insulating to slightly to moderately conducting
How?
>How can you not know this
Easily, I disregarded the warnings, as usual, and tried it anyway. Copper gets green only if you leave flux like for weeks there.
As for conductivity, I figured out that soapy water removes shit really good. But for BGAs and QFNs you have to borrow ultrasonic, because vaseline. Without vaseline shit should be easy to remove.

>LF347
>pic related
>absolute maximum = ±18V
Should I be feeding it a higher Vcc than Vee? Does it matter? I plan a total ∆V = Vcc-Vee = 28V or so.

http://www1.idc.ac.il/tecs/plan.html?fbclid=IwAR2_ymDChipTPjBdxQMF0JgjRW4ur_-Egu4kJeJQepMtECJMp2LG2a-3stw

Are there any good resources like this for RF Electronics/Engineering?

>Chink power supply
Check
>No ground
Check
>No filter
Check
>SMD resistor sandwich in order to fool feedback circuit
Check
>Reclaimed capacitors from 30 years old TV with capacity lower than needed
Check
>Soldered with ZnCl2 flux
Check
>220V HI-POT test
Check
>No fuse
Fail.

So, I modified laptop charger I found in the dumpster into 22V SMPS for T12 soldering station. Or TS100. I don't know yet.
Why 22V? Because power supply is rated for 65W, and T12 will draw 61W at 22V, at least chinks told me so.
Rate my shitty soldering skill, niggaz

>>1685164
>>No fuse
looks like there's a fuse where a filter choke is supposed to be.
I tentatively rate an A but that will drop down to F when it catches Fire
give her a hot supper and see if the smoke comes out.

>>1685172
>looks like there's a fuse where a filter choke is supposed to be.
Silkscreen is misleading.
Where you are supposed to have "F1", you actually have a trace.
Where you are supposed to have "L1", you have one leg shorted via a trace.
>I tentatively rate an A but that will drop down to F when it catches Fire
No fire yet, but caps on output might explode. I needed to swap stock 2x 25V 1000 uF caps for 35V ones, but I didn't had fitting ones, so I used 330 uF from 1990's Samsung TV (made in Korea), and from 1980's TV 220uF one.
On soldering station board I would definitely add more caps.

Also, heatsinks are cast zinc, not even aluminium.

>>1685178
Also, no proper Y1 capacitor, which would make one Czech guy sad.
This is probably why I should ground the tip after all.

bought a little SCART breakout box for audio. but there's major buzzing. seems like a shielding issue? how can i beef up the audio outs so there's no buzzing coming through?

>>1685249
Sure it isn't a ground loop?

>>1685105
are you confusing the inputs with the supply?

>>1685249
when connected to what?

>>1685252
>are you confusing the inputs with the supply?
I assumed it was the supply, since it has a minimum value of ±11V, instead of ±0V like the inputs should be. The output voltage range is ±13.5V typical.

>>1684000

>DIY shithole

man, say what you want, but the number of chink devices I've fixed based on semi-comprehensible Indian you-tube videos is truly staggering.

>>1684295
>He didn;t go with the 9 segment to display proper R's

Disappointing

>>1684067

Diff. Eg. was great because I had this stoned out grad student that was strangely good at explaining things as a section teacher. Pretty easy and the shit we learned was actually useful. might depend on syllabus.

Calc III sucked because most of it felt like contrived busy work, and it was. I basically relearned all multivariable and vector calc in signals and when it became actually useful.

Calc II wasn't bad.

PDEs/Tensors sucked for me. it was a fucking firehose of info in the middle of my capstone year. I have no idea why my school made that a single semester course.

>>1685254
Dude, I bought my laptop after one needful review in Hindi.
Other reviewers were just reviewing number of cupholders and how they don't like plastic.

How do i get outlet 2 to only power up when there is a load on outlet 1 I’m guessing that I need a relay ,just not sure how to wire using AC. If it were DC it would be easy

Bump

>>1685308
Get a master/slave power strip.

