/diy/ - Do It Yourself

I hate building complicated pcbs.
Designing them is fun, but the worst feeling ever is assembling that shit and realizing you made a mistake somewhere, in design or have a faulty component, and will now have to waste hours with a scope figuring out where the fuck up is, and then you realize it's you this whole time

>>1874924
It's ok son, you can go back to your tard cage and play with computer science stuff and programming shit later

Any recommendations for a book /after/ The Art of Electronics? I'm particularly interested in more exposure to a breadth of different circuit types.

I still occasionally run into a circuit in the wild, and feel pretty lost about how it operates.

>>1874932
these books are to be used with a mix of "reading through" and consulting the topics you need. The only way to feel confortable is to gain experience

How does one usually determine if a component is compatible?

I asked for a RBV-602 and the guy at the shop handed me a KBPC610 no questions asked.

Also 602 fits on a heatsink but I don't see how the 610 package would benefit from cooling.

>>1874978
>What are my parameters?
>A,B and C
>Component 1 satisfied those
>Component 2 also does that?
>yes ->use component 2
>no ->use component 1
It seems to be the same component from a diferent maker. Man if you are the guy who asked about an opamp another thread, you really need to learn to read datasheets. Just sayan

>>1874987
Nah, I can more or less compare opAmps, I watched the digikey video about them.
You make it seem pretty straight forward, I can draw parallels from both components but I still lack the confidence to admit their compatible.
Thanks.

>>1874913
I have no background in musical circuits at all, just what I've picked up from the occasional youtube video and from some wikipedia articles. Learnt about fourier transforms and all that at uni. But I learnt what additive synthesis is from this general. No clue if doing it with triangle waves is a good idea, but that's just what I'm using.

hello all, looking for a soldering station. i remember seeing one that everyone agreed was quality, but i cant seem to find it. i mostly use my old iron for putting on pl-259s, so looking for something like that, but im also looking to replace small parts inside radios.

thanks in advance

>>1874994
It's not straightfoward, because defining the parameters is hard. But the sooner you see that only the parameters you define matter the better.
source? Almost lost my sanity choosing materials with a joint project with the design students.

As for your question, I'm pretty sure they are the same device but with a diferent name and manufacturer.

You could try that rn, by listing what made you choose RBV600 instead of the other 10000000 silicon bridge rectifiers.

>laptop battery only charges to 20% now
I can get a replacement online for $15-$50.
Is it worth re-celling?
What is the chance I'll have to fuck with the BMS if I re-cell it?

>>1875018
buy a replacement and use the shitty cells on projects? That's what I usually do.

>>1875022
Did you mean:
>buy a replacement and use the cells from the worn-out battery in projects
Or
>re-cell a shitty replacement battery with decent cells and use the shitty cells it came with in projects
?

>>1875035
I buy a new battery, use that in the laptop. Crack open the other one, use the individual cells for shit and harvest thermistors and inductors.

Does anyone knows a good free gerber viewer that can also read drill files?

I use gerbv, which is good but it doesn't include drill file…

Extra point if you can measure distance directly in the viewer.

Project Question, hopefully this is the right thread.

I basically want to go low output (dim) LED - battery - light sensor. I'd like to use a 1.5v AA (or AA in series) and get the maximum time out of them.

Is it possible to power LED with just 1.5v? Are there specific LED I should look for, or can I just grab a few LED from one of those car lighting strips (all my LED projects have been 12v)? Do I maybe want just a LED diode instead?

Do light sensors draw any power? Is there any benefit to just eliminating the light sensor?

Last what kind of time can I expect from such a system? Could improvements in lighting hours be made by switching to 9v battery over 2-3 AA?

>>1875003
>but i cant seem to find it
did you look under your bed?

>>1875097
gerbv can read drill files

>>1875106
It doesn't for the .drl files generated with Orcad PCB designer.

>>1875116
Works with KiCad generated ones which are in Excellon format.
Does Orcad let you choose different formats?

>>1875121
I believe yes, and it's already in that format. I will check either later or tomorrow and let you know what is the error given by gerbv ?

>>1875099
all LEDs are diodes. your LED strip probably isn't suitable for salvaging. pic related are pieces of shit that you can buy cheap off ebay. they're a fine starting point. there are a large variety of "COB" LEDs that are nicer and more efficient if you want to dig into that.

>Is it possible to power LED with just 1.5v?
yes. how complicated this gets depends on how long you really need the battery to last. at minimum you need a boost converter to increase the voltage. if you were designing a product you'd use a proper chip like the ZXSC380 to boost the input and also drive the LED with a constant current, which is much more efficient than those LED strips which just use limiting resisors. since you're not experienced you'd instead look for a cheap module that already does this on ebay or aliexpress. i don't see any of those modules so unfortunately you may be fucked, which is unfortunate because this should really be a simple issue. i'd rule out a single AA in your case.

>Do light sensors draw any power?
there are different types of light sensors. the simplest and most available is a photoresistor. these draw power, but negligibly so if you know what you're doing. you'll need additional circuitry and some math in any case. the dumbest possible circuit would add an NPN and an additional resistor. there's a lot to be gained in battery life if you think you can handle more complexity. even the dumb solution should increase runtime though.

>Last what kind of time can I expect from such a system?
unfortunately i didn't think about this before writing the post but you shouldn't expect more than a few days at most of runtime out of this setup at the absolute maximum no matter what sort of alkaline you use. one alternative is to trigger the leds off a PIR motion sensor, but then you'd save money if you bought a wall module.

This is a stupid question since I really don't have much electronics experience or knowledge (mostly compsci and programming), but I haven't found a clear answer by searching. I'm planning a little digital instrument based on a microcontroller. For audio output it's going to be connected to an ADC chip via i2s. For connecting the two, I've seen people recommend a direct pin connection rather than wires, with the recommendation to keep the wires short if you do use wires, because of the high speed of data transfer (44.1kHz 16-bit stereo audio).

Why is it that a relatively short connection (<1 meter) is needed for high data transmission speeds? My intuition is that since the signal is digital, noise/resistance from a longer cable shouldn't really matter, so long as the voltage is high enough to not significantly be affected by the impedance of the wire. I saw some people mention it having to do with cable capacitance, but I couldn't find more of an explanation than that, just people affirming that high speed data with long connections doesn't work.

>>1875134
>Why is it that a relatively short connection (<1 meter) is needed for high data transmission speeds?
Look up crosstalk, EMI, and a solution: differential signaling

>>1875138
Given that there are only 3 wires carrying data, all three are digital 3.3V, and I minimize sources of noise nearby, shouldn't I be able to go at least several meters with no problems? What is it about high speeds that makes longer connections more susceptible to error?

>>1875125
Good info, thanks. Unfortunately the wall plug style won't work, what I've got is an ottoman with legs in my kids room which is dark grey so the idea was to tuck the light up under it so it provides a gentle glow at night and I can stop breaking my fucking toe off in the dark. So cording the ottoman or a heavy battery aren't viable. I had thought of motion sensor, since the goal is to prevent me from not seeing it while moving around in the dark but there don't seem to be any good battery options there either.

Thanks for the info, back to the drawing board I guess.

>>1875144
bigger loop area. and I'm assuming you're cables aren't shielded

>>1875146
you can buy PIR sensors as modules. they won't need any more additional circuitry than the light sensor would but you need to choose the right module because some of them waste too much power to be suitable for battery operation. actually mounting one or more under an ottoman seems impractical though.

>>1875144
Can you imagine why a UTP (unshielded twisted pair) LAN cable has no problem with 100 MHz or 1Gb up to 100 m?

>>1875144
top - low clock rate, shielded cable
bottom - high clock rate, unshielded cable

Wires are not perfect - they have resistance, parasitic inductance and capacitance.
The higher frequency you go, the longer wire you use - the more shit you have to deal with

It gets even more interesting when you get into very high frequencies where data takes longer to go through wires than you generate it

For your application - just fucking do it and test it, will be a good learning opportunity if there are problems

>>1875146
>there don't seem to be any good battery options there either.

i have a similar light to pic-related in my bathroom. 4 new AAA cells last a couple of months. if i mod them to use 2 D cells, they last like a year. but i dont use new cells coz these guys work down to 0.8V per cell, so i just use cells that i would otherwise throw away. bought a bunch on clearance at Zellers a few years ago, for like $5.

>>1875146
>there don't seem to be any good battery options there either.

i have a similar light to pic-related in my bathroom. 4 new AAA cells last a couple of months. if i mod them to use 2 D cells, they last like a year. but i dont use new cells coz these guys work down to 0.8V per cell, so i just use cells that i would otherwise throw away. bought a bunch on clearance at Zellers a few years ago, for like $5.

>>1875150
mounting is actually simple as I can just design and 3D print a little bracket and position as needed.

PIR underbed lighting is actually a pretty common thing but as I look could I get two things in layman's terms?
1. what does the MOFSET do?
2. what do the resistors actually do? all the circuits I'm seeing have them with some variation in size. I know what a resistor *does* I just don't understand why you'd drop a 470ohm in a DC battery-light circuit, not vs a 1000ohm but vs not having a resistor at all.

>>1875157
ah nice, I'm going to grab a pair of these and see if I can expand the components to suit my needs- i.e. extend and mount the PIR on the side(s). good lead, thanks!

>>1875152
If wire capacitance stops i2s from being viable past a meter or two, then no, I can't really imagine why. Using twisted pairs helps reduces noise in LAN cables, but if LAN and i2s use voltages that are similarly low, then why can you not achieve similar distances with i2c?

>>1875150
these are trash, i have one and it keep triggering by itself

>>1875198
Meant i2s at the end, not i2c.

>>1875097
I use viewmate.

I know this is a shitty question seeing as I can't supply a part number or photos or anything but I'm not really looking for a definitive answer, just some ideas of what I'm looking at.
So, I was looking at a sine power inverter earlier (I don't have a huge amount of experience with inverters).
Powering the inverter and applying a load seems to work fine but intermittently, the inverter pulls loads of current momentarily and pulls the voltage of the DC supply down. You can also hear the transformer kind of shutting down.
I checked all the mosfets with a diode tester and they seem good and aren't getting hot.
There is one large flat component (or more likely circuit board) soldered onto the main PCB that I'm not sure about.
It's a single inline package and encased in resin, any idea what this might be?
I googled the code on the package but just get a load of Chinkish sites with no info on them

>>1874995
>No clue if doing it with triangle waves is a good idea
I dunno either.
I suspect it won't be quite as flexible as a proper sine wave additive synth, due to the fact that you'll always have odd harmonics.
It'll probably still make interesting sounds though, and triangle waves are much easier to generate than pure sine waves.

