/diy/ - Do It Yourself

>>1560323
pat. pending

not sure I have the dexterity for THT, never mind surface mount
I better buy 10 perf boards just in case I fuck up

>>1560624
A dual comparator version uses more passives but has an actual constant frequency. The design is better for when you have 3 things to PWM at once with a quad comparator like an RGB light, but you can also easily change PWM frequency and duty cycle independantly.

As far as speed goes, a CMOS 555 (or modernised equivalent) might be able to go fast enough, as would a well-selected comparator. Look at some datasheets.

>>1560823
>couldn't wait for page 10
kids these days

this thread's digits brought to you in part by the MAX823 microprocessor supervisory circuit with watchdog and manual reset input. because microprocessors with multiple supplies tend to get very bitchy about their power sequencing

>>1560831
y tho

I got a $25 Amazon gift card for walking so I ordered pic related and some boost converters.

I still want to get a diode assortment, transistors, capacitors, and some other random $10 grab bags but I figure I should probably build something or at least have a plan before I buy more crap.

Also I hate vaginas. I woulda ordered a scope if it weren’t the girl’s birthday in 2 weeks. Kate Spade and Disney > Bench Testing Toys.

>>1560956
>bitch, bitch, bitch
If I hadn't you'd be complaining about it not getting done fast enough.

>>1560990
not I, sir. perhaps others would, but they're whiners

Is this book good? I just want to get first hand experience ASAP.

Can somebody explain to me why the tank circuits oscillates here, despite being connected to a DC source?
Shouldn't it just charge the capacitors, and then keep current flowing through the inductor as a short circuit?

>>1561091
>connected to a DC source
Wrong, there's a DC-blocked source above whose voltage changes as a function of the current through Rc and hence by the current through the transistor's collector to emitter. Each end of the oscillator is effectively on either end of a single-transistor inverter, which is basically what crystal oscillators do in circuits with a schmitt inverter.

>>1561116
oh and by DC blocked I mean the capacitor C0 prevents a net current from flowing

I'm a bit stumped at the moment. Pic related is my circuit, I posted about it here before but I still can't figure out why only one capacitor is giving me the results I want. So I have this cheap chinese 1kuf cap I took from some spare circuit, it's giving my led a very nice dim, all the way from the beginning to the end. However, any other cap regardless of the farad value always turns the led on when I turn my pot so the duty cycle is at least 40% and I don't know why. Without the cap, the pot switch turns the led on immediately and floods it with 110+ma. My cycle range is around 22%-99.5% I accidently bought a 15k pot when I actually wanted a 150k which is why the minimum is so high, still its a nice pot so im keeping it. Anyways, I bought a 1kuf nichicon electrolytic cap to replace the cheap cap but it behaves just like the other caps I tested where the led only turns on at ~40% duty cycle. Anyone know why? Is the cheap cap merely leaking the moment I turn the pot switch on?

>>1561121
Isn't that cap just short-circuited? If you want to use it to smooth out a waveform you'll need to make a low-pass filter with a cap and resistor/inductor. Also I don't understand what's happening at the source of those FETs, please draw a more standard circuit diagram. It's fine if you represent your PWM circuitry with a simple voltage source, similar to pic related.

I just came up with this CD4017 circuit. It uses the counter IC to read the states of 8 DIP switches with three microcontroller I/O pins. CLK and reset are configured as push-pull outputs and the data input is a floating input. Maybe it could be used for reading some configuration settings?

>>1561027
then start your own project and learn what you need to know to make it happen
even a blinky will do

>>1561027
I'm halfway through it and it's okay yes, I'm going back and forth between reading and working on 2 projects. I wouldn't buy it tho, you can find the pdf-ebooks everywhere or wait for the next time humblebundle has them

In a couple of projects I used the totem pole configuration as an output, but on one of them I had to level shift and invert the signal. That's ok right? Just use pic related. I did, and it worked but it results in a big difference in sinking and sourcing of current. Anyone has a better idea of how to do the same thing but without that? I don't want to use a CMOS inverter chip because they occupy too much realstate.

>>1561215
The problem itself is the small source current, not the difference itself, so adding a emitter resistor won't help

>>1561091
There is no 'tank', it's a loop. It oscillates when gain and phase are right.
Colpitts = tapped C, Hartley = tapped L, Emitter-coupled = no tap.

>>1561091
All LC oscilators have the transistor just to keep "pushing" things arround. If you had ideal capacitors and inductors arranged in parallel, they'd start vibrating at the smallest perturbation. Also, the C0 cap blocks DC.

>>1560823
Including the output stage in the feddback.

Hewwo big bois.
Uhh.
How exactly does an oscillator crystal work?
Specifically, how do you see the oscillation in the signal output?
And how does it start oscillating? And how is it kept consistent?

how do I stop getting angry when I don't understand stuff and not even the explanations?
god I hate microcontrollers

>>1561282
grow up

Why is the right half-plane zero something to avoid when designing control systems?

I know right half-plane poles make the system unstable but why RHPZ's?

>>1561282
Find good book(s) about the subject you do not understand. At least, do some research about it with Google/duckduckgo.

>>1561278
Crystals are basically modeled as very high Q RLC networks. Typically a 4-element model consisting of a series RLC circuit in parallel with a shunt capacitance. This particular 8MHz xtal has a resonant frequency of 7.99040MHz and an anti-resonant frequency of 8.009009MHz with a Q of 5,724 which is much higher than the Q of a typical RLC network made out of passive components, even multistage networks. Crystals can actually be used to make very selective filters. Typically in radio receivers.

anyone know of an R2R equivalent circuit driven by open drain outputs? (actually dip switches)

>>1561270
>All LC oscilators have the transistor just to keep "pushing" things arround.
..that's actually a rather poor way of putting it. The transistor is there to provide *gain*, without which the losses inherent in any circuit would mean it'd sit there instead of oscillating; you have to put energy *back into* the circuit.

>>1561278
It might make more sense to you if you look at this diagram.

>>1561282
>how do I stop getting angry when I don't understand stuff and not even the explanations?
Not an electronics question, that's a psychology question.
Do you have ADD/ADHD? Worth asking yourself.
Or perhaps as another Anon suggested, you just need to show a little more maturity instead of getting angry when everything isn't handed to you immediately.

>>1561133
yeah, should work, but resistors are cheaper than diodes, so I might use a 74HC165 instead

>>1561215
that's push-pull, not totem-pole
>0.8mmx0.8mm
>occupy too much realstate
bruh

Just how heat/fire resistant is FR-2 phenolic paper? Like the kind they sometimes make PCBs with. Can I safely build a flame enclosure out of it? Also what are some (cheap) non-flammable adhesives?

>>1561156
I'm reading this book so I can advance my own project (which is making simple noises with analogue circuits, hopefully building into a modular synth). And yeah, right now I'm just trying to get babbies first blinky.

>>1561178
Thanks anon. And yeah I never buy books. I pirate and optionally print (I keep meaning to get into binding but haven't yet).

>>1561282
One solution that worked for me was to expose myself to these situations a lot more. Ultimately, as you know, the only way out of those situations is more careful reading and fuller understanding - which is the last thing you want to do when you're feeling that way (IME).

I would recommend picking up a rigorous math book that's at your level. It will expose you to those situations in abundance so you can powerlevel towards cultivating a more diligent and penetrating mind, and it has a lot of direct crossover to this stuff.

>>1560740
The DC voltages are OK, probably within margin. It is a european amp, so a bit of difference is probably ok. Relative to each other they are very close to the schemaitc. Did not write them down, because they were close. The amp is biased correct, but I tried different settings: Did not change anything. The output tubes are probably not new, but they certainly are not old. I do have a load connected. A self built load box and a 4x12 cabinet. Both of which are proven to work with my other tube amp.

I think that the voltage drop and the output transformer buzz are different issues. Maybe the output transformer is no issue at all. I listened very closely, the sound is only present on certain note ranges and they seemlessy "transform" into lower and higher buzz (distortion) that I cannot hear in the transformer. So I stopped worrying about the OT right now.

What really grinds my gears is the sudden drop in volume. That however, even though it only happens if the amp is overdriving hard, seems to be related to the second gain stage and not the PI.

I alligator clipped a 22nF/630V capacitor from the first gain stages plate to the third gain stages grid while setting the preamp volume to zero. This eliminated the second gain stage from the circuit. Of course I had less gain and distortion, but the sudden volume drop disappears in this configuration. The amp gradually gets quieter, less distorted until no sound can be heard, just the way it should be.

I also chop sticked around the second gain stage and was able to induce some heavy oscillating, high frequent noise which I could not kill in a controlled way via chop sticking.

So I think that there is a problem (it is not the ecc83 tubes, I changed them already) some at pic related

>>1561131
Oh shit. I made that drawing at night while I was tired. The capacitor is supposed to be from the gate to ground. The source is just attached to a NO switch thats in the pot which is then grounded.

>>1561432
hmmmm. the 22n on V1a might be bad, can you easily sub in your tester cap there? do the rails stay up when playing hard? are the V1a plate and cathode resistors in agreement with print values? is the pre volume pot good, and connected to 0V? are the jacks clean?

>>1561432
Hair dryer and scope. If you can reproduce the effect you can locate and fix it.

>>1561459
>can you easily sub in your tester cap
unfortunately, I cannot. The test cap has no legs left :/

>do the rails stay up
will measure this tomorrow. what would the implication be if not? bad filter caps? I already added filter caps by alligator clipping - I did not notice any change in behavior.

>are the V1a plate and cathode resistors in agreement with print values? is the pre volume pot good, and connected to 0V? are the jacks clean?
yes to all

>>1561403
>The SNx4HC165 devices are 8-bit parallel-load shift registers that, when clocked, shift the data toward a serial (QH) output.

Very nice. However, I was just thinking about possible applications for the 4017 outside of typical led chaser/blinker gadgets.
I think the CD4017 would be even more useful if its outputs were open collectors instead of push-pull. That would save a lot of diodes in certain applications. Does any company produce such counter IC?

>>1560823
Output stage now in the feedback path.

>>1561463
got no scope, but why a hair dryer? what kind of fixing tactic is this? I don't get it.

This is only my second tube amp ever (even if I had played this kind of marshall amp before...) and the first for me to actually have to tinker with, sorry for not understanding everything.

>>1561470
Many unstable faults are related to a change of temperature.

>>1561474
ok, I will carefully check that out

>Back to Basics Edition

>>1561432
Your drop in output when playing loudly, which returns as the note fades, sounds like blocking distortion. When a tube stage becomes overdriven to the point that lots of grid current flows, the input coupling cap of that stage can develop additional voltage across it that biases the tube nearer to cutoff.
If you're comfortable with a multimeter, measure each grid voltage to ground while playing softly, then while playing loudly enough to cause the problem. Small variations in DC voltage are ok (volt or maybe two), but much more would indicate a problem.
As for the buzz, I had a Marshall (JCM 900) that would buzz when certain notes were played. It turned out that one of the socket's pin grabbers had sprung a bit, and was mechanically resonant at certain frequencies. I did a little bending with a jeweler's screwdriver, and never had the problem again.

>>1561555
At first the input section looks odd, what drives the second transistor? Then you discover the creative use of the base-collector junction and all's well.

what're the job prospects for Electrical Engineering? Computer Engineering?

>>1561457
It sounds like your circuit could be more efficient. By putting a capacitor from the gate to ground without a resistor before it you're just wasting power, you could get away with a 10nF cap with an appropriately sized resistor.