>>1685251
could be ill look into it
>>1685252
source comes in from left, where you can extract composite, svideo, audio, then out to the right which plugs into your TV

>>1685314
Thats exactly what I’m needing....thanks anon

>>1685253
>minimum
yes, that's a characteristic, not a condition. when the designer satisfies the conditions, the designer can expect performance according to the characteristics
the maximum INPUT voltage range a part will support, either tested or by design, will be at least ±11V. typically the maximum input voltage range a part will support is -12..+15V but it's not guaranteed
supply voltage is an entirely separate animal, Pic related. a "typical" voltage would be meaningless

>>1685256
>not the Rs
that's not what your mother said

>>1685315
yes, and the taps?
I ask because digital computers and devices since years ago force their local """ground""" outputs above actual ground so they can produce negative left/right voltages (wrt """ground""") with a single supply voltage and ground, without requiring a large capacitor on each signal out. if those """grounds""" are otherwise connected to ground via power supplies and/or other ports, you will hear the audio codec and its power supply circuitry "laboring" into the short circuit as irregular ticking or squealing on the signal. it's a particular sort of ground loop that comes up a lot these days when people try to plug their devices into home A/V equipment. I haven't seen exactly that on a device input lately, usually it's on outputs
anyway, if that is the problem, you can work around it by inserting sizable capacitors into both signal *and* ground of each audio channel to break the dc loop. not a bad idea to keep an RCA cable or two hacked like that

>>1685359
Oh ok. Well I've got it on a -12 +15 split rail anyhow because 27V is what my boost converter is set to output, and my ground is made from a L7812. Using the 4th op-amp to add an optional independant gain stage if I want it.

>>1685375
>ground is made from a L7812
probable fail. 78xx can't sink current
if you don't care about exactly splitting the supply in half, use a TL431

>>1685311
>bumping past the bump limit

remember, faggots, wait for page 10

>>1685359
i'll get you a pic after work
its not ticking noises im hearing its just a loud bzzzz like what you hear if you put your finger on an rca plug tip while its plugged into an amplifier. it only happens when there's no sound coming out of the game, oddly

>>1685415
here you go
https://www.amazon.com/BESIGN-Ground-Isolator-Stereo-System/dp/B06XQYN77L

>>1685419
>https://www.amazon.com/BESIGN-Ground-Isolator-Stereo-System/dp/B06XQYN77L
is there something like that but with rca jacks instead?

>>1685423
/ohm/ RULE 1

>>1685431
how embarassing, thankfully this is on its way to page 10 :^)

>>1685359
>that's not what your mother said

Yeah, well, she knows how to multiplex properly.

>>1685432
>single supply audio shit
>having to add a voltage clamp (dc level restorer) at every input
>having to remove dc level at output just to add it again so to avoid saturation in mixers and amplifiers

>>1685447
no idea why I quoted that guy

What's the typical use case for logic level shifters?
Most 5V inputs will read 3.3V as high, and a simple voltage divider lets a 3.3V input safely read a 5V signal.

>>1685447
>having to remove dc level at output just to add it again so to avoid saturation in mixers and amplifiers
Or you know, you can design a proper direct coupled amplifier with global negative feedback and then you only have to bias the input and remove the DC level at the output.
There's no need to capacitively couple every stage.

>>1685456

5V/3.3 V are both TTL levels.

level shifters get used for TTL to ASIC/MCU CMOS logic, especially on low-power chips. There's some very low-power chips out on the market where both "Low" and "High" voltage sit in the low voltage range for a 5V TTL chip. that's generally where level shifters get used.

that being said, many modern chips don't require it much anymore because they'll integrate some form of level shifting or isolation in the package. there's some niche stuff where you might need discrete level shifters, and repairs on older equipment might need it quite a bit.

>>1685467
>5V/3.3 V are both TTL levels.
What exactly is meant by this? TTL means the logic gates are built with BJTs, right? But as far as I know, most ICs today use CMOS transistors and they still operate on 5V or 3.3V

>>1685456

this >>1685467
plus the opposite of this. some CMOS logic likes higher voltage, and some industrial "TTL" logic likes all the way up to 24 Volts. Discrete level shifters can be used in situations like that.

>>1685470

>TL means the logic gates are built with BJTs, right?

yeah, it's sort of a holdover from when TTL was the standby for discrete 7000-series logic chips. Now it's sort of just "discrete 5V or 3.3V logic chips."

>But as far as I know, most ICs today use CMOS transistors and they still operate on 5V or 3.3V
They often do that with the help of level shifters or isolation components. it's not that you can't design CMOS logic that isn't friendly with 5/3.3, it's just that there's no reason to and things are generally smaller and lower power when the voltages are lower. Take the voltages on the CPU you're running right now, for example.