>>1875158
looks like you found a solution but to answer generally, a MOSFET is an electronic switch. it uses a weak signal, like from a PIR sensor, to turn power on and off to something like a LED. it's similar to a relay in that regard (and in that regard only).

>what do the resistors actually do?
batteries provide a voltage. typical circuits take that voltage, and draw only whatever current they need to operate. LEDs are inconvenient because past a voltage threshold they'll try to draw a huge current that will near-instantly overheat and destroy them. adding a resistor in series with the LED will prevent that by restricting the current to a safe value. you'd use 470 ohms as opposed to 1000 ohms because the latter would restrict the current too much and the LED would be dim. that's a specific example. resistors are used for countless other things, and they're also not the only thing that offers resistance.

okay here is USB charging circuit for a little project im working on. do i leave the the usb data lines hanging where are the recommended place to put fuses or should i not put any? any other critics are good. first time ever doing something like this.

>>1875298
The main driver behind using triangle waves was that it would make riding comparators along the wave for phase discrimination much easier, though the same thing could be done for a sine just with a lot less linearity.

>>1875138
>>1875155
It's not just noise, it's driving the capacitance of the wires. A cable is a LPF. You can see that on the waveform

>>1875341
Depends on your power supply. Pull-up resistors on the data lines are for power negotiation with smart chargers. Refer to the USB x.x spec sheet for more info.

>>1875386
yeah for usb 2.0 i just need a 200 ohm resistor across datalines do you think i should put fuses somewhere?

Streaming my first crack at Arduino if anyone's interested:
https://dlive.tv/P3ntiuM

>>1875387
Yes, between the USB connector and C1.

just bought one of these puppies, have never soldered before. what else do I need to order so I can started w/ building modular synth via this playlist: https://www.youtube.com/watch?v=5DKU1m5_b_8&t=1081s

>>1875414
also, i have no idea how electricity works

Brainlet noob here.
Is pic related possible + easily constructible?
(First two pics show how it should behave depending on whether i1 is driven or floating, third pic shows how it might be used as storage)

>>1875419
wrong picture

>>1875420
what is "it"? a switch? a logic gate? an op amp? a drawing of black boxes?
What are you building, describe it. What is it made out of, what is is supposed to do.

>>1875436
it's supposed to be an active digital component with two data input lines and one data output line.
the assumption is that the lines are tri-state: they may be driven high, driven low, or floating / effectively disconnected ("z").
the desired behavior is that if i1 is driven (whether high or low) then i1 should exclusively drive o with no influence from i2, but if i1 is floating, then, and only then, i2 drives o instead.
in this sense, i1 is a "default" line, and i2 is a "fallback" line.
it could also be considered a lopsided form of binary multiplexing, whereby line i1 doubles as the selector line, and both possible driving voltages represent selecting i1, but i1 floating represents selecting i2.
idk what it would be made of etc, silicon i presume. i am stupid.
the motivation is that depending on whether it's possible / how complicated it is to implement, i ignorantly hypothesize it could be a more efficient implementation of computer storage than a flip-flop is.

>>1875419
>>1875420
draw out a truth table, with a row for every combination of tristate inputs

>>1875443
>>1875468
I think it's possible using only simple logic gates and resistors. Maybe 5 gates or so.

>>1875469
so do it then.
If your goal is to beat a flip-flop, you might, but it won't be better than contemporary memory architectures.

Evenin
I have a voicebox for my gas mask and it's not very loud. Any ideas? It's powered by a single 123 lithium battery.
My retained electrical/physics knowledge is roughly V=IR and never touch a sparkin wire.
Pic 1

>>1875523
Pic 2 guts

>>1875524
Pic 3 closer.

>>1875537
Pic 4 inside the mask itself. Not sure what I can take apart (or need to) in here without breaking something.

>>1875542
Strangely the unit would not work when taken apart. I put it back together and it worked right away. Loose wire perhaps.
Pic nil

>>1875543
14.85 ohms and 3.178V on the 2 wire coming out of the box vs 14.39 ohms if I put leads on the 123 battery (read 3.153V) itself.
There was 1.19 ohms when I put the leads on the masks female recepticles.
Not sure if any of that is relevant or not.

>>1875351
You can do that with sine waves just fine.
If you're doing digital synthesis, then compare against the phase accumulator.
It would even be easy for most analog sine wave VCOs, since they usually work by generating a triangle wave and then use diodes to shape it, so there's still a triangle wave you can compare against.

>>1875551
I have no clue what answer do you expect to get with this information. Is it supposed to be louder? was it louder before? have you considere that its just a cheap shit amp? dont expect anyone to do circuite reversing and or modding for you...

>>1875583
I expect nothing, want it to be louder.
I'd asked a few threads ago and was told to post pictures of the guts. I assume it's a small crowd here and he'd turn up eventually.

>>1875543
>Strangely the unit would not work when taken apart.
looks like reed switch in the top left, which means there is a magnet in the mask which triggers it, (no magnet means the switch opens and disconnects power) that is probably causing that

>>1875537
The L6/L7 inductors on the output would suggest it is a class D amplifier with H bridge output. However, the closest chip I can find information on is the NJM2073 which is a linear amplifier, so it could be some clone of that and the inductors are just to filter out some noise.
I can't see where the microphone comes in, but if you can locate the divider resistors you can increase the gain. You might get clipping if the input voltage is not enough.

>>1875472
>so do it then
No I'm not him, I'm the "draw a truth table" guy.

>>1875567
I'm sticking to analog, this is more of a curiosity than a practical circuit, an interesting mix of analog circuit blocks and digital logic ones. Though I'm going out of my way to ensure it doesn't rely on part tolerances or component matching, thanks to the wonders of negative feedback. For voltage AND frequency; all my VCOs will be part of a PLL.
Didn't know that's how sine VCOs work, do they use VCFs for the wave shaping, or do they just have a somewhat limited slew? I'm wanting my "synth" to cover at least two octaves, hopefully closer to 5. Using diy vactrols for the VCOs, unless there's a better but simple idea out there. I'd consider JFETs even with their pricing (I'd need less than 10 total), but my VCOs won't have any of their electrodes be grounded, making it difficult to drive them without at least one extra op-amp (I think).

>>1875607
>want it to be louder
A microphone close to a loudspeaker? There must be limits to the loop gain or it will develop a life of its own. In this case all you can do is sing louder.

Trying to get a parts list for this power supply. The whole process of correctly identifying and sourcing parts is a bit difficult for me so if you could look over my list and tell me it's alright I'd appreciate it.


>12V AC Power Supply
Already have one
>2x 470uF and 1x 1uF Electrolytic Capacitor (35V min),
https://www.ebay.com/itm/1uF-to-1000uF-25V-to-50V-Aluminum-Electrolytic-Capacitors/132414777336?hash=item1ed48927f8:g:n50AAOSwTuJYt8Sh
>1uF Capacitor (30V min)
>2x 1n4001 rectifier diodes
https://www.ebay.com/itm/50pcs-MIC-straight-plug-1N4001-IN4001-rectifier-diode-DO-41-1A-50V/203042952359?hash=item2f464d94a7:g:fxYAAOSw~edfBc3C
>4k7 ohm resistor
https://www.ebay.com/itm/100-x-4-7K-4K7-ohm-4700-Metal-film-1-Resistor-1-4W/182708306266?hash=item2a8a43815a:g:b6kAAOSwhvFZMuxq
>1 led


The tutorial I'm following is right here: https://youtu.be/5DKU1m5_b_8?t=1118, and in it the guy also mentions needing a '7812' and '7912', but I'm not able to infer what this is. Anyway, I'm gonna order this stuff tonight so hopefully it's right.

>>1875811
forgot pic

>>1875811
>>1875812
ok, it actually said in the description that the 7*12's are +/- voltage regulators, so I've got them here:
https://www.ebay.com/itm/10-50pcs-L78xx-Series-Positive-Voltage-Regulator-IC-5V-6V-8V9V-10V-12V-15V-24V/291959479814?hash=item43fa241e06:g:k2QAAOSwj85YQBIe
https://www.ebay.com/itm/Lot-of-10-MC7912CT-On-Semiconductor-Negative-Voltage-Regulator-12V-1A-TO-220-NOS/202101969069?epid=665825780&hash=item2f0e3750ad:g:61sAAOSwbF1Z-2dr

>>1875814
>>1875812
>>1875811
Seems ok man, just make sure to buy them from the same place so you don't spend a fortune in shipping these things. And from what I can tell they are pretty generic stuff, so if you have a brick and mortar store nearby they have it. And if the diodes you find are of a diferent name/model just make sure the voltage requirement is enough.

>>1875827
thanks, yeah there's a cool shop nearby, I think I'll just go there next time I'm going into town.

I want to put as many LEDs as I can in a PCB. How many watts per cm2 can a PCB handle without getting too hot? Google says 1 watt of convective dissipation per 15cm2 (for a 40ºC increase) which is only 7W for a 10x10 PCB.
Also, does choosing matte black vs white silkscreen does anything for power dissipated through radiation?

>>1875836
you're going to want to use aluminum for the PCB and heatsink it.

>>1875839
I planned to put the batteries on the back.

>>1875851
You can't fight physics, anon.

>>1875687
>Didn't know that's how sine VCOs work, do they use VCFs for the wave shaping, or do they just have a somewhat limited slew?
VCFs are one way to do it, although that's not quite as common since it requires the VCF to closely track the VCO, which complicates tuning the circuit.
Usually the triangle wave is instead passed through a diode clipper, using the exponential current-voltage relationship of the diode to approximate a sine wave.
Pic related is about the bare minimum required to make a triangle to sine converter using a diode clipper. Adding an op-amp and changing to the way the diodes are biased can greatly improve the performance.

>>1875839
Copper conducts better. Wtf are you talking about
>>1875836
double side PCB with a vias grid with a heat sink. You have to do the math yourself, just learn how to work with basics thermal circuits. (It's simple, you just need to find good data for thermal resistances)
>Also, does choosing matte black vs white silkscreen does anything for power dissipated through radiation?
Yes and no. It's too small diference to matter in that temperature range. You should ensure good airflow and good heat sink because convection and conduction are the dominant heat transfer in this case.