Hi /ohm/, I was wondering if you guys can give me some tips on how to layout a 2 layer PCB for a university project

>Building a line following robot
>Asked by team to design the line sensor board
>6 Vishay TCRT5000 sensors, analog signal back to the STM micro
>ULN2003 darlington array to turn the IR LED on the sensor on/off
>Only allowed a 2 layer PCB

Is there anything I need to consider when laying out this board? So far I've tried autorouting it and it looks messy as fuck and I'm worried about noise affecting the sensor readings.

>>1561573
In Canada there is about 1 new EE job created for every 3 graduates every year, and this number is expected to go down. Most of the created jobs result from retiring engineers, so they`re not entry level jobs. Personally I`ve applied to 80 jobs to find a co-op internship, didn`t get a single interview. My grades are average (technically slightly above) compared to the rest of my class.

Anyone who I know that has gotten jobs was because they were top of the class, for the average student it was through `networking` (read: nepotism). I`m graduating this time next year, and I`m dreading it, I deeply regret going into engineering with every fiber of my being.
The takeaway is that if you want a technical profession, don`t want to spend your life with socializing, the office politics and networking that most `normie` jobs entail, and aren`t especially gifted, then don`t bother, keep it as a hobby.
Don`t know about other countries, but since we import about 880 new electrical engineers every year when we don`t even have enough jobs for our own graduates, I`m guessing it`s not too good. The field is over-saturated.

>>1561578
Also the 80 jobs figure doesn`t include job fairs and giving my resume to some guy who knows a guy at a local firm and will `have them look at it`, those are just the amount of applications I`ve filled out and kept track of.

>>1561555
now THAT's a totem pole

>>1561577
1. autorouters aren't very smart
2. ground and/or power polygon(s)
3. if noise is a concern, consider voltage followers at each sensor
4. consider SMD. SOT23, SOIC, and 0805 components aren't that hard to solder
5. vias are usually free on prototype boards, within reason. don't be afraid to use them but don't go too overboard with them either
6. find out what your pcb fab's minimum trace/space are, and set your pcb editor's default trace/space to 150-200% of that minimum. high current traces (which you don't appear to have) should be wider of course (maybe 300-400% of minimum), and if you have to squeeze a trace between two pads somewhere you might neck down to minimum for that section of trace. the wider the traces, the less likely defects
7. if the rest of your team won't bitch, don't be afraid to swap/shift connector pins to make routing easier

Where i can find schematic for 1 wire field telephone made from modern and repurposed components? Under 1 km and with call/ring feature?

>>1561577
Ground fill everywhere, decoupling caps close to the micro and motors, and in general keep the traces as short as possible, prioritising ground and power traces, followed by sensor traces.

But what the fuck do you need a microcontroller for driving a line-following robot? Surely 2-3 sensors, a dual op-amp or two, some passives, two power transistors, and geared brushed DC motors is sufficient, no? Is there a reason you have to do this digitally and with so many sensors?

>>1561595
thanks anon

>4. consider SMD. SOT23, SOIC, and 0805 components aren't that hard to solder
interesting, haven't done SMD before but willing to learn. only downside is that the department might not stock the SMD components I want and I'll have to order them out of the small team budget

>7. if the rest of your team won't bitch, don't be afraid to swap/shift connector pins to make routing easier
fuck the team, I've got some chink bitch on my team who this she's hot shit but produces absolute garbage.

>>1561599
>But what the fuck do you need a microcontroller for driving a line-following robot? Surely 2-3 sensors, a dual op-amp or two, some passives, two power transistors, and geared brushed DC motors is sufficient, no? Is there a reason you have to do this digitally and with so many sensors?
That's a good point. I think the project was designed by the department to tie in a bunch of knowledge from different modules (c++/embdded systems dev, analog circuit design etc.).

Has a BLE module as a requirement for robot this year as well.

I think I'll come back in a few days once I have a board design going.

>>1561596

>>1561602
>I think the project was designed by the department to tie in a bunch of knowledge from different modules
Ah, that makes sense. I hope you make it as well as Willam Osman's delivery robots:
https://youtu.be/JUJm2j4E5MQ

Consider buying one of those SMD resistor booklets.

hello /ohm/

I'm a martial artist and I'd like to quantify the force output from strikes using fists, elbows, knees, shins, and various blunt weapons.
I'd like to see force output over time like in this graph, and preferably be able to display it on my computer (linux). do any of you have an idea of how I would go about doing this?

I'm going to strike a semi-solid surface (soft enough to not get too hurt from repeated strikes), account for the force loss through compression of the material, and then initially assumed I would use some kind of mechanical force sensor pick it up and relay it.

one study uses multiple accelerometers embedded in a punching bag. some uses "impact force sensors", and some recommends using piezoelectric pressure sensors. I've considered using a "strain gauge" connected to a voltmeter, as well as filming the movement of fixed points on the strike surface during the strike.

what do you think would be a good compromise between cost, design complexity, and accuracy?

>>1561620
It depends on what you want to measure. If you just want a force-time graph that you can integrate to get impulse then you can hang a normal punching bag from a long rope or put it on linear rails and just have a single accelerometer. F_punch = m_bag*a_sensor and you have your data, and accelerometers and MCUs are cheap as chips. The easiest method would be an arduino + accelerometer module + whatever code library it runs off and just starting a serial plotter on the computer the arduino is plugged into. Copy-paste that into excel and integrate over all the data with respect to time to figure out the total impulse of your strike. It would be quite interesting to check out how this value changes with respect to the length and mass of a blunt weapon.

But a punch or strike also contains information about area, i.e. what pressure you're exerting, and this area will vary depending on whether you strike with your leg or your fist or your palm. To measure this you'd likely need a large strain gauge array, which sounds like a complete pain to arrange, so if you don't need this area data I'd definitely avoid this method. You can also calculate the impact area later on for each implement with a pigment or dye or simple calculations.

>>1561577
Autorouting is shit unless you really know what you're doing with it. Until you've got hundereds of connections to deal with it's usually easier to route everything by hand. Component placement matters, try to keep things in a logical order with the important signals taking a reasonably short path. Try to keep one side mostly ground plane, sprinkle vias around to keep top ground and bottom ground connected.

I have a lot of housefire 18650 batteries from hoverboard laying around...
Can I make spot welder for 18650 batteries out of them? How much amps do I need in order to weld nickel (or copper) strip?

>>1561653
copper tape+hot glue

>>1561578
>>1561580
man I just want to make zx spectrum clones and make synthesizers why does life gotta suck so hard

>>1561602
the nice thing about SMD is not taking up space on both sides of the board. you could fill one whole side with ground and interrupt it (sparingly) with jumpers. that's good for shielding, too, in case you are in an electrically noisy environment

>>1561689
by all means, do. a lack of a degree is not going to bring police smashing your door down and confiscating your soldering iron as contraband
>guarantee does not apply in yurop when using SnPb solder kek

>>1561653
>shorting Li+ batteries intentionally
genius

>>1561689
>zx clones
How are you going to clone the ULA's?

>>1561653
I'd look up a dedicated tutorial for such a thing before coming back. It's definitely possible, but I'd argue that there are probably better things to do with li-ions considering you can make a wall-powered spot welder that doesn't require all that god-awful balancing and charging circuitry.

>>1561703
not that anon but just guessing
https://www.mike-stirling.com/retro-fpga/zx-spectrum-on-an-fpga/

>>1561710
Or a ton of 74 series.

>>1561466
not a counter, but there are open-drain 'HC595-like devices for light-duty power applications

>>1561751
y tho

>>1561710
>>1561751
Isn't there some more modern, all-in-one computer IC you can slap on a board like on a raspberry pi and still keep a barebones <16kB OS on it? Because I looked up the circuit diagram of a zx spectrum and it looks like there's room for improvement complexity-wise.

>>1561770
>modern
>all-in-one
>barebones
pick one, maybe two
also did you even read the link

>>1561772
Barebones is a function of the OS. Surely you could boot an OS that light on an RPi with a little work, no? If not <15k then perhaps <1M?

>>1561775
depends on how much of the hardware you want to use and what amenities you expect that OS to offer. USB? file system(s)? IP networking? visual editor? windowing system? a built-in programming language? with decent libraries?
that said, there are boot environments that come close to offering all the user-facing services of an 8-bit PC with ROM BASIC and accept and run programs from the keyboard, after a fashion (see OpenFirmware)

>>1560823
cheap acelorometer? cheap load cell (50kg)?
>how to not get scammed by the chinks

>>1561784
>cheap acelorometer?

>>1561810
accelerometer

>>1561784
> acelorometer

>>1561814
doesn't get much cheaper than a pair of fuzzy dice

>>1561639
thanks. I'll look into this.
I'll probably just graph with pyqtgraph and do various calcs in python.

>>1561688
No, I'd just solder, but I don't want to solder good cells.
>>1561700
>genius
Yes.
>>1561708
But I don't have any spare microwave transformers or loads of capacitors to make it wall-powered.
200A is enough to weld?

Anyone have sources on building a desktop CNC router without buying one of those expensive kits?
>EE fag
>need to make mechanical parts
>need mechanical parts to make thing that makes mechanical parts
pls

>l298 break out boards are cheaper than a single l298
hmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm

>>1561851
depends what you want to build.
there are plenty of little machininst tricks you can use to help you out.
also you can bootstrap by hand and use your shit machine to make parts better than you can by hand. and keep going.

>>1561880
>there are plenty of little machininst tricks you can use to help you out.
Don't hold back friend, tell me your secrets.
>also you can bootstrap by hand and use your shit machine to make parts better than you can by hand. and keep going.
what do you mean? I have conditions to make a shit stepper driver with discrete parts and some cd steppers here.

>>1561835
The problem isn't the amps. It's the battery going thermonuclear on you. Lithium stuff isn't really made for very large discharges. See if you can find the safe operating area of the battery in some datasheet. Or it's capacity, divide it by ten and start increasing current from there until you stop feeling safe.

>speakers have really hard to reach switch on the back of them
>turning them on/off whenever the PC gets turned on/off is a right pain in the ass
>subwoofer has blue LED thingy that I don't wanna remove cause it matches the rest of my setup
>this combined with the slight white noise means I don't wanna leave them on overnight

My idea was a relay that replaces the speakers' power switch. Where/how should I tap into the PC to get it working and/or is there a better solution to do the same thing?

>>1561968
The usual 5V or 12V output of the PSU will work

>>1561851
Be advised that most CNC routers won't cut beyond plastic or wood. I found this particularly easy to follow
http://www.cncroutersource.com/homemade-cnc-router.html

>>1561985
Are those not constantly live as long as the PSU is switched on?

Can somebody explain to me applications of spark gaps, other than car ignition?
Also, can a spark gap be modelled as an ideal diode? (IE open circuit under a certain voltage, conducts at that voltage)

Can comeone explain how the right part works? The left part is just a constant current source for the capacitor it seems.

>>1562034
Seems to me like when the capacitor charges, current is able to flow through the top transistor on the right, which then activates both the lower transistors, which prevent current from flowing through the 1Mega Ohm resistor for that short period. Someone can correct me if I'm wrong.

How the fuck do I capture lightnig?
Anybody tried it?
What kind of capacitor should be used?
inb4: a giant one?

>>1562039
you'd probably want a tall antenna and a pulse transformer connected to some pretty hefty capacitors or a fuckton of batteries

>>1562039
the energy spread over time is surprisingly small and not worth the effort

>>1562019
circuit protection, high-voltage triggering
>modelled as an ideal diode
that's not an ideal diode. do an honest web search

>>1562005
only the standby supply (purple wire?) is live at all times

>>1561924
there's a machinists' general

Got me some boost converters. This printer has a 2 cell Li-Ion pack, so 7.4V. The charging port says 12V/1.25A.