>>1685442
8/10 multiple keks were had

>>1685457
indeed, the discerning audiophiles don't cap-couple very much

>>1685456
>most
if designed for TTL compatibility, yes
5V TTL logic thresholds are close to 3.3V CMOS logic thresholds, but a 5V TTL output needs a pullup to meet 5V CMOS input specs. close enough for government work, but unsafe above kiddie car speeds, especially if mixing CMOS and TTL devices on the same bus
>voltage divider
nice RC time constant you got there. you're using 0.1% resistors to minimize signal skew, right?
>use case
when you care that your rising and falling edges look good

>>1685470
he's talking out his ass. 5V and 3.3V are I/O supply voltages. thresholds (TTL, CMOS, SSTL, LVDS, etc.) are orthogonal to I/O supply voltages (5V, 3.3V, 1.2V, etc.). usually 25% or 75% of I/O supply voltage for CMOS, usually < ~1 or > ~4 diode drops for standard TTL

>>1685475
>he's talking out his ass.

How exactly

> thresholds (TTL, CMOS, SSTL, LVDS, etc.) are orthogonal to I/O supply voltages (5V, 3.3V, 1.2V, etc.). usually 25% or 75% of I/O supply voltage for CMOS,

Yeah, thanks to output stages. it's not like you're running individual arrays off a fucking 5V rail.

>>1685475
>indeed, the discerning audiophiles don't cap-couple very much
>>1685457
>Or you know, you can design a proper direct coupled amplifier with global negative feedback and then you only have to bias the input and remove the DC level at the output.
There's no need to capacitively couple every stage.
I wish.. It's a phaser pedal. Luckly mine won't have that many stages.

>>1685478
actually I just stop being retarded and make a input buffer with reference to the well.. reference voltage and use that as a virtual ground

Is this a total loss ? If not, what could I do to fix it?

It was holding a 3 prong semi-conductor.

>>1685399
>probable fail. 78xx can't sink current
uuh, can I still use it if I have a really big output filter cap after the 7812? It's 1mF because why not

>>1685480
Yeah that can be fixed, scratch off the nearby solder-mask in a convenient pattern to expose solderable copper. If your (replacement) component has room on the leads you could bend them over and onto this copper for soldering instead of trimming the flush, otherwise a bodge wire or two should suffice.

I'm guessing the problem arose because it can be difficult to pull a 3-pin component out when you can only reflow 1 or 2 pins at a time? For non-plated through-holes it should be quite viable to use a cheap solder-sucker on the copper side of the board and remove basically all the solder connecting the part to the pad, any scraps of which can be picked up with solder wick. Or just wick the whole thing.

>>1685442
oh wow, this had me chuckling for minutes.
I've never appreciated those kinds of jokes but this was golden
well done, anon

>>1685480
>>1685482
Oh and you might not even need to scrape off any solder mask if you solder the leads/bodge wires to that test pad for the middle hole and that through-hole lead for the bottom hole.

>>1685475
>out his ass
>orthogonal

>>1685482
>>1685484

Thank you

>>1685478
>>1685479
Just built half the circuit, working dandy. Now I just need to make some gheto optocouplers and a LFO.

>>1685479
>>1685506
Yeah that's great and all but nobody asked for a play-by-play of your circuit design. I don't care and neither does anyone else. If you aren't asking a question then don't post. This isn't your blog.

>>1685481
Ok so at 1.4mA it drifts by 50mV, at 4.4mA it drifts by 150mV, and at 14mA it drifts by 560mV. None of them are that bad, and are probably fine considering the kinds of impedances this thing will be driving, but for peace of mind (and decluttering some of the circuit) I think I'll ditch my seperate op-amp gain stage and use that op-amp to buffer a voltage divider.

>>1685522
Oh another point, anyone know if I should have normal reversed silicon diodes going to the ±13V rails as my input protection, or go for back-to-back 12V zeners to my 0V ground rail instead? The silicon diodes would be the usual choice, but since the output is a boost converter the only place such an overvoltage condition could send current to would be my filter capacitor. That works fine for medium/high frequency AC, but for DC that doesn't help at all. The zeners would actually short the signal across the terminals it's outputting, so all I'd need to do is add a 1kΩ or 10kΩ input resistor to stop the zeners from exploding, and I think my input characteristics will remain ok.

