/diy/ - Do It Yourself

1206 resistors are hand-solderable r-right?

Was looking at tantalum caps to use in place of electrolytics, but they actually have really high ESR? Like >3ohms. I thought they're supposed to have lower ESR compared to aluminum electrolytics?

Cheap Rail to rail opamp similar to the lm328? (and THT)

I have device than needs 1KW at 230V AC
I have 500W solar panels, what would be the easiest way to combine wall powert with my solar panels so the panels supply whatever they can and the wall power will fill in the rest?

>>2025478
sell 500w to the utility company, buy 1kw from the utility company

>>2025482
they don't buy that shit here

>>2025434
i hand soldered an 0201 last week

If I have 5V but want regulated 5V do I boost then buck/LDO it?

>>2025491
Just back feed your meter

>>2025478
You would need a large battery array that gets charged by solar and the wall through some solar charge controller.
Have the batteries power a 230vac inverter that in turn goes to the "device"
The power output for the solar panels vary too much over time, and using them in parallel with wall power to make up the difference isn't very straight forward. Hence using batteries to store the solar power, and when the batteries get low, the wall power can kick in.

>>2025416
the fuck is that picture

>>2025526
correct. can be done in one stage with a buck-boost, depending on your circuit. much better if you can just live with 3.6v though.

>>2025557
http://nevod.mephi.ru/English/uragan.htm

So I 'm going to run my car's starter motor from a bank of super capacitors, and the rest of the car from a small lipo bank or 18650 or something.
How can I go about charging the supercap bank while ensuring the charger doesnt get fucked when I put a load on the bank and it sags?
like I cant just connect the lipo bank to the supercap bank because when the caps go flat it will kill the lipos.
I also have to raise the voltage up to 15v to charge the supercaps fully, and ideally it will just always keep them at 15v unless driving the starter motor.

can I just do this with a boost converter? or will it explode when the cap bank gets pulled down to 8v suddenly? I could also I guess add a relay that disengages the charging circuit when starting but that is dumb.

>>2025676
you need to add current limiting, whatever you do. this can just be a resistor of sufficient size, but better if the supply itself is current limited.

>>2025696
arent most boost converters current limited?

>>2025466
LM6142? TLC2262?
>LM328
Do you mean LM324 or LM358?

>>2025676
>I also have to raise the voltage up to 15v to charge the supercaps fully
You can do that just by changing the control current of the alternator. So long as you ensure that the rest of the grid can handle 15VDC (probably will), or that the 15VDC is isolated to the capacitor charging circuit, you'll be fine. A boost converter is also acceptable, but might be too slow without a real chonker.

>>2025724
LM324. Dislexia made me mix the quad and the dual package names.
>LM6142? TLC2262?
thanks

>local stores (not english speaking country)
>go to website
>looking for a instrumentation amp
>IC categories
>"black"
>4xxx
>misc
t-thanks
People who buy electronics here are usually self taught hobbists or repair technicians that only buy the exact part number they are replacing (without knowing what it is) so it be like that sometimes

what component would let me connect a heating wire to 120v out of the wall but put a ceiling on the amperage it can pull so it doesnt trip the breaker?

>>2025771
A fuse.

>>2025773
yeah but I dont want it to cut off power completely I just want it to keep drawing power, but not above a certain amperage

>>2025780
A heating element is a big resistor. It limits itself.

>>2025780
a heating resistor of the correct size

>>2025787
What he really needs is a thermostat.

>>2025761
I just search mouser or wherever, see what appears to be common, and then search for that on the more obscure sites.

>>2025771
A thermal switch would probably be good, but only if it's thermally bonded to the heating element, and even then if the heater is sized poorly you could get enough hysteresis that the element is damaged before it switches off properly. There's really no better solution than sizing your heating element such that it gives exactly the power output you need without any extra circuitry.

What should I use to power an LCD backlight https://www.orientdisplay.com/pdf/AFY480272A0-4.3N6NTM.pdf (~16V forward, 100mA peak) from a 5V wall wart? Obviously need some kind of boost converter, but I would like something with current regulation for brightness control (either PWM or a digital interface). I'm having trouble finding a backlight IC driver on digikey that fits this. I thought I found one that would meet these requirements, but it's too small for me to hand solder or prototype with.

Is there a easy structure that allows me to do the following
>amplify the signal (AC audio) if it is too small
>if it is big enough do not amplify or amplify less
I can only thing of this using an OTA or lots of opamps. Is there a simpler way?
>trying to do a pre-amp for something with a weird input

>>2025780
try wrapping it around a pipe and plug it in lol

>>2025780
>not above a certain amperage

you can reduce the max current by limiting the max voltage, as per Herr Ohm: current = voltage divided by resistance.

you can use an (expensive) variac, aka auto-transformer, or a less expensive light dimmer, as long as it can handle the wattage of the heating element.

a thermostat, as others have suggested, is absolutely the wrong tool for the job, as it will pass 100% of current and blow a fuse in the fusebox before the thermostat gets hot enough to turn off the current.

>>2025814
I found this https://www.diodes.com/assets/Datasheets/PAM2841.pdf which looks fine and is in a more reasonable package, but I'm not sure I understand how to size the resistor in series with the LEDs to limit the max current, since I don't see what parameter sets the output voltage/current precisely.

>>2025761
>>"black"
lmao
what even is this? DIP packages?

>>2025853

google ''AGC circuit''
i made a whole bunch of these for audio using JFETs, but i dunno if you can even find new JFETs for sale.

>>2025449
why do you want to replace electrolytic caps by tantalum ones ? i only use the latter in feedback stage for stability since ceramic ones have very low ESR

>>2025861
Hello. I am a brainlet and missing something here. Please explain how a toaster or oven range works.

>>2025919
Google “thermal flasher”. It’s basically a rather slow analog PWM. Could get a similar result by making a relaxation oscillator with a relay and potentiometer. Both cases have their lifespan limited by erosion/spot-welding of the contacts when they break contact each cycle, hence why you want them to cycle as slowly as convenient. 0.1Hz sounds about right for a stove, but you might want faster in the case that it overheats and explodes in only 5s at max power. A phase-fired dimmer does something similar at 60Hz, which is effectively fast enough for anything you’re likely to want to power off 60Hz mains.

While a traditional thermostat won’t work due to the large delay times, you could arguably measure the temperature of the element itself (with a thermocouple or with thermal IR) and switch a relay/SSR/TRIAC via a control loop like a PID. If you want fast response and will encounter a large variety of required heat flows, it’s the best method, but if you just want a heated towel rack it’s a waste of effort.

Whatever you do, add a thermal fuse somewhere in case something goes wrong.

>>2025928
Thank you for taking the time to answer the question in detail and using plain English. Cheers, m8.

Accidentally posted in old thread.

>>2025281
>>2025396
So I went and bought a batch of these things.

https://www.ebay.com/itm/1x-10x-LM2596S-DC-DC-3A-Buck-Adjustable-Step-down-Power-Supply-Converter-Module/223524573272

My question is, is it possible to adjust these with a PWM signal from an arduino or nah? I noticed the little screw you're supposed to turn but I'm hoping I can use one as an adjustable benchtop unit along with that solar project of mine. Don't wanna have to turn the screw and fiddle around with getting the voltage spot on if I could just have an arduino do it for me.

>>2025919
The heating elements in a toaster or range are sized appropriately so they don't blow the breaker. 1500w heating element draws 1500 watts and therefore won't blow the breaker even running 24/7. There's also a thermostat but that's to regulate the temperature not to limit the current.

>>2025416
I have an Ender 3 Pro with a 24V, 15A, 360W power supply that I want to add some LED lights to. I have a 12V strip of lights and some individual 5W LEDs. The strip has its own power supply, so I probably won't use those like I'm thinking. I'd like to add the 5W LEDs to be powered from the printer, some how.
Can I just tap into the power supply, or should I try to find a 5V pin on the board that isn't used? I think there is at least one. How can I go about figuring all of this out? Do I need resistors?
Sorry if these are stupid questions. Maybe someone can point me in the direction of a guide that would help me?

>>2025995
>is it possible to adjust these with a PWM signal from an arduino or nah?

As supplied, no.

You would need to replace the pot with an op amp and control it's gain with the Tarduino. The LM2596 is adjusted by modulating the feedback, as opposed to changing the reference. The latter can be done with a PWM signal and low-pass filter. The former requires an op amp to stand in for the pot.

>>2026028
Also, what size wire would I use for something like this? Or for adding 24V cooling fans? I measured one of my current fan's wires and it was 1mm in diameter, so I ordered some 18AWG. Foolish of me, because the actual wire is 1mm but with the insulation it is at least twice that. Does the thickness of the wire matter that much?

>>2025864
they sorted it by "color". I am not joking
>>2025865
thanks, jfets are expensive tho

>>2025995
Those are trimmer pots, I’d recommend looking at their values (IIRC it’s two 10ks and one 100k) and buying full-sized pots with appropriate knobs. Those trimmers aren’t rated to enough operations to consider using for a benchtop unit anyhow. Desolder the old and solder the new. Probably gonna do the same thing myself.

They’re part of an analog feedback loop, so it isn’t possible to just PWM instead since that would leave the circuit running open-loop. Viable solutions include using a DAC+JFET, a digipot, or maybe an OTA, but I think the analog pot method will be convenient enough.

>>2026012
Well actually a toaster doesn’t have a thermostat at all, but rather a thermal timer. The difference being, the temperature of the bread doesn’t change the trigger time. That said, there is one toaster that does this properly, see the Technology Connections vid on the topic.

>>2026028
LEDs need current limiting. For low power LEDs this means resistors, but for something at 5W I’d see about a switching regulator of some sort. Also check that your 3D printer’s PSU can handle the extra load, another 10-20W isn’t necessarily something to sneeze at.

>>2026028
>>2026035

Get a small buck converter board, LM2596 boards will do just fine, print a case for it and use it to get lower voltages. That's what I did with my Ender 3. 12V might be useful down the line, if you decide to swap your hotend fan for a quiet 12V one.
As for wire diameter, you can't go too big (except for practical reasons). Google an "ampacity chart", as much as it pains me to use that word, to check a conductor's current carrying capabilities.

As for those individual 5W LEDs? You will probably need to limit current going into them somehow, but a resistor isn't a good choice for high-power LEDs. Stick with the strips, there are v-slot mounts for them on thingiverse.

>>2026030
>The LM2596 is adjusted by modulating the feedback
Alright, so if I read that correctly I should be safe in assuming it'll still output the same voltage if the input sags from say 20v down to 15v, right?

>>2026037
Gotcha. I'll look into it once I get them in the mail. It's tiresome fucking around with a huge ATX power supply when I just want to test a circuit.

>>2026057
>it'll still output the same voltage if the input sags from say 20v down to 15v, right?

*assuming I just want the things to output somewhere in the neighborhood of 5v, of course.

But that does raise a further question, am I going to need to also buy a boost converter to charge the lead acid battery if it's a cloudy day and the solar panel output drops to below 13v?

>>2025526
buck-boost it

>>2026061
I’d see about making a dedicated buck-boost charge controller. Possibly with an IC, but probably with an MCU using an internal lookup table or two for feedback. Just to be as nice to your PbH+ as possible. Make it MPPT too while you’re at it, won’t need any more hardware.

>>2026030
>Tarduino
:D

Why is Smartsim such hot garbage?
>https://www.smartsim.org.uk/
I have to use this for my digital logic course, and I swear that nothing in this forsaken program works.
It doesn't even support Ctrl-Z, or any form of undoing previous actions.

>>2026037
I misread my package of LED lights. They are actually 5mm in diameter, 12v and 20mA. So according to Ohms law, that would be 0.24W per bulb. If they work for what I'm planning, I might be able to use a dozen or more. I have 30 of them.
How would I figure out how much extra load my PSU can handle? Just go through and add up all the amps and watts?

