/diy/ - Do It Yourself

>>1679331

>Post your setup

P-Plz no bully

Still learning, but from what I've been able to understand is the following.

>have old PC 4 fan controller that runs on 5v or 12v
>each fan in this are between .2-.5 amp draw
I have a spare 12v 3amp psu in my box of shit. I could re-rig that to a 4 pin molex and run that into the power of this controller and essentially have it running on it's own correct? This is my first time dealing with wall power, so trying to do this safely and will be putting a breaker between it the first time. But from everything I know this would work right and safely correct? Since I'm on voltage point and well within the current.

>>1679355
Yep. If it's the cheapo 3 speed switch style fan controller though it'll only work on high speed mode, the medium/low settings need a 5V rail there to get either 5V to the fan or 12-5=7V.

>>1679361
They are both Scythe Kaze Master II controllers. Pretty decent quality controllers that I used 3-4 years ago and forgot about in storage.

>>1679337
comfy

>>1679355
sure.
here's my somewhat related abomination I threw together during last summer's heatwave using a laptop charger.

>>1679337
nice. what do you use all of it for?
would you recc those helping hands?

>>1679371
Looks like it should work fine off just 12V then

Looking at soldering irons & stations - among others the toy in OP.
4.8/5 stars out of 769 reviews? Is this real life?

>>1679406
It's what everyone uses for a reason. Make sure to buy some tips too.

>>1679406
Stupid button interface, not heater in tip. I'd much rather have a T12 chinkstation.

>>1679396
>>1679376

Thx.

I do mostly HV stuff. A bit hard to explain, but I work with a lot of tubes and klystrons. The rest of my equipment is out in the garage at the moment.

>would you recc those helping hands?

It's just a cheapo Neiko I picked up at a local store. I would go for a proper board clamp plus the QuadHands. I'm gonna be picking up the QuadHands myself soon.

>>1679430
How high are we talking?

I've been working on 20-30kV stuff for work but I haven't gotten to really interact with it much. I'd love to design our new HVPS.

>>1679406
for that price and what I've seen else on Amazon shouldn't there be something better than that? I feel like that's not really that good.

So I'm trying to teach myself electronics and I've got this question.

Is there some shortcut/idiom for reducing voltage (from 12v to 5v in my case) or you always need either a linear or a switching regulator?

>>1679406
A T12 chinese station is 90% as good for a fraction of the price. Get it from KSGER or Quicko on ali or banggood, they're reputable enough.

>>1679455
resistors

>>1679436

I go ~ .5 - 10kV for some more of the pedestrian stuff, but that gold box you see in the upper corner is the core of my 40- 70kV setup for x-ray tubes. it's hacked out of an old Scintillator screen illumination source. I built up a control board for it a few years back and I'm in the process of updating it, which is why it's in my office right now.

By the way - sounds like you work for some HV guys - wouldn't happen to be Spellman, would it?

>>1679459
K thx I'm a dumbass.

>>1679455

Look up voltage divider.

That should really be, like, the very first thing you come across in learning electronics.

>>1679459
This, but remember the output voltage will be dependant on the amount of current you're pulling. A zener diode with a resistor is a simple-man's linear regulator, and can be a better option when you have an unregulated voltage that you need to clip regardless of if you have no output current.

On this note, is there a single component (or component + resistor) that will give the accuracy and ripple-rejection of a low-voltage output linear regulator, but the ability to clip voltage with no load and sit on a potentially high voltage source, so I could use it with something like a capacitive dropper circuit instead of a zener?

>>1679464
Actually I know about voltage dividers and it was the first thing I came across, but I'm still trying to wrap my head around basic load and what the tradeoffs are. I'm trying to do something that's optionally battery powered.
Will report back if I need more sense knocked into me.

>>1679469
What are you trying to do in particular?

>>1679476
I'm making a circuit to control a high static pressure fan that fills a bellows that feeds a small pipe organ. It's easy to just run the fan at 100% and bleed off excess pressure mechanically, but I'd like the fan to be quiet and only use as much power as necessary.
Gonna try to use this P1K-5-2X16PA sensor, an opamp, and a 555 timer (which will all be at 5v) to control a 12v fan.

>>1679480

You're gonna try and do an op-amp PID? or are you doing a wonky PWM with it.

Out of curiosity, did you guys buy, or build your own work stations? Anything you looked for to help you expand or make it easy to use?

>>1679486
My thinking was to use an inverting amplifier, or maybe use a comparator and do a kind of damped bang-bang setup. Then do PWM with the 555.
Are there any good resources for learning about PID?

>>1679494
>>1679486
K I've been reading about PID and this seems like the proper way to do things. Ty.

>>1679458
>>1679406
>>1679449
its pretty nice though

>>1679494
>Are there any good resources for learning about PID
I looked them up on youtube, watch a few of the 5 minute videos, they'll get you where you need to go.

In essence, the P stands for proportional, as it makes a counteractive motion that's proportional to its current position. If it's a ball on a balance, it would be tilting the balance to an angle that's equal to some constant multiplied by the distance the ball is from the centre of the balance, so it always tilts to move it back to the middle. In your case it would be just having your fan speed be proportional to the pressure difference between what you want the pressure to be and what it actually is. A simple proportional control circuit would be easy to do with a single op-amp with a voltage divider/potentiometer on one input, the pressure sensor on the other input, and the output feeding a transistor that turns on the fan. Just this alone might work fine for your purposes, I'd try it anyhow, pic related.
Instead of using a transistor linearly, you could instead feed the output of the op-amp into a comparator (or second op-amp) -based oscillator PWM circuit.

>>1679513
I'm going to keep reading reviews and looking at other equipment. But I'm thinking about snagging these two this week.

>>1679510

No worries. with a proper op amp and motor driver chip, you'll be able to do rather good motor control and not worry about regulation too much on a 12V supply.

>>1679514
>In essence
you'd be surprised.

>>1679514
Cont.
(note that the + and - inputs may be the other way around, I didn't read the datasheet past the word "linear")
The problem with proportional control systems is they have a tendency to oscillate if there aren't enough system losses (i.e. the ball rolling back and forth on the balance). This is why the differential aspect is handy, it slows down the change such that it comes to a rest at the equilibrium position relatively quickly. But if in your system you have enough of those differential (i.e. friction) losses, this isn't strictly necessary. I suspect this to be the case, but it would depend on the rate at which you drive it, which is dependant on your amplifier's gain, which in my circuit is rather high.

The other problem with proportional control systems is they can get stuck (i.e. the ball not moving even though it's on a slope). This is where the integral part comes in, if your system gets stuck at an unwanted steady state for too long, the circuit will continue increasing/decreasing the output power until it gets unstuck. In your case it would be like if you covered the inlet with a piece of paper to stop the pressure increasing, the pressure would therefore continue to increase until the paper is breached. But considering the nature of the circuit I posted the damn thing would go pedal to the metal instantly, thanks to its high gain.

The high gain of using nothing but the pressure as part of an op-amp feedback loop could cause some issues of oscillations and delays, so perhaps using a gain-limited differential/instrumentation amplifier circuit would be a better idea. Probably instrumentation in this case, considering the output impedance of the voltage divider.

>>1679518
what did I miss?

>>1679522
cont.
A proportional feedback loop with too much gain has less of a need for an integral part, but a stronger need for a differential part, while a proportional feedback loop with not enough gain has less of a need for a differential part but a stronger need for an integral part. You may be able to get away with just P, or just PD or just PI. I think op-amp integrators are easier than op-amp differentiators.

>>1679337
Cute HAF912.

>>1679514
>>1679517
>>1679522
Thanks again bro. I'm learning a lot. I think PID is the way to go, but I'll probably test a simpler setup. I'm still bringing up my fundamentals.

Are there inexpensive full packages that help you set up one of PID systems? A quick google is inconclusive. It looks like you can make them out of opamps though.

>>1679523
Ah noted.

Pic related is from the fan docs btw.
K I actually have to go to bed now. Probably ordering my parts tomorrow. Good shit. Ty.

>>1679525
>full packages
From what I've seen you can buy these all-in-one PID machines with a built in display, power supply, and wires for a SSR or TRIAC, usually designed to work directly with a thermocouple for feedback, but they're pricy. I think even some PLCs have them built-in.

I imagine there are standalone ICs for it, for that I'd browse through digikey or someone like that's catalogue.
Wait what the shit I can't find any. Are you telling me there's no package with ~6 or so op-amps in there along with a bunch of pins for passives to work as a PID controller? Or is it because at that point would it just be a quad op-amp IC with a couple of buffers? At least having the PWM circuit in there with an external timing capacitor would be somewhat worthwhile.

>>1679524

Ah yeah, it's actually not a bad case. Got it for like 20 bucks open-box from Micro center. Given what cooler master usually does with cable management, I was pleasantly surprised.

>>1679530

Two quad packages would get him there. the kind of control he's doing doesn't require Hi-frequency for the Int and Diff, which are trivial to do, so
chip 1-
1 set up as a diff amp for error signal

feeds into 3 set up as P, I, and D respectively

chip 2 -

1 set up as inverter

remaining 3 to generate PWM.

>>1679530
>At least having the PWM circuit in there with an external timing capacitor would be somewhat worthwhile.

Basically a DC-DC converter controller. in fact, if he could find one low frequency enough, a DC-DC controller plus a dual amp would do this for him. most even have FET drives on the output to setup a half-bridge.

Got myself a scope meter for 50 bucks, it's old as shit and the batteries are shot, hopefully I can mod it to take 18650s or something similar.

>>1679551
>1 set up as a diff amp
No buffering?
>1 set up as inverter
Don't you need one for summing and one for inverting, or can both be done in the same unit? Actually who cares about inverting when you can just swap the pins of your PWM comparator.
>remaining 3 to generate PWM
Pretty sure that only takes 2 units, so you have some leeway.

>>1679553
>dual amp
Do you somehow mix the feedback paths of a single op-amp to have more than one of P,I,D? Or are you suggesting that he doesn't have to use one of the three feedback methods?

>>1679558
>Or are you suggesting that he doesn't have to use one of the three feedback methods?

This. the Dual amp would just be set up to sum and invert a proportional signal.

Reading what he's trying to do, he basically has a "set" pressure he's trying to maintain, which doesn't really change, and I imagine that most small organs don't really crank down air at a killer rate, so the change in his pressure sensor signal will be a relatively small d(pressure)/dt.

that being the case, going full-balls into PID may not be needed when just proportional would work.

for the DC-DC controller, most have inputs that can be setup to operate as unity follower(buffer) controls. If you look at the block diagrams on most, they literally block out an error op-amp and a switching op-amp.

>>1679558
>Pretty sure that only takes 2 units, so you have some leeway.

two for triangle, one for PWM. cleanest way to do it in a linear fashion with controllable frequency.

>>1679558
>Don't you need one for summing and one for inverting, or can both be done in the same unit? Actually who cares about inverting when you can just swap the pins of your PWM comparator.

and this is a good point

>>1679561

sorry, meant dual amp set-up to diff and invert.

>>1679561
>going full-balls into PID may not be needed when just proportional would work
I'm thinking you'd still probably want another form of feedback because of the delay in the system, but if you tune your diff-amp's gain properly I guess it's fine. I'm not well versed in the use of gain-limited op-amps in feedback loops. At the very least I can imagine situations where the open-loop gain is so high you need PD control, and where the gain is so low you also PI control. No clue whether there's a sweet spot in there.

Also I like to use ~100k - ~1M pots and resistors for tuning but if you want to use a diff amp instead of an instrumentation amp you'd need to use lower values. For an all-in-one IC I'd hope it would have buffered inputs. Gonna design one in KiCAD now and see how many pins it needs.

>>1679562
>two for triangle
One for triangle. If you make the Vpk-pk of the cap-charging wave at ~20%Vcc or less then it's hardly curved at all. Frequency is still highly tweakable (unlike those fucking 555 PWM circuits) just by changing that one resistor.

>>1679569
>Gonna design one in KiCAD now and see how many pins it needs.
Ok so it's:
>2 power pins
>2 inputs
>3*2 for each PID gain element (may be able to cut this down to 3*1 with a different topology whose components go to ground) (may need to be 3*3 for control over both gain elements for more range)
>2 for summing amplifier gain
>1 for cap to ground from oscillator circuit
>1 output
Which adds to 14 (aint that neat)
Could be as low as 11, as high as ~20 for extra pins around the oscillator.