>>1685470
>most ICs today
>they still operate on 5V or 3.3V
many that operate at 3.3V have the option of running at a still lower voltage. 3.3V CMOS is a good convenient compromise logic level standard when you need to leave a system open-ended. but in CMOS every transition involves at least two capacitors charging or discharging, and maybe you want to limit the amount of energy being transferred from one rail to the other for many legitimate reasons

>>1685485
>he doesn't know about orthogonality
w
e
wewew
e
w

>>1685506
come back with the finished project

>>1685517
personally I love seeing finished projects but yeah that's too close to blogging for my taste too

>>1685524
TVS would be the "right" way to go, but zeners might do if capacitance is low enough below the zener voltage

>>1685527
>newfag can't triforce

>>1685527
According to my not-small-signal DMM the reverse capacitance is below 0.5nF. I'll go for the zeners, I can always remove them later if I need to.

>>1685256
i would have had to go QFN to fit enough pins. the board is pretty skinny, the characters are only 0.28".

>>1685164
Ok. This shit makes clicking noises in idle, but it delivers 2,3 A @ 21.8 V no problem. So 50W, which is plenty good.
Caps get really hot though. Need more caps. Also transformer gets warm.
I guess I got chinked, and I would need a fan to cool it.

>>1685533
Wait a minute, zener diodes for clipping doesn't make any sense when I have a differential input. Better swap to normal diodes.

>>1685574
Ok, now it works, and it looks great. I plan on encasing it in resin if a friend has any left over, for now I'll leave it as-is.

>>1685599
neat, now do it with SMD components

>>1685548
I was just giving you shit man. It's pretty cool.

>>1685607

>*Pulls out the copper foil*

>>1685621

Question regarding 3 phase DC motors.

I have two motors a 24v 25w and a 24v 50w motor. If I apply the same same current and same voltage to both, would the winding resistance be the reason the 25w motor takes longer to spool up and down under the same settings as the 50w motor. In particular in relation to the acceleration and deceleration pulses? I'm still pretty new on this subject so if there is some good reading material to cover it I would appreciate that so I can work it out and understand this better. Google hasn't exactly been helpful.

>>1685632
P = V*I
I = P/V
The 25W one will pull a current I ≈1A in operation, meaning it has an impedance of R = V/I = 24Ω. The 50W one will pull a current I ≈2A in operation, meaning it has an impedance of R = V/I = 12Ω.

Torque is by definition proportional to the current (density) through the motor, so a lower resistance will limit this current less. At startup, the motor is trying to apply as much torque as possible in order to accelerate, hence current limiting = torque limiting = slower acceleration.

Note that this is under the assumption that the number of turns in each motor is the same. If there were twice the turns in the 25W motor, the current density would be equal, but I think the greater reverse emf would take on the job of making the motor start slower.

>>1685599
I/V curve of 2.7V zener diode:
digital scopes are shit though, it should look much better when I try it out at home.

>>1685645
Thank you, that's pretty much what I had come to up until the turns in each motor. The numbers I gave were kind of just examples. but really the lower and higher watt motors actually have fairly close amp readings.

Something like 25W, 16v-17v and a readable current rating between 1300-1600mA. While the 35W motor I also read was getting 1200-1600mA, but a 24v rating. Which as I had done the math before lines up fairly well.

The odd problem is that the 25W motor has a longer spool up and slow down time under the same settings as the 35W motor. But the 25W motor takes about 1 second to spool up while the 35W is instant. I appreciate the help as it's going to help me look into a few other things from the team making the motors.

>>1685482
I have a simple 5$ pop-solder sucker but i just hate using it.. i also have the copper braided stuff to absorb solder (cant member the name). But i had trouble using that because the pins were so close together. On top of that my solder gun is kinda shit.

But I havent upgraded, because i dont know if im just really shitty at soldering, or if its just the equipment. So i dont want to spend $80 just to confirm im shit at soldering.

>>1685652
You can always try stalling the motor (holding it by hand should suffice, if it's even necessary) and measuring its DC resistance and inductance, which should give you a deeper insight into what's really going on there. If you lack a way to measure inductance properly, I think a 555 relaxation oscillator circuit can be used with an LR or LC low-pass as its delay element. To measure the frequency without a scope just feed the waveform into a piezo or high-impedance speaker of some kind and record the audio with your computer or phone, free software like Audacity can analyse it and give you the frequency. The frequency measurement wouldn't give an inductance value that's too mathematically precise (± a factor of 2), but all you'd need to do is compare the frequencies of one with the other.