I’m going to step up a regular 5V USB charger to the correct voltage and whatever current. My question being, should I boost this thing up to 12V? Or should I set it to 8.0-8.4V like using the CC-CV method on a power supply?

>>1561992
Well with a cheap spindle + a pump for flood cooling you can cut aluminum. Milling stainless steel, however, is going to be a pain in the ass.

It's not too difficult to design something based on the C-Beam, but with linear rails instead of v-wheels. I'd also design plates so that everything could be potted later on with epoxy granite, which supposedly performs better than cast iron.

>>1562078
My main problem is I don't know the names of the mechanical shit and where to buy them.

>>1562085
I think I'll jury rig some shit with those drawer slides.

>>1562085
-Linear rails , starting with MGN or HGW. Ex: MGN12 or MGN15, HGW20
-Ballscrews, ball screw supports, ballscrew nuts. Normally ballscrews come with a nut which you should never take out because the balls will fall out and you won't be able to put them back in
-Stepper motors + stepper motor brackets

That's the basic

>where to buy them
Aliexpress.

>>1562087
That's a good way to start. If it doesn't work you can burn everything since you'll probably use plywood or MDF

>>1562091
>buying metal stuff from china
top kek, I might sell my organs to pay shipping costs. Mail service here sucks.
>>1562093
Prolly gonna roll with sheet metal from cd drives.

>>1562077
Give it what it wants for fuck's sake, it will already have its own charging circuitry onboard that wants 12V and produces constant current.

>>1562077
it doesn't really matter, you just buy shit you don't need. When you actually build something you should care about voltage levels

>>1562122
>Prolly gonna roll with sheet metal from cd drives.
That's not a bad idea if you fill it with epoxy/granite. I have several empty cd rom driver shells, if I cover one of the holes I can bolt the linear rails to it and fill the whole thing. Hmmm.

>>1562122
much of it's aluminium, shouldn't cost you more than 1.5 kidneys

>>1562214
>>1562219
>giving avatarfags what they want

>>1561943
>It's the battery going thermonuclear on you
Whatever. I'm not too stupid to use it indoors.
And I have a fucktonn of those, so battery current shouldn't exceed acceptable limits.
>See if you can find the safe operating area of the battery in some datasheet
They are noname, but they were intended for use in hoverboards, which have 10S2P and they draw around 15A, so around 7.5A for cell.
Also I've seen quality batteries with Samsung ICR-22P, which allow 10A cont. discharge according to datasheet. (which is pretty cool for ICRs)

I think I can use thin (22AWG) copper wire in water as current limiting resistor.

>>1562214
After thinking about it, that’s kinda what I figured. The 8.4V works but it was really slow.

>>1562219
>tfw my amazon delivery of shit that I don’t need was delayed a day and now I’m bummed
But I actually bought those boost converters with a plan in mind. They won’t give us the portable chargers for those printers so I wanted to make one that would either be a power bank or charge off USB.

Another thought I had- if it wants 12V, could I charge the thing straight off a 12V car outlet with a resistor or something? Seemed too simple.

>>1562224
Some people can’t be pleased. I get rid of the trip and there’s still complaints. I understand this is a smoke-free room, but did you see a lit Newport? Can I not have cigs in my pocket on an airplane?

>>1562311
Note car "12V" can be ~12-15V depending on charge state, and can have big spikes above and below that.

When you use a n-mosfet as a logic level translator like this, how much current can it sink/source? Would it be enough to light up some LED's? I have a little board and I'm going to test it tomorrow but I'd like to know from ohm. To me it doesn't seem like it can actually sink or source any more than the devices on either side can.

I've asked this once but I'll ask it again: if I'm going to drive optocouplers (which can be either common anode or common cathode) from a 3.3v mcu and don't want the mcu to source any current at all, what kind of IC's should I look for?

>>1562415
Thinking about the common anode/common cathode part makes me imagine that this needs to be done with a totem-pole style circuit that can either sink or source current, but it needs to take like 3.3v and source 5v

>>1562415
>it doesn't seem like it can actually sink or source any more than the devices on either side can.
that is correct
>either common anode or common cathode
I have yet to see an optocoupler that has common anything on the LED side, but I'm sure someone knows of one
>don't want the mcu to source any current at all
you could just use a 74HCT gate to do it (74HCT14?). the user will still have to set up whether they want to sink or source current in the active state
or add terminals for separate outputs, one for sinking, one for sourcing, Pic related
or you add a layer of optoisolation and give the user an uncommitted phototransistor to do whichever the fuck they want

>be noob
>making small drivers for small motors
>discrete because the proper stepper drivers will take a while to get here
>need a inverting levle shifter stage to make sure the H bridge has no forbiden states
>basic resistor+npn
>it works ok but because of resistor the sink/source capacity ends up being diferent
>no biggie, the fets won't need heatsink because low power
>still feel stupid because I did not even thought of using active load
1 year doing electronic stuff, I feel I'm getting more retarded with each passing day.

>>1562566
Don't get me started on my soldering skills. I'm studying EE, I never saw most my classmates building anything they weren't told to, but still I think I'll feel like an ass going to a job interview
>so uhh I made this thing that spins and blinks

>>1562360
Yeah I know, I was thinking at the very least I could plug the thing in with the engine off so I’m only getting ~12.5V.

But that’s it? Proper size resistor and it would work?

Help me /ohm/ I fear I might be legitimately retarded.
I tried to build a simple coil gun. I bought capacitors for my capacitor bank, an SCR for switching and a bridge rectifier for the charging circuit.
I figured that I'd do a quick proof of concept before making the capacitor bank and wired up the bridge rec straight to the coil, with the power source being the outlet on my wall (230V, 10 amps).
Turns out it doesn't fucking work. Shouldn't 230 volts be enough to create a strong enough magnetic field to at least budge my tiny projectile?

>>1562569
Make one of those smd business cards

>>1562453
>you could just use a 74HCT gate to do it (74HCT14?). the user will still have to set up whether they want to sink or source current in the active state
Yes but they have such a low current sinking/sourcing capability.

I thought pic related would work but it doesn't output more than3.1v

>>1562673
SN74LVC1G34 seems to do what I need.
>power it with 5v
>accepts 3.3v input signal as high
>can sink or source 50ma

>>1562608
Have you measured actual current flowing in the coil? What's your projectile made of?
Do you have a working coil gun plan to build from?

>>1562608
It depends on alot of thing, show us a picture of your setup.

>>1562675
Suggest version SN74LVC1G34YFP because of the 'realstate'.

How well does electrical tape work for shielding? 5V obviously. I'm getting interference and noise on my LCD and I'm not entirely sure where from so I figured I would rap the LCD ribbon with electrical tape perform a insulation layer.

>>1562047
If I make a shit ton of capacitor the size of a shoe box and connect them together, would they be charged in a couple of microseconds? if not how can one achieve this?

>>1562071
>the energy spread over time is surprisingly small and not worth the effort

I know but what if I create a contraption that attracts lightning and its always in the air?
what i wanna say is what if you collect 20-50 lightning strikes in a few days?

>>1562757
I also have kapton tape if that would work better.

>>1562757
What LCD? What "ribbon"? What "interference and noise"?

>>1562739
Impossible to solder by hand and the sot-23 version is small enough already

I just had a thought right now. Isn't BGA technically easier to solder with a hot air gun than QFN whatever, since it already comes with the solder balls?

>drop the chip onto the PCB
>balls make it self-aligning
>heat with hot air station
>it's done

I think this might be the future of diy electronics

>I think this might be the future of diy electronics
It won't. You need solder paste (as for qfn), but you have no way to check that center boll are properly melted (apart having an xray imager at home). It is difficult to have homogenous air temp. You can solder it with both IR preheatong and airgun, but it is very tricky... QFN are much more solderable at home, and easy to control and/or fix.

>>1562771
If you're not skilled enough to solder QFN then BGA isn't for you...
It's only self-centering if you already centered it perfectly (you're gonna need a print on xour PCB to outline the BGA.
Also gonna need flux, because the balls dont bring this with them.
Also no guarantee that all balls soldered correctly. At least with QFN you can SEE if you fucked up.

>>1562764
LCD came with Kite's Circuit Sword for the Gameboy Zero build. Flickering and screen is noticeably more white than before.

>>1562778
Bga's come with the balls on them, so all you need is flux and a way to do the reflow process. No x-ray, just reflow and assume it's good like people do when reballing laptops.

>>1562779
You can't see under the qfn chips either to see if the thermal pad is properly soldered, or if there are bridges between it and the pins. I can solder 0603 with unaided sight, but I don't have a hot air station or reflow oven for qfn.

>>1562757
>>1562762
Neither will work for shielding. mu-metal tape that's properly connected to your circuit ground might work assuming the noise is magnetically coupled in from an external source. mu-metal tape is relatively expensive though.

If it's electrical noise generated by some part of the circuit though or coupled in from the power source, or is because of a ground loop then shielding won't solve your problems. Techniques for noise reduction in such cases include differential signaling, common mode chokes, isolating galvanic isolation of signal lines (by transformer or optocoupler), etc. Most of these solutions are not easy to integrate onto an already assembled and populated PCB.

>>1562757
>I figured I would rap the LCD ribbon
Have you ever opened an inkjet printer? They have ferrite beads on every single ribbon cable

>>1562566
>1 year
babby-tier. you don't git gud until ~5 years in (postgrad work usually doesn't count)
t.20 years hobbying and still don't know half this shit

>>1562771
sure, if you have a hot air station. if you don't, you can sorta kinda QFN with a pencil as long as you can ignore the die paddle OR have a big enough tip and enough vias to heat the die paddle from the back. in either case, paste still helps

>>1562799
>come with the balls on them
whether that's enough solder depends on each pad's area, which will vary slightly between pads (on NSMD boards) based on trace size or presence, solder mask clearance, etc. a bit of solder paste greatly reduces sensitivity to those differences

>>1562852
alri big clive

>>1562760
My guess is you'd need a shitload of capacitors connected in series (which results in a low capacitance, thus quick charging time, while still being able to hold a large amount of energy), as well as minimizing the resistance of the charging circuitry.
The exact values would depends on how much energy the bolt actually delivers, and the power.
Just looking at some average values, you'd need a capacitor of at most 150 microfarad, maybe smaller, and rated for at least one million volts. I think the most realistic way to do this is by connecting a fuckton of capacitor 'cells' in series. Like I said before, even in this case, you'd probably want a pulse transformer to lower to voltage.

I also don't know much about lightning other than what I just googled, so I don't know how much the values would change in practice.

What variable power supplies do you guys use?

I've been looking for months for a good quality used one to come up on kijiji/facebook marketplace but having no luck so far.

Considering one of these (pic) to get me by for small school projects and whatever.

Appreciate any recommendations. I'm in Canada if that matters.

>>1562871
I'd recommend a power supply that is dual channel which will enable you to have split +/- supplies. You can't do that with a single PSU. I'd also suggest getting one with inputs on the front, so not the one in your pic. It's gonna be a pain if you always have to pull the unit out to connect or disconnect cables. As for voltage/current most reasonably cheap ones will cap you at 30V. How much current you're gonna want really depends on the nature of projects you're likely going to do. For most small signal stuff the average person probably will be fine with 3A or less but 5A is pretty common too. One other nice thing about dual channel supplies I forgot to mention is each channel can be wired in series to get double the voltage or can be wired in parallel for double the current in the event you need more of one or the other than any single channel can supply. You can get into other nitty gritty details like course/fine knobs vs. multi-turn pots or rotary encoders and analog moving coil readouts vs. digital readouts and that all really comes down to personal preference and budget. Neither of those things should be a dealbreaker though.