>>1679574
Wait I forgot the 2 pins for the instrumentation amplifier gain, not sure how needed it is alongside the summing amplifier's gain resistor though. Using a non-inverting summing amplifier could also lead to one less pin, but the resultant gain limitations would be somewhat problematic, as for the other circuits in there.

>>1679337
Looks a bit messy, but comfy. I'd love to have such a setup.

>>1679557
Fluke 97 ScopeMeter battery upgrade to Li-ion
https://www.youtube.com/watch?v=pM8xTWhUnNk

>>1679406
get a gas heating solder instead.

>>1679427
in industry they don't go switching temperatures every 5 minutes

>>1679516
the only thing I dislike about that Yihua is the button temperature interface. otherwise a decent kit for the money, but check https://www.amazon.com/Aoyue-888A-Digital-Soldering-Station/dp/B01NBUJTTS if you too prefer to twiddle knobs

is bump limit different on boards?
i see this thread is up when old one has 330 posts
i remember 750 being the number, apparently its board specific or op is a retard

>>1679614
OP here, I can confirm that I am indeed retarded - but also mainly used to faster boards. won't do it again

>>1679601
>in industry they don't go switching temperatures every 5 minutes
I don't have tens of thousands of components, with similar heat tolerance, to solder to thousands of circuit boards though

>>1679635
>thousands of components
Not needed. You can push buttons and watch display without soldering anything.

>>1679635
right, and that's why the clones with actual knobs are fine for you
btw a full reel of 0402 resistors, 10000 of them, costs $1-$4 depending on value and tolerance

>>1679635
>not having tens of thousands of components, with similar heat tolerance, to solder to thousands of circuit boards
sounds like malpractice
a reel of (10000) 100nF 16V 0402 caps can be gotten right this second on LCSC for $5. need moar cap? scrape off a bit of solder mask and slap that puppy down, just about anywhere

>>1679644
>>1679647
>10000 resistors costs $1-$4
>10k caps can be gotten right this second on LCSC for $5
well that's not something I want or need.

>>1679616
you done good by her, son, 10/10 thread image. the hair-trigger gets better with age

>>1679614

nah, you're just a newfag. /ohm/ does not re-generate based on post count. We wait until the old thread is on page 10. It's on page 5 now, so for a considerable time we will have two /ohm/ theads, and newfags like yourself will post in both of them producing spaghetti conversations.

Proud you should be.

In April 2018 Anon posted this:
>>/diy/thread/S1375447#p1375853
>>1375853
>>400 degrees
>Celsius?
>Fahrenheit?
>Kelvin?

I'd just like to point out that Kelvin doesn't use degrees.
It was also obvious that OP referred to °C and not °F.

>>1679730
>I'd just like to point out that
the name of the unit is kelvin, not Kelvin.

>>1679749
>>1679730

>>1679749
Then I can let you know that while a unit in the Kelvin scale is referred to as a kelvin, the scale itself is referred to as "Kelvin scale".
When someone says that "Kelvin doesn't use degrees" then you should reasonably assume that they are saying that "the Kelvin scale doesn't use degrees" rather than "the kelvin unit doesn't use degrees".
The latter would be silly, especially since I wrote "that Kelvin doesn't use degrees" rather than "that a Kelvin doesn't use degrees".

>semantic satiation attained

>>1679569

>One for triangle. If you make the Vpk-pk of the cap-charging wave at ~20%Vcc or less then it's hardly curved at all. Frequency is still highly tweakable (unlike those fucking 555 PWM circuits) just by changing that one resistor.

Yeah, but amplitude become frequency-dependent. it's not a huge issue if he settles on a proper design frequency, but for playing around with the design, a gain-controlled active integrator is easy to implement and dead stable.

I have a laptop charger that seems to provide intermittent power to any laptop that I try to charge on it. At first I thought it was just my laptop being dumb, so I tried to use it on another laptop and experienced the same issue. Then I figured that it was probably the cord that plugs into the laptop, but when I checked voltage on the end of it with a multi-meter, it had the 19.6 volts or whatever on it that it is supposed to provide. Just to make sure, that the cable wasn't at fault, I also made sure to wiggle it around and crap while doing it. Is there some sort of protection feature on these that would cause this issue that I'm experiencing? It seems that the thing will turn on for a brief second, then off for a while, then on again and continue this cycle indefinitely.

Hey guys im massively retarded and idk how to google this properly. Im planning a project to control a motor with raspberry pi, and i want to tap into my 3d printer power to power this motor, its a stepper motor 28byj-48 5V, im already powering my pi from the printer pau, and would like to run proper test to see if ots safe to power the stepper motor, too.

How would I go about doing that?

this is my /diy/ Powerwall
160 cells (free factory rejects). Rated 11Wh per cell, floating a little undervolted at 16v. so maybe 1.0-1.5kWh. enough to run most of my important things for a while.
rate and hate

>>1679776
>Yeah, but amplitude become frequency-dependent
No it doesn't, the height of the "triangle wave" is set by the voltage tuning of the schmitt trigger part of the circuit, it isn't time-dependant at all. Changing the timing cap only stretches the wave in the time dimension, nothing more.

>>1679802
>some sort of protection feature
They're likely to have a thermal fuse and possibly a PTC thermistor. Not sure if they use thermal switches instead of thermal fuses, but if so they could be causing your issue, along with the PTC resettable fuse. Does this problem happen right out of the gate, or does it take a little time to settle in? I'd get a stopwatch and start measuring the time it takes to cut off, and the time it takes to come back on again. If it's a thermal issue, you should expect some hysteresis. If it's an overcurrent issue, it may be due to your laptop drawing more power than the brick wants to provide, so also check those ratings.

>>1679819
Read the rating on the printer's power supply and estimate what's hooked up to it.
but really it's not much of a safety thing. Hook the shit up, try and run it, and measure the voltage to see if it sags. if it doesn't sag too much and the printer still operates, it will probably be fine.

Why do my tips keep turning out like this? Too hot? not enough solder? cheap tips?

>>1679908
Too hot + shitty solder

>>1679614
It is board specific, /diy/ is set to 300 but it takes so long to fall off page 10 that it's not worth doing new thread just because it's at bump limit.

>>1679911
i have a variable temp iron, what should it be set to?

>>1679919
I usually use ~320 for lead, ~370 for lead free. Kinda depends what you're soldering and how good your iron is at actually maintaining that temp when heating something big.

>>1679919
Real tip temperature should be not hotter than 260 C, even for lead free.
But since your iron is cheap, I'd set 300-320 for leaded, and 340-350 for lead free.
Also, which solder do you use? Chinese?
Anything above 350 damages tips. Though pro-tip. They are steel or copper, which means you can actually scrape their shitty coating off and tin it in plumbing flux with leaded solder.

>>1679901
>rate and hate
not all that pretty but if it works for you that's good enough
the price is right too

>>1679901
Who, that many li-pos.
>free factory rejects
I hope you have asbestos everywhere, otherwise it is dangerous

>>1679923
>>1679921
its a Vastar from amazon, this is what happened last time i tried to use it. porbably had the heat too high plus bad flux. totally fucked up.

>>1679928
That blue thing?
Those have simple dimmer, not temperature adjustment.
This is why it is overheating.
Reduce power of iron. Ignore temperatures on wheel, they are misleading.

Also, that soldering is horrible. I didn't know it is possible to solder that badly.
Get plumbing flux (Zinc chloride), and 50/50 solder for plumbing. Solder is usually good, since it is made locally (though it is kinda shit for electronics, because it takes too long to solidify, compared to 60/40, but once it is solid, it is better, since it provides more strength), and flux is pretty cheap and good, though it is acidic, so you MUST flush it, otherwise it will corrode everything.

>>1679937
>Those have simple dimmer, not temperature adjustment.
>This is why it is overheating.
>Reduce power of iron. Ignore temperatures on wheel, they are misleading.
well shit thats probably my biggest issue then, my last soldering gun was a craftsman 40w fixed temp and i never had issues with it, but with this i'd set it to "350" which was probably wrong and too hot

>>1679938
My advice would be to rip out the PCB from iron, and replace variable resistor with pedal (resistive).
> fixed temp
Those still can overheat, but at least you can move tip out a little bit, so it is cooler.

>>1679942
Rip from PCB and put in pedal*

>>1679937
>simple dimmer, not temperature adjustment
Are you sure? I assumed that they had a simple thermistor-based temperature control due to the markings on the knob being in °C and not %. I should look for a teardown online I guess.

>plumbing flux
>50/50 solder for plumbing
A good slathering of rosin paste will work with any solder, no need for that acidic garbage. I can't speak for structural solder, but I've never had solder joins fail before the wires being soldered or the pads themselves, both with RoHS and 60/40.

>>1679968
>A good slathering of rosin paste will work with any solder
No. It doesn't work with shitty chink solder. Not enough oomph.
>but I've never had solder joins fail before the wires being soldered or the pads themselves, both with RoHS and 60/40.
I have seen this a lot, even with 63/37. Fatigue and shit. Pads unironically hold up good.
>>1679968
>Are you sure? I assumed that they had a simple thermistor-based temperature control due to the markings on the knob being in °C and not %. I should look for a teardown online I guess.
100% sure, chink trick. Triac, diac and couple caps.

>>1679978
Aw man. Is there a simple analog temperature-controlled cheap soldering iron that I can rec to anons who don't want to go for a more expensive digital one?
Never used shitty chink solder before, but I doubt my solder is of high quality. If everything's cleaned with IPA beforehand and covered with liberal amounts of flux I never have any troubles with solder not adhering. Pulling off old solder with flux and wick can help with the cleaning process.

>>1679983
>Is there a simple analog temperature-controlled cheap soldering iron that I can rec to anons who don't want to go for a more expensive digital one?
I had CXG / A-BF ""90W"" iron.
It was OK and cheap.
Uses Hakko 900 tips.

>>1679601
>twiddle knobs
I am a fan of dials and knobs. So the one you linked interests me greatly. Thanks fancy anon.

>>1679903
>the height of the "triangle wave" is set by the voltage tuning of the schmitt trigger part of the circuit, it isn't time-dependant at all. Changing the timing cap only stretches the wave in the time dimension, nothing more.

Okay, can you please draw the circuit you are thinking of?

>>1679990
This one. Oddly enough the amplitude does drift a bit at higher frequencies, no clue how that's supposed to be happening, but I bet that can be fixed with a slightly more optimised circuit, and I can't see it being terribly useful to PWM at higher frequencies than 1kHz unless for particularly fast loads.

This might sound like moving of goalposts, but why was a variable frequency a requirement in the first place? For a self-contained circuit that you can just drop into anything else you'd probably want a variable frequency, but in that case it would either be some sort of monolithic package or a PCB covered in tiny packages where in both cases it doesn't matter if it takes a few more op-amps, and in the case that it's an anon building it for his own project out of discrete quad op-amp ICs he'll be selecting his own passives so a frequency-dependant triangle wave amplitude wouldn't be an issue aside from in the initial design phase.
If I know what I'm talking about, you'd have your summing amplifier able to output a voltage range significantly greater (i.e. 2-3 times if not more) than the voltage range of the triangle wave anyhow, so a 10% or even 50% drift wouldn't make a big difference. The negative feedback loop would lead to the summer's output always being forced into the range of the triangle wave anyhow, and it could be compensated for in the gain of the circuit elements themselves. A gain difference of 3dB isn't likely to be terribly significant here anyhow. As opposed to if you were controlling the PWM without feedback (e.g. a knob), which would also make a nice little drop-in monolithic circuit. Yes I made that mistake recently, where turning the knob all the way to the left still had the LEDs dimly lit.

>>1679766
are we talking about color temperature or thermal temperature? only for color temperature do we call the unit a kelvin. the latter is "degrees Kelvin"

>>1679901
6/10 cute build, fire insurance paid up?

>>1679983
the 936 with the power brick is probably as good as cheap gets

>>1679990
not all hysteretic comparators are Schmitt triggers

>>1680007
>Unlike the degree Fahrenheit and degree Celsius, the kelvin is not referred to or written as a degree
>The kelvin is the primary unit of temperature measurement in the physical sciences, but is often used in conjunction with the degree Celsius, which has the same magnitude.
>https://en.wikipedia.org/wiki/Kelvin

>>1680006
>no clue how that's supposed to be happening
overshoot happens and opamps are usually slower at lower supply voltages. some opamps are also pretty slow coming out of the saturated condition. what's your time step?

>>1679990
Wait how does your 2 op-amp circuit that isn't frequency dependant work? I was assuming you were using an integrator, but those are definitely frequency dependant.