>>1685667
Delaminating pads sounds like an issue of being too hot for too long. If you don't have a temperature-controlled or variable iron there's less you can do about the former, but the latter can be fixed by just getting in and out quickly. Even if it's a cold joint, get in there and get out after a second or three and give it a little time to cool down. Often the cause of having to have your iron on a spot for too long is that the thermal conductivity from your iron to the component/pad is too low, this can be improved by having a tiny blob of solder on your tip just to meet with the component and make for a much larger area through which to transmit heat.
The only solder sucker I have is a $5 aluminium one and it works well enough for what I do, but I've noticed a fair difference between the cheap solder wick and the name-brand (Goot) wick. The difference may just be whether there's flux in them. On that note, get some external flux, at the least it will eliminate one unknown, and at best you'll never get a dry joint again. I use rosin, never had much luck with the no-clean pens.

>>1685667
>pop-solder sucker
they're based, just pre-position the sucker so you can tip it over onto and flush with the joint once the joint is well above liquidus temperature. there should be no more than 1/3 second from the solder tip breaking contact with the pad and lead to activation of the sucker. if a partially-sucked joint is giving you grief, add fresh solder to the joint and try again. desoldering practice is a fine use for random dead THT assemblies
don't forget to remove the tip end of the vacuum chamber and clean out the tip when it has solder shards sticking out when the plunger is fully pushed in past the locking point
>copper braid
solder wick. these usually come impregnated with a dry flux. don't care, add more. if you're really cheap you can maybe visit some music stores in town and see if they might cut you a deal on some stale rosin for violin bows. thin it with IPA to desired consistency and put it in a needle-tip bottle, or into a small lidded tin for application with a toothpick. viola, cheap rosin flux
>solder gun
>kinda shit
for small work, you are correct. they're much better suited to large work like antenna building, shielding cans, thicc wire, etc. hit up ali for a 936D clone and a tip assortment so you have a more repeatable platform to learn on. $25 isn't too bad. also, find a good light source to illuminate the board very well while you work
>shitty at soldering
it is a skill that can be acquired with practice and proper tools. some things with a lot of mass are just really hard to solder or desolder without more heat than an iron can bring, e.g. multilayer boards
check out jkgamm041's youtube channel. start with the tip care video. then watch this master do some soldering. bring a tissue. and lotion

>>1685599
>>1685648
>tfw the 1000µF capacitor is too beefy for my computer's USB port to handle the inrush
I shoulda seen it coming. I guess 100µF will be ok? Should I go down to 22 just to be sure? It's after the boost converter so I didn't expect it to get charged up that quickly, but I guess this is a pretty beefy boost converter.

>>1685687
>see if they might cut you a deal on some stale rosin for violin bows
Never thought it went stale, might give it a try when I'm near a music store next.

How do I increase efficiency of switch mode power supply?

>>1685690
What do you have control over? Changing base frequency, switching algorithm (i.e. bang-bang, variable duty-cycle, variable duty-cycle and frequency, and with what response curve) replacing MOSFETs with ones with better switching characteristics, using larger heat-sinks or stronger fans, even changing thermal grease will all have some sort of effect.

>>1685648
>>1685689
There we are, 2.7V zener in CRT beauty.
Trying a negistor based on a BC547, it's hitting a normal reverse breakdown at 0.3V or so in one direction, but I'm not seeing any kind of negative resistance curve or any difference in the send/return paths from one another. At least not at a low frequency, at higher frequencies every low-current path acts like it has a capacitor in parallel. The threshold frequency for this reversed BC547 is about 1kHz, which is low enough that it might not be capacitance giving me this activity.

>>1685695
IN CRT BEAUTY

>>1685692
>What do you have control over?
Let's say everything, I'm just curios how do they achieve 90% efficiency figures on low power supplies.
It is flyback laptop power brick. Cheap one. Which is probably 70-80% efficient (I need to measure it).
Meanwhile brandname bricks (LiteOn, Delta, etc.) have efficiency in 90% I guess.

Fast examination revealed transformer is getting hot (core itself, not windings), and output capacitors (well, they should since I replaced those with shitty ones).

>>1685699
Because I can't tell the difference judging by PCB. Same power part, similar PWM IC, similar MOSFET, similar Чeткий diode. Sure, filters don't exist, but I doubt they have to do anything with efficiency.
Better iron core material? Better transformer calculations?

>>1685599
that's a big capacitor