>>2026042
Those buck convertors are pretty cheap, I found 10 of them for less than $20. I was thinking about the kind with a display and then making some sort of control box with maybe a switch for the LEDs and possibly a pot of some sort for controlling the fans. I don't know if the pot idea would he OK since the printer can already put them at a fraction of full power. I'm guessing it'd be OK. Probably should just start out basic though, I tend to come up with grand plans and ideas and then not follow through, leaving myself with yet another project box...
I'll check for the mounts, thanks for the info. I really want to put something on my hotend so I can actually see what's going on down there. I ruined my eyes when I was a kid with lasers so I have a really hard time seeing in dark or contrasting light, even in the shadow of the hotend carriage. Probably the strip wouldn't be good for that. I know I actually have some decently bright Cree LEDs that I was going to use to build an aquarium light with around here somewhere. I even bought the lenses for those.

how can I can i confirm the green objects to the right are in fact LM386 amps?

>>2026161
ok I'm realizing the green objects are likely not IC let alone LM386's. I think my chips and solder are both in storage, will get tomorrow.

At least I tried :^) really wanna learn this stuff, I think once I get past the noob barriers I'd really enjoy it.

>>2026161
compare # of pins, for one

>>2026158
>5mm in diameter, 12v and 20mA
Post pics? 5mm LEDs are usually specced at 3V, you'd put 3 in series plus a resistor to run on 12V. That's probably what your 12V LED strip already does.

>>2026161
>>2026172
They aren't LM386s, they remind me of those shitty class-D amp modules that use RC filters on their output that are aimed at blinkies. Read the part number on the IC itself to see what you ended up with.

>>2026161
look at ic, read part number, google datasheet. with a bit of circuit analysis you can come to a conclusion whether or not they are what you need
>inb4 fake semiconductors

>>2026206
>part number
Yeah that's how I realized, thanks. I order some cheap stuff from AliExpress but got impatient and made another order domestically. The domestic ones looked proper and had LM386 written on them. Not sure wtf these are but they were only like $4 shipped.

anyway, I ended up soldering a 10v power supply to a 12v DC motor to make a stir plate. It was my first time both just fucking around with live wires and using my soldering iron

>>2026222
>Not sure wtf these are
well what does it say on them? not like they're useless, might even be better than an lm386.

>>2025814
My monitor uses this

>>2025917
They're smaller.

What is the best way to remove the solder mask of a very thin trace?, it happens that the circuit board is cracked and the trace is too thin to use sand paper or someting like that. I can't use a bridge either because it connected directly to some chinese chip blob. The trace is about 0.5mm (0.020 inch)

>>2026301
I use needles

Apparently all footprints in kicad were drawn in imperial units. Should I redraw everything in mm?

Hi, I'm making a desulfator. I want to make a couple inductors from salvaged toroids. As far as I know, there are two main types of toroids, lossy ones with large hysteresis loops used for filtering and efficient ones used for power electronics, right?

How could I tell them apart?

Why the output does not go to the rails? I think it is a LTspice setting somewhere because it does not matter which opamp or comparator I use.

>>2026294
>They're smaller.
Oh, sweetheart, that's not how you do things in electronics. Different components are designed for specific purposes. You can't properly swap out electrolytic for tantalum without making sure the circuit can work with it.
Now, tell us what project you're working on, hmm?

>>2026331
>How could I tell them apart?
If you are salvaging them from old electronics, you should take note of what they are being used for in the existing circuit.
The bare torrid typically doesn't have really good markings on it, so you won't have a good datasheet to check out specs.
You can try to wind various lengths of wire around the torrids, and try reading charge/discharge currents to get an idea for their capabilities.

>>2026357
I replaced them with ceramics

>>2025676
but why tho? and what kind of car? is your car a late model with ECM monitoring of battery voltage / ECM alternator control? ECM will not like even a couple tenths of a volt difference. Does your vehicle have many high-current accessories? Will they function off your small LIPO bank after key off (seat return to position / power door) ? will you keep replacement LIPO bank / replacement boost converter / replacement supercaps on hand? because O'reilly's won't have them in stock when your car won;t start and you need to get to work. If your car wont crank and supercaps get drained, or alternator fails how will you jump start? If alternator fails at night and you need to limp home a few miles on battery only, how long will LIPOs last running fuel injection, ecm, lighting etc? Are you prepared to do all your own maintenance re: anything electrical / ECM related since dealer / shop will take one look at your mods and say "there's your problem"? If you sell your car will you convert it back so new owner can get it serviced?

>>2026321
Embrace the inch, my child.

>>2026354
Is it an open collector comparator?

Any recommendations for a program to design circuit diagrams? I have a copy of NI Multisim, just curious if there is anything a little more lightweight and presentable.

>>2026331
Wind them up with 10 turns of whatever, put a resistor in series, throw them on the function gen and measure the two voltages with your scope. Should be pretty obvious which one is lossy and which isn't.

>>2026354
Use an ideal op-amp/comparator model if you're worried. In general, rail-to-rail op-amps/comparators get kinda nonlinear towards the end of their travel, so I'd advise against designing circuits that require them.
Definitely use a comparator of some sort in this situation though.

>>2026568
KiCAD

>>2026321
There's a bunch of footprints that are metric. Anything with a 1mm pitch for starters, a reasonable amount of switches and pots and other hardware too.

I sell DIY circuit board kits and getting all the components sorted is becoming a pain in the ass.

If I order ten resistors from digikey they send me a bag of ten resistors. If I order 100 resistors they send me a bag of 100 resistors. Is there any way to get them to send me ten bags of ten other than making ten individual orders? Not digikey specifically. I'm looking for any supplier that will do this for me.

>>2026612
If you're not doing SMDs where you'd just cut the reel at the correct length, then you're probably shit outa luck. Are the resistors in a long strip, or are they loose? If in a strip, cut them to length, if loose I'd weigh them out.

>>2026424
no, my car isnt gay and doesnt have any of that bullshit.
it's a track car so it makes perfect sense to ditch the 15kg battery, all the mounting, cables, ect, then fit the supercap bank near the starter with short and thin/light cabling, and mount the lifepo4 pack somewhere else.

I was planning on including a boost converter with a type-c cable which I can plug into my phone and charge the supercap bank should the main cell go flat, then including a 3 position battery cutoff switch so I can isolate the discharged lifepo4, running the whole car from the supercaps and then alternator when it starts.

>>2026528
no
>>2026570
fiddling around with simulation steps and compression improved it a little. I mean I know that the real circuit DOES NOT do that so I am safe.

Another question. I am modulating two waves using PWM. Why does this filter works well when the amplitude is larger? It does not make sense for me.

>>2026631
I decided to use my brain for a while and it has nothing to do with the filter. It is just that the PWM is with D = 1 at those times.

>>2026620
> If in a strip, cut them to length
That's what I've been doing thousands of times a week. I was thinking of building a machine to automate it but my order must be already running through machines like that at the supplier side.

>>2026634
yes machines that are called chinese hands.

>>2026331
those black ones are usually the lossy kind. You could always make a fast edge pulse generator, make an LC tank, and compare the ringdowns (short = lossy)

>>2026681
>a fast edge pulse generator
Or just charge the C up to 12V or whatever and close the switch. Assuming your scope does single-capture, that is.

Can anyone help a beginner out? I have a line out signal from my computer, and I want to drive speakers with impedance of 4 ohm. I figured a common collector setup would be best, as I don't need amplification, I just need to drive the speakers from the wall rather than from the line out. However, via simulation I realized that 4 ohm impedance is very low, and all the voltage is being dropped by the last capacitor. I know you're usually supposed to drive a high-impedance input with a low-impedance output, but I can't change my speakers. What can I do to circumvent this problem?

I'm building a relay setup with an arduino for heat humidity control for a mushroom (the tasty kind, not the dude weed kind). I need temperature and humidity sensors that I don't have to worry about letting the smoke out and killing a batch. Any recommendations on mfg's, or do Chinese dollar bin peripherals have a good track record?

>>2026809
You could look into a DHT11 or DHT22. Haven't really heard any bad from them besides them feeling flimsy as fuck.

Just run a couple redundant ones for each batch for basic sanity checks to make sure they're reporting relatively accurately, they're cheap enough.

>>2026799
at 1khz that 2.2nF capacitor is a very large impedance. You should increase it to something like 10-100u
>>2026809
can't go easier than those dht11 things you find. I have used them for a growing box once.

>>2026799
Bro, 2.2nF cap has a reactance of like 72k at 1kHz. You've created a voltage divider between your 72k cap and your 4 ohm load. Of course you have no voltage. It'll take 39uF to match your 4 ohm load probably like 470uF-ish to look like a low impedance to your 4 ohm load.

>>2026817
>>2026819
Right, I forgot to factor that in. Thanks for your help.
I'm getting some distortion on my output (this is using 100 uF). Do I have to amplify the signal first (via a common emitter or opamp) then put it through the common collector?

first time solderer here, my weller wes51 was working then all the stopped, the pen doesn't get hot at all now. still powers on fine, there's just no heat anywhere on the pen. what do? tried turning it off for a couple hours, along with resetting the power strip it's on.. nuthin

>>2026850
btw, i was running it on max :( please don't tell me i fried it or something

>>2026841
you must decrease the value of the emitter resistor. Try something like 10-100 ohms

>>2026858
Also, driving speakers from an emitter follower is a TERRIBLE idea. You will waste 4x more power on the resistor and tranny than the speaker

>>2026859
and you have not said anything about the speaker power rating

>>2026850
>>2026853
Check the resistance between pins 2 & 6 on the iron plug (normally 11 Ohms). Open the base up and probe for 24AC from the transformer secondary, 5VDC from the regulator, then check the fuse.

>>2026858
Oh, that works beautifully. Thank you very much.
>>2026859
I will consider a push-pull follower instead, thanks.
>>2026861
I'm actually not sure... I got these from a broken Panasonic surround sound system. I'm just going to hope I don't blow them out.

>>2026859
>Also, driving speakers from an emitter follower is a TERRIBLE idea

it's not a terrible idea. it's wasteful, and the dude needs to lower all the resistors, and force the transistor to run hot, but it works, and is as simple an amp as you can make.

as for efficiency, who cares? when i turn off the lights in the living room, there's about 20 LED power lights lit from various pieces of equipment, each piece wasting 1W to 5W 24 hours a day. since we're drowning in wasted electrons, a few more wont matter.

(also the dude is confusing common emmiter with common collector, so your comment is gonna confuse him)

>>2026859
>Also, driving speakers from an emitter follower is a TERRIBLE idea

it's not a terrible idea. it's wasteful, and the dude needs to lower all the resistors, and force the transistor to run hot, but it works, and is as simple an amp as you can make.

as for efficiency, who cares? when i turn off the lights in the living room, there's about 20 LED power lights lit from various pieces of equipment, each piece wasting 1W to 5W 24 hours a day. since we're drowning in wasted electrons, a few more wont matter.

Turd herder here
Are there any resources on how to upgrade from 200A to a 400A residential 1 phase panel at my house?
I'd like to do as much as I can and then get a sparky to sign off on it so I'm not paying one $100/h to pull cables and screw a pannel to the wall.
I would probably just add a second 200A panel as long as it doesn't have to have a 3' offset from the 1st panel.

>>2026885
just bypass the fuse m8

>>2026885
not wiring general. Fuck off to another thread

>>2026799
In addition to using a low-reactance capacitor, you’re going to need that emitter resistor to be somewhere in the ballpark of 4Ω, in order for your amp to be able to sink sufficient current. At 12V, with a 4Ω emitter resistor, with no input signal whatsoever, your amp will be wasting 1.5A / 18W. 9W for each the transistor and the resistor, which will both need heat-sinking. This would be made somewhat better with a class-B or class-AB solution, maybe with op-amp feedback for linearity. But considering you’ll probably have SMPS noise anyhow, a ready-made class-D amp board from eBay is likely the most practical solution. That or a nice big audio transformer.

>>2026878
>yes, all my resistors are ceramic 20W, how did you know?

ohmnii-chans, what do I make as a beginner project? I looked at the rolling list and I was like nani
ok but seriously, is a digital clock a good beginner project or is that also too high level? (im not going to use any microcontrollers or anything)

>>2026931
yeah, clock dividers are super easy to use.
trickiest part would be getting the 32.768kHz to oscillate, if you're using discrete transistors or smth. And then just wiring it all up to look nice.