I personally use a GW Instek GPC-3020 which is a dual channel 30V 2A meter (60V 2A w/ channels in series; 30V, 4A channels in parallel) with a dedicated 5V channel as well. It's useful. I also have a cheaper generic ebay 30V 5A PSU but I generally prefer the GW Instek and only use the other one on the rare occasions I need the extra current. It has come up once in prototyping a flyback converter with a large boost ratio which resulted in a relatively high peak primary current in order to get the output power I wanted.

>>1562852
>babby-tier. you don't git gud until ~5 years in (postgrad work usually doesn't count)
How much did you screw up on my stage? Because screwing up and organizing the mess I made is what I'm doing.

>>1561215
The only way your n-channel is permitting current is through the 1k, through the NP junction, and into the 10k or into the 1.8nF. I'm surprised it works at all.

>>1562871
I have an LM317 attached nonconductively to a small heatsink, and a little digital LED voltmeter and pot. I found I do actually have a negative voltage regulator and I should have another voltmeter, so I'm thinking of just repeating the circuit for that. I have an adjustable laptop supply, 12-24V 4A. Doubt I've pushed it pass 1-2. Few wall-warts/SMPS here and there, I want to wire better pins to my raspberry pi 3 and see if that clears up the voltage warnings. If I get to it, I can just bump the voltage slightly on a 5V supply and see what happens.

>>1562894
that's about right. not much later on, I did have a mentor living 5 minutes away. if you don't have one of these, you might try finding one

>>1562871
meh I just grab a shitty little buck or boost module out of the box and feed it with a chinkshit 12V open-frame switcher
for when I need ≤3.3V at up to 1.5A, I built an LP3906 breakout with a micro USB input (really, this was as much for a soldering practice board as anything). I can hook it up to I2C to change output voltages as needed. I keep meaning to try out some power supply glitching attacks with it
I don't usually need - supplies, but when I do, I have Sean Connery bring an ICL7660 module over and in exchange I listen to him bitching about Alex Trebek

I'm the guy asking about making sawtooth waves for pwm things. I was not happy with the stuff indians built with a timer 555 because even with the math saying the frequency doesn't change at all or much, it does vary quite a bit. I guess if you are using a 555 for that you don't really care.

So I decided to put my noggins to work and made this. I'm still testing but managed to go up to 1Mhz just fine with 5 Vpp. I've made it with the LM353 comparator, which is an old fart but managed quite nicely.
In the real circuit, the output passes through a clamp to remove the negative part of the transient, as it messes up the PWM if you use single supply (as I am doing).

>Timer 555 PWM
>kinda kinky routing because apparently whoever made the 555 thought those pins were a good idea
>can't reliably go to 100% or 0% duty cycle (usually one or the other)
>frequency varies quite a bit

>Analog discrete version
>frequency set by current and capacitor, will change with temperature but not a lot. (-2.2uA/deg for a 1k resistor will not change the charging time by much I think)
>goes from 0% to 100% reliably (If I have problems with this I'll use a opamp peak detector)
>same ammount of solder
>routing is less idiotic
>same dip 8 package

>cons
>open collector output limits output current by resistor
>planning to do a push-pull output so I'll need a buffer or something, increasing part count

I'll make a permanent version when I can and post. This is just a random "tester" I'll use for motors and power converters (The motors work fine with the shitty 555 versions I've made, but with power stuff the changing frequency really mambo your jambos). Anyways, end of my blog.
>tl;dr
>analog stuff is cool, try it out

>>1562934
I found a lot of my frequency drift was because of poor power supply regulation. On the other hand, charge/discharge is not very 1:1, so I had more trouble with getting a 50:50 duty cycle.I did the simple thing and used two diodes+two pots with the charge path coming from the output. You're really adjusting t_on/t_off when you change the pots, but I also can socket different caps to make gross adjustments to the frequency. I used 100k pot+diodes in series with a 10k resistor. Could've probably gone higher, and I'd get decent resolution but I didn't have anything between 100n and 1uF.

>>1562942
I was using one of those chonky 5V regs with all the caps and stuff, I don't think power supply was the problem.

>>1562942
your analog circuitry needs decoupling too, especially where rapid output changes happen. did you decouple that Schmitt buffer and the op amp? the current source?
also, that discharge transistor might have caused ground to bounce unless the transistor was connected directly at the cap. 1.2uF is kinda big
since you have a second op amp in that LF353 package, I'd probably have used the classical integrator + comparator (or Schmitt) triangle generator, as in the Falstad circuit library

PCBgogo ads have been popping up on Youtube as of lately.

They're offering PCBA for 50$, which doesn't seem to be expensive when you think about assembling even 10 boards by hand.

What do you guys think?
https://www.youtube.com/watch?v=AIgPpNTXrig

>check schematic
>over 20 SMD resistors
how much are those vacuum component holders again?

>>1563159
Fine tip tweezers are much better.

>>1563159
can confirm, vacuum holders aren't so great for passives, there's only so small you can make a suction cup at chinkshit prices. go down to your local beauty supply shop and get some pro-level hair plucking tweezers, if you don't want to wait for mistah ali

I am trying to revive a dead laptop battery by replacing the 18650 cells inside.
The original ones had contact plates that were spot welded onto them or something like that. I managed to pry them off without damaging them too much.
Now I want to attach the new batteries. Is soldering safe? Do I have a chance to damage the battery while trying to solder them?

>>1562982
The guy you made the post you replied to wasn't me >>1562934
The lm353 is a comparator, everything is decoupled, I did not put everything inthe falstad sim, just wanted to show the idea.

Thanks for the feedback anyway

>>1563159
Buy stainless steel surgical tweezers. Theya re cheaper and stronger than the ones they sell for electronics

so I am a hopeless DIYer but also a compulsive tinkerer and my projects always go wrong

I took apart this shitty 1970s Sony stereo receiver (just like the one in the picture) and fucked it up royally

details:
>one channel quiet/intermittent
>swear I'm only going to open it up, spray some contact cleaner on the pots, and put the top back on
>notice the front perspex is loose and there's dust in the display
>take the front panel off to glue the perspex back on
>do a horrible job of it
>blow the dust out
>try to get access to some of the blown lightbulbs
>accidentally knock the tuner dial cord/string off its pulleys
>no idea how the hell to put it back on
>struggle for hours, can't get it quite right
>want to cry and pour beer in the fucking thing

okay, who's had this shit happen before? what can I do?

How bad of an idea is to add an output choke or a LC filter with very low cut off frequency on my PSU?

How come an SSR can only be used for only AC or only DC?
Normal relay can be used for anything

>>1563248
Some use thyristors and other fets

>>1563248
>How come an SSR can only be used for only AC or only DC?
https://www.allaboutcircuits.com/technical-articles/basics-of-ssr-solid-state-relay-the-switching-device/
explains it all
You can get AC&DC mosfets too

quick question
i have purchased this arduino relay pair, but the problem is that the linear regulator on my esp32 is shit and when the wifi turns on, the relays do CLICK CLACK sound because the there is not enough power left to keep them on while the wifi module is doing stuff.

So I was thinking, the esp32 is powered by a 5V wall wart, i could hookup the GND and VIN pin on the relay directly to the wart so it has enough power, BUT will I still be save to connect the two logical pins on the relay to my esp32 which operates on 3V3 logic without frying it?

>>1563269
check what these are, if they are optocouplers the inputs are physically isolated from the rest of the relays.

>>1563272
They have 4 legs and there is text for ants on them that reads dp 817c 1812
Google seems to think they are octocouplers but what does that mean for me?

The input and VCC grounds must be common (only one ground wire is used) so the relay power can't be totally isolated from the logic power

>>1563275
>can't be totally isolated
It can. Get the circuit diagram.
Remove the jumper and connect +5V to the pin JD-VCC.
These things are active low, so if VCC is only 3V3 you have 1 IR diode + 1 LED + a resistor in series. Ground an input to test its function.

>>1563275
>vcc vcc gnd relay side
>gnd in1 in 2 vcc side
>logic side
If you use the same ground the MCU will not be electrically isolated, but that will not cause problems for what you are doing.

>>1563281
>Get the circuit diagram.
this is from ping pong store, so no such luck

so if i understand this correctly,
vcc goes into jd-vcc through a jumper and then into the relay coil, so by removing the jumper and powering the jd-vcc from 5V, the 3pins on the left and 4on the right become completely isolated from each other right?
So does that mean the vcc pin and the gnd pin on the right now don't need to be connected? or do i still need to keep the connected to esp32 so they can power the logic circuit in the relay?

>>1563284
Try to see with a flashlight or something the area under the optocouplers. The board probably will be clealy have two portions that do not touch ever. If that happens, they are isolated, if that does not happen, they wasted two couplers. Please draw what you are asking because at this point I have no idea of what you are saying.

>>1563288
Not sure if it helps but it has this on the bottom

>>1563219
Have tried to solder them but the damn solder just falls off the cell's terminals
Wat do now?

>>1563290
jesus m8, it's literally drawn out for you. Go to arduino general or something if you don't understand, prolly someone there has used those

>>1563269
take that jumper off the jd-vcc and vcc and you're set

>>1563292
Use plumbing flux.
Don't overdo tho.

>>1563305
Just small dab would be enough to tin nickel-plated battery.

>>1563284
Here you can see why VCC=3.3V may or may not work. If not, short the LED.

What is the proper name for this part? I need to buy more of them but don't know what they are called.
It is the thing that plugs into EU outlet.
If i seach for "eu plug" i just get results for adapters from burger power cables to eu outlets

>>1563315
https://www.worldstandards.eu/electricity/plugs-and-sockets/c/

>>1563316
I know that, but searching for power plugs like that just ends up in results for like power cables adn stuff like that, i am looking for something i can't embed into a 3d printed case, like the pic above, so i can plug that case directly into an outlet, without having to to attach a redundant power cable to it

>>1563317
>can't
can

>>1563266
>You can get AC&DC mosfets too
I meant to say mosfet ssrs. Surprised nobody autistically pounced on that little mistake.

>>1562859
thanks for the info
I'll look further into the topic

>>1563313
well the relay works on 3.3v i already tested that it's just i need the external juice since the power supply on the esp is such a piece of crap. or maybe it could be the usb port too
but that schematic makes it clear it should be safe to use the external power with the jumper removed. hopefully connecting the 5v and 3v grounds won't do anything bad

>>1563319
Mosfets are bi-directional :^)

>>1563221
>>1563198
>>1563163
will do thanks

>>1560823

Sorry if it doesn't entirely belong, fampaitachi, but I suppose you lot do the most soldering out of anyone out here...
Anyway, does the Hakko FX-888D iron have a silicone cable?
I couldn't find out ANYWHERE. I do have some really nice and flexible silicone 5-core in case it didn't, but replacing it would void the warranty, and it's not some chinkshit station where I wouldn't give a shit about something like that...
It kind of looks like plain old PVC but why the fuck would anyone put a PVC cable on an iron that costs $100 alone?

>>1563219
>Do I have a chance to damage the battery while trying to solder them?
yes

>>1563219
take a car battery and rig yourself a spot welder instead

>>1563404
Okay, so what other options do I have?
Ideally I want to keep the small contact plates that were welded on the old cells.
I have already performed soldering trials with the old cells and the solder doesn't stick to them, so that option is out.

>>1563408
Wouldn't that actually heat up the cells more than soldering?
Also how the fuck do you even spot weld on a cell? I thought you needed to have the spot you want to weld between the 2 electrodes.

>>1563416
Welder is faster, thus it doesn't heat battery much.
This is why I want to make a welder of shitty 18650 to weld good 18650s with it.