>>1680010
>time step
Whatever the default is. I figure it was overshoot, but it's surprising how big it is even at 20kHz with a MHz op-amp.
>slow coming out of the saturated condition
I replaced it with a 280MHz comparator (LTC6752) and now it's working better, though with a bit of a skew to the wave even with a 1k pullup. Not that that matters as a sawtooth would still work just fine for PWM. But I'd prefer that this sort of circuit be made of cheap and ubiquitous parts.

>>1680006

>0.24 V peak to peak on 5 V rail

That circuit is literally operating past the RC cutoff. How is he supposed to run that against a proper error signal?

The reason your amplitude is changing, but not by much, is because you're so far down the tail of the cutoff that you won't see much change.

>>1679331
How can I learn how to generate a carrier signal, modulate it with an arduino, and then amplify and broadcast it?

What can I read to learn about that?

>>1680021
>operating past the RC cutoff
Past the what?
>How is he supposed to run that against a proper error signal
With a properly tuned gain from the summing amplifier. I assumed ~-10dB wasn't going to be a problem, since it would only be affected by in-circuit noise, or so I assumed. As I said, the negative feedback of the system would pull the summing amp's output towards the triangle wave. If I'm missing some sort of threshold or slew voltage that's intrinsic to op-amps and/or comparators please elaborate.

>>1680024
There's a lot of possible ways to modulate a signal, and it depends on whether you're doing analog or digital data. From the rudimentary research I've done it feels like amplitude-shift-keying (or even on-off keying for maximum S:N) with a heavily filtered signal is a good method to get a decent bitrate at a low bandwidth, and is relatively simple to make with just a PLL-based sine generator/frequency multiplier (assuming the VCO's output is smooth enough) that feeds a mixer of some kind. The actual amplification and impedance matching of that signal is something that would better be answered by the ham radio guys.

PLLs are cool, look up how they work.

>>1679908
you are also not coating the tip with solder after you're done with it, causing it to oxidize.

>>1680029
>Past the what?

You circuit is the passive version of a 2-Op Amp triangle wave generator. square wave to drive a passive integrator. This would be fine, if the way it was configured didn't require you to be past the RC cutoff point for it to operate.

>since it would only be affected by in-circuit noise, or so I assumed.

yes, this is the problem.

A more sound approach it to take error signal - ostensibly quite small, I don't think pipe organ bellows is going to see 20 psi pressure swings - and amplify it until you are within 10% of cutoff for all amps in circuit. you do this before it gets to PWM generation, and you generate a triangle wave at similar levels - so that the PWM output is as clean and unambiguous as possible, and as resistant to changes in component tolerance as possible.

> I was assuming you were using an integrator, but those are definitely frequency dependant.

Not below cutoff, they aren't. they're almost perfectly stable up until that point.

the cutoff of the RC stage you use to drive feedback. generally cutoff is given by 1/(C*R) for the resistor and cap you use for feedback to the inverting terminal.

I'm just starting out in a computer engineering degree and want to get some kind of breadboard kit to play around with and try to get ahead of my classes. Any suggestions on what I should get to start out? Kind of aimless right now, honestly.

>>1680024
>arduino
1. >>>/diy/arduino
2. >>>/diy/ham

>>1680043
okay school but what's your prior experience? is there a software or hardware project you're proud of?
>comp
ardui-
>eng
oh, you need non-toys then
if first-year then you probably want a kit that you can use for logic and programming. since you're doing comp eng and not mechatronics, you probably won't need a whole lot more than LEDs and switches for I/O
for the logic side, you could buy a hundred bucks of assorted CD40xx and/or 74xx logic ICs and try to build a simple instruction processor and/or stack of peripherals out of them, but I'd just buy a reasonable smallish FPGA board like the Lattice iCEstick. Verilog isn't bad, it's just a bit weird at first if you're coming from other programming languages due to its "task" model
for the programming side, you could configure a soft CPU core for the FPGA, but that takes up a lot of logic resources and you can probably save that for later. I'd start with some kind of board with a hard processor that puts you relatively close to the metal. AVR is a good platform for that. you can buy very low-cost "arduinos" which are usually just the MCU, a regulator, and (sometimes) an onboard serial port interface for the host. just stay away from the arduino IDE and stick to vanilla avr-libc tutorials and guides
once you're using the AVR's resources pretty well (especially learn to use interrupts and its serial buses) you can hook it up to the FPGA with a complementary serial interface configured, and have some nice three-way chats between them
what's your schedule like? what's your reading list like?

>>1680058
Cool, I'll look into this stuff. Thanks

As for experience, I don't really have any at all in hardware except playing with some Radio Shack breadboard kit when I was a kid for like a week. I do have plenty of software (C++) experience, though. Never touched assembly.

>what's your schedule like? what's your reading list like?
Right now the only hardware/actual engineering class I have is Intro to Digital Systems, which is my first engineering class. The class is over combinational/sequential logic, boolean algebra, digital design logic, and some assembly language. Haven't gotten to assembly yet, and my first lab isn't until next week, but we're starting out doing stuff on an IDL-800. Books are pic related. I only got Computer Organization and Design and Schaum's. I don't really like the first one, Schaum's has been way more helpful to me in learning the material.

>>1680058
Looked into the Lattice iCEstick. I'm not really sure what I'd be doing with this? The class is currently using 74LS gates, so I was thinking of getting a box of those, a small wire kit, and a breadboard. A box of 74LS is 30 bucks, but 74HC is about 20 bucks. I have no idea what the difference is, honestly, but the lab says we're using 74LS. How exactly would I use the iCEstick?

having some issues with transistors. im still trying to understand how they work in this circuit (70s VCR)

issue: solenoids are not activating. I belive the brake solenoid should be ON during playback (pulls away brake pad from roller)
issue at the moment is TR924. (S is standby, P is play mode). in P, i get 0V from C leg. E and B read fine. I took the trans out and tested it and it "seems" ok.
on the schematic, i am getting 23V from pin 10-3 to the Solenoid.
If the C leg is not getting the 12v how far back in the line do i need to go?

>>1680098
oh, hold off on the iCEstick for now. the deal with FPGAs is that you write Pascal/C-*like* code describing the operation of your logic design, instantiating it, and connecting instances to I/O pins and to each other, then "compile" it into a binary file which gets loaded into the chip somehow, and then (hopefully) the FPGA connects its "gates" and latches and pins together according to the file, and thus realizes the logic function
for now you've got the right idea. HC and LS families are approximately comparable and are interoperable in practice. the HCs work well at lower voltages and don't consume as much power at steady-state. you'll probably want small bags of LEDs, current limiting resistors (470 ohms is a generally safe value), DIP switches, and tact switches (so you can learn about switch bounce first-hand). for power, get a 5V 2A adapter. the 700mA "chargers" are mostly shite
you might consider a cheap logic analyzer (a $25 Saleae Logic clone is good enough). it's a bit slow as LAs go but it will *help* you see into more complex designs at home, and it is cheaper than a scope
I haven't read CO&D but I did read their Computer Architecture: A Quantitative Approach. CO&D must be the what and how upon which CA:AQA's when and why are meant to build, but I still managed a couple of neat machine designs out of CA:AQA alone. Hennessey and Patterson are tough and meaty authors. you should definitely know digital before expecting to make much sense of them. I learned all about the gate level from Ken Roth's Fundamentals of Digital Logic Design, probably ~= your Schaum's
C++ hides a lot of the details from you, which is not necessarily great for learning. get to the point where plain C pointers are effortless for you, and the assembler course later should be a breeze
just curious, what assembly are they teaching later this year?

>>1680110
What's the difference between C++ pointers and C pointers? Honestly, I tended to avoid the OOP stuff in C++, I more or less used it like it was C. I've done some mildly low-level stuff like OpenGL, but not embedded-level stuff. We're going to be using SPIM / MIPS.

>>1680038
>square wave to drive a passive integrator
The RC circuit is what creates the square wave in the first place, it provides the hysteresis that allows the schmitt trigger circuit a delay before triggering. I'm not sure if I'd classify that as a seperate square wave oscillator that makes a triangle wave via integration, it's closer to a triangle wave that makes a square wave. Still unsure what you mean by RC cutoff point, unless you just mean the corner frequency of that RC circuit, in which case the frequency of the oscillator is set by that very RC circuit in the first place.

>>1680058
>oh, you need non-toys then
Based.

>>1680110
>$25 Saleae Logic clone
I'd recommend one of the $8 alibay logic analysers, I've never had any issue using them with pulseview.

When some one designs the power supply voltage dropping network in something like pic related, what is the process? Do you use the quiescent currents to calculate resistance and treat the triode sections like resistors? Is there a fast way to do this?

I am an absolute noob. I am very poor and would like to learn how to build electronic noise makers out of scrap electronics I find in the garbage. I am not entirely concerned with them sounding fantastic, cause I would eventually hook them up to a computer and other electronics to change how they sound. I also want to make diy synths and effects pedals, but I dont have money for that right now so I'm going the recycled materials route. Right now I have a small CRT TV that's probably broken I can use, and I might be able to get my hands on a few digital alarm clocks. Tomorrow I can post some pictures of the junk I can find around here as possible building materials and maybe someone will have ideas where I can start. Once I start my job in a few weeks I can start buying additional components i would need to modify the garbage, so long as i keep it cheap. I have an old radio too maybe I can add an output to to hook it up to shit. Feel free to tell me I'm an idiot cause I'm sure this notion is misguided.

>>1680112
C++ pointers are the same as C pointers

But be aware that C++ also supports references which C doesn't. Autists argue endlessly about the differences between pointers and references.

>>1680112
>MIPS
Hennessey would, wouldn't he. it is a pretty clean design, beats the hell out of a machine with an 8-bit native address bus width and some wacky bank-switching system
Microchip made some MIPS-based microcontrollers under their PIC32 range, check out Pinguino for cheap hardware

>>1680114
I was tempted to rec a plain old $4 EZ-USB FX2 board as both a mediocre LA and a microcontroller platform but I have a strong allergy to MCS-51

>>1679331
what's that kid eating?

>>1679983
936 clone, but as with any chink device you need to replace the garbage aluminium mains cable for a copper one ASAP before it crumbles apart (this doesn't matter for the internal wiring, which never gets flexed)
buy a reel of H07RN-F mains wire and put it on all your tools, it's floppy and burnproof, you won't regret it.

>>1679331
>Post your setup and/or your favorite equipment Edition
dynascope is olev. wish i could afford a mantis though.

>>1680124
>treat the triode sections like resistors?
Are they not closer to current sinks?

>>1680141
leaded solder

>>1680158
I always heard they were closer to JFETs than to BJTs.

I have a T12 soldering station and want to modify it to have a DC socket so I can use it on the go. The output stage of the flyback has a big fat TO-220 schottky diode on a heat-sink that's larger than the switching transistors', which feeds a 1000µF cap. This cap then feeds a low-pass LC filter made of a fairly coarsely wound (~12 turns around a ferrite rod) followed by a second 1000µF cap. There's also a small ceramic cap across the rectifier, and a large 1kΩ resistor across the 24V output, I guess for a constant 24mA load.

My first question is whether I should solder the jack before or after the filter (probably before unless anyone has any drastic revelations)? The main problem being it may be difficult to get a bodge wire into that first part of the circuit, compared to just throwing a switch and DC jack in the power-board to control-board jumper. I'm guessing a car battery is pretty low-ripple as-is, but I'm unsure if throwing it before the filter would be bad if I were to leave it turned on while the alternator was running. I've heard bad things about commutators, perhaps I'd even need to add my own filter on top of the existing one.

Also, can anyone can tell me what voltage the OLED controller on this can go down to? I get no luck when searching for the datasheet of the on-board switching converter, and searches for this model of controller plus "12V" only get me results with the LED 7-seg models.

In the end I'll see about wiring a barrel-jack to cigarette-lighter-socket or alligator-clip cable. Whichever is more convenient. If it's gutless enough on 12V I may wire up an external boost converter to fit between the battery and the station itself.

>>1680177
No BJTs in this context.

>>1680151
beautiful. what do you make using all of this?
how are those chinkstations working for you?

>>1680195
I mean BJTs act like current-controlled current sources, while JFETs act like voltage-controlled resistors

>>1679331
Spoonfeed me please. The best battery spot welder and solder station?

Building an ev. Soldering copper wire just aint gonna fly anymore

>>1679458
Would this work or spend 25 bucks on ksger?