>>2026939
If he’s using FFs or SRs for clock division then he should be using Schmitt inverters for the crystal. Or running it off optocoupled mains. What I see more issue in is turning that binary number into a BCD one to feed the 7-segs. I think the only viable is to run BCD counters in parallel with the binary counters, with proper div/60s and div/12/24. I saw what was required to turn a binary number into a decimal number by watching a Ben Eater video, and it required finite field algebra on shift registers in order to perform long division.

>discrete transistors
Complete waste of time, unless you’re really into discrete transistor/diode logic.

>>2026871
well, I guess this is as good a first project as any... It stopped while I was using it, while it was on max heat. I was almost done, too. My multi broke so I'll have to get a new one tomorrow, but thanks for the advice, I will be rereading this tomorrow in hopes of getting it going again.

>>2026953
>max heat
Is it a temperature controlled iron? You want to run those around 300-350C, running them too h out burns off the flux and leaves you with dry joints. Running an element too hot can result in it burning out sooner than normal, they’re usually only rated up to 400 or 450 or so.

>>2026321
all kicad parts are drawn in metric. it just defaults to imperial in the grid size

>>2026951
>turn a binary number into a decimal number
just count to 0-9 and wire 10 to the reset pin

>>2026931
>ohmnii-chans
incorrect, we're "ohmlets" here.

Is it possiblel to isolate the signal ground and power ground in an simple opamp amplifier? I see that some power opamps have a differente ground for signals and I was wondering if the same was possible in a common opamp to redudce noise from a ground loop

>>2026985
w-wait why
>>2026951
>Schmitt inverters for the crystal
whats that...
>>2026939
>wiring to make it look nice
haha

>>2026984
For a clock it works fine. Actually now that I think of it, you don't even have any redundancy either. I was thinking that you'd use binary counters to get your h,m,s values, then add extra BCD counters to get the values for the 7-segs, but you can't even get minutes or hours with binary counters in the first place. The same counters that are doing the clock divisions to get the next longest increment are also doing the display, i.e. no redundancy.

But if you have a binary value sitting on a register or in memory, say 10110100, and need to get that to a 7-segment display as decimal 180, how would you do that? You need to figure out how many times it goes into the hundreds place, how many times the remainder goes into the tens, and how many times the remainder of that goes into the ones. In other words, long division. Considering the number of clock cycles required, it might actually be no slower to just have a series of cascaded decimal counters.

>>2026987
If you have a split-rail power supply, say ±12V, then the 0V rail will usually not be used by anything but your signals. If you're running some amplifier topology that needs to source or sink a significant amount of current into or out of the middle rail, then that might cause some noise or unwanted feedback, then I can see wanting to seperate the two. You'd want to keep the amplifier ground the same as the power supply ground for efficiency reasons, since that's where all the current is being drawn through, so I'd just take the power ground, feed it through like -60dB of low-pass filters to get rid of the ripple, then buffer it with a spare op-amp.
Never seen an op-amp with any ground pin though, always just Vcc and Vee. Unless you're talking single-supply.

>>2026990
A schmitt inverter (inverting schmitt trigger) is often used in a Pierce oscillator, due to its positive feedback preventing the system from reaching equilibrium. Schmitt triggers are comparators with hysteresis, read their wikipedia.

>>2026996
Screw the 7seg display, I'd just use a ring of leds for hours

might be a retarded question but is there a such thing as a "cheap" pic'n'place machine? How much do they generally cost to rent or buy anyway? Thousands I'd presume.

>>2027039
You can order assembly service from pcb fabs.
You can also DIY a PnP.

>>2027057
I wanted to try to reduce costs but pcb fabs service might be neccessary, and I could only imagine a DIY PnP would be really a nightmare to maintain and troubleshoot,but I could be wrong

>>2026996
https://www.youtube.com/watch?v=BA12Z7gQ4P0

a eprom with byte write you can do it without the arduino. the guy even did the transistor logic version. the chips arent expensive but there arent a ton that still have byte write. its like atmel cornered the market or some shit

so basically the address lines can be configured to have a set on off logic vs input pins mimicking the transistor logic. you have to look on the datasheet and see byte write

Will you guys scold me if I ask circuit analysis questions...

>>2027010
You mean 12 or 24 LEDs? Requiring that many individual outputs wouldn't be that much simpler than a 7-seg model, might even take more ICs. Assuming you'd use a demux circuit for the LEDs. Counting to 12 or 24 in BCD is 4 or 5 bits respectively, so you'd need at least a couple of 74HC138s.

>>2027039
3d printer + vacuum pump + some extra printed hardware to hold reels of parts and some steppers to move them along.

>>2027072
Are you talking about converting a long binary number to BCD? If it's an 8-bit EEPROM you'd feed 8 bits from the counter into the address pins, and you could fit both of the BCD outputs at the output. Assuming 8 bits will be enough, might need some extra gates. Actually a reasonable way of tackling a clock problem, since you just need one EEPROM with at least 8 address inputs, or 256kB/2048kb. Need three EEPROMs (plus six 7-seg decoders) for H:M:S, plus the counter ICs. One or two ICs to get down to 1Hz, then 17 bits of binary counters, say five 74HC69s. Not sure if it would be any better than just using BCD counters, but it's still an interesting approach. Writing the code to generate the EEPROM data would be a small fun challenge if nothing else.

Course that isn't a general solution to division, it's effectively just dividing by 10.

>>2027101
Depends on how stupid the question is. Does it require Millman's theorem?

>>2027108
>Does it require Millman's theorem?
Uhh, it's questions that involves small details that I cant wrap my head around like why a voltage divider doesn't include a resistor far away on the other end that is parallel (and separated by an inductor) in its denominator, ill have a picture up tomorrow! it's just these super small details that makes me question my life

>>2027119
If it's a DC analysis, then the inductor shouldn't matter and the two resistors should be counted as parallel.

Hoping to get a better answer here than in /sqtddtot/. I need a trusted brand of micro drivers for RS work. I’d like a kit with seperate drivers for each size so I don’t lose the tips in a field or on a roof. I mostly use phillips and slotted but if it comes with torx or hex too I won’t complain. As long as its in a case. Price is not an issue and I’d like glorious Nippon or Kraut steel. If there’s a US brand that’s fine too but I don’t know of any. I am sick of my cheap HomoDepot micro drivers getting twisted and fucked on RS lugs because the last guy partially stripped the screw with the wrong size and using retard strength to crank it down. I read here recently that Wiha is made of shit now. I found Wera but I don’t know where these are made. I’m lost on this.

Sup /ohm/, i don't have a lot of knowledge on circuitry but i want to build some simple synthesizer components and something i'm looking to build is a component that takes an alternating input signal (like a piezo element in a guitar) through a line in input and puts out the amplitude envelope as an analog voltage, so when the string is struck the output will be a relatively high voltage that decreases as the amplitude of the string decreases and i have drawn this simple circuit using 4 diodes that i think will work for this, are there any mistakes i'm making or would this work fine?

>>2027275
That circuit is not even close to right. What you want is something called a peak detector.

The diode allows the input (on the left) to charge the capacitor quickly when it's at a higher voltage than the capacitor. The capacitor can only lose voltage slowly through the resistor. Imagine it like the audio waveform is a set of mountains you can run up then hangglide over the valleys between them.

An OP-amp buffer is needed on the output in most applications. Without that whatever you're connecting to acts as another resistor and changes the discharge speed.

>>2026960
yes, it maxes out at around 426C, which is where I was using it at the time. it was working fine, then just straight up stoppped. I'll have a look at it applying the advice I was given above later today, i hope i didn't burn out the tip/element.

>>2027275
did you know? people have already done this before, look at google image results for an "envelope follower" or "envelope detector". the simplest circuit is this pic >>2027422, but there are plenty of more advanced circuits using an op-amp or two, some even use two or three sample+hold circuits in parallel and switch between which one is dominant by resetting them out of phase.
what circuit you use will depend on what requirements you have. the plain diode detector shown in that image has the problem of a continuous voltage drop from the input to the output of 0.7v or so, meaning it won't work very well if your signal amplitude is below 0.7v (1). the single op-amp models fix this, which is what i'd usually recommend using (2). for slightly faster response there are "full wave" op-amp rectifiers, which give you both sides of the waveform (3). but all of these use a resistor to slowly drain the capacitor. this is fine if your input frequency is a lot higher than your required dc voltage variations (an am radio detector being the archetypical example) but if you want that output voltage to change quickly, or to have less ripple, you'll want to use sample+hold circuits on clocks (4).

>>2027548
lmao your element is fucked, buy a replacement element for the pes51

>>2027275
>>2027275
my good siiirs you nmust do the needfull and research amazing american tehc called "envelope follower " on search engine google sir please do the needfull and then you will find you answer sir

How do I find cheap BT audio modules/ICs (not that cheap really, just not a premium price)? All commercial products seem to have an IC that is unavailable to me in any way and exclusive to china, and I can't seem to find any cheap solution. I would like an internal ADC (probably going external, since even TI's IC's don't have one that meets audio requirements) and I2S output if possible.

>>2027422 (checked, based and redpilled)
Thanks man thats exactly what i'm looking for, the circuit i drew is one isaw used in a 230 AC to 12 DC circuit in a book about electronics on yachts and i figured this would work as an envelope following ac to dc converter as well

>>2027663
>ADC
So you want to turn an analog audio signal into airwaves? BT transmitter ICs are a lot less common than receivers, since they’re usually bundled with something else, like a generic RF IC inside a phone. Where is the audio coming from? If it’s from a computer or CD player or other digital device, I’d want to just output a digital S/PDIF signal directly.
>I2S output
Wait which is it? Both audio receiving and sending? Might end up doing something like an esp8266, where you have to do some programming to get a general-purpose IC to do what you want.

Also consider LCSC for part sourcing.

>>2027682
A full bridge rectifier is slightly different to what you drew, because A: the grounds need to be floating, making it a pain to use for an audio circuit, and B: you put half the diodes backwards.

>>2027275
You drew a diode ring. Great way to make metal robot noises

i understand and use bode plots just fine but my brain fucking shuts off when I try to understand nyquist plots or transfer functions, in anything besides simple filters. i've read a ton of whitepapers and articles where every sentence makes sense but the overall concept is complete nonsense. has it "clicked" for anyone else? any suggested resources?

>>2027875
transfer function's just the resistor divider eqn on steroids, right?

>have been using zener diodes as flyback diodes
haha
oops

>>2027882
yes until you add a BJT

>>2027786
>You drew a diode ring. Great way to make metal robot noises
Are you referring to a diode ring mixer, with transformers and such? Or is there some non-isolated use of a diode ring used for audio effects?

>>2027875
Nyquist plots are difficult to decode because a coordinate on them doesn’t equate to any actual frequency. You only use them when you’re interested in certain features (can’t remember, it’s been 3 years). I think they were useful for something to do with poles and zeroes, though the intuition of them never really clicked for me. And I only ever used transfer functions and plots of circuits with passives and op-amps, no transistors.

I suspect Nyquist plots are easy to draw on a scope in XY mode, maybe. Designing a circuit to generate live plots would be neat, alongside a curve-tracer circuit. My function generator shares a ground with my scope, so I need to use a transformer for a curve tracer.

>>2027904
That’s really bad, assuming the zener voltage is sufficiently low.

I think I'm retarded. Can someone explain/link me to something that explains how oscilloscopes measure voltages in circuits without a common ground between the scope and the circuit? I figured the measurement would need to be a differential measurement with the scope converting it to a common ground before having the ADC sample it (unless the ADC has a differential input I suppose), but I was looking at these open source scopes:
dpscope.freevar.com/files/DPScope_II_V2_06_Schematic.pdf
https://reference.digilentinc.com/lib/exe/fetch.php?tok=a3dd5f&media=https%3A%2F%2Fs3-us-west-2.amazonaws.com%2Fdigilent%2Fresources%2Finstrumentation%2Fopenscope-mz%2Fopenscope-mz-sch-revg.pdf
And neither seem to use a differential input with an instrumentation amplifier. Wouldn't connecting the scope ground to the circuit's through a probe potentially connect two different potentials together? Also if you have more than one channel wouldn't you only be able to have one ground connection, otherwise wouldn't you potentially connect two different points in the probed circuit to one another through ground?