>>1562771
..no, it's not that simple.
You need the proper equipment, not just 'a hot air gun'.
For any given PCB you need a heating profile that will get everything hot enough without destroying parts or the PCB itself.
You have to have a template to apply solder paste to.
Since you can't see what's going on underneath the component you have no way of telling if all the balls are adhering properly or if there's solder bridges between balls; they have very expensive X-ray machines for that and even then they're not 100%.
If you screw up (bigger chance than you think) then removing the part may destroy the part, the board, or both. This isn't as much of a problem in a mass-production environment since they allow for defective units but if you are making 1 PCB you have to have 100% success and you just can't with BGAs.
For these reasons BGA designs are pretty much out of reach for the hobbyist.

>>1563436
BGA needs some practice for sure, but I don't think it is impossible, but I'd not use it, unless it is significantly cheaper

>>1563315
I've found this
https://www.aliexpress.com/item/dragon-Dildo-king-size/32909658646.html
https://www.aliexpress.com/item/EU-dildo/32968184652.html
But beware, those aren't compliant at all.
You better remove your ass from seat and go to nearest hardware store thing and buy a plug.

>>1563416
>Okay, so what other options do I have?
google: diy battery spot welder

>>1563453
some parts just can't be gotten in friendlier packages, sadly, thta guy's Pic related being a prime example which I happen to be working on right now. routing DDR2 on 4 layers is a bitch, folks

>>1563436
none of that seems to bother the guy at the local cell phone repair shop
there are people whose hobby is to climb into machines worth tens or hundreds of thousands of burgers and send them through the air. to into BGA, one merely needs to be a bit higher level of a hobbyist, which means being a little less risk-averse and willing to expend some sacrificial material for testing

>>1562893
Thanks for the response. I've used this exact psu at school quite a bit and its nice. It's also for sale for $350 CAD used right now but its kind of out of my price range. I might skip the $100 Chinese psus and save more.

>>1562901
Lol ya, I've been getting by with an lm317 on a breadboard/wallwarts. Getting tired of it though.

>>1563465
>google
>click first video
I had no idea that was how they got that shit onto batteries. Also was pretty cool that the guy was doing it with a motorbike battery
>bust out a UPS unit
>remove battery inside
>connect 2.5mm solid copper wire for instant spot welder
Not the most elegant solution but it got the job done. I got about half of the cells welded nicely. For the rest things didn't go so well because those contact plates don't seem to like being rewelded. Thank god the battery pack case is tight and that keeps the plates in contact with the cell poles.

>>1563315
https://www.tme.eu/en/katalog/#id_category=113600&search=plug&s_field=accuracy&s_order=DESC&visible_params=2%2C6%2C613%2C5%2C77%2C412%2C18%2C418%2C419%2C7%2C2555%2C9%2C1322%2C82%2C1247%2C416%2C1182
Pick your poison

>>1562934
Working dandy at 200khz
>ran out of 5v regulators, had to use lm317 and that wasted a little bit more of space
>memorized the lm 393 pinout wrong and wasted an hour finding the error
>500mA push pull output

>>1561282
Listen/Read the expliantion 3 times in a row
deconstruct each step/component/part in a written format
Go take a walk while thinking about it
Sit back down and try again
if it fails, sleep on it, tackle it first thing the next morning

works 99% of the time.

sometimes the explanations are shit or incomplete and you need to synthesize an answer by looking at several different sources.

>>1563566
can confirm, practical electronics is 1/2 physics, 1/2 library science

I want to make an electric kick scooter battery.
What is safer: LiFePO4, or typical 18650 ICRs?
ECU gives zero fuck about input voltage.

ICRs have 5A discharge current, so I'd use at least 3 in parallel, while LiFePO4 has like 70A discharge, which is scary (and I don't believe honestly, since it is aliexpress).

>>1563577
I'd go with the LiFePO4s. Safety is hit or miss, unless you bought batteries with integrated protection chips. I'd choose the LiFePO4s because their charge profile is more consistent over time. The 18650s would decay pretty quickly over many uses. At the size of "scooter" you'll need big batteries. I think although the Fe's might charge slower, it gets them even better life. Lead-acids were always a thing for scooters, and if you're DIY-ing it, you could put in gel batteries or whatever, stuff that lasts better over time and deeper discharges.

I don't frequent these threads at all but I'm needy and I have retard question anyway because I know nothing about EE. Am I going to brick a device that runs on a regular 9v battery by hooking it up to 9v 1a dc adapter?

No. They're voltage source driven devices, and you're using voltage sources.

>>1563660
>>1563662
You.

Technically, a 9V wall-wart will have a higher voltage, perhaps 10.5V, but an alkaline 9V could also be over 10 at full charge. Unless it was like a 9V hearing air from tiny watch batteries and you have a 9V 10A supply, I wouldn't be concerned.

>>1563663
PEACHY.

>>1563514
>>bust out a UPS unit
>>remove battery inside
>>connect 2.5mm solid copper wire for instant spot welder
This is what /diy/ should be instead of buried shipping containers full of imitation crab me.

Glad it worked out so easily for you.

>>1563514
>>bust out a UPS unit
>>remove battery inside
>>connect 2.5mm solid copper wire for instant spot welder
This is what /diy/ should be instead of buried shipping containers full of imitation crab meat.

Glad it worked out so easily for you.

Newfag here. Got a Corsair CX 600m power supply. Anyone see anything obvious in the pic that looks wrong/ out of place? Probed the 24 pin with a dmm, no voltage from anywhere there.
When jumped, It typically turns the fan on for like 0.75 seconds and stops.
My immediate thought is bad caps, but they seem fine. I'll have to probe the big ones.
My school has an O-scope that I can use freely, just haven't used it before.

>>1560823
>designing a cheap amplifier
>not bootstrapping the VAS
but then
>SIGNAL-LEVEL TRANSISTORS IN OUTPUT STAGE WITH AN 8 OHM LOAD
>N O G L O B A L F E E D B A C K
shiggity diggity holy fuck

>>1563677
Wasn't that like 5 years ago? Didn't they spoil by now?

>>1563640
>Safety is hit or miss, unless you bought batteries with integrated protection chips.
I'd have a BMS with balancer anyway.
>I'd choose the LiFePO4s because their charge profile is more consistent over time.
Hm... Typical Li-ion is annoying on low charge, since it goes like really slow, yet 30% of power is still there.
>At the size of "scooter" you'll need big batteries.
>lead acids
It had 10S2P 18650 original LG batteries, but 6000 mAh is too little.
Official extended battery consts an arm and a hand, so I bought universal ECU...
Putting different cells in parallel is sketchy, so I decided to make a new battery.

>>1563710
I had a power supply go the same way I think. In my case it was feedback from the psu that took it out, or the motherboard went when the psu failed 'silently'. I've encountered a power supply before where a 5V regulator failed. I'd imagine it would auto-shutdown the PSU if an overcurrent was detected.If you hold the power and the fans keep spinning, you can let it sit there a min and then disconnect it and see if any of the regulators are hot. I don't know if the schottkys closest to the output have high voltage (they probably don't), so just keep your fingers away from the HV side, or discharge them quickly by burning up 2 moderate kohm resistors.

>>1563728
You could put 3x 12V batteries in series. Ironically, although the charger should certainly be changed, 13-14.4V per 12V lead-acid is an ideal "charge whatever" voltage, though certainly less current is preferable. 15V 2A is recommended for actual charging, but you could just drop 3-4 volts from 48 with a linear regulator and have good charging on 3 batteries. You could get 12V 7Ah batteries for relatively cheap too: https://www.amazon.com/BATTERY-REPLACEMENT-ENDURING-6-DW-7-UB1270/dp/B007ROUBYC/ref=pd_cp_23_1

>>1563729
The first mobo that I paired with had some pretty fucky circuitry, so I wouldn't be surprised if it overloaded the psu in some way. I'll look into your advice.

>>1561469
Shouldn't the first-stage gain be TRASHED because you're bypassing Ro by having C3->C6->GND act as an AC-path short.

>>1563727
that meme is shelf-stable

>>1563364
> It kind of looks like plain old PVC but why the fuck would anyone put a PVC cable on an iron that costs $100 alone?
Because people will still buy it, just as you did.

>>1563721
>autism

How do I properly design multi-stage amplifiers where each stage is direct coupled to the next?

I wanna do differential front end into class A gain stage into class AB power output stage but I've been having trouble designing and biasing it right.

god bless oshpark for their board verification. if only they weren't so purple.

>>1563501
I'm a bit surprised it's that expensive. I got mine for something around $70 I think. I got it on eBay and I believe it was refurbished but it was tested and confirmed working and I've never had any issues.

>>1563736
Not every capacitor acts as 'AC-path short'. R4||C3 sets the upper frequency above which the gain is reduced (about 5kHz), typical for a voice amp. If you want 10kHz, reduce C3 to 15nF.

Where can I read more about power electronics and smps and regulators? Google only gives me indian people

>>1563721
2N2222/2N2907 is on the edge of medium power. Ic is rated at 1A so should drive that speaker just fine on the shitty 5V supply called for
but yes to all the rest

>>1563780
ayyy longboard

>>1563826
semiconductor designers' app notes
or, Switching Power Supplies by the late Abraham Pressman et al.

What's some quality motor driver for brushless "DC" motors?
I mean something that won't give me hell for troubleshooting. I don't care so much about price.

>>1563830
What sort of motor? What sort of application?

>>1563832
Not decided yet on the exact motor, but not so much power. 100W an overestimate.
The application is some kind of force controlled claw. End goal is to do it without force sensor in a first time (only encoder), then without encoder by using the hall effect sensor.

>>1563771
You forgot to post your circuit diagram.

>>1563845
This is an old rev that worked well in sim but had really bad crossover distortion when I actually built the circuit IRL

>>1563845
>>1563849
Ignore the disconnect between the sig source and C4. That should be connected obviously idk when that got deleted. I'd assume most people would be smart enough to realize that but I know some asshat is gonna point that out.

>>1563836
You need to pick a motor and then a driver, not the other way arround.

>run out of relays and dupont cables
>only even bought them from china but don't want to wait a month so go to a local store

>one logic level ready relay $4
>one dupont cable $0.5

holy jeesus fuck, is this what normies pay all the time when doing projects?
this is insane! for the same money i can have 10 relays from china and 7 dupont cables

fucking thiefs, imagine what insane mark ups they are putting on those for that sort of pricing, specially when they get them even cheaper in bulk from the same chinks i do

i hope they go all out of business

If a power source turns AC into DC and it only has two AC wires coming into it, it always means it is isolated right?

>>1563763
It doesn't melt at 400°C, so no, it's not PVC.

>>1563886
No, it could have just a bridge rectifier and a few capacitors. If it has a transformer then it's always isolated.

I'm designing a arduino "shield" for an educational robotics kit.

Since it uses RJ12 (6p6c) cables for both motors and sensors, it's possible to do certain wrong things, for example:

- Connecting a motor into a sensor jack and vice-versa. I've dealt with this by making the motor use pins 5 and 6 of the jack, and the sensors and sensor plugs have the 5 pin not connected to anything, that way a sensor never closes any kind of circuit when connected to the motor jack, and neither does the motor when connected to the sensor jack. Same when someone tries to use connect a cable between a motor output and a sensor input.

Now, what if someone connects a cable between two adjacent rj jacks? the 5V and GND would be tied, no problem, but one pin could be an output set to high, and the other one could be an output set to low, which would cause a short. If I add a 150 ohm resistor to each signal line, then one pin would be sourcing 16 miliamps and the other one sinking 16 miliamps, which is OK, but exactly what would happen to signal from analog sensors with this 150 ohm resistor in series?