[US$18.53 7% OFF]STC-T12 DIY Kits Electric Unit Digital Soldering Iron Station Temperature Controller Kits for HAKKO T12 Soldering Iron Station DIY Kits with LED Vibration Switch Professional Tools from Tools, Industrial & Scientific on banggood.com
https://banggood.app.link/0GrrqcfZRZ

>>1680236
Well yes probably, but you need a power supply to go with it. A case too I guess. They run best at 24V so you'll only be getting 63% of full power with a 19V laptop charger. Also they're probably cheaper on aliexpress. Also that isn't an OLED model, just a 7-segment model, the OLED ones have multiple things shown on their display.

Soldering stations recommendations by price-range

>Yihua 936, analog ($15-$30)

>KSGER T12 ($30-$60 )

>MINI TS100, iron (~$55)

>Weller WES51, analog (~$85)

>Hakko FX-888D ($100-$130)

>>1680236
>>1680237
Forgot to mention that some KSGERs have power supplies built in, and some don't. Some have OLEDs, some have 7-segment displays.

>>1680240
>MINI TS100, iron (~$55)
If you have a power supply for them.
Good list though. Perhaps include that iron with the temperature buttons on it if there's a common model number (>>1679985). Maybe the Aoyue clones or whatever too.

>>1679901
Rate 10/10 good work

Hate 10/10 i want free cells.

Or cheap.


Hell, is it possible to get half ass dece t china cells for like a buck each?

>>1680237
>>1680240
>>1680243

Very nice. Thank you. Is there a model number? Sorry I'm noob as fuck. I cant figure out what ones have power supplies and oleds.

>>1680109
full board page if it helps

>>1680246
https://www.aliexpress.com/item/32948842049.html or https://www.aliexpress.com/item/32836043839.html are the KSGER ones I’d be likely to go for, they’re the ones at version 2 or 2.01. The quicko brand ones would be the T12-952. The quicko 942s require an external 24V 3A PSU, as do some of the KSGERs. The ones with the 7-segs are cheaper (951 and 941?), as are the ones with the plastic case instead of aluminium, so you might want to consider whether you want the aluminium case. Quicko also looks like it might be cheaper than Ksger, but Ksger have models with the latest hardware and firmware versions.

>>1680225
>JFETs act like voltage-controlled resistors
Really? Curves are almost parallel, not much change in ∆Vds/∆Id. Compare to ∆Vce/∆Ic.

>>1680283
you can use them in team with op amps to get some pretty precise constant currents

>>1680246
>I cant figure out what ones have [...]
read about it on the product pages and reviews wherever they are sold.
just search the brand + model name. also look it up on warosu.

>>1680283
For use as a voltage-controlled resistor, you should be looking at the linear/ohmic region (left-hand edge, where the curves are steep), not the saturation region (right-hand side, near-horizontal lines). For Vds<1V, Id~=R*Vds with R controlled by Vgs.

In the saturation region, they act like voltage-controlled current sources (although the Ids/Vgs relationship is far less linear than a BJT's Ice/Ibe).

when looking for equivalent transistors, how exact to do the values need to match? I need to find a equivlant of a 2SD389B
could I use this? https://www.el-component.com/bipolar-transistors/2sc2075

>>1680283
>not much change in ∆Vds/∆Id
I thought the point was to change Vgs? In which case, flat response regardless of Vds would be desirable

>>1679455
resistive divider with a voltage follower transistor

Anthing else is a toy

>>1680418
You're missing the context of that post (voltage-controlled resistor). A resistor follows Ohm's law.

Is there any way to use the Reset pin of the 4017 as a "chip select" signal for connecting >1 switch reader circuits? Basically, I'd like to make the Data output become high-impedance whenever the Reset pin is high. PNP transistor at the output?

>>1680134
I couldn't post pics yet cause I'm helping clean out my deceased grandmothers house now but no one has advice for this? I can post a pic of the TV when I get home.

>>1680134
>>1680480
>advice
start learning about how to build electronic noise makers out of scrap electronics.

>>1680474
>Is there any way to use the Reset pin of the 4017 as a "chip select" signal

nope. you gotta add extra circuitry, like a latch, to the outputs.

>>1680474
>Is there any way to use the Reset pin of the 4017 as a "chip select" signal

nope. you gotta add extra circuitry, like a latch, to each output.

if you're only concerned about providing HIGH outputs, one possible solution is to put a PNP transistor on the Vdd pin to cut off power to the chip.

>>1680474
>Is there any way to use the Reset pin of the 4017 as a "chip select" signal

nope. you gotta add extra circuitry, like a latch, to each output.

if you're only concerned about providing HIGH outputs, one possible solution is to put a PNP transistor on the Vdd pin to cut off power to the chip. but this may not work, as CMOS chips can power themselves from siphoning power from a HIGH input. they're tricky!

>>1680484
Well shit do I just open up the TV and start fucking around? Aren't there any books or something i can read?

>>1679928
Not hot enough and maybe too little flux.

>>1680505
>open up the TV and start fucking around?

CRT TVs are not to be fucked around with. there are killer voltages and currents inside. the only thing you can do with it is pull parts out of it: caps, resistors, diodes, transistors, transformers, a speaker, and possibly a chip or two to be used elsewhere.

typical ''modern'' TVs use highly integrated chips that perform many functions, and are useless to you. so expect slim pickings.

>>1679331

how do i get my electrical signal into the air?
i want to broadcast my morse code signal on some frequency. how can i make this happen?

>>1680523
See the pixie circuit, on-off keying, CW radio, PLLs, envelope detectors, Fourier transforms, Signal:noise, Shannon-Hartley, frequency mixers, crystal oscillators, and the amateur radio regulations and frequency bands for your country
It’s illegal to transmit Morse code on an amateur radio band without an amateur radio licence, without one you’d either be transmitting at a low enough power than it gets mixed in with electronic noise, or at a low enough frequency that it’s unregulated (without running into a time calibration frequency).
Should also learn Morse code, try some sort of software Morse trainer, I’ve heard good things about the Koch method

>>1680196
99% of what i do is replacing phone/tablet screens, batteries, usb and headphone sockets. the other 1% is re-capping and appliance repair. haven't built anything just-for-fun in years.

chink stations, the 936 cost £22 in 2013 and i'd rate it 10/10 at the price. it turns on, melts solder, doesn't burn the house down.
the 959's only a week old but it hasn't done anything weird yet. i'll check back in 2025 to let you know if it's any good.

>>1680505

>>1680523
check /ham/

I have 11 (eleven) power bricks I stole from university trash. Various power output (most powerful is 19V 3,95A, so 75W).
How much power can I expect from T12 soldering station, or TS100 soldering iron at those voltages?

>>1680523
Ignore the babble, CB radio is free. Just whistle your dits and dahs into the microphone.

I love goosenecks. I want to use them for diy helping hands, LED work lights, small fume extractors, et c. Does anyone know where I can get some empty gooseneck metal "tubes" or "cables", whatever they're called? Some other place than Ali*?

VCA/VCO/VCF anon here again
I must have downloaded at least 25 books now on analog computers and op-amp design

STILL can't find ANY mention of them, let alone any elementary/introductory level information on their principles of function, or implementation

what the fuck
granted I think I've fallen in love with analog computers now so at least there's that

Do you need to use extra resistors with led tape lights, or are the built in resistors good enough?

Pictured is the higher quality/more expensive ones, but I'd probably be buying the cheap shit from banggood/Ali/wherever

>>1680666
It is not necessary to use extra resistors with tape lights

>>1680523
>i want to broadcast my morse code signal on some frequency. how can i make this happen?

here's a much simpler, cheaper way to do it than previous answers: car FM transmitter. used ones are like $2 at the thrift store.

>>1680631
amplifiers, oscillators and filters aren't op-amp designs, they're op-amp applications. aim lower! get an introduction to opamps, and an opamp application guide/cookbook.

>>1680114
>he RC circuit is what creates the square wave in the first place, it provides the hysteresis that allows the schmitt trigger circuit a delay before triggering. I'm not sure if I'd classify that as a seperate square wave oscillator that makes a triangle wave via integration, it's closer to a triangle wave that makes a square wave. Still unsure what you mean by RC cutoff point, unless you just mean the corner frequency of that RC circuit, in which case the frequency of the oscillator is set by that very RC circuit in the first place.

I think you are missing some theory here.

this anon >>1680038

is correct, the circuit you are presenting is a passive integrator being driven to create a triangle wave. what he is saying is that the way the circuit operates - by necessity - pushes the RC stage past it's cutoff frequency in order to achieve both switching and triangle wave generation.

also, by cutoff I think he means -3dB point. it isn't really "corner frequency."

it is not necessarily the case that it exists to create hysteresis, and indeed, it doesn't really - that is more the role of the resistor fed into the non-inverting terminal. The RC stage itself has virtually no hysteresis.

You are right that the frequency is set by the oscillator,it's just that the frequency also happens to occur very far down the tail of the magnitude curve. the circuit is essentially setting itself into a highly attenuated state, hence the reason why this isn't used very often, and the two Op-Amp circuit is almost always preferred.

Even then, using op-amps for this sort of thing is a little dodgy, but for the "organ" poster's application, it's appropriate.

>>1680672
Sorry that was poor wording on my part, I meant introductory op-amp application books. I'll keep looking though for more intro to op-amp books

>>1680672

Have you read Horowitz and Hill?

>>1680674
I guess that makes sense. But the whole reason it was slated as improper for this instance is because the amplitude varied with frequency right? e.g.
>>1680038
>> I was assuming you were using an integrator, but those are definitely frequency dependant.
>Not below cutoff, they aren't. they're almost perfectly stable up until that point.

The op-amp integrator one's amplitude also varies with frequency, because the slope of the triangle wave is dependant on the square wave's amplitude (which is 5Vpk-pk and constant), but the interval between the slope's changing is varying, hence the amplitude of the triangle wave is inversely proportional to the frequency. That's how integration works mathematically too.

Hence I'd say this integrator-based triangle wave generator is a much more inappropriate selection for an enclosed or monolithic circuit where you want to be able to vary the PWM frequency with few external passives. I still don't really see much of a problem with using the single op-amp method, the wave should still be stable enough once your filter capacitor is decided on, no?

>>1680405
that depends entirely on the design of the circuit around it, whether it's to be matched with some other transistor in a pair, etc. is this repair? a new build of an old design? beware, old designs may depend on transistor pairs manually matched as to particular specs (e.g. V(be) at chosen current). performance of the design may suffer if your transistor's specs aren't matched to its mate
the proposed replacement doesn't seem able to dissipate as much power as the original. also, f(T) too high may increase tendency to oscillate, and lower h(FE) may distort outputs
anyway, for a straight replacement that's likely to work: all absolute maxima should meet or exceed the original. nominal dc current gain should match (±50%) the original. nominal f(T) should be as close as reasonable to the original. other stuff may or may not matter depending on the particulars of the circuit

>>1679455
not really but
>idiom
shows you're thinking at the right level
I mean, the general design goal can be stated as to just eat up 7V or so (Kirchhoff's voltage law). there are many ways to do it, of varying accuracy and practicality. without converting power, the two main voltage regulation topologies are series and shunt: respectively, the series element passes current from the input to the output depending on the output voltage, and the shunt element diverts input current from the output depending on the output voltage. a negative feedback of some sort will be required to maintain regulation, in the series case, usually a reference voltage and a difference amplifier, in the shunt case, usually the slope of a diode's I-V curve. the series element can pass current continuously or chop it PWM-style. so can the shunt element but that's rare. also, the load is effectively a component of the shunt resistance and needs to be accounted for

>>1680474
74LVC1G66

>>1680568
enough

>>1680666
it'll be fine, satan

>>1680683
yes, VCA/VCO/VCF are not covered in the Art of Electronics

What's the best book, or guides to learning schematics and drawing diagrams?

>>1680695
>enough
How much?
45W at tip?

>>1680698
we're talking controlled amps, oscillators and filters, right? there's anons discussing a classic op-amp VCO in this thread right now that's an H&H staple. I know for a fact it discusses VCF's, albeit in a more fundamental way, and gives a few example circuits. They break down an entire op-amp schematic and go through it's detailed behavior.

I think part of the problem you;re running into is that people don't really call them VCA's, VCO's, and VCF's in most applications. VCO's don't appear outside of microwave frequencies much anymore, and a VCA in beasic terms is literally an Op-Amp and a FET.

>>1680686

As a question, why are you using 1Mohm and 100nf for that circuit? those are unusual values to pair with an active filter. cutoff frequency for that circuit is ~1.5 Hz.