>>2028057
Differential probes are addons for scopes, somewhat expensive ones at that. If you need to use one, then use one. Making your own differential amplifier circuit to do so for you is also an option, so long as the voltage difference isn't more than your battery voltage; this is the sort of thing I'd consider doing for a high-side current shunt or measuring a high-side transistor biasing circuit or whatever.

In normal operation, you'll almost always be just measuring the voltage from one point with respect to ground. But, assuming your measurement source has no protective-earth connection through the wall to your scope, you still have one degree of freedom. In most situations, just having one probe tying the ground reference to wherever should give you the measurements you need, only if you need two or more simultaneous measurements will you possibly need a differential probe. Because ground clips can cause unwanted shorts, and because you technically only need one ground connection unless you're dealing with high frequencies, I tend to just connect one ground clip to the circuit, and probe everything else floating. If you have a large DC offset on a low-amplitude AC signal, then swapping to AC-coupling can prevent reaching for the differential probe.

But not all circuits are floating with respect to your scope. A non-isolated mains appliance will have its neutral connection to worry about, meaning you should connect the signal of the probe(s) to the circuit but leave the ground unconnected. If you need to measure on the other side of a bridge rectifier, then every cycle the negative rail will switch from 0V to -340V/-170V with respect to ground. So for this, you either need a >340V differential probe, or to slap some kapton tape over your scope's ground lug (generally not recommended if you don't know what you're doing). But thankfully, most mains circuits are galvanically isolated from neutral via a transformer, be it flyback or otherwise.

>>2028057
>>2028061
But as I said in >>2027943, a curve-tracer is a circuit where a differential probe can be necessary. You need to measure the voltage across two different components that are in series, and apply a signal through the two. In my case, my function generator and scope both plug into the wall and both have ground references. So I can either:
>connect scope ground in the middle of the two components, probe on either side and invert one channel, and use an isolated function generator
>connect function generator across both components with ground at one end, put one scope probe from ground to the middle, put second scope probe from middle to top with a differential circuit in the middle
Since differential probes are expensive and cheap circuits can be noisy, it's much easier just to pipe the f-gen through a cheap isolation transformer.

That said, this is a somewhat rare example. For all of the standard work I do, mainly probing digital signals jumping from 0V to 5V (or -12V to 12V RS232), and probing audio signals varying about my 0V ground rail, there's no need for any differential measurements. Just slap the ground clip on the GND rail and everything else will follow easily. Note that I do have a digital storage oscilloscope, so in a pinch I can just save a waveform with one ground reference, and superimpose it on my measurements with a second. Or save them both and zip them to my computer to process with python or whatever.

In the case of current measurement, I'd consider buying/making a battery-powered current probe, using one of those hall-effect ferrite clips.

>>2028061
retard here
can I just use an LM339 or suitably high speed compatator, hook wires to the inputs, and use that as a diff probe?

>>2028067
Depends? If you want to get differential measurements on a digital signal, and you can power the circuit with sufficient voltage to bridge the maximum gap between the two grounds, then I guess so, sure. But if it's a digital circuit, you'd probably be better off with an optocoupler or transformer-based solution. Or even just capacitively coupling your signal with the AC-coupling mode. Can't really think of a situation where such a comparator circuit is necessary though.

Differential probes are mainly required when you have an analog signal that you don't want to distort, which requires a rather linear circuit. An instrumentation amplifier would be your go-to so long as the maximum gap between the grounds isn't above a sensible op-amp power supply voltage (±18V or whatever).
One method for higher isolation is using an analog optocoupler, with mock-feedback. They have one LED and two phototransistors, one on each side of the isolation gap. You use one to get the signal across the voltage difference, and the other to feed into an op-amp for feedback. So long as the two phototransistors are well-matched, the circuit should be pretty linear. Something that uses the hall effect could work too, though both would require input buffering to bump the input impedance back up. No clue what kind of methods proper HV differential probes use.

>>2028057
>Wouldn't connecting the scope ground to the circuit's through a probe potentially connect two different potentials together?

this is perfectly fine 99% of the time, or even 100% of the time if you never fuck with live voltages, which you shouldnt ever scope anyway. (use an isolation transformer if your circuit uses 120Vac)

>Also if you have more than one channel wouldn't you only be able to have one ground connection

yeah, which is also not a problem coz just about every circuit has a single ground.

(you're seeing non-existent problems, which may have been inspired by a certain EEVblog video which made so many people unnecessarily paranoid about scopes. fuck Dave for doing that.)

>>2028101
>which is also not a problem coz just about every circuit has a single ground
Doesn't mean you won't want to measure voltages not referenced to ground. Say, if you've got a class-AB amplifier, you might want to put a probe to measure the difference between the voltage at the two output transistors' bases. Or maybe you want to measure the voltage across a high-side biasing resistor or current-sense resistor. Maybe you want to measure the input offset voltage of an op-amp, or the gate-source voltage of a high-side MOSFET. They're not common, but just because your circuit has solid and useful ground connections doesn't mean they remove the need for an occasional measurement departed from those grounds.

>>2028105
>measure the difference between the voltage at the two output transistors' bases

easily done. put one probe at each point and put it in X-Y mode. gives you the differential signal.
various ways to accomplish this. in my case, i choose ADD in vertical mode, and push the INVERT CH. 2 button.

>>2028105
>measure the difference between the voltage at the two output transistors' bases

easily done. put one probe at each point and put it in X-Y mode. gives you the differential signal.
various ways to accomplish this. in my case, i choose ADD in vertical mode, and push the INVERT CH. 2 button.

https://www.youtube.com/watch?v=zsy-pxpfB4o
(advance to about 7min mark coz this fucker just rambles and rambles)

>>2028057
osciloscope probes have ground connections and clips. You use a spring to connect to high speed shit and regular clips for slow stuff

>>2028128
>put one probe at each point and put it in X-Y mode. gives you the differential signal
If you only want to view one trace at a time, then of course using your scope's subtraction mode will work for that. But that's unnecessary, just clip your ground clip to one of the two points and clip the signal wire to the other. Assuming the circuit isn't ground referenced, that is. But as I was mentioning in one of my earlier posts, you only have one degree of freedom. If you want to analyse two different signals at the same time, with different ground references on each, you need a differential probe. Say, analysing the input signal at the same time as the transistor biasing. Or the gate-source voltage of both the low-side and high-side FETs. Examining two signals at the same time is a reasonably common thing to want to do, I often find myself wanting more than two channels. I also said that usually you can get around this, and a digital storage scope helps somewhat, but it's still not unheard of that you find yourself unable to probe two desired signals at once. Such as my example of a curve tracer: >>2028065, where you need the two traces for XY viewing.

>in my case, i choose ADD in vertical mode, and push the INVERT CH. 2 button.
Same with my scope, not that I use that feature often.

>>2028129
He's asking about differential probing and ground references, not the kind of probe hardware you should use.

>>2027882
The transfer function is H(s)=Y(s)/X(s) where X(s) is the input and Y(s) is the output, both in the s-domain. This stems from the fact that the Laplace transform is a convolution transform, i.e. the transform of the convolution of two functions is the product of their transforms:
[eqn]
{\mathcal {L}}\{h(t)*x(t)\}={\mathcal {L}}\{h(t)\}{\mathcal {L}}\{x(t)\} \\ Y(s) = H(s) X(s) \\ \implies H(s) = {Y(s) \over X(s)}
[/eqn]
The behaviour of any linear time-invariant (LTI) system can be described as the convolution with the system's impulse response (a unit impulse aka Dirac delta is an identity for convolution), and concatenation of LTI systems is convolution. So in the s-domain, you just multiply their transfer functions to get the overall transfer function.
Y(s)=H1(s)H2(s)H3(s)X(s) => H(s)=Y(s)/X(s)=H1(s)H2(s)H3(s)
For an amplifier or attenuator (divider), the transfer function is just a numeric constant, y(t)=k x(t) => Y(s)=k X(s) = > H(s)=k, but the general concept also encompasses integration, differentiation, time shift and arbitrary FIR and IIR filters.

>>2028150
yeah, that

>>2027875
only time I have used nyquist plots for anything was for system stability when tuning control systems. I do not think you use them for filters.
>how to make it click
study classic control

>>2027904
So they behaved exactly like a normal diode? what

>>2028141
>He's asking about differential probing and ground references, not the kind of probe hardware you should use.
most scopes are ground clips are grounded to the outlet ground so if your circuit is not ground referenced you can just do whatever

>>2028166
most scopes have ground clips that are grounded*

>>2028165
they exploded

>>2028190
Zener works like a normal diode dum dum. You probably just shorted the supplies by connecting a zener in the "wrong" way

>>2028197
or used a zener with a voltage lower than the supply and shorted it

>>2027275
Check out the envelope follower on the MFOS website

>>2028315
what software you guys use to make block diagrams? (BLOCK diagrams, not circuit).
I need to make a small presentation on a project I have and I want to do something nice.
>inb4 download 50gb of matlab just to make blocks in simulink

>>2028327
non intended quote

>>2028327
draw.io

>>2028327
Just download 50GB of matlab to make blocks in simulink.

>>2025416
Hodoscope?
I've built one before.

>>2028327
Download 50gb of matlab just to make blocks in simulink.

>>2028327
I just use KiCAD for everything. If I need to do something abstract and not very complicated then I'll do it on paper.

>>2028414
Photomultiplier I take it? I think the one I found for the OP pic is from a GM tube based detector.

How the fuck do you prototype on BGA packages if you're a small guy? All the good ICs are only available on shit like BGA and QFN, and it really makes it difficult to develop something capable that competes with consumer electronics.

>>2028327
I have heard Visio can do stuff like that. Don't know how good it is really.
Just don't do them on TikZ or something tedious like that.
You could also try some vector graphics software like inkscape.

>>2028497

>>2028327
>software you guys use to make block diagrams?

- LucidChart (free) for drawing online.
- Visual Paradigm can be used either online or as a downloaded app for windoze.
- Microsoft Vizio - dunno if it still comes with new versions of office
- Gliphy.

>>2028497
>All the good ICs
Only like half of them
>QFN
That's solderable by someone without a reflow oven or hot air, just extend the pads out a bit. Unless it's got a pad underneath too. A reflow oven isn't that big of an investment, you just need a toaster oven and a PID controller with programming to follow a custom temperature curve.

>>2028414

Tech ingredients, is dat you? Is recognize that extruded aluminum anywhere

>>2028497
>How the fuck do you prototype on BGA packages if you're a small guy?
ha ha haha ha you suffer

Will audio filtering capacitors put like this interfere with eachother?

I don't understand how these things work together to produce logic operations? Is a tube basically a giant transistor, which can switch states between passing a high current and a lower current to represent holding a 1 or 0? And if you arrange them in the correct way you can build OR and AND gates? How can you use this same component in tube amps for music? Also, I know nothing about electronics.

>>2028522

no

>>2028523

when used for audio, you drive them in linear mode. when used for logic, you drive them in digital mode, meaning they are either fully turned on, or fully turned off. aka as ''saturated'' and ''cut-off''. never ''marinated''.

>>2028352
x2
Good opensource project

>>2028529

so when is it good practive to put them 90ª from eachother?

>>2028537

the signals traveling through the wires have the force of a hurricane, while the signals traveling thru the ether have the force of a schoolgirl fart. so, position is irrelevant.

>>2028489
No, cathode strip chambers. They are read out by propriety electronics which make use of charge-sensitive amplifiers.

Also, photomultiplier tubes (PMTs) are not a 'basis' for a hodoscope but rather an element of a scintillator-based hodoscope. They convert light to electrical signal. For instance, if I have a grid of scintillator fibers, I could read them out using PMTs or SiPMs, but I don't say that my hodoscope is a PMT or SiPM hodoscope, it is a scintillating fiber hodoscope.