>>1875877
If you plan on putting the batteries on the back maybe you can diy some heat sink with scrap aluminum/copper pipes in a way that allows you to take the heat away and still keep the batteries. It'll be shit tho and the batts will get hot

>>1875877
>>1875880
You can find the heat resistances for vias pretty easily in the internet. I used that when I made a flyback for school and there was a specific temperature requirement.

anybody know where I could find a 10K lin or log pot with these terminals? Not having much luck on eBay.
Stereo could work too because I could just trim the excess terminal.

>>1875917
I think those are called linear potentiometers.

>>1875917
It's called a slide pot, and it could be linear or logarithmic.

>>1875940
meant for >>1875928

>>1875928

>>1875943
>linear
>it's actually curved
wheew

>>1875943

>>1875870
Oh, basically a passive log amp. Not as good if you need a constant output amplitude, but I'm not even in that position myself.
Since it's that simple, I may add the option to run this circuit as a sinusoidal additive synth, just by soldering different parts in place. I think it's rather simple to modify it to be a sawtooth additive synth too.

>>1875877
Copper is a better thermal conductor, yes. But if people used 1mm of solid copper on the back of your heat-sinking PCB instead of aluminium, they'd be significantly more expensive. Aluminium is a comparatively cheap metal, and it also has much less heat capacity, for what that's worth.

>>1875943
What are those even used for?
Seems super niche.

>>1875949
I though those just used a rotary switch.

>>1875961
>Oh, basically a passive log amp.
Basically. There's more advanced versions around.
>Not as good if you need a constant output amplitude
If you can get a constant amplitude triangle wave, then it's not hard to get a constant amplitude out of the sine wave converter.

>but I'm not even in that position myself.
How are you generating the triangle wave? It's fairly easy to make a constant amplitude triangle wave generator.

New to electronics. Is this the right truth table for a mosfet?
s = z, g = z: d = z
s = z, g = lo: d = z
s = z, g = 1: d = z
s = lo, g = z: d = z
s = lo, g = lo: d = z
s = lo, g = hi: d = lo
s = hi, g = z: d = z
s = hi, g = lo: d = z
s = hi, g = hi: d = hi
(b always = x because it's just a drain terminal for g)

Connecting two caps like +--+ or -++- to catch both pos and neg AC waves, is that retarded? What if connecting them in anti-series(?) sends a charge back to the source, would a diode be needed?

>>1876108
>Connecting two caps like +--+ or -++- to catch both pos and neg AC waves, is that retarded?

it's a standard technique to turn 2 polarized caps into 1 non-polarized one.

>What if connecting them in anti-series(?)

that's what you just did above.

>sends a charge back to the source

WTF you talkin' - it's just a cap, not Purolator.

>would a diode be needed?

for what!?

>>1876048
MOSFETs are not logic devices.

>>1876131

they are if you're autistic enough, like if you're the king of the autists on planet Autistica.

>>1876133
> Why isn't it powering my arduino? I'm feeding it 5V and it's turn-on voltage is 5V!

>>1876131
>>1876133
>>1876136
thx you two i fucked around on a circuit simulator and it is not as simple as i thought i feel even dumber now time 2 study

>>1875992
>then it's not hard to get a constant amplitude out of the sine wave converter
I guess so, but the tempco of the diodes could make it vary more than just the tolerance of the resistors.

Ok, I have to drive this MOSFET:
>IPP023N10
And not just one of them, three that are already in parallel. Story is, my existing gate driver IC (IRS21850) died somehow, so I need to bodge in a totem-pole circuit of some sort. It's going to use low-side-switching, so that's not an issue. The old driver was capable of putting out 4A, so I just need to match (or improve on) that. I hope the MOSFETs aren't broken either. I have a 12V rail (from a linear regulator) at my disposal, which should do the job going off the datasheet so long as I use an appropriate electrolytic cap, though there aren't any graphs of Rds over 10V Vgs. Gate charge is 54nC from the spec table, but a graph shows the gate charge getting up to 180nC above 10V.

I plan on modelling my circuit in spice just as an idiot-check before building it in the physical world, and in both cases I'll be using capacitors to represent the gates of the FETs. Do I just go "C = Q/V" and throw in 6nF or 18nF or whatever?

I'm also wondering the best way to get my square wave from a 5V digital IO to the 0-12V signal to feed into a totem-pole. A pullup resistor feels inferior.
I'm guessing standard small-signal transistors won't work for the output stage, not even 2N3904s, though there doesn't seem to be a "safe operating area" I/t graph.

>>1874995
>>1875687
How's THIS for a triggering waveform? No risk of overshooting my pots with this. But it's too fast for the slew-rate of most op-amps, so I'll need to leave it as-is and hope that it doesn't get pulled down too much by my comparator inputs. All 16 of them. Are there JFET comparators?

>>1876240
fg

>>1876240
On second thoughts, I think I'll try buffering it with a totem-pole.

>>1876207
The actual important characteristic is how fast you need to turn the FETs on and off, and that depends on how often you will be toggling it and how much current will be passing through around that time. Too slow and your FETs will blow.
>there aren't any graphs of Rds over 10V Vgs
more is better as long as you don't exceed the maximum rating
>Do I just go "C = Q/V" and throw in 6nF or 18nF or whatever?
Yes, that will give you an approximation that is OK sometimes. If you look at the gate charge curve you see that it isn't a straight line like a capacitor would be, but you can pick a point on the line you care about and do the figuring. The "shelf" on the curve is the danger zone, so best pick a point above the shelf (that will give you a higher capacitance value).
>I'm guessing standard small-signal transistors won't work for the output stage, not even 2N3904s
Well, a 2n3904 certainly won't put out 4 amps for very long. This all depends on what you're doing with it. For example, if it's a battery protection circuit where the FETs will be on 99.9% of the time but turn off and stay off under fault then a pull up with smallish transistor may work fine. If it is a motor driver that will be getting PWM and running many amps that will guaranteed blow your FETs in short order.

>>1876278
>Too slow and your FETs will blow
The actual switching speeds are sub-50Hz, for whatever reason. So even if I only pull 200mA in and out of the FET gates I shouldn't overheat them. But I'm going to shoot for the 4A peak that the old driver had, just to be safe.
>more is better as long as you don't exceed the maximum rating
There was a 15V zener diode on the circuit for limiting Vgs (which wasn't used since the driver was on the 12V supply, just for safety's sake), which I could reuse. Otherwise, the only other convenient power source is a 12V regulator, since the battery voltage is above the 20V maximum for this MOSFET. I figure there won't be much difference between the two, and using the zener might put more stress on my totem-pole.
>a 2n3904 certainly won't put out 4 amps for very long
If all goes according to plan, it's only two short spikes each cycle, each spike being <1µs wide. And with such a slow PWM frequency, that shouldn't be an issue for the BJTs heat-wise. The resistors would get hot before that would happen. Not sure about damaging the silicon itself though.
In reality, the resistors in series with the FET gates limit the current the driver needs to handle to 2.4A, and hence the current per FET to 800mA. That's using 20nF as my gate capacitance, lower if I use 12V instead of 15V.
>blow your FETs
With high gate-currents? I assumed the gate would be mostly proof against that, but maybe my assumptions are wrong. I'd hope that they're fine against 800mA at any duty-cycle.

The application is PWM for an existing DC motor speed controller, with prior damages.

>try to make a schmitt trigger with a comparator
>accidentally use negative feedback instead of positive feedback
>wonder why my circuit keeps oscillating at 100MHz

>>1876285
>>blow your FETs
>With high gate-currents?
No, with insufficient gate current. Power is dissipated during turn-on and turn-off, so a weaker gate driver means more heat is generated.

>>1876289
Oh that's what you meant. Due to the low frequency I'm pretty sure I'd be fine with up to as high as 100µs. Not that I intend on doing so, 10µs is the maximum I'm setting for myself, and I'll try to get as close to the <1µs that the original driver would have been able to push through the gates. I messed a bit about with driving capacitor loads with a BC547/BC557 totem-pole earlier this year. The transistors were obviously heating up too much and not switching as fast as they should, but that was driving them at quite a high PWM frequency, so I figure with something meant for >500mA continuous I'll be able to get >2.4A peak. Like the 2N2222A, and whatever transistor is close enough to be its complimentary. Unless there's some other factor I'm missing.

first schematic looking for tips/review

>>1876651
No current limiting on the LEDs, mo bypass caps for each IC, and the mismatch of resistor/pot symbols are all I can see.

>>1876651
Wait, are you grounding BT+ right under R6?
Also it’s just the mic without a bypass cap, which it should have since it’s current changes as a function of which leds are on.

>>1876668
LEDs have internal resistors for 5V, I have C8, C9 on the top to decouple U3 pin 7, 20 should I move it closer to U3? The other ICs were drawn from datasheet. For the mismatch I created a custom symbol for FSR but I can re-create it using international resistor for consistency

>>1876673
Oh, didn’t notice those up there. IIRC there should be symbols for both us and eu resistors and pots, doesn’t matter which as long as you’re consistent.

>>1876670
>Wait, are you grounding BT+ right under R6?
yeah I'll remove that thanks
>it’s just the mic without a bypass cap
not sure if you saw >>1876673 if you did I am confused where to add additional cap

Probably a simple question but I am merely a humble farmer.
I’m looking at an led array which can take a max current of 1400ma.
As I understand it I shouldn’t be running it at that because it kills longevity.
If that’s true I’d be better off buying a driver that puts out less than that right? What about 1050ma?

Hi guys,

I would like to ask for advice regarding multi-channel pots. I have to generate 3 channels of PWM signals for an RGB LED strip. I have 3 linear potmeters to set some opamp/comparator reference voltages, but I want to implement a fourth potentiometer inbetween the opamps and the previous 3 potis to further divide their voltages while keeping the proportion between the colors.
Think of setting R/G/B and Brightness individually.

I've ran into a brick wall though. I can't find any linear potentiometers that have at least three gangs AND don't cost 20-30 bucks to buy & ship to my country. I'm wondering if there are other solutions for this issue with a minimal component count and without breaking the bank. Plus I'd rather not be forced to add a 7805 and an ATTINY/ATMEGA for such a seemingly trivial task.