I don't know where the guy you were responding to was, but I'll try to get something simulated and posted. give me a bit.

>>1680710
Art of Electronics has a section on the dos and don'ts for schematics.
The Circuit Designer's Companion is a really good book IMO for understand the practical aspects of circuit design.
Usually think of your signal path as going from left to right, and your positive voltage rails above, common/0V below, and negative voltages still further below. Most companies that provide a representative schematic for their ICs follow this style. You'll see this in e.g. most 555 schematic diagrams.
Use schematic capture software, don't use babby tools.

>>1680720
>don't use babby tools
I believe some clarification is in order

>>1680568
63% of the maximum 24V. Actually my T12 runs a little hot at closer to 25V, so it could be even less. But I've no idea how often the station is actually running at 100%DC aside from startup. In any case you likely won't need the full 24V unless you're soldering onto a thick ground plane or aluminium-backed PCB.

>>1680717
Just ensuring that the cutoff is as low as possible but still have a reasonably high input impedance with the 100n. The 1Ms don't determine the cutoff of the filter, that's mainly the 10ks' job. AFAIK the 1Ms are just for stability, though I know there's a more conventional and rigorous rule to use when sizing the components. Using a 1n cap causes the amplitude to get far too high and the triangle wave clips, a 1µ causes bad stuff to happen.

>>1680727
He probably means don't use that arduino circuit design thing or that one black and green simulator that may or may not actually be kinda ok
kicad is a good (FOSS) choice for laying out circuits and also for turning them into PCBs, not sure how good its built-in spice is though

have a circuit with an LED indicator I don't need.
if I want to disable the LED, should I just replace it with an approxmately equal resistor, or resistor + diode?

>>1680747
disconnect the LED and see what happens

>>1680739

Alright, took me a while to get to it, work sucks -

I'm pretty sure this is what the earlier anon was trying to get you to look at. Instead of passive integrator, active integrator. you change a single resistor - frequency changes with virtually no effect on gain.

vs. the single amp circuit you were using, you add a one more amp and another resistor. in return, you get full, surprisingly independent control over gain, frequency, and offset while maintaining a decent noise immunity.

you'll have to excuse the use of Falstad's sim, it's what I have available at the moment.

>>1680749
I've done it before.
What's the "general practice" for it? I don't want to mess up the flow to other components.

>>1680739
apparently it's about as good as anything else that links to lbngspice2
>hating on falstad this much
actually, that's my preferred tool for drafting and qualitative "design engineering". at least it's not Fritzing. MicroCap otoh has a good front-end with a nice XP-ish UI

>>1680766
you probably won't. the LEDs are usually just turned on and off, there's no real sense in checking whether they're there
* except for a weird power bank controller I read about once

What the fuck i had no idea the laser diode would be this smol
How do I even solder something that is nanoscale

>>1680784
ig you make a pc board for it?

>>1680773
>you probably won't.
alright, cheers

>>1680784
>milliscale
>small
use LED magnifying glasses, a fine tip, fine solder, and pair of steady hands

>>1680751
Oh shit that's actually really good. Is there a name for that circuit? Because I'm pretty sure I didn't see it when searching for "op-amp triangle wave" on google images.

>>1680773
>hating on falstad this much
I did say that it "may or may not actually be kinda ok", it seems to be pretty alright but it takes me 10 times longer to make anything in it than in ltspice or in kicad if only because I'm not used to it (and because ctrl+z inside a browser can be trying to do multiple things at once). Fritzing is the devil.

>except for a weird power bank controller I read about once
I've sometimes used indicator LEDs as high-impedance pulldowns or pullups in my circuits where I can get away with ditching a passive where I already need an indicator, there's nothing wrong with it. Maybe some cheap circuits do the same thing, but if so it would be rare.

>>1680786
>pc board
That's a new one.

>>1680797
this power bank IC tried to push a little current through its indicators to see how many LEDs the designer had connected to the bar graph outputs. pretty sure it was one of the chink chipmakers, part number was lost to the sheer wew of it all, but ISTR it was kind of a common one as power banks go
>new one
mmm, metal-core

hey /ohm/ what are your go-to brands for electronics?
I'm looking at all kinds of things:
>multi-meter
>digital calipers
>pliers
>wire cutter and stripper
>solder tip thermometer
>small heat gun (for shrink tubes and phone displays)
>rubber work mat
>magnetic screw/parts tray
>ESD wrist-strap & work mat

any recommendations on what to buy, or what to NOT buy?
I'll probably just pick up most of it from HobbyKing or Banggood

Anyone here try using negistors? I'm looking for an AC audio clipping component that enters negative resistance with only up to ~7V or so as its breakdown voltage, and apparently a backwards BJT (negistor) can do this for one cycle, so I'm pondering putting two in antiparallel. How would I find the one with a particularly low breakdown voltage? V_CE_Reverse isn't a commonly found statistic.
Gunn or tunnel diodes in antiseries or whatever could also suit my application if they come in low enough breakdown voltages.
UJTs also look neat, lambda diodes too.

>>1680810
what's so special about it vs. Zener diodes in antiseries?

>>1680811
Zener diodes don’t have a negative resistance region; they clip to their sender voltage but don’t go any lower. But a tunnel diode or diac or spark gap have a breakdown voltage, and after they hit this the voltage across them can go down and current will continue to increase. A sine clipped by a zener (or normal diode) would just have a flat top, but one clipped by a negistor or gas discharge tube would look more like some sort of phase-fired control deal, reaching that threshold and getting immediately pulled to 0V until the next 0-crossing.

>>1680812
got it
it seems like it wouldn't be a performance effect as it could be pretty hard to keep rein over, but might be worth exploring in post-production where you can control the signal level going into it. you might consider a quick DSP simulation and/or prototype
the trouble with most negative-resistance devices is that they require a bit more power to trip than your typical low-level audio signal offers. interesting results might be found in the 100mW-10W range, less than that wouldn't hold the NRD open long enough and would probably sound like a loose cable. more than that seems possibly explosive

>>1680815
I’ve designed a circuit that lets me freely modulate the gain of an amplifier before the clipping circuit, followed by another amplifier on the second row of the same potentiometer such that the total gain is always unity. There will also be an option to turn one of the amplifiers into an integrator and he other into a differentiator, and swap the two around, to see what clipping those waveforms would sound like.

I hadn’t considered the fact that they might take more current to trip than the 10k resistor in series will let them have, perhaps I’ll need to run this circuit onto a daughter-board and do it with FETs and comparators instead. But before that I’ll see about testing with my BC54Xs to see how they behave. I think last time I messed about with negistors the 2N2222A worked but the BC547 didn’t so I may need to get myself some of those.

>>1680695
>transistor
thanks for the info. again in this case the 12V leg of the transistor was not reading any power so it maybe an issue along the line and not with the transistor itself. at this point i opted not to replace them. this is me >>1680109 the transistors in question are used to power the solenoids that run various rollers and brakes for the heads.

>>1680829
I don't see any D389Bs on that drawing. show relevant part pls
if they are just using it as a switch, power capacity is less pertinent because it'll be saturated or off almost all the time, and you could certainly try that replacement. might even be able to drop a MOSFET of your choice right in there

>>1680832
right sorry im having two different issues here:
1. the solenoids (Last post)
2. maybe the 12V regulator transistor isnt sending the power to the HV board where the solenoids are.
here's the part in question, on the power supply board. seems to be a 24.5v line and a 12v line. There was a crack on the pcb where the 12V generator(?) was which was causing the drive motor to not engage but that was fixed and I didnt see any other issues.

>>1680835
the KSD363 (Fairchild/ON Semi's issues of JIS parts are prefixed KS) might be a close replacement. a little slower and a little less gain but comparable in other respects. might have to nudge VR1001 after installation. 93 burgercents on Digi-Key rn

>>1680714
What do most people call them?
Also, do you know what page or section in AoE they're in?

>>1680843
thanks man, i'm still learning the best way to find equivalents for transistors. not as easy as cap replacements. again at this moment im hoping its just a 12v issue and not a transistor issue.
now, back to the HV board here >>1680109
TR927 (top right) 2SC1384 did have a broken leg, which i put a bit of solder which seems to work, but there's a chance i'll need some new ones down the line. thoughts on an equivalent?
https://alltransistors.com/transistor.php?transistor=11830
>digikey
gotta love it. $2 for a dozen parts and $9 for shipping

>>1680848
>2SC1384
a good old 2N2222A would fit there
I am simply grateful Digi-Key no longer has a minimum order

>>1680853
much appreciated. if I get this VTR working again it will be a fucking miracle.
I did find however someone is selling a brand new never used vr-1000 on ebay for over $700 bucks. if it were cheaper i'd highly consider it
https://www.ebay.com/itm/Quasar-VR1000-Mint-Condition/223461550242?_trkparms=aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D20160908105057%26meid%3D5d0ac7d2712c485fb6987111b4504cad%26pid%3D100675%26rk%3D1%26rkt%3D15%26sd%3D223461550242%26itm%3D223461550242%26pmt%3D0%26noa%3D1%26pg%3D2481888&_trksid=p2481888.c100675.m4236&_trkparms=pageci%3A9b363909-d4b2-11e9-b850-74dbd180cd97%7Cparentrq%3A2130f05316d0a9cb719da551ffb81357%7Ciid%3A1

>>1680879
monetised post?

>>1680898
just pajeet front-end skills

>>1680898
>>1680915

no idea why ebay links are so fucking long.
if you wanted, just search vr-1000 quasar on ebay

>>1680495
>>1680695
I tried connecting a PNP transistor at the output but it didn't work at all. The counter even stopped working. I was hoping there was a simple way to do it. 74LVC1G66 seems fine, too bad it's only available in SMD.

>>1680712
I'm confused now.
Will T12@45W tip perform in same way as classical 900M based station @ 60W?

Also, what is better: T12-based soldering iron, or TS-100?

Do CAN bit segment counts and sample point matter when two nodes are running on at the same baud? I'm going near mental trying to get a nucleo-f767 to play nice with a canable and pycom controller, both of which are able to transmit at 125kbit, 250 500 etc. I know the controller is capable of messaging because i managed to get two of them communicating with eachother, but no matter the settings i give it, it just will not talk to anything else.
At this point i dont even know if it's an electrical problem, clock problem or in the code. Even the example code from stmcube won't run. Any suggestions would be highly appreciated as i'm just lost here

>>1680797
>Oh shit that's actually really good. Is there a name for that circuit?

I picked it up from an old-timer who just treated it as the default way to do compact triangle wave generation, I don't really know if there's a name. there's a couple iterations of it, but you're right, they really don't come up much when you're looking for things on google.

speaking of which:
>>1680631
>>1680680
>>1680698


Pic related is literally the textbook op-amp VCO, and a variation of what we've been talking about. this is given on page 240 of H&H, under a section titled "Voltage-controlled-oscillator"

it's a small section, but this is the general idea behind most VCO's. H&H actually talks about it being borrowed for the book from some manufacturer datasheets - there's a bit of a hint there. datasheets will often teach you more than a textbook.

>>1680938
The PNP transistor resets the counter via its base current. Exchange R701 and R? or make sure the reset signal is really low, not just open.

>>1680977

sorry, i didn't mean to indicate

that you >>1680797

were >>1680631

bad wording, just . . .

never mind. I need to go to work now. it's going to be a long day.

>>1680977
It's the first circuit on the LM358 datasheet.

>>1680978
Thanks for the tip. Need to test it tomorrow.

I also came up with this idea: when the counter receives a reset, its Q0 output becomes "1". Then only that particular output would need to be switched off with a PNP transistor. All the outputs have diodes, so aren't affected if another similar circuit wants to drive the data output "high".

>>1680989

are you VCO/VCA/VCF anon or another?

>>1680977
>>1680986
and stop typing like a retard

>>1681002

how am I "typing like a retard"? you mean properly spacing posts?

>>1681010
>properly spacing posts

>>1681030

oh no

>>1681030
>>1681030


does this make it difficult for you to read ?

>>1681030


Maybe you should get some help with that

>>1681072
>does this make it difficult for you to read?
that wasn't claimed. I asked you to stop typing like the retard you are.
since that's not something you're going to do ... I guess there's not much more to it.

>>1679557
Anon, Fluke moved to a lifetime warranty some years back. Call them and find out if that unit qualifies.
I've had my old Fluke DMM repaired a couple of times for free.. breddy gud

>>1681083
>>1681030
>>1681002

what the fuck crawled up your ass and died?