>>2028509
No. 80/20 is a commercially available framing system, which is widely used.

>>2028523
Like transistors, they can operate in the linear region or in the saturated region, depending on how they're biased. I don't know about vacuum tubes, but transistors are usually governed by a simple relationship, like I_collector being proportional to I_base. I'd recommend learning how simple BJT amplifiers and logic circuits work before moving on to vacuum tubes, if only because there's a lot more material on them.

>>2028545
I think he might be referring to microphonic characteristics. The electric field from even a parallel-plate capacitor would be negligible at distances significantly greater than the plate separation.

>>2028546
>cathode strip chambers
Not sure what those are, I'll be reading into them.
>a grid of scintillator fibers
Oh the fibres are actually the scintillating parts! That makes a lot more sense than my assumption that they were ordinary glass fibres. I'm not well versed on scintillation and photomultiplication, we only briefly touched on them in my digital logic paper of all things.

>>2028352
>>2028417
>>2028489
>>2028497
>>2028503
thanks
thanks

>>2028523
apparently they're just like big jfets

>>2028523
yes they behave "like" transistors but most do not have a nice equation that describe their behaviour. you usually work based on experimental data and stuff the manufacturer gave you

I dun goofed. Can I bend this back? I was thinking of using a screwdriver.

>>2028734
Something softer like plastic or wood

>>2028734
lol don't bother. Tried fixing a thumb drive that got run over: it "worked," until I pulled it out and half of the laptop's connector came out with it

Anyone know much about BLDCs? I'm considering changing the stator-poles to rotor-poles ratio, at the moment it's 24 coils on the stator and 28 magnets on the rotor. I'll be building a new rotor anyhow, and finding magnets that fit with nice spacings is difficult. My motor drivers should be able to handle some variety in count, but I suspect I'll end up changing the Kv and smoothness.

>>2028758
I found that 20x6x2mm magnets will fit my rotor reasonably well, with 88% filling efficiency. They're a bit long, but everything is either 10 or 20mm.
I think changing to having 26 magnet poles would make the cogging extra smooth, but I don't think it would impact Kv or torque, from what an article I'm reading says at least.

>>2028734
It's over, Johnny.

>>2027786
Ye i have heard about ring modulation and i think i am still gonna build this circuit for fun noises but i expected it to work like an envelope follower with some voltage drops where the wave goes below zero

>>2028892
diodes were in the wrong direction for that

>>2028908
direction matters?

>>2028917
For diodes, yes. Diodes only let current flow in one direction, as indicated by the symbol which is basically an arrow.
Some components, such as resistors and non-polarized capacitors don't care about direction. Other components, such as diodes, and polarized capacitors matter which way they are hooked up.

If I am making a pedal that uses those large TRS connectors but is a mono pedal, I should just short the T and R right?

I have 9 comparator outputs of 3.3V that needs to go into an FPGA with I/O levels of 2.5V (this is retarded, i know, don't worry, it's meant as a temporary interface to an evaluation board)

They are strictly inputs to the FPGA. I can add a voltage divider to each of the channels, but do any of you know of a simple IC that can solve this without having to add 18 resistors to my board? A voltage translator seems overkill since it's just a step down

>>2028917
>direction matters?
Imagine if Quentin Tarantino directed The Goonies.

>>2029023
>I should just short the T and R right?

it cant hurt. if you use a 2-wire male cable then it's pointless. if you plug in a 3-wire stereo cable, then it's useful.

>>2029023
>I should just short the T and R right?

it cant hurt. if you use a 2-wire male cable then it's pointless. if you plug in a 3-wire stereo cable, then it's useful.

>>2029050
>3.3V that needs to go into an FPGA with I/O levels of 2.5V

diff is 0.8V which is about the same as the drop from a diode.

How do I measure the dissipated power in a boost regulator?
I don't think my tiny TPS61023 can handle the 1.8A@5V from 3.2-4.2V I'm wanting squeeze out of it.
Is it just 1.8*5*Efficiency?

>>2029472
>Is it just 1.8*5*Efficiency?
yes. switchers are around ~90% efficient and the losses that do occur are distributed through the inductor, diode, and switch. i always do the math personally but you're near the border of it not mattering.

>>2029486
to be clear the losses are 1.8*5*(1-efficiency), i guess that should be obvious though

>>2029488
oh right, my mistake
>>2029486
damn, that sucks
thanks anyways

>>2029486
>i always do the math
What do you mean by that? Calculating the switching losses from the FET’s datasheet, then the other losses from the inductor’s ESR and the diode’s drop? Or do you make (or sim) the circuit and measure the voltages and currents and make calculations that way? If the former, I’d be quite interested in how you calculate switching losses. Do you have to integrate down the Vgs/Rds curve with respect to time? I’m a bit lost on how to design my own switching controller such that I maximise efficiency while minimising ripple.

>>2029542
simulate to get ideal V/I waveforms. it's trivial to calculate diode and inductor and Rds-on losses. then there's Qg, Coss, and switching transition losses. you can ignore the first, and the second at low voltages. switching losses however are a rabbit hole with no conclusive bottom. desu for my last switcher I just picked the gate charge at an arbitrarily high gate voltage and compared to the driver's peak current plus some margin. it worked out in practice on a fairly aggressive design.

also important are loop stability and capacitor ripple ratings.

>>2029590
>loop stability
You mean something about the controller's speed to respond compared to the ripple frequency, causing potential output oscillations? Or is the delay from the inductor feeding the capacitor more important? I'm actually considering putting a ∏ (pi) filter after the rectifier in order to drop the output ripple, because it's for a geiger counter and I plan on varying the voltage to see if I can get proportional counting, and want no false detections. Maybe even an LC snubber, tuned to the frequency of the oscillator. Not sure if I should even attempt variable duty-cycle with an MCU as my controller, or if I should just go for an existing controller.

>>2029591
sadly the power stage, output filter, and load have to be considered in tandem for loop stability. if you're doing a geiger counter though i assume it's portable, so you're wanting to boost <12v up to >150v? if so you should figure out your basic topology before worrying about this stuff.

>>2029597
>load
Basically a small slightly lossy capacitor in this case.
>i assume it's portable
Not intrinsically, but I'll almost certainly be using it with a low voltage like that yes.
>basic topology
Don't need anything isolated so I'll be using a simple boost topology, with an IRF840, up to about 400V in standard operation. Voltage divider for feedback. Haven't figured out what filter I'm using because I haven't calculated the ripple voltage (nor do I know what kind of signal strengths I'll be getting at the various voltage levels). To know the ripple voltage I'd need to settle on a frequency and whether I'll be doing variable duty-cycle or not. Which all depend on what parts I'm using. My first proof of concept was using fixed duty-cycle (just a 555-timer with a feedback comparator hitting it on the enable, no FET gate driver), until I shocked too many things with my converter and it just died. Never got it to the point of radiometry. Ripple was somewhat high, and tuning freq and duty cycle to get it boosting up to 400V without issue was also a bit of a hassle. With any luck, being able to vary the duty cycle will give me a lot more freedom. Or do I just want to vary one of the times? I've also heard that using a transformer topology is more efficient than a HV boost converter, so maybe I should be using that $2 flash lamp autotransformer I got from jaycar instead of a 220µH radial inductor.

Replacing caps in 1973 kenwood stereo receiver.
Has 4 axial e caps 1000µF 63V.
Maybe it's autism but, the thought of using radial instead of axial feels niggersome at best.
Digi/mous have the same axials available. I found 3 options that weren't too expensive. Shown in chart on picrel+linked below. Am I wrong that even these should easily meet/excede the originals for this type of circuit design? Or is radial the only way to go?

https://www.mouser.com/ProductDetail/?qs=rBx5S%2FKsFjzOV6PLF12s8w%3D%3D
https://www.mouser.com/ProductDetail/?qs=3tP%252BN51vMXeM%2FCsz2SE3sQ%3D%3D
https://www.mouser.com/ProductDetail/?qs=3tP%252BN51vMXc9gvKytV%252BPeQ%3D%3D

>inb4 picrel is unreadable
havent tried stitching together hi res pics before, lemme know if anything is unreadable

>>2028734
If you choose the alternate lifestyle choice of being a faggot who tries to reshape it enough to be able to force it into the port then please keep me updated on what you break

>>2029641
Can't see any reason why not. But just to be sure, I'd look into the kind of specs electrolytic capacitors for audio circuits should have. Chances are someone's written a blogpost on it, if not there's always stackexchange. Are they power supply filter caps, or are they AC-coupling caps for the audio output? Because for the former it likely doesn't matter nearly as much as the latter.

>>2029641
>Maybe it's autism but, the thought of using radial instead of axial feels niggersome at best.

this is a phenomenon i see too much of on here: people being stupidly picky about inconsequential matters. they're like 99% the same under the hood.

>>2029472
current in (rms) * voltage drop across

>>2029697
Well, they won't fit very well on the same PCB, if nothing else

>>2029718
it's average current*average voltage. Sorry I just woke up

>>2029697
>>2029719
op anon here, agreed. Normally would just use radial but using one here I think i'd need to use a proper adhesive, lay radial on side, tack on an extra lead or wire to one leg in order to span the gap. The 3 trannys and the two resistora keep the area hot and its on bottom of chasis, caps facing towards ground. I think if I use axial I wont have to modify/adhesive. Hotsnot will fail quick here, would need some kinda electronics safe caulk.
>>2029688
Thanks fren, this part of the circuit is the power supply and doesnt have anything to do with audio path. I think its a linear power supply design, unsure.

>>2029767
If you use radials, put each one on a piece of perfboard. Run one leg through the main PCB hole to anchor it, and solder a wire from the other leg to the other PCB hole.

>>2029641
Isn't the metal top shorted to one of the legs? could probably just solder a lead there

>>2029767
>>2029788

An "8ohm" speaker measures 8ohms at 1kHz, and self-resonates at 260Hz. That means:
>the inductance of the the coil is 1.274mH
>the parasitic capacitance is 294uF
using X_L = 2pi*f*L and 2pi*f = 1/sqrt(LC)

are these numbers right? 1.274mH seems doable on a big transformer, but 294uF sounds way too high for parasitic capacitance

Are relay pads and footprints kinda the same or is it a free for all?

>>2029836
the second order nature comes from the mechanical aspects not from a capacitor

Is this ok for a bypass? I am making a tremolo and there is probably enough board space left for a distortion/gain stage.

>>2029854
does the mechanically induced resonance somehow induce a virtual capacitive reactance on the coil? So the coil appears like a 290uF cap to an outside device?

>>2029862
from frequency response perspective yes. That is why you see "without a baffle" and probably "free air" elsewhere if you measure that in a enclosure or with a horn then the frequency would change too.

>>2029872
elsewhere on the datasheet*

>>2029611
boost converters aren't well suited to >10:1 ratios. not that you can't do it, but a transformer is just better regardless of isolation. there's an open source counter out there that uses one of these cute little flash transformers.

if you want to sperg out over ripple you could consider an ~unregulated boost converter with a discrete linear postregulator.

>>2029611
you can make an auto transformer if you do not care about isolation on the boost converter.

Does anyone have any resources on explaining what the components in an op-amp VCO do?
I know the very basics, one half is an integrator and the other half is a schmitt trigger
but beyond that I don't really know.
I'm more looking for how they choose the component values, and what overall effects they have on the circuit as a whole.

(something tells me there's not gonna be any one readily available resource but it doesn't hurt to ask)

>>2030024
post what circuit you are using

>>2030026
I think what it's going to come down to, is me just learning what the individual components in each distinct circuit block does
as in learn how the capacitor values and resistor values affect the charge rate of the integrator; and learn how the resistor values change the threshold voltage of the schmitt trigger

Would it be possible to build a fully electric clock using no ICs or transistors at all?

>>2030068
no, never has been attempted.

>>2030068
You could build a mechanical wind-up watch if you're Swiss enough.

>>2030068
relay logic

>>2029836
8Ω isn't just the reactance, it's the total impedance. The DC resistance plus the imaginary reactance. So before calculating anything, you need to know its DC resistance, and calculate from there the reactance.