>>1563916
Forgot pic related

alright, retard here whos been shocked too many times. I have a toroidal transformer. I hooked up AC from the wall to the primary side and nearly burned out my eye sockets. what is the procedure for connectinv these things? if the primary is just a long wire, wouldnt it just short out and light shit up again?

save the safety talk for another time. I learn better this way.

>>1563881
It gets worse
>shop that does sell parts
>consumers are dropping because they can get them 1/10th the cost overseas, should we reduce cost or sell in larger quantities to compete?
>nah just raise the price so we bring in more money
???????

>>1563927
>if the primary is just a long wire, wouldnt it just short out and light shit up again?
It essentially DOES short for a moment until the core gets magnetized, that's the inrush current. Afterwards the current drawn by primary from mains is dictated by the current drawn by whatever you put on secondary, and limited by the size of your transformer - if a 230/12V step down transformer can deliver 10A with shorted secondary, then the current through primary will be limited to less than an amp, depending on the efficiency and other factors.

>>1563932
thanks anon, that really helps me understand. will the inrush current be enough to make sparks or trip the breaker, or is it no big deal

>>1563934
>will the inrush current be enough to make sparks or trip the breaker
It might, depending on the size of the transformer and your breaker. But most likely not if there is no load on secondary. Generally, the inrush current can be an order of magnitude higher than what the transformer is rated for. NTC thermistors are used to limit inrush current, or I guess you could connect it through a strong enough power resistor and then bridge it.

>>1563934
>>1563934
Depends on the size of the transformer. Commercial large ones use control systems to only connect it near zero crossings.

Are you using one that was made in a factory or you wound up yourself? Because transformers need A LOT of turns, if you don't make enough, the ratio might be correct but the coil impedance is very low and makes magic smoke.
>>1563929
Here in brazil all business do that because of greed and culture, even before china was a thing. It really sucks.

Which one of these of this dryer wires is ground? I believe it's the middle but I'm asking just to be sure

>>1564050
you can easily check for continuity with a multimeter

>>1564050
>wires is ground? I believe it's the middle
The middle wire is connected to the L but it may not be 'ground'.
It's possibly wired to Neutral and not connected to 'Ground'

>>1564053
This, but with a 9V battery and your tongue.

>>1564070
Or with the outlet.
If you have two phases, the one which doesn't cause you a fibrilation is earth.

>watch most big clive videos because I learn a lot when he rambles, and chinkshit is amazing
>several about those things you connect to your car
>he talks about the "canvas" network
>search far and wide about that protocol
>no dice
>only today I discovered it's actually Can bus
Jesus fuck anglos, get your shit together.

>>1564075
>he doesn't have thicc skin that can stand testing live wires
top kek, go back to arduino general fag

>>1564077
Try it with your tongue. I can get zapped with 127v and remain conscious most of the times

>>1563733
No, that shit is too heavy.

I lit up an LED strip using a 12V 1A DC supply. Then I switched to a 12V 2A AC supply and now it flashes on and off. Is it supposed to do that with an AC supply? How do I stop it so it's on all the time?

>>1564100
You'd rectify it and filter it with a cap, but if it's 12V RMS then that will put it at 17V, which will kill the LED strip. In fact by putting such a 12VAC supply across it you're probably overcurrenting the thing even without rectification. Basically, get a new DC supply or deal with the lower power of the old one.

>>1563849
>3 diodes
>between two e-b diodes
R2 looks 50% too high, did you try adjusting it for reasonable quiescent current vs. crossover distortion?

>>1563917
about 0.3% inaccuracy, depending on the sensor

>>1564076
it's the cannabus network, he was just out of weed

>>1564050
>>>/qtddtot/

>>1564077
this, next OP please change the "mains" verbage to "home improvement" or "houses & appliances"

For a capacitive spot welder, is it better to use higher tensions instead of buying super caps?

>>1564179
High voltage is beneficial, since you can use powerful thyristors, which are cheap. By adjusting voltage you can adjust chooch factor on weld.
Supercaps... Why not get a lead-acid battery for this purpose?

>>1564184
Because I have several 70v 2200uF caps and zero lead acid batteries. Also
>space

>>1564179
uh, sounds like an impedance matching problem

>>1564148
It's not two b-e junctions though. Look carefully, the top is a Darlington pair the bottom is a Sziklai pair. This was done because I have no complimentary NPN-PNP pairs of power transistors in my parts boxs so I wanted both output power transistors in the push-pull arrangement to be NPN. This is called a quasi-complimentary class AB as far as I know.

>>1564234
fug, that'll teach me to design review on limited sleep

>>1560823
goddamnit I think I fudged a couple op amps; is there a reliable way to know when they're fried?
or alternatively are these fault indicators
>if the inputs are > a diode drop (when they're supposed to be internally clamped)
>output is 0 when v+ > v-

>>1564247
Build a follower circuit by connecting output and neg together. See if it follows.

>>1564247
what >>1564253 said but there are some other tests too. If you apply power and ground to the op-amp and nothing to any other pins and you feel the op-amp getting warm/hot then something is shorted inside and it's dead.

>>1564247
Measuring it's gain at 0Hz is also a pretty good idea, which I think you can do by making a follower circuit and trying to measure the voltage between the - and + inputs. It should be really fucking low. That circuit may not work though, I'm unsure.

>>1564196

>>1564247
The sanity check I do with op amps is test them as comparators. Ground one of the inputs, apply test voltage to the other, see if it works. If it doesn't it's ded

>>1564100
Just rectify the 12VAC, no cap and all's well. 120Hz is fast enough and the peak current is short and not a problem.

>>1563849
Your unfinished diagram shows no means to adjust the quiescent current. Did you ever do that? R2/R3 = exactly 3 looks suspicious to me.

>>1564321
I've played with some different values for R2 and R3 and I've even swapped the vbe multiplier out for discrete diodes no effect. But there is no need to adjust the current because that whole arrangement is driven by a constant current source, Q9 tapped off the current mirror there provides a 240uA constant current to that whole branch.

>>1564318
>get an op amp with diodes across the inputs
>*pop*

>>1564344
Why would a opamp have that?

>>1564337
Where do you think the crossover distortions could come from?

legit question here, no bullshit.

how much electricity will fuck you up? I shocked myself with 120vAC and it felt like a ticklish tingle in my hand. people always made it seem like wall outlets would cook you instantly. is it down to the individual and how conductive their bodies are?

>>1564353
If you are typing then you are ok. Don't get shocked.

Is it a bad idea to use a SDR to measure signals which aren't radio signals?
E.g. I have a H-field probe and want to measure magnetic fields. I could put it on a SDR but I could also simply use a scope. The frequency is in the range of the scope and the scope also has higher sampling rates (1GS/s vs 20MS/s).

>>1564363
Dave from EEVblog has a video teaching you to make an H field probe, it's really easy and the results are very good for the buck.

>>1564364
Well I already have a decent probe, I'm just wondering if there's any point in using it with a SDR.

>>1564353
the volts jolt, but the amps cramp (see also "let-go current", "ventricular fibrillation")

>>1564365
if you want to see in the frequency domain, it might be, but generally speaking, probably not

fuck me DDR2 is a bitch

Is it normal for a closed relay to make almost inaudible whispering noises?

>>1564408
You must not stop taking your medication, and add hysteresis to whatever condition you are using to toggle it.

>>1564349
zero drift amps almost all seem to have them, much to my chagrin. i assume it's to keep the internal chopper as low-voltage as possible.

This is in the datasheet of this opamp. Can someone explain to me how it works? I tried solving it but results made no sense.

>>1564642
Fuck, cropped it wrong, It is suposed to be a current monitor. TI gave no output equation

>>1564649
the op amp will try to make its two inputs equal. the inverting input is equal to Vin-R1*i_load. to make the noninverting input equal it pulls current through R2 until Vin-R1*i_load = Vload-R2*i_sink. you can see that the amount of current it pulls through R2 is equal to i_load*R1/R2. that current has to go through R3 which per ohm's law will have a voltage across it. so your current monitor output = i_load*R1*R3/R2. plus some error from the bjt's base current.

also it's possible they have the +/- inputs swapped. i'm sleepy so i don't care enough to simulate it.

I have a 2008 Panasonic Viera plasma tv that I really like, HDMI2 died recently. I want to open it up and see if I can just solder it back to the board, what do I need to be wary of?

I am absolutely retarded when it comes to electronics and I need help. Please tell me why I'm retarded or answer my question.

I have 2 3V LEDs and one 3V motor running off a 3V coin battery. The lights are too dim with the motor running and the battery doesn't last nearly long enough.

If I use 2 1.5V AAA batteries or even 3 1.5V AAA batteries, will my lights be brighter and the motor probably be enough resistance to keep the LEDs from blowing out? Do I need to figure out exactly how much resistance my motor provides and then pick a voltage based on that? Can I use a 9V battery and a bunch of resistors to give me more battery life?

>>1564701
You can use a 9V battery and hook up your LEDs and motor in series, so you don't waste energy on the resistors.
I think coin batteries have somewhat high internal resistance, so even your two 1.5V AAAs in series should work better.

>>1564701
>coin battery
you can easily find datasheets for them that show just how much energy they store, and it isn't much at all, certainly not enough to spin any appreciable motor
>series circuit
good, it shows you understand Kirchhoff's voltage law
but now also consider Kirchhoff's current law. how much current does the motor consume at 3V? it's very high at startup, could be an order of magnitude higher than no-load running current. the current consumption also increases with mechanical load. in either case, your LEDs will probably be called upon to pass a lot more current than they can handle, and they go pop
so how about the LEDs? the V-I curve of an LED is fairly well known, Pic related is typical. since the motor has the same resistance as basically wire when starting up, pop
best bet for the noob is to power them separately. on the LED supply, add an extra battery to allow some headroom for dropper resistors (to prevent catastrophe due to too-fresh batteries). on the motor supply, just give it the 3V it wants
>brainlet
keked and saved

>>1564740
The motor is tiny - 13 mA at no load, 150 mA at stall load, uses .2 watts. It's geared down to 9 RPM and only turns a small, decorative set of 3 gears that have no load on them other than the mass of the 2" diameter plastic gears themselves. Does that change the answer? Is >>1564711 a bad idea or just a different idea? If I did use two separate circuits for the lights and motor, would it be ideal to use a coin battery for the lights and 2 AAA or AA for the motor?

Also this is the kind of answer I was hoping for, thank you my dude

>>1564701
Your coin cell (2032?) is not intended for applications requiring high amounts of current, like a motor. You need a bigger battery.

>>1564749
>13 mA at no load
Like I said before: a coin cell isn't designed for high current draw applications, and for a coin cell, 13mA is a big load. Try a couple AAA or AAAA cells in series instead.

>>1561639
An accelerometer would not capture nearly the information needed without a shit ton of testing done on that particular bag with that particular amount of weight - the deformation of the bag alone would make this measurement extremely inconsistent.

>>1561620
One option would be to embed a force sensor or series of force sensors between two rigid surfaces with a firm pad on the side that you punch, and put that up against a concrete wall or something with similar rigidity. Of course you'd need to do some minimal testing for calibration, but the measurement should be relatively consistent after that is sorted out.

Another option that could be easier on the joints is to use a set of springs instead of force sensors and measure the distance that the pad on the rigid surface is able to compress the spring. F = kx for a spring, so the calculation should be relatively easy and the testing/calibration shouldnt be crazy tough. Might want some lubed up guide rods for this as well.