>>1681106

>>1681138

reply to that dude, it's his jimmies getting rustled

>>1680946
>>1681143
>>1681138
>>1681106
Stop shitposting and better say, if T12 @ 45W is comparable to 900M @60W soldering iron?

>>1681158

I personally don't like tips with integrated elements. they might be a bit more responsive but i don't think it's worth it.

>>1680879
Jesus just delete everything after and including the first question mark. Everything else is your session ID.

>>1681166
Worth what? The price of each tip? Because as far as I know that’s the only downside.

>>1681166
Thing is, I broke 3 heating elements, just because soldering iron fell on the ground...
Also, chink tips are $3/piece, which is not bad.

>>1681183
https://hobbyking.com/en_us/soldering-iron-tip-set-10-pcs.html?___store=en_us
I get my Hakko chink tips for $.6 a piece

>>1681178

>Worth what? The price of each tip? Because as far as I know that’s the only downside.

Yes. the performance isn't really much better. at the power levels you're using, you'll notice the 60W vs. the 45 W still.

T12's on genuine Hakkos are mostly for the 70W series as far as I know, and most of the Hakko handles that use 900M's tips are 50W, so it sounds like you're looking at buying clones. Is that the case?

>>1681183
chink integrateds will break when they hit the ground, too. nothing ruins a session like hearing that ceramic tail with the ground. and

>>1681185
900M style tips are cheap as shit and basically as good as the OG tips.

I need to construct a Wheatstone bridge to see how the output changes when I change the value of a potentiometer from 0 ohms to 10 kohms. How do you choose the smallest value for R2 so that R1 and R2 work reliably? I have the data sheets for the resistors and potentiometer that I'm using but I'm not sure what to look for.

>>1681197

pots all have a wiper resistance listed on the data sheet. that will be your effective "minimum"resistance for your pot.

the other resistors usually have a percentage tolerance.

what question in the lab manual are you trying to answer?

>>1681202
The question asks to find the minimum resistor value R2 so that the pot (R1) and R2 work reliably. That's verbatim from the manual. It says to refer to either the power, current or voltage ratings in the data sheets. I'm confused as to how those values would get me the minimum resistance of R2. The pot datasheet says the absolute minimum resistance is 2 ohms or 1% (whichever is greater)

>>1681191
>900M style tips are cheap as shit and basically as good as the OG tips.
It is hard to find good ones. I ended up stripping all to bare steel and tinning them with acid flux.
>chink integrateds will break when they hit the ground, too. nothing ruins a session like hearing that ceramic tail with the ground.
True. But with T12, you just swap the tip and continue.
>>1681185
Sure, they are dirt cheap.
>>1681187
My soldering 900M iron had 60W heater..
And I consider getting T12 control board and handpiece, just because I have 11 19V power supplies laying around, because I hoarder junk.
Or maybe even simple relay and push button, because temperature control is overrated.

>>1681209

power rating = heat generation = resistance value drift. same with current rating. if your resistors are too small, you'll suck too much current, things are no longer accurate.

This is really an ohms law question disguised as a Wheatstone bridge question. pot a minimum value will suck the most current through that circuit. how much current will that be? how much will go through Leg 1 vs. Leg 2? what does that tell you about R2 based on the datasheet ratings?
they're trying to get you to maintain minimum current through the bridge while still allowing you to get an accurate reading across the bridge.

that, or they're referring to the 15V source, in which case this is more a Thevanin/Norton equivalence question. depends on where you are in class. you probably just started semester, so that's unlikely.

>>1680739
>>1680720
Appreciate the info and clarifications.

I wanted to do remote telemetry for some water tank gauges, they're 200 metres away, but over a hill or in a gully and through dense forest (no LOS). What would be a good wireless network to use? currently looking at a bunch of pages (LoRa/Zigbee/6LoWPAN) but not sure what would be a good fit. Any suggestions? Was originally planning on using ESP32 to do it, but haven't bought any parts yet so can go with anything.

This little guy was the shit for games and programs that didn't run at native resolution or color depth. Not sure why Microsoft killed it off in Windows XP onwards.

>>1680835
another question about this..
if it says i should be getting 24.5V and i;m only reading ~23V on my cheapass Centech multimeter, is the MM just miscalibration?

>>1681095
wtf that's based

>>1681284
LoRa might work. Or if you don't mind paying a bit for service GSM/UMTS

>>1680938
SOT23 is pretty easy as SMT goes. if a SOIC-8 would be easier, there are also 2G66 might want to git skild in dead-bug assembly style because components aren't getting any bigger
try a MOSFET instead. 2N7002 should work

>>1680949
sounds like you need a logic analyzer or a scope or the candump utility or anything visual. my first guess is that the bit rate set on the devices matches but is off (system or peripheral clock mis-configuration?)
also check: wiring polarity, termination (which is also bias), address format, interface configuration

>>1680977
and application notes will, where available, teach you more than datasheets

>>1681010
>"""proper""" spacing
>reddit
pick any two

>>1681284
oi you got a loicence for that telemetry?
>network
y tho, it's not like Amazon needs to know your tank levels. aim lower
I'm leery of recommending high UHF for this application, but nRF24 gets pretty decent range in open areas, has auto acknowledge and retry, and they're cheap
some regions may reserve a unlicensed HF or low VHF ISM band with relatively large radiated power limits (13.56MHz, 27.12MHz for example). those would shoot over the hill and through the woods easily, but will probably require larger antennas. there are also few highly integrated "SPI-to-antenna" converters and you may have to roll your own protocol or borrow from ham homebrew (APRS?)
in any case, field testing is definitely required before a final design. your local conditions may require a repeater and a fairly directional antenna in order to stay within regulatory power limits

>>1681302
23.5V is within 5% of 24.5%, I wouldn't sweat it

>>1680989

yeah, no shit. hence the hint.

>>1681332
>>network
>y tho
definitely not a requirement, it just seemed to be what was being suggested when I searched. I will look at what you've suggested, thanks

>>1681367
>emitter in the middle
10/10 design for kekability

>>1681363
I mean, you will need some sort of organization for all those nodes and you will need to place them within some system so that they can talk and be heard as needed amongst themselves. strictly speaking, that qualifies as a network, but if it's just you and the tanks, and the tanks don't really need to talk to one another, the network layer can be very thin indeed. if using nRF24 series chips or modules, a bit of manual address assignment might be all that's required

>>1681214
>temperature control is overrated
never speak again

>>1681367
>>1681374
CBE is objectively the best pinout and should be standardised
unless anyone else has opinions to the contrary?

>>1681367
Kek
>>1681377
>never speak again
It is overrated.
I used pic related (not exactly this, but DIY jobie too, based around halogen bulb transformer), never had overheating problems, while with Hakko 900M I had problems with temperature setting. Set to low - only tiny SMD joints will melt. Set too high - SMD pads will lift.
With momentary iron, it was simple. Push in short pulses for small through-hole and SMD, push continuously for ground planes.

Also, I can't drive automatic cars too, gearbox confuses the fuck out of me.

>>1681380
I basically want this, but with nice hakko -k tip. And their microwave tip.

>>1681380
You're not going to get high power and low heat capacity out of a standard iron like you can with one of those guys. And a temperature that only melts solder on tiny SMTs? Sounds like a problem with your iron not putting out the power it could, or a poorly shaped tip. Use extra flux, get as much of the iron tip in contact with the joint as possible, and set it to somewhere between 300 and 350 for ~60/40.

>>1681377
>being this JEDEC
>2019
Japanese and European semicon companies seemed to have opinions to the contrary from the beginning
the collector or drain is almost all of the body of the chip. therefore the collector in the center is the only sensible option for power devices with exposed tabs
BCE isn't everything))), it's the only thing
(ha ha... łittle Russian joke to go with borscht)

>>1681380
pads don't lift by themselves. git gud

>>1681385
>300 and 350 for 60/40
lolwut I've been using 330 and a 1mm bevel tip for general SAC305 SMD work without damage. gotta turn it up a bit for ground planes but not a huge flow-breaker. what is a bit of a flow-breaker is shutting down and swapping back to the 1.6mm screwdriver to do more mechanical stuff like DPAK tabs and pin headers. maybe I need another 936 station just for that

>>1681385
>You're not going to get high power and low heat capacity out of a standard iron like you can with one of those guys.
One I did was quite shitty and took around 10-15 seconds to heat up (depended really badly on electricity in mains. At 220V it would heat up in 10, but at 215 it would take 20 seconds). Mainly because my SMPS was modified by me when I had no idea about SMPSes.
T12 seem to heat up in same interval.
> And a temperature that only melts solder on tiny SMTs?
280 C, leaded solder.
> Sounds like a problem with your iron not putting out the power it could, or a poorly shaped tip.
I suspect shitty heat transfer and slow chink electronics, because it was overshooting badly.
T12 has no heat transfer problem, and electronics seem to be faster.
> Use extra flux, get as much of the iron tip in contact with the joint as possible, and set it to somewhere between 300 and 350 for ~60/40.
Well, adjusting temperature for each kind of joint helps, but this is not the thing I was expecting from temperature controller iron. It is basically the same experience as dimmable iron. (pistol thing with button and diode inside. Normally it runs at half-cycle, and once you press button, it shunts diode, and heater runs full retard).
This gets annoying when you're repairing motor controllers for e-bikes. In the beginning I tried using 60W hakko knock-off, and it was pain, I was switching between 450C (those mosfets are mounted on heatsink and I can't remove it without unsoldering), and 280C for gate drivers that are based around tiny SMD bjts.
I eneded up using my momentary shit for MOSFETs. (I was running transformer well above its 45W (130W measured after smoothing caps), I have no idea why BJTs there didn't exploded and caused blackout in entire city)
>>1681389
>pads don't lift by themselves. git gud
Well true. But burnt table happens on its own every time. Especially since I don't like soldering iron stands.

>>1681389
>the collector or drain is almost all of the body of the chip
Oh, should that be changed? It makes a big difference when it comes to D2PAK or those other back-to-board SMTs.

>another 936 station
Or just a station that uses T12 tips that can be hotswapped, if I had a silicone mat I could just grab the old one from my T12, pull it out while it's still hot, and throw another one in.

>>1681390
Yeah when working on ground planes that beefy it may be worth swapping to hot air. Ideally, the temperature sensor would have a significantly smaller thermal resistance to the tip and joint than its thermal resistance to the element, but without increasing the thermal resistance from the element to the tip. This way you'd get negative feedback not on the temperature of the element but on the temperature of the joint.
Perhaps a smart IR thermometer camera pointing at your work could meme that into existence.

280 is tempting a cold joint, if you have 300-320 you should be able to get in there quick and out again, but then again I have little experience with SMTs.

>>1681392
>Yeah when working on ground planes that beefy it may be worth swapping to hot air.
Unless it is BGA or similar package, I prefer soldering iron.
>Ideally, the temperature sensor would have a significantly smaller thermal resistance to the tip and joint than its thermal resistance to the element, but without increasing the thermal resistance from the element to the tip.
900M have thermal couple in ceramic heater. And there is a giant air gap between heater and tip.
T12 has everything in one package, so it should work better.
>280 is tempting a cold joint
Nah, I know what I'm doing. Anything hotter makes joint weaker, because too much intermetallic material (which is tin-copper, lead is not involved, since it doesn't wet copper that well) is formed. Which means joint is ruined, and it will fall off and burn another set of MOSFETs. And you need to let it cool, add more fux, and use lower temperature, or lower time (which is hard).
>, if you have 300-320 you should be able to get in there quick and out again, but then again I have little experience with SMTs.
Works for normal joints, with tiny SMDs it doesn't work that nicely.

>>1681390
>mosfet tabs
fair enough, you've found one of the legitimate applications for soldering guns in electronic assembly and repair: pretty much anything that can't be bent with a fingertip is heavy enough to require extraordinary heat

>>1681392
>should that be changed
no
1. the collector is the most lightly doped part of the transistor. it makes sense for the more heavily doped parts of the transistor to constitute a smaller part of its volume rather than to try to counter an existing bulk-doping
2. the collector is the part of the transistor where the heat is dissipated
>T12
meh. I drop shit too often. maybe when I get a larger workbench

>>1681396
Well, not tabs, but MOSFETs mounter on heatsink. Middle pin (one that connects to positive in N-channels, always forget how it is called) conducts heat good enough, and it gets annoying quickly, because base and ground pin don't take that much heat (330C deals with them without any problem), but middle ones require me cranking shit to 400C, otherwise it takes forever to remove the fucking fet.
Also, not just soldering gun, but ghetto soldering gun. Saved my ass not even once. Most reliable soldering iron I ever had.