>>2029981
>boost converters aren't well suited to >10:1 ratios
I've heard that, yes. I already have one of a different sort of flash transformer, which I'll see about using for my next experiment. Ratio might be a bit extreme for what's a 80-30 times conversion, but if it's acting as a buck converter that won't effect the efficiency like it does to a booster, right?

>>2029998
Which is equivalent to just soldering two wires together on an isolated transformer. Not anything wrong with it for my current purpose, but the result is something with the same downsides of a transformer-based circuit: bulkiness and inability to select a customised inductance ratio.
Be nice if there were little inductors wound on thin fibreglass tubes, that you'd either slide onto a core by themselves to have an inductor, or slide multiple on for custom transformers.

>>2030031
The first op-amp (the integrator) slowly charges up the capacitor until it reaches the threshold of the second op-amp (the schmitt trigger). At this point it activates the NPN transistor, pulling the integrator's input to 0V, causing the capacitor to slowly discharge. The time it takes for the capacitor to fill up is dependant on the current going into it, which is dependant on the input control voltage. I highly recommend simulating it and watching the traces as you change the control voltage. The important thing is that the integrator's non-inverting input is set to Vc/2 by the two 51k resistors, which the inverting input must also equal. When the transistor isn't active, the current into the capacitor is given by I = (Vc - Vc/2)/R = Vc/(2R). When the transistor is active, it's I = (Vc - Vc/2)/R + (0 - Vc/2)/(R/2) = Vc/(2R) - Vc/R = -Vc/(2R).

>>2030068
vacuum tubes,

>>2030068
use vacuum tubes

>>2030109
get an inductor core, make a inductor and put a tap in the correct winding to get the ratio you need. It is very different and much easier than making a transformer.
>>2030068
yes. With a synchronous motor. Old clock radios used them and were pretty bad because line frequency oscillates with load.

>>2030109
>8Ω isn't just the reactance, it's the total impedance.
at resonance it is.

>>2029981
Transformer based converters aren't really suited for 10:1 ratios either. Any calculator for isolated topologies tends to give transformation ratios in the realm of like 2:3 at like 150 turns primary 450 turns secondary on a tiny ass E16,8,5 core. Like how the hell am I supposed to wind that? With 40 AWG wire? How will that handle to 16A peak primary input current?

>>2030253
>put a tap in the correct winding to get the ratio you need
Oh sure, if you're winding your own coils, there's nothing wrong with that at all. But that's bit of a pain in the arse even for 1-offs if you don't have a coil-winder, not that I haven't done so myself. If you have a project that needs a bunch of them, or you're doing a limited production run, then it's just not feasible without some way of automating it. My suggestion was aimed at the rapid-prototyping and limited production markets. When you're just messing about with time on your hands, there's nothing wrong with winding your own coils, or looking about for existing products that might suit your cause.

Also open-core inductors/transformers aren't the best thing to default to, toroidal cores give better performance and less noise. And boy are toroidals a pain to wind.

>>2030271
But it's not 8Ω at resonance, it's 8Ω at 1kHz, which is 4 times higher a frequency than resonance.

>>2030273
Just drop the number of turns and increase the frequency.

>>2030284
> And boy are toroidals a pain to wind.
just use a needle bro

>>2030273
>Transformer based converters aren't really suited for 10:1 ratios either
are you stupid? do you think everyday shit you use work on what? magic? coaxing the tension to lower itself through kind words?

>>2030285
And what increase the switching losses? Increase the complexity of the circuit because you need faster controllers and MOSFET drivers. Decrease stability because of the RHPZ?
Absurd compromise.

>>2030324
They have machines that can wind large numbers of turns on tiny cores cleanly. It's too fine for me to do by hand though and I can't tap it good forbid I wanted multiple taps. Even if you do bifilar windings and use that as a tap all it does is half the number of turns you can have on the core since you effectively double the thickness of the wire, plus your coupling isn't as good and efficiency drops. Plus you have to like solder the tiny ass ends of the wire to the pins because there's no mechanism for attaching it. There's nothing to keep the cores together other than tape or glue. It's all a real mess. Never work with transformers.

>>2030109
>8Ω isn't just the reactance, it's the total impedance
so what's that from, just meters and meters of copper? hysteresis of the core? is it counterwound somewhere?

>>2030362
And you know what, while I'm ranting how do you even design a proper switching transformer without expensive lab equipment anyway? Even using the practical transformer model and KVL/KCL you still don't know a lot of information. Often times you have random cores sitting in a bin and you don't know the specs. The core area the inductance factor or anything. You have no way to measure the size of an air gap if you require one or any real way to control it outside of shimming some paper or tape in there. How the fuck are you supposed to determine the leakage inductance. That's a very important parameter in a lot of converters and you're essentially just forced to guess! There's literally no way to measure inter-winding capacitance or core losses or anything! It's impossible to make a good transformer yourself because there's literally no way to characterize it! It's all just guesswork and I hate it.

Sup, /ohm/lets. I want to use an old phone display (Nexus S 4G Super AMOLED 480x800 24-bit) with a SBC. Any off-the-shelf controller boards that you know of?

>>2030375
>That's a very important parameter in a lot of converters
you can pick a conservative value - say .95 coupling coefficient, depending on how you wind it
>It's impossible to make a good transformer yourself because there's literally no way to characterize it
true but don't let that stop you. i posted my diy bench supply last thread and i've fixed like 4 issues with it so far, with a few more to go. some things can only be determined experimentally but if you know what you're doing you can account for those on the fly. the same shit happens at corporate on $10M projects.

>>2030395
That's all well and good but I'd rather not be un-winding and re-winding a 600 turn transformer on a pea sized core a hundred fucking times. I'd like to be able to do all the calculations and wind it right on the first shot.

>>2030398
i sympathize, magnet wire is ridiculously expensive. therefore you should find a transformer on digikey and develop your project around whatever properties it has to offer.

>>2030284
>And boy are toroidals a pain to wind.
pretend you're an old cat lady knitting something while watching television

>>2030379
>with an SBC
many SoCs have controllers built-in, which may or may not be easier than an external controller. there may be panel-specific initialization, which may or may not be easier to leave to Linux than to try to do yourself. MIPI DSI is highly proprietary and kind of a bitch

>>2030398
you should build an inductance meter. at least you can get the A(L) value to within 10% or so

>>2030403
also the 25 cent phone charger transformers on aliexpress are well-characterized

>>2030362
>And what increase the switching losses?
Any design is a compromise. In this case, you can reduce the number of windings (and reduce output ripple) in exchange for increasing the switching losses. If you select your transistor properly, you can keep those losses insignificant. Unless you care about energy efficiency a lot (e.g. you're using batteries) or you're near the limit of your current thermal design (e.g. will need to swap to a bigger heat sink), there's nothing terribly wrong with making your converter a couple percent less efficient. As for any RHPZs, that's up to the designer of the control loop to make as effective as possible, up to you as the part selector to ensure that your parts are appropriate for the end product, and up to you as the circuit designer to read the datasheets and ensure you're within the stability requirements.

>>2030369
As I said, it's a combination of both reactance and resistance. The reactance from the inductive (and capacitive) elements, the resistance from the metres of copper. There was a rule of thumb commonly used about ~80% of it being from the DC resistance, if I recall. Don't forget to add them at right angles.

>>2030375
>he doesn't use FOSS magnetics simulation software

>tfw first pcb worked as expected

>>2030451
now you can get lazy and start designing boards that almost work as expected

>>2030454
sounds like too much work I'd rather just almost design a board

Theoretically could I replace a 12v car battery with 9 1.5V batteries im series and call it equivalent?

>>2030576
8 my bad

>>2030576
Well, two small things:
they're closer to 14V when running
they need to source 70-250A of current to crank the starter motor

>>2030580
So you're telling me I gotta pull the D cells outta my ghetto blaster? That's rough, man, I need my tapes to keep me sane.

>>2030576
you can jumpstart your car once or twice before they die I guess

>>2030704
you may have difficulty getting 70A out of a few series alkalines

>>2030752
that pajeet electroboom started his car with them.

>>2030776

No, he used them to charge the car battery up to the point that it could start the car itself. It took 10 minutes.

There's more than enough _energy_ in a few AA cells to start a car, but they can't supply anywhere near enough _power_ to actually get it done by themselves.

>>2030597
>D cells
Gonna need more than a few. A high-end AA battery might be able to push 5A or so, and that's the real top lithium primary cell kind of battery, $5 each if not more. A D cell of the same chemistry wouldn't be more than 4 times higher, so you'd need 10 in parallel and 10 in series, 100 cells. If they even make high-end D cells, last I saw they had neutered capacity compared to an AA. With normal AAs you'd be looking at multiple hundreds.

Use a lipo instead, a single lipo I bought from aliexpress can output up to 130A. Of course, you need to ensure it's being charged properly and safely.

In kicad what should I do if I want to mount a device in the casing? (Pots, leds etc) Should I change the footprint to a male/female connector?

>>2030927
Depends on how you'll connect the component. In a professional setting, you'd use some sort of connector like JSTs, which you'd naturally just introduce the footprint of. If you're just using jumper wires like a pleb, then I'd just use the footprint for a 0.1" pin header.
There's also the option of mounting the PCB right up next to the faceplate.

>>2030941
Yeah I have pin headers available. It is not possible (or atleast my brain can't think of a way) to connect three of those large TRS female things with several leds and pots in a casing that does not look like dogsheet

Hey guys, I'm trying to repair my WES51 that stopped heating while in use. It was my first time using it, but it was purchased used from an honest and reliable friend.

I have read the user manual section on troubleshooting this issue, but need some help. I've noticed in pic rel that I only have 5 of 6 pins.

Having a hard time typing on mobile so I'll just upload picture and drop manual link:
https://www.scu.edu/media/school-of-engineering/pdfs/maker-lab-resources/Manual_WellerSolderStation.pdf

>>2030967
>dogsheet
>sheet
this is your brain on kicad kids

>>2031020
BAsically, I don't really know what's what when troubleshooting, barely know how to use a multimeter beyond continuity test.

For example, in reference to manuals first step for troubleshooting this, what are "line chord blades"?

>>2031024
>line
the source of power, mains
>line cord
the wire assembly between the control unit and the power source
>blades
male pins that are more flat than round
put it together and you have
>line chord blades

>>2031033
thanks! is it possible to fry my device with a multimeter?

ok guys, weller wes51 fag here, really having trouble with narrowing down where the problem is. it's not getting hot, the green light just stays on, so I'm assuming it's in a wire within the pen itself. i've got my multimeter, but im not sure how to use it a manner that would narrow things down. there's 5 pins, should these all be displaying continuity to the tip of the pin to the tip of the soldering pen? they are not

>>2031045
yes. consider these two popular ways of blowing up a circuit under test:
>sticking probes into a circuit to measure voltage, only to find the meter was set for current and had shorted out the circuit under test
I'm paranoid, I glance at the meter before I make contact with the second probe, just to make sure
and
>shorting due to probe slippage
for fine work it helps to put the tip of your little finger on the table to stabilize your hand, and carefully choose the direction from which you probe or apply contact force

>>2031080
Not him but I've also wondered: Are there ICs or other components so delicate that using a continuity tester on their pins can cause damage?
I figure the meter has to be passing some amount of current between the probes to register continuity, just wondering if this is ever enough to damage anything.

>>2031068
Most likely candidate is the element, wires don’t just disconnect by themselves. So disassemble the pencil and remove the element, and measure its resistance. Do an idiot check by ensuring the probes give close to 0 resistance by touching them together.

>>2031068
Set your meter to measure resistance. You'll have to find the pins yourself (one pair at a time). The temp sensor should read ~22 Ohms and the heating element ~12 Ohms. If you're not getting anything, replace the iron.

>>2031092
and probe the pins on the end of the iron, not the base.

Ok, so I believe I'm failing at this step. When I have multi set to V_AC 200, I get a reading of 0.04. I'm not sure if I'm taking the right reading with the meter though, but I'm sure that I';m placing the prongs where they need to go.