>>1564778
Momentum is conserved; the deformation of the bag would smooth out the high-frequency peaks of the force-time graph, but the total integrated impulse would be unchanged. If you need to you could also fairly easily amplify the higher frequencies in post to a reasonable level as the bag (both its hanging mass and its spongy surface) will act as a series of band-pass filters. In fact, you could mathematically compensate for the pendulum-like mass of the bag itself, leaving only the deformation issue. From what I've seen those things don't deform by more than ~8mm or so, so it might not even be an issue, depending on what you're trying to do. You're right that calibration would be a pain (due to the high amount of damping in the deformation "oscillator"), but you could measure the acceleration of your fist itself with a high-speed camera, which would help.

>>1564778
>>1564792
The spring idea is pretty similar to the hanging bag idea, they're both harmonic oscillators of some kind. In general their oscillation periods are so slow that they'll act like low-pass filters.

You'd need some form of deforming damping material so you don't mess up your hand regardless, so that will always be something you have to compensate for. Not to mention it will always be a factor in whatever you're planning on hitting anyhow, so it shouldn't matter. You could put the accelerometer inside your boxing glove or hand wraps if it matters though.

>>1564792
If you punch a bag in the middle where there is less sand, it will accelerate quickly. If you punch in the bottom where the sand is pooled, it will barely move. You also have to account for where exactly the impact occurs, which is dependent on both the type of strike and fighter height. Measuring the acceleration of any point on the bag is related to the changing state of how much sand is pooled up where. I think any measurement taken with a bag would be extremely inconsistent. Trying to math around the problem is way more work than it's worth and I think it would likely be incorrect over periods of time due to the bag shifting weight around. You also have to find out precisely where to put the accelerometer to ensure it stays secure.

>>1564794
>accelerometer inside the glove
Doesn't account for the force, just the acceleration. Part of throwing a good punch is being able to put your whole body into it. You could have extremely quick punches with very little force because you didn't distribute your mass properly during the punch.

>>1564749
>150mA
the absolute maximum values for a device are those beryond which the device is likely to be degraded, damaged or destroyed. your typical garden-variety LED is rated for an absolute max current of 50mA for fast pulses (probably faster than a motor start), somewhat less for steady duty. look inside a clear LED body with a magnifier and you will see very fine wires. if the LED chip doesn't melt from the heat first, the wires will
> >>1564711
actually a bad idea
>a coin battery for the lights and 2 AAA or AA for the motor?
actually a decent idea

>>1564814
You'd have to standardise what part of the bag you hit due to the radius to the pivot point (an X of tape would suffice), but I assumed that the bag would be so full of sand that there wouldn't be a discernable difference between the top and bottom. I've never hit one so I guess I'm not the best person to ask about how punching bags work. In any case, the use of a "bag" wasn't necessarily what I was suggesting, but rather the mechanism within it. If there's something more rigid that would suit that purpose then that would certainly be a better idea.

As far as the accelerometer in a glove goes, you'd have to use it in conjunction with the one in whatever you're hitting and use them to account for deformation and nothing else I suppose.

I am trying to get started with verilog, to do this i am trying to make conways game of life on a little led matrix with my fpga. Now I am trying to basically do the same calculation for every bit in a 64 bit array, in a normal programming language this would be a for loop taking up multiple clock cycles but I really want this all to just be logic which can happen in a single clock cycle. Am I just supposed to write a formula for every single LED or is there a way to cycle through all the LEDs in verilog which gets compiled as something that runs in a single clock cycle?

Discrete PWM guy here. After making it work, and being happy with results I`m moving to greater objectives. I already have a bench supply I made for myself, it`s the quick and dirty linear regulator one, and I want to see if I can make a buck converter (switched regulator?).
Ok, question time with light loads the perfboard unit I made gets a toasty flyback diode. (Because the duty cycle is low).
To solve this I've been thinking of instead of changing the duty cycle, I could change the frequency instead. (Because low frequencies would mean less current in the diode). Has anyone tried to do this? What should I expect from using this control method, instead of pwm? If someone has some reading material about that I'd be glad.

>>1564964
create a helper module that calculates the next state from 8 current states
then, in your top module, use two for loops to instantiate the calculation module using the appropriate bits of the current state and outputting to the next state
then, always @(posedge clk) current_state <= new_state;
further reading: look up Verilog2001's generate/endgenerate block

>>1564998
Tnx. Allthough I think I have already solved it with a for loop which apparently works exactly how I needed it to work.

>>1564997
>Has anyone tried to do this?
sure, almost every buck converter IC vendor. typically they switch between PWM and PFM mode depending on load (or called-for duty cycle). you should expect a lower minimum load
another way this is sometimes done is to skip pulses (at the normal frequency) when the load is low. either one might require high-side current sensing
>toasty diode
are your inductor and operating frequency appropriately matched? is your error amp gain reasonable? do you have any low-pass filtering on the error amp? is your output cap rated for high frequencies?
consider grabbing a datasheet for the MC34063 and running through the inductor and output cap selection equations. weirdly, smaller minimum loads seem to call for larger inductors

>>1565003
ebin. how many LUTs/FFs did it take?

>>1564827
>>1564711 here, can you explain why it was a bad idea? Not disagreeing or anything I just want to know

Is this good for SMD prototypes?

>>1565061
bit overkill

Accidentally lost a resistor while installing a PS1 modchip, so I consulted with a service guide to determine what I needed to replace. Would this be suitable for it?
https://www.mouser.com/ProductDetail/KOA-Speer/SG73G2ATTD22R0D?qs=sGAEpiMZZMvdGkrng054t7z4BkURc4Lz%252BjFEEtNJGESDJP0VHJ0ZzQ%3D%3D
The service guide which includes the motherboard schematics & more is available here for reference:
https://gamesx.com/wiki/lib/exe/fetch.php?media=schematics:service_manual_scph_9000_3rd_ed.pdf

>>1565062
Well I like it that it also has a soldering iron so I can retire my own. How about this, then?

>>1565066
Man that's an expensive resistor. Is it precision military grade or something like that?

>>1565074
>>1565061
From what I heard the iron on that station is crap.

>>1565079
It's got 0.5% tolerance. I don't know much about SMD resistors or their prices, but I do know that's a pretty tight tolerance.

Can anyone help me figure out freq compensation? I'm working on a CC/CV power supply and having trouble stabilizing it. Is the most common way just popping in capacitors to see what works best or are the circuits actually broken down to a control system view?

For example, I get the reason for this cap (miller, roll-off) but I wondering why they placed it on the red dots versus the yellow for example

>>1565074
You could get a KSGER T12-style hot air station. Not sure how good they are and they go for 80-90 dollarydoos, but they're an option.

>>1565149
>diy hot air station
That doesn't look bad at all!

Is there such thing as a flip flop circuit or IC that will remember its state even if it's not powered and isn't an eeprom or flash memory chip? I need only a single bit

>>1564649
>Back to Basics Edition
I_L*R1 = I_R2*R2
I_R2 = I_L*R1/R2
Vo = I_R2*R3 = I_L*R1*R3/R2
diagram example:
R1*R3/R2 = 1Ω
Vo = I_L*1Ω = one volt per ampere

People always say that you need to avoid ground loops because they act as antennas and whatnot, but that got me thinking: exactly what is an antenna? Is it just a loop that produces a current inside a magnetic field? Would it be possible to light up an LED using nothing but wifi signals?

>>1565004
>buck converter
thanks for the info man

>>1565275
>MC34063
And btw, checked the datasheet. I'm amazed that I managed to get to a solution so similar to the commercial one (in idea atleast) and it kinda worked, usually I'm very wrong about everything.

>>1565222

>>1565222
https://www.maximintegrated.com/en/app-notes/index.mvp/id/4624

>>1565256
Anything conductive is a antenna, but not necessarily a useful one.
WiFi is pretty low power, harvesting any usable amount of energy from it is difficult.

Hello /ohm/ I have a simple goal, tell me if it is possible and how to do if possible.

I am trying to connect 12 devices (via I2C) over ~20ft/16M total of cable. Even if I go with only 100Khz SCL I am worried about ringing and rise/fall times. Would a differential output and twisted pair for SCL and SDA help with this issue?

I plan on using 4.7k resistors pulling to 3.3V on the bus and 110 ohm series resistors for each SCL/SDA pin if that matters.

>>1565352
Is the distance from the single host to a bunch of devices located closely together at the other end? If so the differential i2c extenders like PCA9615 should work alright, can use regular cat5 cabling. Direct i2c would really not be suitable.

Another alternative is run RS485/RS422 to a remote micro that then speaks i2c to the devices.

>>1565302
That little IC4 has an EEPROM in it

>>1565355
>>differential i2c extenders

Oh nice, I had no idea those were a common thing. The difficult thing is, that the sensors are spread out non-uniformly, some are only 1ft away from the host, a few are 5ft away. The network looks kind of like a smashed up flower.

>>1565054
see
>>1564740

>>1565061
it would do, functionally
Aoyue seem to have good build quality and design. no doubt they should have a version of the same

>>1565149
>le T12 meme
controls are shit. where's the fan speed

>>1565352
like other anon said, an I2C bus extender like the 82B715 is the easy way to go about it. there's also another chip, whose type number evades me at the moment, that turns an open-collector I2C-style line into separate Rx and Tx I/O, which you can hook up to your choice of wired-AND transceiver (RS-4xx, CAN, etc), in case you want a really long I2C bus

>>1565377
Can have PCA9615s daisy chained, with termination resistors on the first and last.

>>1565399
>where's the fan speed
Shit, really? Not even in the nested menus? If I get one I guess I'll be making a PWM circuit for it then. All the blower handle units look the same as that one, I'm guessing there's no issue with its design?

>>1565222
>>1565297
There's a video of a guy making an SR latch with a magnetic core and it works well, but proper magnetisable cores are damn hard to obtain. I'd see if I can go with a latching relay instead, and power it with a half-bridge, that's not too difficult. For more than a bit or two, perhaps an obscure EEPROM IC that you can address with a shift register or two would be a good choice. If you can get an 8bit parallel EEPROM IC of some kind then that's extra useful, though not exactly likely. I don't think it would be possible to address any kind of flash memory IC without an MCU, but you may be able to obtain those floating gate transistors in individual packages.

>>1565481
Oh I googled as to the availability of individual FGTs and it doesn't look like they exist, but storing the previous state in a capacitor that you charge and discharge with 2 FETs looks quite possible and even recommended, with a third transistor for reading purposes.

>>1565222
Relay autism maybe?

>>1565487
https://www.electronicdesign.com/analog/relay-based-onoff-flip-flop-remembers-state-during-power-failure

>>1565488
Here is one with a non latching relay. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1338590/?page=1

>>1565484
How long would it retain the charge when power off?

>>1565222
there are many, but something like an ATtiny10 would seem to be the most practical and least susceptible to environmental perturbation

>>1565517
probably tens of hours, depending on local humidity and temperature and whether you got a dud capacitor

How difficult is it to make something like pic related, but with separate camera modules and video "combiner"?

Are there circuits that combine more than two feeds?

>>1565518
> something like an ATtiny10
'10 doesn't have eeprom memory. '13 does though.

When dealing with eeprom one should consider limited amount of write cycles and slow write speed. That is, it is very possible that if the latch changes state right before power is gone, it will be left in undefined state.

I personally would go with sram memory and a coin battery that would probably last a lifetime.

>>1565488
Will it oscillate when you keep S1 closed?

>>1565543
good points
I was thinking in terms of the package size and went lazy on checking the datasheet. those bastards

>>1565542
9/10

>>1565563
>9/10

I'm not trolling. It's a genuine question. I've learned that the combiner is called a compositor ( https://photo.stackexchange.com/questions/26193/is-it-possible-to-combine-multiple-cameras-for-improved-detail).).