Thanks in advance!

>>1876735
Rotary encoder + 3 xc90 digital pots. GadgetReboot did a video on doing this.

>>1876735
do like this

>>1876735
jfets, op amps, OTA's

>>1876731
Depends on the LED. 1400mA is only 1.4A, and there are LED arrays that take many times that. For example, a 50W 12V COB LED will take 4.2A, and they're relativiely common IIRC. Using a couple of lower-current LEDs in parallel (probably with balancing resistors) would work too. So you happen to have a driver that outputs 1.4A? It might be possible to modify the driver to output less current, though likely not feasible for beginners.

>>1876771
This is actually a nice and elegant solution, provided the PWM generator doesn't feed current back through each pot.

I'd also like to see a design for a HSB-based LED dimmer, as opposed to RGB. I suspect it's possible with a bunch of summing amplifiers, though maybe it needs OTAs too.

Will this tip (1.6mm) be sufficient for soldering PCBs for guitar pedals? Or should I also get a tiny pointed conical tip?

>>1876930
1.6 should be plenty small for most work. I've done 0402s with larger.

>>1876651
what's with the crystal, why voltage divider?
I'm just curious

>>1876935
The crystal looks normal to me, no voltage divider near it at all.

>>1876937
noticed my mistake, not used to net names being on the end of trace

>>1876940
Oh that. Be nice if you could put them diagonal and/or have a small dot on the node it's attached to.

>tfw I need 8 SR latches with complimentary outputs
>that's 2 4043/4044s and 2 hex inverters since none have not(Q)
>or just 4 quad NOR/NAND gates
Guess I'll go for the second option, since the 4043/4044 are 16-pin while the 4001 and such are only 14 pin. Even if there's a 7400 solution out there, I don't plan on having a 5V rail and will need a full-scale output voltage from my logic gates anyway. While I'm here, I might as well get schmitt input NOR gates, since I won't be driving them with a totem-pole output, but rather a capacitively coupled rising edge.

Hey guys, RGB LED dimmer dude here from a few posts back.
Thank you everyone for the suggestions, I'll have a look into them! I was a little scared of digital pots/JFETs I admit, because I've never worked with them before lol

I also really like the master dimmer pot solution that was suggested here >>1876771 for its elegant simplicity.
It's actually similar to my original idea, and I don't know why I haven't thought about it sooner.

So thanks again everyone, I'm going to dedicate my day to fiddling around with this project now :)

>>1876974
I have like two years of electronical experience but usually a digital pot is the last thing I use. Usually there is always a simpler thing you can do ,with a bit of thinking, to achieve the same result.

>>1876974
I'd agree that that simple idea is by far the best solution. Usually I don't consider cascading pots like that without some sort of buffer (presumably related to the op-amp suggestion by that other anon) since the impedances will affect one another, but in this case since they're potentiometers the current going through each color pot remains constant regardless of how the color pot is varied. But note that the master pot will be somewhat nonlinear if you don't buffer it. Buffering it would be a pain since you'd want rail-to-rail action, but it's probably not a big deal, especially not if you make the master pot a 10k and the other three 100ks.

>>1876977
Anon who suggested the digipot here. If you want a 5 or 10 gang potentiometer without paying out the arse or confabulating some sort of mechanical solution, those cheap digipots are a pretty elegant and linear and cheap method with high-mileage that doesn't require an MCU or any kind of DDS. Plus they're multiturn. And unlike JFETs or vactrols (not sure about OTAs) they have that 2-sided potentiometer action (as opposed to a rheostat).

So hardly the ideal use here, but definitely worth considering in some cases.

I have to run a fairly long function during an interrupt for timing reasons. Which is fine since for timing reasons the function has to disable interrupts.
This might be a stupid question but how does a chip handle interrupts during register updates. Like say for instance say the interrupt happens around the same time as x++. If I try to read the value for x in the interrupt do I have to worry about garbage values?

>>1877019
It might depend on the exact processor you're using, but generally they should complete all issued instructions before servicing the interrupt.
That means that if the interrupt occurs while x++ is executing, then x++ will complete before the interrupt service routine is called.
If the interrupt occurs before x++ is issued, then the interrupt service routine will be called immediately.

Buying my first soldering station and need to buy tips. I will be doing work on PCB for audio equipment. Are conical tips between 1.5mm - 2.5mm about the right size?

I have a circuit board that has a microphone on it for controlling lights.
What changes would I need to do to replace the microphone with an aux input to just get direct audio?
I mean I might be overthinking it, but I feel that the circuit will have an amplifier for the microphone and plugging audio into it would just boost it further, right?
Or just go for a resistor on ground for the audio jack and pray?

Which book/chapter/guide do I read for basic power safety?
I have a synth I want to mod, I'm thinking of just splicing a 9v wall wart directly onto the power tracks, but this strikes me as a bad idea

>>1877105
2mm chisel tip is best IMO.

>>1877137
Panel mount DC barrel jack with a pigtail lead to the synth power leads on the PCB.

>>1877134
Try bypassing the mic pre-amp circuit.

Hi, total noob here. Which one of these DC jack/ regulator schematics is correct? If neither of them are correct, how so? Many thanks in advance!

>>1877182
Neither are correct. The polarity of the DC jack in the left drawing is center-negative and the grounds should be tied together. The right side DC jack is shorted (pin 3 should be tied to pin 2).

>>1875018
buy a new battery for the lappy, take the old battery apart and see if any of the cells are still good.

>>1877192
Thanks man, does this make more sense? The DC jack symbol really confuses me to be honest. I don't know which pinout is standard

>>1877217
Just make sure to note the polarity of the DC power supply that plugs into the jack. That schematic is center-positive, so your power supply should match.

>>1877217
...and yes, it will work.

>>1877182
I'm not aware of 2.0mm, normally it's 2.1mm or 2.5mm. There are center (+) and center (-) versions, the latter is mostly used for stomptronic devices that also run on a 9V battery. I would leave the switch (pin 3) unconnected unless you know what it would be good for. Leave off the silly 'Power Input'.

>>1877134

a mic puts out about 1mV, and AUX is about 0.1V. so, you need to reduce the AUX signal by about 100 times. very approximately. you can use a fixed voltage divider (google it) or a pot, which acts like a 1-component voltage divider. a lot of values will work, but a 10K logarithmic pot is ideal for every situation known to man.

>>1877269
This, but with a DC-blocking capacitor as some microphone inputs have a low-voltage phantom power on them.

>>1875414
Look into a cheap signal generator, ebay has tons from china in a case for less than $15. Good practice for soldering and will give you a couple oscillators to play with. From there study opamp filters. Then study amplitude and frequency modulation. Frequency synthesis...sounds complex? It can be but you can make a silly bleep bloop synth with that cheap siggen and then adding a resistor network in place of the frequency adjust pot corresponding to the musical scale you want. You may need multiple to cover the whole audio range, I don't know, similar to an old organ with 2+ keyboards. Not impossible to hack something together but you will need to be diligent, curious and persistent.

Some faggot has been stealing money from the cash register at work. We are setting up a camera to record the register, but it is going to be a bitch to sit through the 10 hours of footage every day to check the camera. I was thinking that I could set up something that would give an alert when the register gets opened. Maybe just a red light out of view of people's eyes but in view of the camera. Or something more sophisticated that would actually keep a log of when the register was opened and for how long that way I could compare that to the actual footage.
Do you guys have any ideas on how I could implement this? The red light I think could be achieved with a magnetic reed switch wired inside the register with the magnet mounted to the drawer. When the drawer opens, the red light comes on. It's not foolproof. Basically I just want something that will make it easier to look through 10 hours of footage every day.

>>1877487
Make a cheap esp8266 wifi datalogger that uses a hall-sensor or LDR or whatever in order to detect when the cash register is open. Configure it to send a ping to a device connected to its wifi network whenever the drawer is opened. Arguably easier than running a bunch of wiring from the drawer to somewhere the camera can see but the thief can't.
Or just add an extra small covert camera in the ceiling, like one from an old laptop webcam, wired up through a few metres of twisted-pair to an old laptop or raspi or whatever.

>>1877487

(1) you can get an indoor cam with 180-deg view and internal storage and set it up to record only on movement, and only in certain areas of the screen. that cuts your time down to 45 minutes.

(2) you can get security cams with alarm inputs to connect your reed switch, and set it up to only record on alarm. that cuts it down to 15 minutes.

(3) if you go analog, you can get DVRs with both alarm inputs, and motion detection. usually each recording will flag which event triggered the recording.

>>1877493
>>1877521
Thanks for the ideas. Gonna figure out something and soon. Kind of worried that our ceiling is too high for a small laptop camera, it's probably 15 to the ceiling. The owner of the shop wants me to install two Nest cameras because that is what he uses at his house. Kind of sucks we have to resort to this shit. We are a small locksmith shop with only three employees, excluding the boss and his wife. We hired a new guy about 3 weeks ago and since then, the till keeps coming up short. Not hard to put two and two together, especially when the two new guys are obvious hard drug users (meth mouth, twitches, etc.). At least one of these guys is going down, since they won't know about the cameras. I'm so fucking pissed about this shit, I hope we catch them red-handed.

>>1877487
>Some faggot
it's you isn't it?

>>1877546
>Nest cameras
nest cameras are traaaaaaaaash for many reasons i am too lazy to go into, any cloud services cameras are.
get some normal iptv camera that outputs h264 stream

you can also leave an opened laptop on the desk next to the register with some excel spreadsheet open on it while the camera in the lid will be secretly recording

>>1877546
>nest cameras

completely the wrong application for this

Nice to see others on this track. Everyone seems to want to do CS. But there's something so comfy about angry pixies and what they can accomplish. I really wish I had money. I want to work on repairing more electronics but most people just throw things away. Seems like buying eBay junk would be okay but how do people figure out what to flip so it isn't a loss. I want to use my skills some how.

>>1877597
Anything on as large a scale as ebay has more people who can see the true value of things, the smaller the market the more likely you are to pick up something undervalued (to repair or otherwise). Be this a local equivalent to ebay or craigslist, local facebook marketplace groups, or a brick and mortar 2nd hand or junk store. But you're also more likely to come across something overvalued by looking at smaller markets. So it's really key to do your research. Oftentimes, you'll be able to buy things and just get a replacement part off ebay and they'll be as good as new. The actual opportunities to get in with the soldering iron are diminishing as equipment becomes more dense and hard to troubleshoot.