>>2031080
Ok, so
>sticking probes into a circuit to measure voltage, only to find the meter was set for current and had shorted out the circuit under test
only happens if you have your meter set to measure continuity, right? not ohms or voltage? really worried I fried my station now and made a bad situation worse.

>>2031101
Make sure your probes are plugged into the correct sockets on your meter. One is for current measurements, the other for everything else. Black to COM, Red to V/Ω/mA.

>>2031089
Bob Pease wrote in one of his books of some sort of delicate signal transformers. incoming inspectors were found to be checking continuity through the windings, by which test current the cores lost their designed magnetic properties. they were thereafter instructed to do nothing more to these components than count the leads
today, it is possible to break high-frequency transistors or IC inputs with ohmmeter voltages and currents. RF JFETs for example

>>2031099
blown triacs are also a common failure mode of low-voltage soldering irons. it might be hard for you to replace that without a working iron tho

>>2031101
not continuity, current. continuity is a special case of ohms

>>2031091
I don't think these pencils are able to be disassembled without permanently modifying (i.e fucking up) their housing. Would it be fine if I were to just remove the tip and measure the resistance from each pin to the heating element?

When I touch the two probes, I get 0.03 ohms. sound right? I barely know how to use a multi.

>>2031092
I have it set to "200" within the ohm setting. What do you mean by
>You'll have to find the pins yourself (one pair at a time).

I've got the pencil unplugged and I'm trying to find the resistance of each pin to the tip of the pencil, but the values are bouncing all over the place.
thanks a bunch for the help anons, really wanting to get this working and move some projects forward this week.

>>2031112
so if i have my meter set to ohms and I'm just poking around, i can fuck stuff up? do you think this could happen with an item like a soldering station?

>>2031121
yes, it's possible
no, it's very unlikely. when the ESD warning labels on component packages start to double up, then start to worry

>>2031119
>What do you mean by
Measure the pins until you find the pair for the temp sensor (22 Ohms) and the pair for the heating element (12 Ohms). You won't fry anything because it's not powered when disconnected.
>>2031121
Measure resistance only when a device is off, not when powered.

weller wes51 retard mobile posting with a pic, the multi is how I have it set up to measure ohms. With one prong on the heating element and one on a pin, I was only getting resistance readings from the bottom two pins, and the top one farthest to the right. The readings were just bouncing all over the map, also.

>>2031129
Ok, so there's a near 0 chance I friend my station by poking to the tool receptacle pin-holes whether it was off or on, no matter the multimeter setting?

>>2031132
forget pic, sorry for shitting the thread up, don't have anyone else to ask

>>2031139
Put the black probe in the middle socket (COM).

>>2031148
thanks, i was trying both but still not getting values close to 22 or 12 on any of the pins when measuring from both pin to heating element and pin to pencil tip.

^ it's really hard to get the values to stabilize, is that normal? i think it's because im moving slightly, but they're jumping all over from like 1-130

>>2031160
Ok, switch the meter to 200 VAC and probe the base pins for 24V.
>>2031161
Yes, if you don't hold still the reading will fluctuate.

>>2031161
Also, ensure that your probes and the pins are clean, and apply even pressure.

>>2031170
>Ok, switch the meter to 200 VAC and probe the base pins for 24V.
finally some readings that make sense: with the ground in receptacle 2, I was reading 26.5V from receptacles 3 and 4. this was with the black prong still in COM.

im going to see if i can get a more stable reading on the pencil, I have two pencils, and both aren't working.

>>2031171
is there a significant difference between measuring the resistance to the pencil tip versus the heating element?

>>2031178
The pencil has two components: the temp sensor (22 Ohms) and the heater (12 Ohms). One pair of pins per component + ground (ignore the ground pin).

>>2031131
ok, initially i thought you meant measure the resistance with a prong on a pin, and another prong at the tip, or at the element inside the tip, of the pencil.

Measuring between pins I get either roughly 12 ohms or 0.5-1.5 ohms, which falls inline with what the manual states
>https://www.scu.edu/media/school-of-engineering/pdfs/maker-lab-resources/Manual_WellerSolderStation.pdf

no pair on the tool was 22 ohms, is that the problem?

>>2031216
>no pair on the tool was 22 ohms, is that the problem?
No, you're right. The sensor reads good and so does the heating element. The problem lies elsewhere.

and on the other one, I don't think it has continuity on a couple of the pins. i can't get a reading of 12ohms at all on that one, only 0.5-1.5.

so is the solution to buy more of these pencils? have i confirmed them as broken?

>>2025416
Anon, thank you for the links. Loved pop mech before it became woke meque.

>>2031223
>i can't get a reading of 12ohms at all on that one, only 0.5-1.5
That means the heating element is bad.
>so is the solution to buy more of these pencils?
It seems like it. You've got good 24V from the base station.

>>2031220
hmmm... To reiterate, I have two pencils. I didn't realize when I first posted that they both work for this model. I have a WES51 and a WSD81, but I don't have any pencils for the WSD81 and don't know if it works.

Anyway, one of the pencils reads good apparently, and said pencil actually gets warm, but not hot enough to melt solder. The indicator light for it (which implies it's heating) only goes off sporadically.

The other pencil was working fine until my nooby self had it cranked on max trying to fix a wallwart power supply to a DC motor, when it suddenly stopped melting solder. I've been keeping both machines outside, where I work (covered area so no rain and I keep them in moisture proof totes), and it's been below freezing lately, not sure if those factors may contribute, but that's the backstory and where I'm at right now.

On a side note, I think I'm gonna exchange my WSD81 for a WES51/WESD51, so my pencils are compatible between the two. Bad idea? I couldn't find much info about the WSD81.

>>2031229
That makes sense, because I probably burnt it out or something when running it max, and when that pencil is plugged the indicator light on my station is always on, implying it's heating up.

If this isn't something that's (easily) repairable, I'll just order a new one right now. Thanks for the help :-)

now to figure what's up with the buggy pencil #2

>>2031233
The WSD81 is 80 Watts and the WES51 is 50 Watts. Only use the 50 Watt iron with the 50 Watt station.

>>2031020
>drop manual link:

dunno what your problem is, dude. these are very clear instructions: (1) test resistance of heating element, (2) test resistance of the temp sensor. if you dont know how to measure resistance, there's 1000 Youtube videos on the subject.

the only possible tricky bit is the pinout of the connector. it's not clear whether the drawing is for the male or female. so you may have to flip the image left-for-right.

>>2025434
0805 is a nice case size that is still easy

>>2031020
>drop manual link:

dunno what your problem is, dude. these are very clear instructions: (1) test resistance of heating element, (2) test resistance of the temp sensor. if you dont know how to measure resistance, there's 1000 Youtube videos on the subject.

the only possible tricky bit is the pinout of the connector. it's not clear whether the drawing is for the male or female. so you may have to flip the image left-for-right.

>>2025434
0402 might be easier than you think, once you unlearn the advice not to load the solder onto the tip
at any size, good tweezers, a stable pc board holder, and paste flux make SMT comfy and normal

>>2031243
I don't have any irons for the WSD81, thankfully, or I may have made that mistake. What's the use case for an iron of 80w over 50? Asking because I'm thinking about selling it in exchange for a back up WES51, so if it shit breaks down again I can still work on projects.

>>2031266
Where I work we have good microscopes, vices designed for holding PCBs, and really good JBC soldering irons with all sorts of tips for SMT work plus loads of flux and anything else you could want and even with all that I still cannot solder 0402 packages.

If I am actually lucky enough to get a component onto the board it's almost certain it is not flush to the board. Usually one side is elevated on a mountain of solder and the other side is angled down into the solder. I have the same issue on 0805 and 1206 but at least with them I can usually reflow it and clean it up a bit. With 0402 I usually just end up destroying the component.

>>2031270
The higher output irons are better for boards with heavier ground planes and thicker wires/junctions, and lead-free soldering. They output more heat continuously so when you touch the iron to the surface it doesn't drop the tip temperature as quickly (better thermal recovery). Think of the heat generated by a 40W light bulb vs. a 100W bulb, and the time it takes for each to reach maximum temperature.

>>2031279
that's sort of what I was anticipating. maybe I will keep the other one around for when higher wattage is more suitable.

>>2031282
I'd keep it for the sake of versatility. Just make sure to match the proper iron to the station. Check the PDF for more:
http://www.farnell.com/datasheets/19755.pdf

Anyone know what these connectors are? I need the male part of them :/

>>2031287
Cool, I think I'll order a pen for that too then while I'm at it. Still not sure what was up with that one pen for the wes51 not heating up adequately despite the ohms checking out so hopefully the replacement pens work

>>2031323
Something like locking IDC audio headers.
>>2031324
Try bringing it inside for a test. Just for shits and giggles.

>>2031240
The pencil should have been designed to be disassembled to have a replacement element put in, so I’d scour any user/service manuals for such a feature.

Irons with cartridge tips are better though. Even if it’s just a backup, something like a T12 chinky station is reliable enough for me and also easy to fix if something goes wrong. And it’s $50. If you want something from a name-brand with a cartridge tip, not sure what the price would be but I’d guess $150 at least.

>>2031275
Sounds like the one thing they don’t give you is good tweezers.

>>2031323
Molex SL series
https://www.digikey.com/en/products/detail/molex/1719710004/9564229

>>2031348
thanks so much anon!

>>2031337
No, I had good tweezers as well.

I'd tin both pads and melt the solder on each and let the component just slip into place but it was never flush. Eventually I figured out tinning one at a time might be better but it seemed I could still never get parts to sit straight.

>>2031363
Use solder paste so you can squish it down cold.

>>2031363
>tin both pads
Not a good idea, impossible to get the component level, at least not with an iron. With hot air it's a different story. I've barely soldered SMDs before and found that you only want to tin one pad, if any at all. The technique I was using was to tin one pad, put the resistor flat on the other pad, and cover the surroundings with flux. Then I'd reflow the soldered pad, and push the component into it with tweezers. Some remelting and pushing the component is somewhat often required, after which soldering the other joint was trivial. Holding it in place and feeding solder into the joint while heating both the component and the pad up with a soldering iron is even better, but it requires having three hands. If I had a good PCB clamp with some sort of pogo-pin on a gooseneck (or maybe a lockable dial-indicator stand) for holding SMDs down, that would work.

If I had solder paste like >>2031369 suggests, it would be even more trivial.

>>2031337
>The pencil should have been designed to be disassembled to have a replacement element put in, so I’d scour any user/service manuals for such a feature.
yeah, issue was it that it was only $10 more for a new pencil entirely, and dissembling the existing pencil didn't look very clean. I'll keep the cartridge tip idea in mind, but I believe these pencils should last awhile.

i'll double check prices for elements, though. admittedly i just did a cursory search.
If anyone might have input left, I was wondering if keeping my stations outdoors, in below freezing temps, may have contributed to some of the problems I've experienced? They are kept covered in totes, but not cartridge tipsreally buffered against temps.

>>2031399
>it that it was only $10 more for a new pencil entirely
Sounds like a better idea then, I agree.
>I believe these pencils should last awhile
They should, so long as you run them below 380C.
>outdoors, in below freezing temps
Being outdoors may have exposed it to internal condensation, which could possibly cause an issue. Also thermal shock in general is what would break the heating element. Lots of hot-cold cycles is a problem even for structural metals, to say nothing of whatever alloy they use for the resistive element. Once left a drill bit in a chuck, came back after a cold night and it was shattered, due to the differential thermal expansion coefficients between the two different metals.
Note that heating the tip up to 400C and back down again would likely be far more significant than the freezing outside temperatures.

>>2031418
>Being outdoors may have exposed it to internal condensation, which could possibly cause an issue. Also thermal shock in general is what would break the heating element. Lots of hot-cold cycles is a problem even for structural metals, to say nothing of whatever alloy they use for the resistive element.
i worried about this too given the sensitivity of the instruments, I'll start keeping them inside. space is pretty limited as I live in a small shack but i can make it happen. i use them outside in a covered area, so it's nice to keep them where i use them. but thanks for the advice, I'd really like this new pencil to last at least a couple years.