>>1681395
ah so you are down with http://publications.npl.co.uk/npl_web/pdf/matc85.pdf then?

>>1681402
N-no I guess. I didn't know leaded solder isn't best for rework. Now I will get leaded solder with silver in it.
About that intermetallic thing I heard from oldfag who was soldering aviation related shit in USSR and who stole more than 9000 gold-plated IC.
And it is true. Overheated joint combined with vibration (e-bike has plenty of those) crack joints and this burns the MOSFETs, because gates are not discharged fast enough. Mainly because chinks can't set their reflow ovens properly.
Joints themselves look fine (besides cracks you barely can see). If I didn't know that shit, I would probably say MCU is dead.

>>1681409
I've heard issues of too much gold dissolving into solder joints. Maybe they should be washed with mercury beforehand. Or just wire-wrapped.

>>1681414
Gold is good AFAIK. It is expensive, but they figured a way to make ICs with copper leads, not gold-plated.

>>1681398
>always forget how it's called
drain

>>1681409
if I read it correctly, the gist of the report is that mixed soldering alloys are generally no more prone to failure than pure soldering alloys, kosher-tier separation of Pb tools from no-Pb tools isn't a danger to solderability, and that mixed alloys are mostly no big deal from a practical standpoint, only the lead nazis in Brussels to worry about

>>1681415
gold is great for electrical contacts, like pins that need to slide into sockets, or switch contacts that rub a little when making contact. it helps cushion vibration and doesn't form non-conductive oxides under standard conditions. but it is well-known to make SnPb joints brittle https://www.dfrsolutions.com/gold-embrittlement-in-leadfree-solder

>>1681425
>only the lead nazis in Brussels to worry about
Honestly, they are victims of lead poisoning apparently. Normal people can't be that retarded.
Their RoHS protects pajeets form lead, and this is main problem here. EU doesn't recycle.
>but it is well-known to make SnPb joints brittle https://www.dfrsolutions.com/gold-embrittlement-in-leadfree-solder
Huh. Interesting.

>>1679331
God I hate that stupid soldering iron.
Literally all they needed to do is add ONE FUCKING TEMPERATURE CONTROL KNOB to make it usable.
Instead the executive-bots saved $0.62 and all users of this iron have to press the tiny "up" dimple button 175 times to change the temperature.

fuck hakko.

>>1681540
details are left to the student

>>1681561
Children love button pushing, especially when it beeps.

>>1681561
Temperature control is overrated anyway.

>>1681561
all you need to do is add a knob to it
you do know how to solder, don't you?

>>1681561
Mickey Mouse design deserves a knob!

>>1681332
I decided to go simple and drop the 3-state output capability. I can use a multiplexer IC if I ever need to extend the number of switches. Currently I only need to read 5*3-position switches + 5*2-position switches for the front panel of my game.

>>1681593
cool, it's basically an NES controller

>>1681600
It's for a control panel simulating a simple CTC panel with 5 switches & 5 signals.
https://en.wikipedia.org/wiki/Centralized_traffic_control

anyone ever lose a whole fucking day due to a bad clip lead?

>>1681613
neat
my bad, wrong chip

>>1681623

I lose days to bad BNC's all the time. especially after kiddy-intern-months at work.

clip leads i'm paranoid about after having lost many days.

>>1681332
hey thanks for the info, the can controller clocks were off indeed. Fixed that but something went wrong with both microcontrollers - used an oscilloscope to diagnose them and saw that they're transmitting garbage. Might be sync bits that aren't getting an ack or sth i'm not sure, will check. The usb-can thingy on linux was sending data nicely.

>>1679985
Hakko fx601 costs $50 leaf bux on amazon

>>1681700
I'd still prefer a china T12 station but hey the price ain't bad for name brand.

Did old arcade crt chassis use some sort of white goop to pot down the components? Or are the caps here leaky & dead?

The only thing that comes on when it's powered up is the degaussing coil. And the only fuse I can spot on here seems good, at < 1ohm of resistance.

Got sent here from /qtddtot/
Does anyone know of a book similar to the mechanisms and mechanical devices sourcebook, but dedicated to electronic components?

>>1681764
silicone goop around caps to hold them in place is pretty common

>>1681623
defective clip leads, loose breadboards, intermittent probes and blown ceramic fuses should be widlarized at the soonest opportunity; don't let that garbage sneak back into the supply chain.

Gonna make a mains-powered ozone generator based off big clive's tutorial (the one with the spike, ring, and CW multiplier), the purpose is to de-smell an ex-flatmate's room before we get another guy in here
wish me luck

>>1681765
No clue what that book is, you'll probably want to elaborate what you're aiming to learn

>>1680244
>is it possible to get half ass dece t china cells for like a buck each?
maybe. especially if you're buying reject cells that have been shadily resold.
Expect overrated capacity as well as more serious issues like electrode overlap and inadequate separator that causes metal lithium plating and eventual short circuiting.
fire is not too likely as long as you don't abuse them. Charge slow, under-charge(important), don't deeply discharge.
still, large lithium packs from sketchy cells is not a good project for those who believe they have a lot to live for or a lot to lose in a fire.
just to note, I spent literal months charging and (somewhat) discharging my big battery because I trickle-charged it, measuring often, and never let it run unattended.
After half a year and only one puffed-up cell, I'm finally comfortable float-charging it full time and leaving it on standby.
still thinking of rigging up an automated fire extinguisher above it. And adding another 16-32 cells (I've got something like 300 total)

>>1681623
>anyone ever lose a whole fucking day due to a bad clip lead?
Worse, I lost other peoples' days due to bad jumpers.
they were the cheapshit ones made of thin iron wire, they started to go high-resistance a little while after being installed.
very insidious. Whenever I get leads now, I check them with a magnet, if the wire is magnetic, into the trash it goes.

>>1681809
Wait shit, 340Vpeak mains means I need caps and diodes rated to twice that, but I only have 1N4004s and 400V 1µF capacitors.
I guess I'll double-up on all my diodes and caps then, probably knocking my number of full stages down from 15 to 10. What a blunder, and I bought these to make a big CW multiplier 3 years ago.

>>1679901
>powerwall
>lithium polymer
>factory rejects

>>1681765
none that i know of. imo electronics isn't as suitable of a field for such a book as mechanical engineering is. EE is a little more open-ended and the math behind its fundamental building blocks is easier to remember, eg if you were to compare the design of an op amp circuit to the design of a gear tooth or seal gland

>>1679901
>using an inverter when 90% of household electronics run just fine on 100VDC

>>1681861
Maybe he is 220V fag?

>>1681862
Even then, basically all SMPS devices can run on 100-240VAC, and have an internal rectifier. This way they don't have to be designed differently to work in different countries, and why those power socket adapters don't destroy thousands of appliances per day.
So I guess that's actually ≥140VDC he'd have to step up to, but still a lot easier than an inverter.

>>1681861

Any transformer-based device or adapter.
Most PSUs with active PFC (mainly computer power supplies).
Any kind of induction motor.
Clocks that use mains as their time standard.
UPSs won't like it.
Light ballasts.
Anything with a capacitive dropper.

Probably a couple others I'm forgetting.

>>1681877
Most of those can be fairly easily selected around. Not many things use transformers these days (aside from microwaves), it isn't terribly difficult to get washing machines and the like that have brushed universal motors, the clock thing could rather easily be fixed with a little tinkering, and ideally you'd be using SMPS constant-current LEDs of some kind. But if most normal computer PSUs have active PFC, that would be a problem.

Though if he is on 120VAC, swapping to 120VDC might mean that his SMPS devices don't start at all since they're under ~140VDC. As a denizen of the 240V world, that isn't much of a concern.

>>1681668
that would be weird, CAN doesn't really handshake with or address individual nodes (per se), just an ACK slot and a NAK slot for any connected node to use
I'd check the terminations and the bus common-mode voltage but aside from that no big ideas here

>>1681765
I have seen electronics sourcebooks aimed at book club subscribers, once owned one but can't remember anything much about it. Horowitz and Hill may be pretty close to that at a lower level, the numerous electronic project sites run by plagiarizing pajeets are something like it at an even lower and less trustworthy level

>>1681779
found it, I now have two clip-to-Dupont leads that test good. but man with the shitty rotten copper wire on those chink clip leads I won't be passing them down to my godchildren

>>1681878
>Not many things use transformers these days (aside from microwaves)
sometimes the transformer is cheaper than cap dropper + extra insulation. there is a sweet spot in the 1-3W isolated range where a little rectifier transinductor is just the thing as long as it doesn't try to self-identify as a capacitor

>>1681813
Ok, ready to go, safety squints on:

>>1681885
Ok it seems to be wired up correctly, since it's making a faint hissing and there's a tiny purple glow at the tip when I turn the lights off and close the curtains, and I might even be able to feel a faint "ionic wind" out the end. But it's way too weak, so I'll try moving the needle closer to the hoop. And maybe look harder for my sewing kit so I can find a needle that's sharper than a piece of thin capacitor wire.

>>1681887
Moving it closer helped a little. Also here's something strange, it still works even when I turn it off at the wall; I guess the current is so low that capacitive coupling through the switch is enough to keep it going. But I found a safety pin, going to see how much of a difference that makes.

dronefag here, not an electrician

whats the best way to waterproof/weatherproof a circuit board and the connections to it while saving weight?
I've seen some boards with a coating of some clear resin or epoxy i assume was on it for waterproofing, what is it and where do i get some?
i was thinking rubber cement might work but it might also melt if it got somewhat warm

>>1681889
>it still works even when I turn it off at the wall;
Never mind, it's just due to the long RC constant of the 1µF capacitors. I tried using a safety pin but it made no difference, so perhaps voltage is my problem. But I don't particularly want to add another 5 or so stages.

I have a question. Will TSA search my rectal cavity if I have soldering iron with me on plane?

>>1681809
>mains-powered ozone generator

a half-gallon of bleach is way cheaper and more effective. just quickly go over every surface with a sponge mop dunked in a 1:4 solution.

>>1681623
>anyone ever lose a whole fucking day due to a bad clip lead?

not anymore. the trick is to make sure that the wire is held only by solder, so if the solder connection goes, it breaks. it isnt held in place by the sleeve.

>>1681890
the thin layer of stuff is "conformal coating", the deep goop is "potting compound"
leave a blown IC and twenty dollars under your pillow; the radio spares fairy will bring you a can while you're sleeping.

Looking for a term and google is not helping. For a wiring drawing/diagram. When I get to the edge of a wire, say 5354 and it hits the end of a page, and it has another number there 9955, I flip over to the next page and it has lines 9925-9960 listed on the left, I can then look at 9955 and see that 5354 continues from there. What is that number called? the numbers that list where you can find a wire or set of components? Not the wire number itself.

>>1681917
Nevermind, I did eventually find it as just "line numbers" I just started picking random names that I felt they may be called and came across them.

>>1681877
Most active PFCs will run fine on DC, they just become a boost converter

>>1681912
>bleach
Kinda hard to do that on carpet

>>1681923
not if you dilute it properly. Ozone isn't exactly non-oxidizing either

can anyone recc me a decent basic multi-meter?

>>1681930
The problem is more it soaking in and concentrating as it evaporates. That concentration doesn't happen with something that's a gas to begin with (the air diluting the ozone doesn't disappear, rather the ozone itself does). Also I want something that penetrates more into the filler of furniture (and a mattress). I'd use one of those UVC ozone lamps if they were readily available or didn't take a month to get here.

wait what the shit you can get UVC LEDs (270nm-285nm)

>>1681931
Aneng AN8009 from aliexpress for $30 or so, because I'm a shill.

>>1681938
>concentrating as it evaporates
https://pubchem.ncbi.nlm.nih.gov/compound/Sodium-hypochlorite#section=Density
>Vapor pressure, kPa at 20 °C: 2 - 2.5

https://en.wikipedia.org/wiki/Vapour_pressure_of_water
>Vapor pressure, kPa at 20 °C: 2.3

In any case, you can't expect the concentration to increase as it will readily react with various substances in the material.

>I want something that penetrates more into the filler of furniture
Well then. Good luck with the O3 generation.