Also be sure to check smaller markets for oscilloscopes and other tools that would otherwise be pretty expensive. The less common and/or the more time since production, the less concrete the perceived value will be. For example, CRT oscilloscopes can certainly go for $50, but I've never seen a fluke DMM under $180.

I want to build a VT100 style tty but can't seem to find a suitable video generator - In fact I don't even know what to search for.
Target would be composite video.
I tried the TVout lib for arduino but the resolution is too low

>>1877621
Thanks but I've got most of what I need besides say a good microscope. Few odds and ends missing but nothing I really need. Hot air station is kind of ass too but it works. Hell bought a tektronix 2213A that was broken and got it working good. Just got all this stuff and nothing to actually work on. I get things rarely and make them work again which is very satisfying but it comes around so rarely it's pretty frustrating.

>>1877580
As a locksmith, I would like to think I have a bit more integrity than that. I'm in line to buy this shop, why would I jeopardize that? All for around $100? I have a savings with many times that, I don't need $100, it won't make me or break me.
Other than the register being short, we actually had a break-in yesterday. Well, it wasn't a break-in exactly. Someone came through the front door, which was apparently left unlocked, went to the back office, went to the exact spot where we keep the cash register drawer and emptied it. Then the thief went to the exact spot where we keep the deposits and stole those. And then they bailed. All in the space of one minute (alarm goes off after 30 seconds). When the cops arrived just a few minutes later (they are on the same street a few blocks down), the thief was gone.
I think it was an inside job. No one outside the shop would know where we keep the cash drawer when we close and no one would know where we keep the deposits. I can see their line of thought, unlock the door but don't turn the alarm off so it looks like someone left the door unlocked and some random thief came through the door. The perfect crime? Maybe to the drug-addled mind. All of this happened at 4:55pm yesterday. The register being short has been happening for over a week though.
>>1877581
Yeah I kind of figured, thats why I will install my own cameras parallel to the bosses. It's going to be a huge pain in the ass because I can't let anyone know what I'm doing per the owner's request so I can't get any help with it. I think he thinks it is suspicious too, why else would he ask me to install the cameras and not tell the other guys about them?
>>1877592
Tell that to the boomer owner. I think he just likes the fact that he can view them remotely, he isn't exact in with technology. Hell, he doesn't even know how to use the transponder cloner in the shop. Give the man a safe to be cracked and he can do that though, he is a registered safe-cracker.

>>1877638
>In fact I don't even know what to search for.

https://en.wikipedia.org/wiki/Motorola_6845

>>1877733
>I think he just likes the fact that he can view them remotely,

all the options in >>1877521 can be viewed remotely. it's insane how much functionality you can pack in a tiny package today: camera + encoder + storage + web server

>>1877733
>I think he just likes the fact that he can view them remotely,

all the options in >>1877521 can be viewed remotely. it's insane how much functionality you can pack in a tiny package today: camera + encoder + storage + wifi + web server

>>1877733
>It's going to be a huge pain in the ass
wot?
wait for the shop to close and employees to leave and then literally just screw two screws into a wall to attach the camera and plug a network cable. And if the camera supports POE you don't even need to connect a power cable. Literally 5 minute job
But unless you hide the camera there is no point nobody is dumb enough to steal shit in front of a camera, might as well instal one of the fake ones for $5

>>1877521
>internal storage
*steals the camera*
your move boomer

are there any high quality chink breadboards on ali? i am tired of hunting down bugs in my project just to find out it was a loose jumper wire

>>1877734
OP probably wants something in current production.

>>1877783

you can only get a genuine look from genuine parts.

Are the chips usually flashed and then soldered, or the opposite?

I've got a mass produced PCB which still has pin headers on it. What are the chances they first soldered the components (few microcontrollers) and only then flashed the firmware on them using those exposed pins?

>>1877826
>pin headers on it

means you can correct the more obvious software bugs before the product ships. coz the software guys have been working 20-hour days with the help of meth.

>>1877740
I think whatever I set up, as long as it has an easy enough web interface, he will probably be OK with it. I don't know, I haven't got to talk to him in person about it yet. I imagine Ill get to tomorrow though. Kind of dreading it because I was the last person through the door on Friday, but I know I locked it. I double check everything I do because of the fear of fucking up. Our front door lock is a Mul-T-Lock so I really don't think anyone picked the lock, that's why I think it was an inside job.
>>1877742
Our shop is essentially a two story building with a single story shop inside of that. The building is over 100 years old. I have to get up above where the ceiling is to run the cables and shit. I've only peeked up there before on a ladder, I'm not sure how safe it is actually. I haven't seen the cameras he wants me to install. My thought was to install them above the ceiling and just make a hole so the camera can see through that, I doubt they would have panning and tilting capabilities so that should be OK. The boss doesn't think the other guys are observant enough to even notice the cameras though so he might just want them installed straight to the ceiling or wall. I'm leaning towards getting some harvested laptop cameras and just using those, they are tiny enough to hide wherever, connect them to an old laptop above the ceiling and just let it run.

>>1877733
Once you narrow down the suspect list get some basic spyware onto the phones of those people. It will show the entire history of the phone and if they had location turned on it will place that phone at the scene of the crime at the time of the crime.
Really simple to install. It sends the information to whatever email you want to send it to.

>>1877734
whew, that reads promising, yet, I doubt I'm up to that (yet).
As I read it, I'd need additional circuitry to construct a video signal, correct?
>>1877783
preferably, yes. but salvaging would also be an option

Do you see anything overtly wrong with this design?

>>1877777
Probably not, but I suspect the ones with round holes are better than the ones with square holes.

>>1877882
Your BJT is backwards for starters, the emitter should be pointing towards GND. Also why is there a resistor in series with one of those connectors but not the other?

>>1877902
Derp, thanks.

There's an LED connected to U9 and a solenoid valve getting 12V connected to U3. Does that look correct?

>>1877906
Yeah that sounds fine, so long as there's a freewheel diode for that solenoid valve.

>>1877914
Is D1 not sufficient for that?

>>1877743
Hide it. Cameras are so small now it's not hard to integrate it into an everyday object.
>pic unrelated but what does a pepperoni pizza sound like? Would the toppings sling off at 78rpm?

>>1877941
d1 is relay flyback but you also could use one with solenoid, technically as you are switching solenoid with relay it is not be mandatory (might be in some cases but not yours) but its always ok to have one, same applies for base emitter resistor

>>1877777
I've used shitty breadboards and really shitty breadboards, but never a good one. I recently got a wire wrap tool and wire and have been using that with perfboard instead. Works pretty good and once you get it how you want you can solder to make it permanent.

Hi,

I'm a jaded SMT manufacturing tech who got laid off, is becoming my own designer going to be its own Sisyphus task or should I just succumb to being a solder jockey for hire

>>1878008
I find even shitty breadboards are fine for prototyping, so long as you use thick enough wire.
For 1-off permanent production, I’m taking a liking to manhattan, at least while my home-etch setup isn’t up and running.

>>1877105
You can find that having a variety of tips will come in handy as you progress. Big chisels for bigger tasks like removal, or big components, and small conicals for doing fine work.

>>1877105
Conicals aren't very good in my opinion, I much prefer bevel, edge, or screwdriver tips, depending on the scale of the work. I find that pointy conical tips only make good thermal contact when one side is held flat against the pcb, making it difficult to solder without existing components getting in your way.

Going to be making some audio PLLs, which calls to question how I should do the filtering after the phase comparator. I'd like this synth to get into the bass region. Is there some phase comparator that isn't an XOR gate that would remove this issue, or should I look for more exotic filter topologies, or approximate voltage-controlled filters with vactrols in them like my oscillators? Even with voltage controlled filters, down at the 20Hz region I'll want the resulting wave to be stable down to relatively small keypress times, which feels difficult without a very high-ordered filter or some sort of notch filter targeted at the (moving) fundamental. From what I've seen it isn't too big of a deal to end up with ripple on the VCO input, but I've no idea how much ripple is too much ripple, and I'm pretty sure it will distort my triangle wave more than it would a resonant oscillator.

>>1878008
I use wire wrapping for production sometimes, for example to connect pins on ic to other shit, when i'm too lazy to hand make a PCB, but i don't really use it for prototyping because you still have to solder shit to perfboards and add pin headers and shit and even though you can wire wrap directly to component legs in most cases it gets messy and it's too cumbersome. Nothing beats simply plugging jumper wires and components directly into a breadboard.
If i could get my hands on a quality bread board and jumpers it would be the ideal way to prototype. But the chinkshit i order from ali is just cheap low quality crap that works, but i get the problems i complained about.
And buying locally makes no sense because what they sell is the exact same chinkshit except for three times the price.

>>1878288
Oh, and any phase comparator design I pick has to be sufficiently simple compared to an equivalently effective filter (e.g. number of components, precision of components, etc.). I'm going to need 8 of the damn things, and just chucking a couple 4000 series quad-XOR gates on the board is about as simple as it gets for this. On the filter side, I'm starting off looking at a 4-pole butterworth filter using 4 op-amps. It gets a 12V 50Hz ripple down to 1.3V with a 10Hz cutoff.

Maybe the fact that I'm simming with a butterworth shows my inexperience, might have to look up chebyshev coefficients or whatever.

>>1878290
I actually fidn much much better to just take a copper clad, scratch a grid with a knife (sometimes very fine) and just solder the components. When working with complex things bug testing on a breadboard is hell and most of the time is some bullshit you've looked at 3x times and missed. It's more resistant and I don't think running a 60w iron for 4 hours a week costs anything

meanwhile, ben eater releasing another video of a MHz circuit made entirely on breadboards
iirc he only ever had one "bad breadboard connection" issue

>>1878288
>>1878291
>look up analog phase comparator
>uses two double-balanced mixers and sums the result
Ok that's kinda bad. I'm guessing since I'm already using vactrols that the easiest way to multiply two signals is with a variable-gain amplifier, though I'm pretty sure the vactrols would have to be matched well in order for the 2f components to cancel. Also no clue how to use an op-amp to get both positive and negative gain just by changing a resistor, so maybe it needs to be a single-quadrant multiplier with that ground just a volt or two away from a rail. Using log, sum, and exp amplifiers also has the single-quadrant issue but it takes even more op-amps. Might be more well-matched though.