>Note that heating the tip up to 400C and back down again would likely be far more significant than the freezing outside temperatures
good point.. but doesn't this undermine the idea of keeping the devices indoors? the elements and sensors will always be subjected to crazy temp swings, so i guess an extra few degrees below freezing temp isn't much really compared to 750F.

>>2031335
I actually thought the same and had it indoors for like two hours today, no magic unfortunately

>>2031382
>pogo pin on a gooseneck

>>2031487
That might work too, at least for SOICs. Similar concept to one of those weighted welding clamps with a spike on the end. But you need a really small point for 0603s or 0402s, and the force should be as vertical as possible. In your case, I don't think there's any horizontal force, but the whole assembly gets in the way a bit and is easily bumped. I think it would be ideal to have one of those dial-indicator articulated arms that you move into place and lock for that reason. A pogo pin could provide force in such a situation, but so could a needle with a weight on it.

>>2031505
probably want to use really loose gooseneck

>>2031511
Problem with goosenecks is they have backlash. Hold them into place, and when you let go they relax into a more stable position by a mm or two. A dial indicator mount would be great. But I'm also thinking, if you had your PCB in one of those horizontal frame vices, you could have an X-Y or pantorouter-style articulated arm holding a vertical pin that you'd lock with a wingnut or cam or whatever. Shouldn't be any backlash in that, though it might be a pain to get your iron into place. Once I've started earning and have a place with enough room for proper workholding, I hope to try that sort of thing.

When you have an IC with several "units" such as a quad opamp or an octo shimitt trigger. is there a way track of which IC has which units assigned besides manual checking?

>>2031631

question makes no sense.
are you asking about pinouts, like you find in the datasheet?

>>2031633
I have a design with 16 opamps and more comparators. how do I keep track of which one is in which (physical) IC? I have been doing it manually but maybe there is an easier way I am not aware (because it takes some time to go through each subsheet when you have several). Having some of them on the same chip would make layout much much easier.

>>2031639
The individual opamps within one package are usually labeled with a letter starting with "A"
So a quad opamp IC would have an "A" "B" "C" and "D".
I've also seen numbers instead of letters on occasion.

>>2031639

still failing to see what problem you're trying to solve. every circuit starts with a schematic. you start at the top left and label each op-amp U1A, U1B, U1C, U1D along the path of the signal. then move up to U2A, etc, when you get to the fifth op-amp. using these designations, you then add the appropriate pin numbers to each stage.
since your circuit is logically laid out in sequence (right?), this sorta makes sure that the layout will be clean, without traces flying all over the PCB.
but if that's not the case for one or two op-amp stages, then you can always flip 2 stages as they are all identical.

Fuck
is there a single easy way to make a HF 90degree phase shift network with:
>less than -2dB insertion loss
>wide bandwidth (covering 1MHz or more
>low component counts and no ICs
or am I just gonna have to cut 1/4 wave's worth of wire?

Some PCBs have nice smooth edges, how do they do it?
Just sanding? Burnishing?

>>2031665

you can specify cuts and rounded edges in the gerber file if your software has the functionality.

>>2031660
>since your circuit is logically laid out in sequence
hehehe back to the drawing board

I'm trying to come up with a way to digitally control parameters on voltage controlled (oscillators,filters,amplifiers)
That means I'm going to need a bunch of DACs.
What first came to mind was that I could have a master-slave setup, where my master mcu (an stm32) would receive parameter data, and then shuttle that down an SPI bus connected to a bunch of attiny10s, which have two PWM outputs each.
The specific attiny the SPI packet was addressed to would receive the command, and set its duty cycle in turn.

Alternatively, last night it came to mind that that might be cumbersome and would be better suited to an FPGA. Of the two options, which would be the better case?

>>2031672
single DAC and a multiplexer. there are 4xxx series ICs that do that. (analog switches)

>>2031666
>satanic trips
what about ones I already own?

>>2031680
they use a milling machine. you might be able to rig up a jig similar to a router table to get a similar effect, or sandpaper backed with a flat block if you're really desperate

>>2031672
>>2031675 has the correct answer
pic related from a 40-year-old design using CD4051. you'll want the speed of a real DAC rather than PWM

>>2031680
couple minutes with your file I guess

>>2031662
>90degree phase shift
>wide bandwidth
Pick one. The greater the phase shift, the more different the result will be at different frequencies, hence the more distortion you get at high bandwidths. I'd advise starting with a couple of cascaded filters, and do the mathematics to see what the values would need to be to get 90 degrees at two different frequencies. Then look at the resulting transfer function to see what it's like in between. I doubt you'll get anything decent looking with less than 4 cascaded filters. If you're fine with something approximating a true delay, then 1/4 wavelength of coax would be a reasonable way of going, so long as it's closer to the 30MHz end.

Here's a janky method: get a dual-gang tuning capacitor, wire both ends up to identical filters. Use one to phase-shift the input signal, use the other as the delay element of an oscillator of a PLL. Connect a DC motor to the PLL's output such that it controls the tuning capacitor. Could also do it with two matched varactor diodes, which is definitely more sensible, but also requires more signal conditioning circuitry.

i accidentally gave a fan lots of volts and that made it sad. it'll be cheaper to just replace ZD1 and U1 than to buy a new fan. sadly I can't identify the chip (the dremel scar doesn't help). in fact I can't find a single 8-SO fan controller on digikey with an integrated hall sensor. anyone mind giving it a shot?

>>2031891
ZD1 is a zener diode.
U1 is https://www.melexis.com/en/product/mlx90297/12v-pwm-single-coil-fan-driver-configurable-speed-curve

>>2031935
thanks anon, MLX90287's pinout lines up with what's on the board too. shame mouser is the only one that seems to carry it.

>>2031964
nevermind digikey just didn't want to show me their stock at first

>>2031966
there is always aliexpress

>>2031969
Not for obscure parts. If a part has more listings than total purchases, it’s probably not a good price.

Is there any book anywhere that describes common electrical design best practices for automotive systems? I'm familiar with many circuits that I see in boards and know they are general best practice but I'd like to more deeply understand why. Picture relevant, so I know the two capacitors in series on the VBATT are supposed to play a part in bulk input capacitance and I think transient suppression but I'm not 100% certain why.

>>2032065
prolly to get cheaper caps at lower voltage rating

>>2031680
>what about ones I already own?

when had no tools at all, i'd use the sidewalk to sand down corners on my PCB's. then i bought a $5 sanding disk for my drill, which made it much easier and faster, but awkward to hold. then i bought a detailing sander which i tie-wrapped to a work table: heaven to use, but too much work setting up. then i taped a sanding disk to a dead hard disk drive, and it is perfection.

Ohmbros, is there an electrical circuits analysis textbook you'd recommend? I don't really like the one by Alexander and Sadiku...

Ideal diode question:
Can someone tell me why for the last step of trying to calculate for V, they add in -3? I can get the 1/3 * 6 part because that's just V = IR but I don't get why they decided to add in -3. This is for part (a), I tried to erase part (b) out. The assumption for this current one is D1 = off, D2=on because before, it was assumed that both were on but D1 was in reverse bias so the assumption was wrong as written in the picture.

>>2032232
Haven’t read any of it, but considering the units are V and there’s a -3V source on the end of the 6R resistor, could that be it? Don’t know why they have that negative voltage reference though. Never seen that circuit before, so I’ll take a better look at it once I’m back home.

>>2032233
I honestly have no idea, I've never really dealt with -3V source either, shit is confusing. Here is someone else's work in the possibility that it might help, I still don't get why they're adding/subtracting

>>2032232
the equation (1/3 * 6) gives the voltage across the 6k resistor with 1/3mA current, i.e. assuming the lower leg is at 0V. but it is actually at -3V. therefore add the voltage offset at the lower leg of the resistor with respect to ground to get the voltage at the upper leg wrt ground

>>2032268
> i.e. assuming the lower leg is at 0V. but it is actually at -3V.
Oooooooooooooooooooooh, thank you, that's actually super cool, my professors had only ever done 0V so I never knew how to deal with stuff like this where the lower leg is not 0V. Thank you!!

Quick question, why is it that when I detach the load resistor and the diode to do the Thevenin conversion, the two 10k resistors are in parallel rather than series? I suck really bad at Thevenin questions because of this particular issue that I always come across.

>>2032285

step 3 of the process = Open current sources and short voltage sources.

when you do that, 5 volts and ground become the same, so they're in parallel since the 2 ends of the resistors are tied to that same place.

According to the schematic, a 1V dropout LDO with a 3.3V output is enough for a 4.2V battery, did they even test this stuff?

>>2032289
Like this? Sorry for the bad drawing, I can't really write that well with a mouse...

>>2032298
many LDOs are PMOS, where that 1V dropout only applies at rated current.

>>2032299

yeah, but add a wire shorting 0V and 5V to make the illusion come to life.

>>2032302
Uhhh, like this?

>>2032304

when you short the battery then things which were not in parallel, become so, and this lets you calc Rth.

>>2032307
oooooooooooooooooooooooooooooooooooh, that's what you meant! sorry for taking so long to get it, and thank you for taking the time to help me with this... I'm going to need to mega review my basic circuit analysis stuff again after this

How many working engineers do we have here

>>2032301
unfortunately that wasn't the case.
After further observation, it seems there is a 5V input considered in the schematic which is set with a jumper, which is stupid considering it would not work properly with a battery.
I went with a 100mV dropout regulator at 3V, which is close enough to 3.3V anyways.

>>2032323
obviously not enough

Anyone got any advice for making an ATX power supply into my first bench supply? I scored a nice barely-used 500W Antec but I've also got some 200-300W cheapies lying around because people just kinda give me their old computers. Alternately I've got a modular 500W iCute that's been hanging around for a while, but I've heard they falsify their power ratings.

I was thinking I'd just get some of those cheap Geekcreit digital V/A meters and put them on one each of the 3.3, 5 and 12 volt lines, but that'd leave a spare 12v (the spare 5 would be used for internal stuff). I was thinking I could get a boost converter and replace the little adjustment pots with lines to front-panel pots so I could get one output that was adjustable from about 13v upwards. Opinions?

>>2032342
>>2032342
>>2032342
NEW THREAD

>>2032335

there's a million how-to's on the interbutts. but it's a really really really bad idea given that you'll be sending 30A into your breadboard every time you plug in a transistor backwards, causing fire, smoke, and exploding transistors parts embedding themselves in your eyeballs.

>>2032361
What about some kind of current limiter, polyfuse, etc?

>>2032369
>What about some kind of current limiter, polyfuse, etc?

transistors and chips die in microseconds. fuses in milliseconds or seconds. you'd need to add instantaneous current limiting, as seen on every decent $75 chinese bench power supply.

>>2032376
Noted, thank you anon.

>>2032369
Sure, but they'd have a fixed current, you'd need to swap to different ones depending on what kind of work you're doing. Instead I'd put a MOSFET on the output, with a current measuring circuit. Feed that current value to an op-amp to amplify it, and feed that into a comparator alongside a variable (i.e. using a potentiometer) reference. Then have the comparator trigger an SR latch, which turns off the MOSFET and turns on a warning LED to say that your circuit tripped. Then have a button to reset the latch once you fixed your circuit. Also probably have a capacitive circuit that automatically unlatches your latch on turn-on, though maybe some smart resistor work will do that for you.
Do it well, and you can get it to trigger faster than most parts will die.

>>2032376
>$75
Maybe if you want linear CC regulation, but if it's just an overcurrent cutoff there are cheaper options. For CC regulation just get an LM2596 module with 2-3 pots, in most cases you won't care about output ripple when you're doing CC.

also there's a new thread

>>2032378
Also noted. I'll look into that.

You guys are super helpful.

>>2031662
wide bandwidth is a useless figure without knowing your center operating frequency.
1mhz bandwidth at 5mhz center frequency is huge but at 30ghz it isn't.

>>2033178
>>2031662
>>2031866
also depending on your operating frequency you could do it digitally in DSP and just have two RF dac's output two waveforms 90deg shifted.

I love you fag and none of you will see it :^)