>>1681941
Turns out the other flatmates didn't like the idea of leaving a mains deathtrap in an unoccupied room overnight, can't blame them though. Previous flatmate was a smoker, who suddenly left after being convicted of rape so we're out 1 flatmate that we've got to get into this smelly fucking room. Maybe I'll go for a battery-powered solution to ease the hazard and see if I can boost the voltage even more. Flyback transformer I guess. While I'm there I might go for the AC methods, that have two electrodes on either side of an insulator to get mega coronal discharge. Might get one of those <$2 10V UVC bulbs just in case it gets here before the flatmate does. No clue if they're any good though, will have to look into that one.

>>1681943
Hmm. Interesting scenario. I would probably surface-treat surfaces using diluted bleach, and either toss the furniture/mattress or put them in an enclosed space and generate the O3 in there. UVC will (as you probably know, because wavelength) only work superficially. I would probably avoid O3 generation anyway due to the health & fire hazard.

>>1681946
The UVC lamps make O3 gas that should diffuse through the carpet and mattress to at least some extent in the hour it takes to decay. The health hazard hence shouldn't be an issue if nobody is in the room for the next hour or two, and I've never heard of it being a fire risk. Some air purifiers have UVC lamps in an enclosed container from which no light can easily escape, and with a fan blowing air through them, which deals with the optical hazard. To be honest we're probably going to toss the mattress anyhow, both because of the nature of the crime, and because the police hadn't the foresight to cover it with something before inadvertently covering it with fingerprint carbon powder as a part of the search warrant.

Wait did we hit bump limit before the old thread expired? Never seen that before.

>>1681949
>UVC lamps for O3 generation
Oh, I assumed the use of UV directly to destroy in-air and superficial odors.

>health hazard hence shouldn't be an issue if [...]
Just FYI: it can cause acute respiratory issues at just a few ppm.
https://www.nap.edu/read/12032/chapter/11#187
The half-life is a day or so iirc.

>never heard of it being a fire risk
Well, it's an extremely powerful oxidizer. There's not that much to "hear" about it beyond it being an inherent quality of the substance.

Is this an acceptable way to do this? I'm an absolute beginner so want to make sure I'm not doing anything retarded.

>>1681981
Some other questions: While soldering this, my fingertips got quite close and Im sure the diodes and resistor was exposed to some heat. The soldering iron was hot. Possibility they were damaged?

Second, does it matter which way the resistor is facing on the circuit?

did I create this correctly? Will it work? I'm the precious two posts

>>1682012
What's the diode for?

>>1681903

You not supposed to carry it in your ass anon

poo is bad for the tinning

>>1682012

No high-side diode with an inductive load. Just put some sort of fly-back suppression, cap and a fast diode in parallel with motor, cathode towards positive supply.

>>1682043
Motor has electronics, nothing is flying back.

>>1682016
It's not really needed but an anon got me alittle paranoid since it's a cheap chinese fan. A different anon said it didn't matter where I put it in the circuit so I just put it there (have no idea what I'm doing).

>>1682043
It'll take me awhile to understand what you're telling me, but I'll get it eventually

>>1682046
will my circuit work? I've already built it but don't want to test it yet

>>1681887
speaker cable is an excellent source for fine wire strands. sometimes power cables are good too

>>1681912
the wire itself was bad. anyway it's in a better place now, in the back of a garbage truck

>>1681951
OP, being a faggot, put up the new post while the old one was on page 3 instead of 10. a bit of a head start, that

>>1682061
quit stalling and do it faggot

>>1682046
ah, you're right. didn't read, just saw DC and something involving "brush" In that case, it's fine, it'll work. BJT isn't ideal, But it'll work.

>>1681943
>suddenly left after being convicted of rape

this wouldnt by chance be a guy named Kevin, self-styled DJ and tattoo artist? if so, small world.

>>1682099
>>1682046
To be clear... There is really NO real reason at all to have a diode added to the 5v wire leading to the transistor and DC brushless fan, even if it's a cheap aliexpress one?

>>1682143
The more diodes in series the slower it runs.

>>1682143

No. you're working with a regulated source and you have no heavy load requirements. the brushless motor in a cheapo fan like that isn't able to act as a generator through the commutating electronics, and you don't have any kickback to worry about.

>>1682143
If you want to do something useful, calculate the otimal value of the base resistor.

>>1681881
yeah so it would seem, the controller was trying to get a pls respond from somebody but the data didnt even make it to the canh/canl lines. at that point i almost lost all hope.

turns out the controller and the clock were at fault. Slapped on a similar one. slowed down the system clock and voila, 1mbaud stable. Now the controller that used to work fine doesnt, same problem. Fkin electronics i tell ya

I need to build or buy a damn workbench. I'm going to run out of room tomorrow.

>>1682128
no

>>1681981
Soldering could be better, but its ok. Cover it up with heatshrink to be safe.
>>1681983
What kind of soldering iron are you using? Chances are they are just fine, components like those are very resilient.
Resistors can face either way, it doesn't make a difference.

I'm an idiot and didn't realize I had to take the capacitors out of the circuit to measure the capacitance. Once I got them out, they show pretty close to the correct capacitance: 3 nF instead of 2 nF but that's all the resolution my meter has. This also explains why the circuit pretty much worked as expected despite everything being covered in leaked capacitor oil.
They have lost quite a bit of oil out the back (pic related shows where it came out, the upright one looks the same on the bottom) I'm guessing most of the oil is out based on how much ran down the inside of the machine. What should I do? use them like they are? Pop the plate off, pour in some mineral oil and fill the back with epoxy? Can you even buy capacitor oil? What does the oil do, just cool the plates inside? Thanks as always for your advice.

>>1682307
>mica capacitor
I don't think that these have any oil in them, unless there exists some type of Mica capacitor that does contain oil that I am unaware of. Mica capacitors are generally known for their stability and generally do not require change. They are used places where stability is super important since they do not change value with temperature change. The oil in capacitors that use it is used as a dielectric.

>>1682307
>everything being covered in leaked capacitor oil

https://nj.gov/drbc/library/documents/PMPWorkshop_0107/dupont_electrical.pdf

https://www.aliexpress.com/item/32904490215.html
Has anyone used one of these things to mess around with tube circuits as the B+ supply?

>>1682327
>Application:
>1.Pressure test power.
>2.Hunting,eradication of mice.
>3.Capacitor charging,electromagnetic guns power supply.
>4.The power supply for your electronic device, according to your system can set the output voltage value.

>2 (Dead).Hunting,eradication of mice.

>>1682332
holy shit

>>1682332
>2 (Dead).Hunting,eradication of mice.
lol, I didn't even see this shit.

>>1682325
well if that's the case, I 'managed' those PCBs all over my fingers. Either way, thanks for the link.
>>1682316
I don't know enough about capacitors in general, but the owners manual says 'oil filled' and I read at least one testimonial where I think a guy was in the same situation - capacitor leaked oil but registered a reasonable capacitance value and he 'refilled' them somehow, recommended capacitor oil, but said mineral oil will work in a pinch but I don't know how he did it from a mechanical perspective.
I did get 2 doorknob caps with the same nominal values as the ones in the pic, but am on the fence over installing them vs. trying to resuscitate the originals. My main concern is how would the ceramic doorknob caps fail? Short or open circuit? I fear if I short it I will fry the HV transformer and be shit out of luck, which is why I lean towards resuscitating the originals. Again, the original circuit seemed to work with the original leaking, but capacitative capacitors in there.
Here's a pic with the OG caps and the sparklers running that shows a bit of the oil / PCBs leaked out. Seems like the leaky capacitors were in their element once things warmed up a bit.
Any recommendations based on new evidence appreciated.

>>1682337
I bet the local hams hate you lmao
From what I remember, a shorting capacitor wouldn't do much harm, could you repost the schematic just to be sure? Jam a fuse in a low-frequency part of the circuit if you're worried.

>>1682337
do I remember correctly that your ceramic caps are also rated at twice the voltage of the original micas? thus failure is far less likely

>>1682375
They also look bigger and dryer (no oil), hence less prone to thermal issues. I say go for it.

Thanks guys, sounds like I will just go with the new caps since I have them.
>>1682371
>could you repost the schematic
pic attached, I believe its 115VAC across the two wires coming into the left side of the transformer.
>>1682375
>rated at twice the voltage of the original micas?
Yes, I am just paranoid about there being some magical properties of the original caps that would be lacking in the new ones.
>>1682381
>They also look bigger and dryer (no oil), hence less prone to thermal issues. I say go for it.
Sounds good my man, I will go for it later today just need to get some screws to fit the new ones and I'll be broadcasting all up and down the valley.

>>1682450
I made a mount for the new capcitors I think I will drill some new mounting holes in the orange piece so it moves everything over to the left about a half inch to give more clearance to the spark gap radiators but I will have to make longer copper straps for the electrical connections. I have copper flashing and copper pipe that I could hammer flat. Is one better than the other and should I do anything to prepare the surface where the bolts go to make it more conductive?

>>1682463
ceramic caps do tend to fail short. but so do mica caps. the only real magic in the old caps that isn't in the new caps is that the old caps could have used the backplate and oil fill as a heat sink, so there is the small matter of heat dissipation to be wary of. I would gently break in the HF at first then check the cap temperature (C2 will discharge thru R5/T3/T4 when idle, so no shock hazard). it is possible that you might need to limit the use of HF in cont. mode if heat is a problem. it shouldn't be because you did verify the current ratings were close when buying, r-right?

Curious, anyone have any suggestions for how to draw out a two level terminal block? I'm going over some materials, and it has everything on it except for the terminal blocks over and over, and they have jumpers that go from the 1st level to the 2nd level, however no documentation for that. And I feel this shits pretty fucking poor in that perspective as it leaves you guessing where it goes.

>>1682492
Thanks that is all really great info and stuff I was worrying about. The old caps were rated 5 KV and 5A at 3MHz the new caps are rated 10 KV and the only current rating I found was 10A no frequency specified. What kind of temps do you think I should start worrying? Thanks.

>>1682509
70-80°C or so is where I'd start to worry. to be most sure, you could cement a thermocouple on one of the caps and monitor while you run the HF to see if the temp levels off anywhere. if there is forced air, heat should of course be much less of a problem

Holy shit i can't believe it worked so well on the first try!
I present to you a fully milled one stop PCB, no toner transfer, no lasering, no gross acid etching, no photosensitive uv curing crap.
It also does THT holes, i just haven't placed the pcb on a piece of wood so i couldn't let the mill drill through it.
Literally put a blank pcb in, and two minutes later pull a finished, fully drilled, ready to solder pcb out.
That shit is so precise that after its done etching, i will be able to apply solder mask, cure the whole thing with UV (no need for printing out template cover pads) and then mill the mask paint from the pads. I can even let it mill in silk screen letterings into the mask.
Best $500 i ever spent

>>1682553
>etching
*milling

>>1682553
that's sexy

>sending pcb orders without verifying availability of components
why do I keep doing this
>>1682553
fml

>>1682553
any trouble with registration (lining up the back side of a 2-sided board)? are you doing those little press-riveted vias?

tried to add a headphone jack to a toy keyboard. the jack has a "normal" connection meaning that when nothing is plugged, the normal and "out" are connected, otherwise the in goes to the tip and out is the sleeve.

what happens is that the keyboard is silent UNLESS i insert a wire and bridge the tip and sleeve on the other end. if i plug in a head phone, the speaker is outputting sound but quietly.

could the jack be broken (short circuit somewhere)? or could this be messing with the rest of the board somewhat?

doesn't matter if i open or close the gate before the jack (meant to put resistors there bc the output is 9v).

>>1682619

you wired it wrong. keep trying diff combinations until it works.

>>1682622
ait, ill just keep at it until it works or breaks i guess.

long shot, but do you know if it's reasonable to believe the main chip of the thing outputs the audio one bit by bit on 8 pins? and then they are summed/turned analogue in a resistor ladder?

that's what i believe from looking at the pcb. i know nearly nothing (can you believe it?) but the sound is supposedly 8-bit, and 8 pins are wired to the ladder, and the output goes to an audio amp chip. so... maybe?

>>1682623
>and then they are summed/turned analogue in a resistor ladder?

this is a perfectly respectable way to do digital-to-analog conversion.

>>1682625
ah! thanks! im learning :)

new thread when?

>>1682636
omw

>>1682638
>>1682638>>1682638
>>1682638
>>1682638

>>1682638

>>1682636
go back to /g/ and stay there

>>1682643

some fags seem unable to comprehend that, once the tread is past the bump limit, the thread doesnt jump to page 1 so noobs wont find it and we get LESS EXPOSURE, and fewer funny dumb questions.

>>1682653
everyone uses the catalog

>>1682653
go back to /b/ and die there