>>1878303
I've managed to do a gilbert cell for audio frequencies. I used it as a tremolo effect. I think shitty audio shit is the only time when doing one of those with discrete components cna give something usable.

>>1878320
yeah but that's high part-count, especially for 8 of them.

>>1878303
Drat, the two-mixer method requires putting a 90° phase shift on each input wave to get quadrature inputs (reminds me of the SDR signals), which even with square waves is a bit of a pain. The single-mixer method requires a filter.

But if I use an XOR gate (or the single-mixer method, for an arbitrary waveform), half-assedly low-pass-filter the output to pull it away from the rails, and then just add an ideal rectifier + capacitor to capture the peaks of those, I should get a relatively flat line. Not very linear, but it should be perfectly fine as it's part of a negative feedback loop of a PLL anyhow. That way I just use 1 op-amp and 1 xor from two quad-chips, plus a handful of passives.

>>1878300
>iirc he only ever had one "bad breadboard connection" issue
Only one that he showed.
>MHz range
>1Mhz to 1 Ghz
very specific huh.
Go to the end of the app_note to see how to breadboard Highspeed circuits.
https://www.analog.com/media/en/technical-documentation/application-notes/an47fa.pdf

>>1878355
>damn highspeed prototypes look like THAT?

>>1878357
That appnote is one of Analog's "classic" ones. Really helped me understand measurements.

>>1878357
i just vomited in my mouth

i would like to remind you that /diy/ is a safe for work board and posting gore is not allowed here

>>1878381
>t. plebeian who only works if he has a board made and shipped for him

>>1878357
>>1878359
god i love manhattan
that's a bit disorderly though

I can't figure out how do I get this voltage regulator (https://www.ti.com/lit/ds/symlink/tps745-q1.pdf)) to output 1.2V. Is this done by manipulating the Rpg resistor?

I'm using a TI TMS320 processor and it has 1.2V core voltage and 3.3V for the digital I/O. Now I've got the 3.3V line and would like to convert it to 1.2V. I think I can use this regulator to get to 1.2V (as far as I understand, it converts AND regulates the voltage).

Also more generally, do I need a converter or a regulator here?

>>1878588
>Typical Application:Adjustable Voltage Version
>Page 1
R2 and R1 set the gain for the feedback. Probably Vo=Vref(1+r1/r2). Jesus that's a shitty datasheet, even for TI

>>1878602

they stick the most relevant fact on fucking page 20 of datasheet. these writers need to be hung by the balls.

>>1878606
They didn't know that stupidity is the new normal.

>>1878602
Yeah, that's what I thought. What is "fixed-voltage version" then?

>>1878606
You think page number indicate importance?

>>1878606
>>1878606
Uh I retract what I said, it's an ok datasheet. I missed that part for some reason.
>>1878614
Yup. But I had some bad experience with TI datasheets a couple times and it was a very shitty experience.
>>1878628
Voltage regulators (and almost everything) have a reference voltage.
>if output is above Vref do X to make it go down
>if output is bellow Vref do Y to make it go up
If you put a voltage divider in the feedback path, it adds gain. (Because the regulator will see the output X times smaller you need a bigger Vout to trigger action).
>>1878646
Not the person you are quoting but I said it was a shitty datasheet because I somehow missed the part detailing that, now I see it's included so it's a perfectly ok one. Not on par with the old AD app notes as out friend pointed out here >>1878355

Should an ESR mega part identifyer be able to identify voltage regulators?
If giving it a BA50BC0 and it tells me "no, unknown or damaged part" and to be fair it might be damaged, or it might not be able to measure voltage regulators

>>1878686
I would be somewhat surprised. For one, it needs >5v to do anything and the microcontroller runs on 5v. I'm not sure how it applies power, but I imagine they would be hesitant about applying more than 5v to mystery parts.

>>1878686
I have roughly the same identifier and tried a few regulators. Some read as pnp transistors, others as damaged. None read as regulators.

I can't seem to find much information online for it but how much current can a castellated via handle about or how would one go about calculating the max current for one? Thanks in advance

>>1878710
>castellated via handle
Try googling keywords like specification

http://circuitcalculator.com/wordpress/2006/03/12/pcb-via-calculator/

>>1878714
So it's calculated as if it were a normal via yes?

>>1878710
With the big blobs of solder you use with them, I imagine they can carry more current than a 1mm wide trace, if that helps. If you’re planning on putting more than a couple of amps though I’d lean towards an actual connector, not that stacking PCBs is a terribly good idea in that situation for thermal reasons.
>type in “1mm wide“
>autocorrects to “1mm wife”
qt pick and place robottos when?

>>1878729

I'm trying to fix my dishwasher. I'm in a location where "lockdown" stops me getting someone in. The issue is the program lights continuously cycle and you can't select one to start the dishwasher. The P button has no effect when pressed.

>>1878867
I put an ohmmeter on the button and when pressed the circuit closes so I don't think it's the button

>>1878868
I don't know what to look for but this is the circuit board. Seems fine to me. The cables coming out the top seem kinked so maybe a broken cable? I don't know. Any help would be much appreciated.

>>1878870
Measure the voltage across the buttons instead. Are you fine with the machine being plugged in and turned on when you've got a DMM in there?

>>1878875
Already got zapped once so why not. What do you mean across the buttons?

>>1878875
>>1878877
I was zapped from touching the inner body of the dishwasher. I think I know what you mean by voltage across the button. Got it working after playing with the wires which I suspect are damaged. Not really sure if I'm up for cutting the wires out though. Might just try and keep it going for the next 6 weeks.

>tfw responsible for someone shocking themselves via the internet
not sure how to feel

yaaaaas it's finally here!

time to find out if i just wasted 10 yuros

One spot looks darker, but it conducts electricity between all three solder spots.
1/2

One spot looks darker, but it conducts electricity between all three solder spots.
1/2

I also read that the component that's on the other side (one of those white 899B thingies) often fails and causes the specific problem I'm experiencing (dishwasher stuck on the last minute forever).
So I was thinking of replacing that component, but I can't find the exact same component (899B-1C-F-C E 6V). What would be a suitable replacement?

>>1879157
>>1879165
I meant that the white component connects one leg to the dark spot. It also reads overload Ohms from that leg to all the other legs (unless that's how it should be, I don't know, but I'm assuming that it means it's broken).

>>1879165
this is a relay, in a common form factor. part numbers on electronics packages are shit so you have to be creative with your search terms. "899-1C-F-C" gives results. the characteristics you care about are: 6V coil voltage, 277VAC 10A switch rating. as long as you've met those criteria and get the same pin footprint then you can drop in just about any relay. there are other characteristics pertaining to reliability and standards compliance that you can ignore unless you're anal.

i haven't read your other posts but if one of the relays does turn out to be the issue then you can replace it for cheap if you have a decent soldering iron, flux, and a solder sucker. this link is to an almost identical relay in the same series that looks right at a glance:
https://www.mouser.com/ProductDetail/Song-Chuan/899-1C-F-C-6VDC?qs=0rbOLB9uh%252BAJTo%252BeJJFsDQ%3D%3D

>>1879176
>there are other characteristics pertaining to reliability and standards compliance that you can ignore unless you're anal
So I should ignore that 899B is "Basic series designation with insulation barrier" and 899 is "Basic series designation"?
I'm also asking since I actually found this listing but not only does it have this slight difference in number, it also has expected date of delivery at the end of October.

>>1879184
the "B" option is a compliance feature. you can ignore it if you can't find the exact part, it has marginal implications for safety and reliability which only matter if you're shipping a hundred thousand units.

the JS1F-6V-F appears to be a suitable alternative. if you end up needing it then i hope you enjoy paying $7 shipping for a $2 part.

>>1879188
Would https://www.conrad.com/p/finder-361190064011-pcb-relay-6-v-dc-10-a-1-change-over-1-pcs-503066 also be acceptable? This has better shipping to my country.

>>1874921
>(embed) (embed) (embed) (embed) (embed) (embed) (embed)

>>1878996
what is this?

>>1879239
a punch for crimping thru-hole rivets.

What happens when you run a 7400 IC at 12V or so? Do the outputs end up shorting the rails or something?

>>1879190
sorry got distracted by a worthless meeting. that relay will work.

retard here, I know when a microcontroller outputs 5v it's generally necessary to put a 220-330 ohm resistor between an output and the anode of an LED, but what if the controller is outputting 3.3v? also, if I'm using a shift register to control, say 8 LEDs with 1 microcontroller output, could I put a single resistor between the microcontroller output and the shift register's input, or would I have to hook the microcontroller output directly to the shift register and then have 8 resistors between the shift register outputs and the LED anodes?

>>1879346
until you really understand what the resistor is for, you always need a resistor for every single led. it needs to be directly in series with the led itself. you'll need to size the resistor differently for 3v3 than for 5v and not all led colors will be practical to drive at 3v3.

>>1879355
a resistor for every LED then, got it. I'm thinking basic red LEDs at 3v3 with 100 ohm resistors, what do you think

>>1879359
if your shift register outputs are rated for 15mA then go for it

>>1879361
ah, no. they're ±6mA

>>1879362
270R or higher then. in practice the sr pin output voltage will just droop when overloaded so it'll work, but the datasheet won't tell you whether that might overheat the chip until you try it.

>>1879365
yeah it's for a product so I'm trying to keep it by-the-book. Thanks

>>1879333
You would fry it for sure but how exactly is a crap shoot; transistors fail both open and short.

>>1879471
>transistors fail both open and short
That implies that the transistors will be damaged by voltage itself, or that their breakdown voltage is a little over 5V. But that sounds really unreasonable, compared to either thermal overload from current sources dropping too much voltage, or some sort of push-pull transistor combination (likely with at least one common-emitter) shorting the two rails because of failed biasing. I'm thinking the latter, since it reminds me of what can happen with a class-AB amp, and these logic ICs often use output stages that aren't as simple as a class-B.

Pic related, though I don't see any immediate problems. Looks like TI actually have some of the most informative 7400 datasheets. Shame they're almost all just scanned without proper OCR.

>>1879534
why no pnps?

i just can't do this shit anymore, it's too stressfull
every time i build a part of a circuit for an hour and then connect power to test it i am shaking terrified it will start smoking and i will lose expensive bugs and also will have to start from scratch remaking it
or when it just doesn't do anything, oh shit, is the design bad? is it just a loose wire? is it a bad component? i will not havo multimeter and oscope that shit for an hour to find out and fix it, and it's just too fucking nerve wracking and tedious and fuck this shit, electronics makes me want to kill myself, so why the fuck do i keep putting myself through this hobby?

>>1879570
>it just doesn't do anything, oh shit, is the design bad? is it just a loose wire? is it a bad component?
But that's the best part, anon.
Also you don't have to get too much better to know you won't smoke a component. Or always buy 3 extra?
It shouldn't be stressing you out, that's your personality, not the hobby. You're talking about the must exciting parts of it, bringing up the circuits for the first time. It's fine to be unconformable / nervous / stressed. Just switch that to "excited" and enjoy the rush.

>>1879570
It seems your problem is not with electronics man. Things can go wrong like that with cooking, carpentry, fishing..

>>1879580
when was the last time you had to replace your whole house because you hammered in one plank wrong?

>>1879581
>make a table
>fuck up one of the joints
>whole thing is shit
>waste one hour to find out what I have fucked up and another one to fix it

>every time i cook a part of a meal for an hour and then turn on the stove to cook it i am shaking terrified it will start smoking and i will lose expensive ingredients and also will have to start from scratch remaking it

>>1879581
Just use a current-limiting benchtop power supply, a dead short of 5mA through any IC shouldn't kill it. Just don't turn on any LEDs and use your scope (or logic analyser) for analysis. Should be perfectly fine for simple analog or digital circuits.

>>1879581
>>1879588
And after you gain some experience with electronics you'll get a feel for what kind of values will cause magic smoke to come out or simply not work in most situations. I think you have anxiety. It's not "electronics" fault.

>>1879534
There is no telling what silicon is in the chip and a quick look shows me a absolute max Vcc of 7V, that's all I need to know. It could be as little as busting the max Vce and it's EOL. No sense in trying to reverse engineer a circuit when it's active components are unknown unless you are in an absolute dire need.

>>1879572
>most exciting parts
>yay i have to waste another hour by hunting down a cold solder joint
so much fun indeed

>>1879613
get /out/ then faggot

can somebody here help me out? i have to tear down a shed that has electricity and im not sure what to do with the old cable once the shed is gone. the new shed will be built in 1-2 years so the cable has to sit there for a while. would it be enough to use pic related on its own or does each wire need its own end cap and then the big end cap on top?`sorry for my stupidity.

>>1879681
I would put plasti-dip or plasti-tape to seal the end, then I would stick it into an outdoor rated waterproof junction box
Itll be good for a year or two

>>1879681
Disconnect the other end from mains, or remove the breaker from the panel if that's the only run on the circuit, until your new shed is built.

>>1874921
Design a circuit to turn a LED on/off.

>>1879738
epic challenge bro

I don't know shit about high power (>1/4W).
Will ~2W of RF going through a T37-6 toroid heat up or radiate a shitload of noise everywhere?

Since couldn't find /MCG/ I guess I'll ask here:
I've dug out my LEGO NXT, thing is, I can't seem to be able to install LeJOS.
I want a way to program it, is there a way to interface an Arduino with everything else in the NXT?
Alternatively, is there a way to use the NXT's MCU with Arduino directly?

>>1879875
What frequency? Usually high power RF shit uses aircore or just use stubs and open circuits to make stuff.

>>1875812
that is the fugliest schematic I've ever seen
regardless if you haven't already bought these items I bet it would faster and cheaper if you went to Tayda
all looks fine though

>>1878359
i <3 Jim
would've loved to have called him up to help get my 1Hz to 1MHz V-to-F running ;(

>>1879905
~30MHz, nothing crazy. Aren't aircores even leakier than toroids?

>>1879938
Air cores have no *ahem* core losses

>>1879941
Also 30 Mhz is larger than 1m, I don't think you'll have problems with a small toroid, but I don't have a lot of experience with RF shit.

>>1879941
to my knowledge/experience, cylindrical cores in general behave like antennae. Put your finger next to an LC resonator with a cylinder coil, and the freq will go wild.
toroidal shapes are self-shielding; magnetic field lines always want to loop back on themselves, and a toroid guides them to do just that.

I'm worried about both power losses (air is superior) and noise radiation (toroids are superior), but idk if air-core toroids are feasible.
As long as the device doesn't burn out, I guess it's all good

30Mhz needs a PRETTY big inductor to start working as an antenna.

Is EasyEDA really shit at importing bitmaps to PCB components in comparison to KiCAD or am I doing something wrong?

>>1879882
There might be communities still based around LEGO NXT, be it on the /toy/ board or on other sites.
You could at least ask >>>/toy/lg

I read that you can make a voltage inverter using a switching buck regulator if you swap the output and low side switch pins.
Is pic related dangerous? This is a SOIC package with four pins normally grounded and the functional block diagram doesn't show a configuration I recognize. Using those four pins as negative voltage output seems like it would cause some kind of thermal issue under load?

I was given an Amber 5500 distortion analyzer, very interested in getting it going. Power supply has a failed regulator IC, going to test the rest of them but high end old test gear gives me such a hard on; it consists of seperate modules connected to a common backplane. All top notch parts, and some very interesting engineering. No options unfortunately but the ability to measure THD down to 0.0005% and IMPTE IMD has my pp fucking tickled.

>>1880040
I'm confused, did you make that second schematic yourself? If it was in the datasheet it should work. Diodes inc aren't retards. It should at least not ruin the chip.
In general though, I think there is usually a couple caveats like startup, etc. I'm not wiling to check to see if that would actually work but it can definitely be done. Google AN19 linear tech, pg 44. I've built that and it's worked for a year with tons of shorts applied and no problems!

>>1880096
also that chip looks like the classic lm267x series, I think they have a schematic for negative regulator also

>>1879546
at the time, they only had lateral PNP which were shitty (low gain, low speed, low current). But if you're saying why no pnps on the output stage, it's because a) they're logic devices so only needs 0-Vs, b) the cascode configuration is faster (npn + npn)

>>1879333
just regulate the voltage with a 5v1 zener. unless the rest of your system needs the HV.
Even so, there's ways you might be able to use the chip with zeners and or transistor combos. try googling around

>>1880106
that answers both of my questions thanks. got any guides for building cascodes? copied a circuit off google images but it was slower than a pnp + npn totem pole

I can feel it bros. She's dying.
My o-scope takes a full 5 seconds to start up, and lets out a sickly whine when powered down

which caps do I test first?

>>1880236
Probably the LVDC ones right after the PSU. Though the input caps and the HVDC output caps are also worth checking. First clean the PCB in case some metallic dust or whatever is causing issues.
Transformer might also be going somehow.

>>1879975
>air core toroid feasibility
nigga just bend a regular air core coil wtf

I thought I'd try simulating a 555 circuit I read online, but I can't get the simulation to do anything

>Pin 1 is connected to ground (negative side of the 9-volt battery).
>Pin 2 is wired to pin 6.
>Pin 3 is connected to one of the transducers.
>Pin 4 is connected to the positive side of the 9-volt battery.
>Pin 5 isn't connected.
>Pin 6 is connected to one side of C1 (the other side to ground), one side of R1, and one side of R2.
>Pin 7 is connected to the other side of R1.
>Pin 8 is connected to the other side of R2 and also the positive battery terminal.
>The other transducer wire is connected to the ground side of the battery.

The "transducers" mentioned in the circuit are just big flat electrodes you wear on your shins.

>>1880096
Second schematic is an ms paint edit of the first, sorry. It's based off of my notes from reading a pajeet article on the subject.

It looks like the an19 document is about a chip with.... different topology? I'm not sure if I'm correct but these two block diagrams appear to be functionally different in some way...

>>1880305
Your battery is backwards.

>>1880314
I flipped it and followed the instructions. I fucked-up somewhere.

>>1880321
You are connecting Vin to gound.

>>1880322
Every diagram of a 555 I see has pin 8 as Vin and the instructions say to wire pin 8 to positive. Either way I wire it it doesn't do anything.

>>1879614
>faggot
jaded cunts like you creating toxic work environment are the reason why nobody wants to go into EE except pajeets and retards
just the thought of having to touch a circuit together with an obnoxious shit like you makes me sick

>>1880326

pin 7 is used to discharge the cap to ground to reset the timing cycle. if you wired it like you have it in real life, you'd burn the chip up, as it would short V+ to ground thru a transistor inside pin 7. seems you cant follow simple instructions, so maybe move on to a diff hobby, like guns.

>>1880326
dude just post the picture of the circuit you are trying to make please.
>>1880327
ok, I'm a gaming man myself so you are just over reacting on a anonymous congolese IC recycling forum

>>1880331
>dude just post the picture of the circuit you are trying to make please.
There is no picture, just instructions, which I posted already. I got them from here: http://www.textfiles.com/bbs/KEELYNET/BIOLOGY/sleepind.asc

>>1880326
You fucked up where you put that 1k resistor. It should be between Vin and dis. In the discharge state the dis pin is connected to ground. The 555 will stay like that until the capacitor has discharged to 1/3vcc. That's never going to happen in your circuit because the capacitor is gaining charge through the 1k a lot faster than its losing it through the 2.5M.

Also I'm not sure what you're trying to accomplish but with those component values this is going to have a period of about 30 seconds; obviously too low to make sound.

>>1880344
Click the circuits tab on the top, go to 555 timer and choose the astable one. It's what you want to do. Compare that with what you are doing

>>1880329
Learn to read a schematic. He doesn't have the discharge pin connected to ground. The traces jump over each other.

>>1880352
Astable=square wave generator

>>1880326
>reset pin close to Vcc
>control pin close to ground
wow if only the actual fucking IC was built like that

>>1880352
>>1880355
Oh, wow. That seems to be exactly what the circuit was trying to do. I changed the components to see what frequency it should be putting out, but I can't really read the scope, and considering that the electrodes will be attached to a body, I'm not sure exactly how that'll effect the output either.

>>1880356
What? Why?

When doing a continuity test, if there's a regulator between two points will there be continuity?

>>1880363
"if I'm measuring the size of an object, will it be sweet or savory?"

>>1880360
>Why
the actual pinout of the IC makes routing traces a real pain, especially on a single-sided-board. Same for a breadboard if you're trying to get away with just one of each rail.

>I'm not sure exactly how that'll effect the output either
Read the datasheet

>>1880402
>Read the datasheet
I don't know how.

>>1880413
>I don't know how.
Left to right, top to bottom.

>>1880413
I think you want to know how much current the output can drive, or some approximate output impedance.
https://www.ti.com/lit/ds/symlink/ne555.pdf
Page 4 says ±200mA, which sounds ok. But at 3.3V output voltage (Typical, from Page 5, High Level Output Voltage for Vcc = 5V) you might not be able to drive many loads.

What's the circuit for?

>>1880415
>jumper wire under the IC
what the hell

>>1880419
not him but whats wrong with the wire?

>>1880415
Where's this datasheet you speak of?

>I think you want to know how much current the output can drive, or some approximate output impedance.
The circuit just said a 9v battery, but didn't specify, but I'll assume it does a standard 50 milliamps. The bigger problem is output impedance, since the load is human skin, which changes in impedance.

shit man, how do i attach oscope probes to a dip socket? i tried jamming a breadboard wire in it bu the connection is shit, i can never tell if i am not getting a signal because the contact on the dip is not recieving any or if the jumper cable i stuck in it is loose aagain

whats the least autistic way to connect say 1 button to a 3v and 5v chip.

Sup /ohm/, newbie here. Since I wanted to get started in electronics I got myself a small chinesium signal generator that you have to assemble yourself, solder, and an iron. I just finished soldering it, it was my first soldering job, pic related. I'm pretty satisfied, but I have two problems:
>the solderings look dull and not shiny
Did I fuck something up (temperature maybe?) or is the solder wire I got from amazon shitty?
>I used tons of flux (got that from amazon too) and now the whole board is flooded in it
I tried to remove it with alcohol but there's still a lot left. How do you clean your boards from all the flux?

>>1880519
Get liquid flux and real solder like Kester 44 and use lower heat. To clean the board, heat the flux and wipe it away, then scrub the remaining flux with a toothbrush and isopropyl or another mild solvent. Alternatively, put it through a cycle in your dishwasher.

is there a trick to open one of these up? every guide just says to unscrew to open but there are no screws cause its ip44. just want reaattch the strain relief

>>1880519
Where did you buy that excellent 63/37/0.8/2% solder wire?

>>1880519
nothing wrong with dull. doesn't have to be shiny.
some fluxes make it shinier. but it doesn't matter other than the looks. as long as you apply enough heat and don't use excess solder to avoid cold joints.

>>1880519
cheap solder tends to get dull like that. not great, not terrible.
as long as it's not tacky and lumpy, it'll do

>>1880525
>>1880553
>>1880565
Thanks
>>1880531
No need to be sarcastic anon.

>>1880568
>sarcastic
Obvious misunderstanding. My local dealer no longer sells 60/40 type solder wire in favor of 'green' lead-free crap. That's why my question.

>>1880576
Everybody in the hobby should pretend that lead-free solder doesn't exist.

>>1880419
>>jumper wire under the IC
>what the hell
More typical in older gear but it happens.

Could I split the output from this 6V solar panel:
https://www.adafruit.com/product/3809
to a pi zero and this 6V solenoid:
https://www.adafruit.com/product/997
to actuate it and make it fully solar-powered?
I have the pi running off the solar right now but it's got the solar charging kit and a buck converter to regulate to 5V. I just don't wanna split the cable from the solar panel if it won't work.

>>1880580
Pretence is a form of self-sabotage.

>>1880519 here. The little shit doesn't work: it just outputs the 12V DC that I put in. I tried to look for shorts, but I haven't found any. I feel like I got chinked hard.

>>1880576
I got it from amazon, but it's crap: not shiny and doesn't flow very well despite the copious amount of flux I use.

>>1880615
So is lead-free solder.

>>1880430
If the wire is on the front of the board, then it's going to push the IC upwards. If it's on the bottom of the board, then the solder joint won't be as strong, with a higher risk of delamination, especially if the ends of the wire don't stick through the holes. Not really a deal though.
But it should have been possible to route the board such that the jumper was somewhere on the front.

>>1880580
Works fine for me, barely any difference to solder if you know what you're doing. Not like we don't have our temperature-controlled-irons set to over 100C above the melting point anyhow.

>>1880594
Those are bodge wires, not the same as a dedicated jumper. Dedicated jumpers are usually of a more solid wire.

>>1880660
>barely any difference to solder if you know what you're doing.
wrong
>Not like we don't have our temperature-controlled-irons set to over 100C above the melting point anyhow.
retard

>>1880668
>wrong
I've soldered with both in recent history, didn't notice a single difference, but maybe that's just indicative of my soldering style.
>retard
63/37 Pb/Sn solder has a melting point of 183°C. I barely ever see anyone soldering below 280, it's usually between 300 and 350. Pure tin, which is the main ingredient in lead-free solders, has a melting point of 232°C, so any lead-free solders are going to have a melting point below that. Tin also wets better than lead, though IIRC a lot of fluxes don't work as well with lead-free solder. And when you're cranking the temperature up to 400C when working on aluminium-backed PCBs, a 50C difference in melting point is completely negligible, which to some extent applies to more regular soldering too.

>>1880688
What's the power rating of the iron you're currently using?
>working on aluminium-backed PCBs
More reason to use leaded solder and maybe a pre-heater.

>>1880699
80W I think? Maybe 60W.
The aluminium-backed PCB already had lead-free solder on it, so I still had to melt that stuff. Don't have a nice setup with a pre-heater.

>>1880711
You could whip one up with a PID controller + hotplate & thermocouples, or halogen lamp + fresnel lens.
60W is pretty low for lead-free work. The higher powered irons are far more thermally efficient, so you don't need to crank the temp to 11 and the tip recovers a lot faster.

>>1880717
>whip one up
If I had a hotplate (or better yet, a ceramic element) and enough bench space, I'd do exactly that. But I've got a clothes' iron I can hold between my knees in a pinch.
>60W is pretty low for lead-free work
Checking now, it's a little over 70W. I have integrated tip+element+thermocouple tips (T12) so I imagine the element-to-workpiece thermal resistance isn't as high as it could be for more conventional tips.
At the very least, I've never had power troubles, it barely ever goes up to 100% duty-cycle besides startup. On the aluminium-backed PCB, when up at 400C it was taking maybe 1-2 seconds for the solder to flow nicely, if I recall.

Audio transformer

Audio Converter:

Hi /ohm/,
I'm currently trying to create a 8ohm 8w Input to Line level output converter (-10dbV).
Here's a link to a Pro Audio version:
https://www.rdlnet.com/product.php?page=666
I understand that I will need a 10kohm potentiometer, however the Pot is rated at 0.5W so how should I get 8W DC down to 0.5W DC for this application?

>>1880758
The transformer steps down 12:1 or thereabout, and the secondary has a 10k pot across its terminals to trim the output to ~.5W.

>>1880782
What type of transformer would I need for this? I was looking at a buck converter.

>>1880788
>I was looking at a buck converter.

wat! you want a good old audio transformer, like in the pic. however, i see no need to isolate the signal with a transformer in normal circumstances. you can just a simple 2-resistor voltage divider (4 resistors for stereo) like what they use in stereos to drive the headphones. in this example, of the hitachi sr-2004, it uses 270 ohm resistors and 32 ohm headphones to create a 9:1 divider.

>>1880788
https://www.ebay.com/itm/Jensen-JT-DB-E-Premium-Direct-Box-Transformer-12-1-/222529366630
The schematic you posted is for a passive device.

>>1880798
Yes the Speaker I'm using is a line level RCA connection. It has it's own DC power supply separate from this audio converter.

>>1880797
Ahh okay so if I use 2 resistors in series before the Pot this would work. What resistors rated at 8W can take 8W down to 0.5W?

>>1880797
Also I will be making this mono

>>1880812

assuming 8W is the output power of the amp into 8 ohm speakers, that makes the max output voltage equal to 8V. (P=E^2/R)
line level RCA is about 0.1V to 0.2V, so you wanna divide by 53. take a random value for the bottom resistor of 100 ohms, makes the top resistor about 5.6K. power rating of 5.6K is ~11 milliwatts. so, any kid of resistor will do. and any kind of pot as well. no worries about anything burning up.

>>1880839
Thank you so much. Cheers

Im building a project for an stm32 and I think I finally got it working. Only thing is it only builds an elf file. Im used to seeing bin and hex files. Does the elf serve the same purpose?

>>1880000
Yeah, it's shitty

>>1880872
No, you are missing the final step somehow
AD9833_f103.hex: AD9833_f103.elf
@echo 'Invoking: Cross ARM GNU Create Flash Image'
arm-none-eabi-objcopy -O ihex "AD9833_f103.elf" "AD9833_f103.hex"
WHat IDE do you use?

Just soldered my first SMD component today, any way to test all the solders are good without plugging it in?

>>1881069
Just do it, faggot.

>>1881069
Looks OK, just make sure to clean off the flux and you should be good.

what crimper can I buy for Molex 254 that doesn't cost $500 and isn't cheap Chinese shit? I've never crimped rectangular connectors. I checked the engineer pa compatibility matrix and it's not listed. Would be a bonus if it works with dupont or others too

>>1881293
looking more I found Molex 0638111000 for $90 CAD and 0008550129 is listed in the datasheet so I don't mind if it makes a difference in quality

>>1881299

WTF amerikka! $1 for a single pre-crimped 6-inch wire.

This isn’t very electronics related, but I’d like to see a hydraulic sandwich press. One that turns your sandwich into an extruded mess AND cooks it at the same time. Actually, a hydraulic press in the kitchen could be used to vastly simplify the process of rolling dough flat, and also for flattening stacks of layers like croissant pastry.

Oh and someone made a new thread.

>>1881327
that's reasonable for the entire assembly.

>>1881042
no ide. just making the project. But running
>arm-none-eabi-objcopy -O ihex "AD9833_f103.elf" "AD9833_f103.hex"
does the job. No idea how to add that to the makefile but its not that important.

>>1881153
Damn, I used none

might as well link to the new thread
>>1880918

Guys do you have any recommandations on books with problems on diodes, op amps, and bjt transistors? Thanks