>>1565551
with S1 held on, both caps are being continuously discharged through the relay coils and the diodes. S1 has to be off for a while for the correct cap to build up enough charge to pull in the other coil

>>1565574
and that was a genuine reply. if I thought you were trolling I'd have r8d on the b8 scale
assuming there were chips to do exactly that off-the-shelf, you still would have to deal with designing the board (possibly entailing trace length tuning, impedance control, etc.), configuring the chip (possibly entailing some C programming, SDK learning, serial flash burning, in-circuit debugging etc.), dealing with fine pitch component interconnects, etc. which individually or in combination put that project into the black diamond class
if there were not, and you wanted to do this in hardware, you might probably use an FPGA, possibly with external RAM, to capture the camera data, store it, perform whatever undefined computations you want to on it, and reformat it as a USB video stream, which requires knowing all those protocols fairly intimately, plus all of the above
if you didn't mind latency, what's probably the easiest way is to use software on some ARM dev board with a USB-OTG port to read two cameras, merge/compose/whatever, and feed the finished data to the host via the USB video class gadget (not super obviously documented). but it's hard to do all that both fast and easily

>>1565551

>>1565599
BTW If you want to search more about relays google ladder logic. It's still used today (even when no relays are involved) because it fit PLCs in a nice way.

>>1565599
>at power-on, K1 is actuated and K2 is released
>user closes switch
>both relays actuated through diodes
>K2 is also held actuated through its own contacts
>user opens switch
>K1 is released
>time passes
>users closes switch
>K1 NC contacts and switch short power input to ground
>sparks fly
>K2 is released
>surviving user opens switch
>power supply recovers
>K1 is actuated through K2 contacts
now this is bait

>>1565598
>and that was a
Thanks for clarifying. The idea is to compose videos from lots of pinhole cameras with carefully drilled holes, using the cheapest camera modules I can find. Realtime processing isn't a necessity but it would have been nice.

>>1565625
I'm curious about your project. Do you just want synchronized video streams in a single stream (CCTV style)?

>>1565628
Not so much a project more of an idea, to improve the quality of covert (pinhole) cameras. The use case is lifelogging without the danger associated with causing offence. Most people don't appreciate being filmed even if the motive is innocent. Steve Mann was assaulted in a fastfood restaurant for example.

>>1565399
God I'm retarded, I always forget to consider maximum load
Thanks.

How dangerous would it be to build my own variable power supply if I have some theoretical knowledge (EE student) but never actually built anything that complex before?

>>1565639
don't create any voltages higher than 50V and don't stick your fingers in it when it's plugged in, and you will probably avoid killing yourself

>>1565639
it's ok, I just did the same thing. It's ugly as fuck but works. 0-24v - 1.5A

>>1560956
Anyone got a LaTeX style file for such documents?

>>1565436
>>1565399

Thanks for the solution guys. I think I got this figured out now.

What's a good DC adjustable power supply for under $60?
I see a few on eBay but don't want to get one that will release the blue genie when I power it on

Can you buy RF shield boxes for PCB's in standard sizes or do they always have to be custom made?

Also, how do you actually use them? Do you just put them over the circuitry you want to make immune to RF and then solder it to gnd?

>>1565793
metal bawkses

A friend of mine needs these arranged in a certain way to bypass an O2 sensor code (we don't have emissions in our county, anyway).
What I'm curious about is if I can replace these with much, much smaller SMD components, since I would rather have that to put together. I'm basically just trying to make the install as cheap as possible, since I'd rather tape up an abomination that was only an inch wide, versus this.
Sorry for the spoonfeeding request: I just don't even know how I would start calculating if I can just willy-nillily replace components.

>>1565896
and they're all made out of ticky tacky and they all work just the same

>>1565781
they all do that... eventually
if you're that concerned, buy pre-owned

>>1566120
mounting to the surface provides mechanical support that the thin, delicate terminals of SMD need, so you'll need a bit of copper-clad laminate to mount to. I've done well carving out channels between sections using a #11 pen knife then soldering the components/wires down to that board. I've also had some luck using slivers of adhesive copper tape and solder to connect devices together
components in those values could be gotten in (inch size) 1206 or 0805 packages quite readily, but not necessarily in small quantities. also, you will probably want a temperature-regulated soldering iron to avoid damaging the components and to get a good joint. otoh 100 of those resistors cost about a penny each and the caps cost about a dime each in 25s, so you have spares in case you fuck up
if you don't want to fuck with that and just want a smaller, more reasonable capacitor, you really only need about a 50V rating, which might make it a lot easier to find one that fits your size budget, see https://www.digikey.com/product-detail/en/illinois-capacitor/105MWR100K/1572-1024-ND/5343954
>for off-road racing use only, you must confirm that laws applicable to your jurisdiction allow for modification of emissions components in your circumstances

>>1566131
Holy shit. Your post is so helpful that I feel that familiar "am I getting trolled?" feeling that someone gets when someone's genuine.
Thanks a fuckton for the help. Damn.

Speaking of the "fucking with that" part, is the copper-clad laminate the SMD analog to the through-hole "prototype board" that people would usually mount these to?

>>1566120
What voltage does the cap need to be? Because the one pictured looks like a med-high voltage cap, so a standard SMT ceramic wouldn't be sufficient, though that's not to say you can't get SMT ceramics of that value and voltage rating.

>>1566138
Copper clad laminate is generally what you'd etch a PCB into with photoresist and all that, but using a dremel or knife to carve out isolation gaps is pretty similar. A CNC router can also carve these slots, though more precisely. If you want to mount SMT packages to a small custom PCB without going into the trouble of etching one or getting one made by a fab house, then you can either go the slot carving route, or you can buy SMT breakout boards and lay out the circuit on normal perfboard. The former looks to be a good option for less than 10 parts, even more so for only 2 parts. Though I doubt space saving is that much of a concern, and you'd certainly have an easier time soldering and testing the circuit if you used THT parts.

If you can use a small, lower voltage (maybe even electrolytic) THT cap, I'd do so.

>>1566138
CCL is just blank pc board that is typically used for home etching. instead, I cut and peel up slivers of the copper to create islands insulated from each other. Pic related is an example from my scrap box viewed through my potatoscope
I have seen proto-board for SMD passives, see https://www.aliexpress.com/item/SMD-Benis-Adapter-XDD/32673482747.html
either one can be cut to size with shears. or you can use tin snips, at the risk of bending the substrate a little bit, which is not a huge deal if you haven't mounted the components yet

>>1566142
I'm really uncertain on the cap voltage. I guess I could link the thread where I found the hack in the first place.
https://www.yotatech.com/forums/f160/elvotas-3-4-swap-128285/index13.html#post50717045
The idea of SMT was that everything was small. Typing it out, I realize how presumptive I have been, and should have just generally asked "is there a better way," but wanted to look like I put in some thought.
>>1566143
I'm rolling at that mnemonic slug. Good one.

>>1566145
Actually, looking at that circuit layout, I'd be tempted to just freestyle it without any board at all. Cover the joints in a dab of epoxy/hotsnot and you wouldn't have an issue with it. It's going to the ECU so I can only assume it's ~14V maximum, so a 16V cap would probably be fine. Bonus points for covering it all with heat-shrink.

>>1566145
I think the cap can be substituted pretty broadly, O2 sensor working voltages are on the order of 1 volt. try a 0.1µF disc cap of 50V or more, which might still be smaller than the wires. I wouldn't use electrolytics here because they can be a bit touchy under temperature extremes

>>1566149
Was thinking about something like this (especially the heat-shrink).
>>1566152
That answered my other question. If the operating voltage is 1V, why go for the 50V capacitor? I really need to read up on my electronics theory again.

>>1566156
50V is standard value for many capacitors, especially ceramics.

Is there a component that works on 3v3 logic level with at least 3 legs or pins that can do this:

Sets output PIN1 to low
Waits for input PIN2 to be pulled high. waits for this to happen for a maximum amount of 5 minutes (ideally adjustable)
If the PIN2 was pulled high at any time during this time, sets PIN1 back to high,and just goes back to step 1 and loops forever like that
If the PIN2 was not pulled high during that time frame, it sets output PIN3 to low for a second or two, then sets PIN1 back to HIGH and goes back to step 1 and repeats this cycle forever

I am currently doing this with arduino mini, but it seems like a huge waste of money, space and resources to use arduino for something this simple

>>1566172
any small microcontroller with an ADC, e.g. ATtiny5/10

>>1566173
>ATtiny5/10
I never used anything other than arduino, is there anything like that compatible with the arduino ide? Or how are these things coded and then the code uploaded?

>>1566174
http://www.technoblogy.com/show?1YQY

>>1566178
neat, chinks sell this thing, this is great

>>1566168
I see.
I forgot to ask:
In the extreme temperatures inside of an engine compartment (hot in use, cold in winter), or in any conditions, what makes a difference in the longevity of large through-hole devices versus tiny chips?
Furthermore, how are tiny chips able to do the same thing in most cases that large resistors, etc, do? Or is that why things are usually moved down to smaller voltage levels? Has the technology to make analog-level parts smaller stagnated because most development is in digital/SMT stuff?

>>1566180
I can just program it, then cut off the usb tongue and have a tiny compact chip with a built in voltage regulator so i don't have to fuck around with buck convertors and it even has labeled pins i'm going to order a full loin cloth of them

>>1566183
mostly things like coefficient of thermal expansion. fatigue cycling is a pretty typical failure mode. ECUs, for example, are typically built on flexible polyimide film instead of fiberglass
>tiny chips able to do the same thing
the substrate (pc board, copper, etc) assumes some of the functions of the larger through-hole packages, notably heat dissipation
>why things are usually moved down to smaller voltage levels
the ability of insulators to separate voltages is measured in volts per unit length. faster and less power-hungry ICs are 1/2 of the motivation, miniaturization and materials cost reduction is the other 1/2. all of those imply lower operating and signalling voltages
>Has the technology to make analog-level parts smaller stagnated
not sure what you mean by analog-level, but passive components can be gotten so small these days (1/50" x 1/100") they're considered inhalation hazards. every digital chip needs one to hundreds of capacitors to keep the power input solid and clean. inductors have gotten smaller and more available than 20 years ago thanks to the ubiquity of switching dc-dc converters
as for analog ICs, a through-hole package is some 99+% of the mass of the device, with the chip being on the order of 0.1% or less. now chip-scale packaging, where a device is little more than a chip with some solder bumps on the bottom, is a normal option in chip vendors' catalogs. there's just no real point in making THT smaller, since holes in boards are expensive and take up a lot of space that could be used for components

>>1566184
nah you just have to deal with the big old 7805 on the board. don't burn yourself

>>1566192
>7805
It is powering attiny, it probably won't get even warm

>>1566196
And the ATTiny is going to drive whatever loads that are connected to it's pins.

>>1566197
The pins will only be handling logic level shit
You ain't gonna set the house on fire with milliamps

>tfw payday finally rolls around
>normal people take out their gfs for a fancy dinner or party with friends
>tfw i am sitting on ali picking out high end chink crimping tools
does R = V/I stand for Roastie Vaginas Inaccessible?

I'm bad at this but what voltage regulator should I use to get the output voltage of 4.5volts and 3 volts?

>>1566213
lm317

>>1565599
>12V "or more" through an LED with no current control
>12V shorts to ground when relay 2 is energized and user presses toggle
nothin' personnel, kid

Fuck. I just accidentally zapped myself with 230V when reaching for my mouse. Guess having your multiple 230V powered projects loosely hanging on their power cords from your desk lamp for weeks at a time isn't such a great idea.
That shit hurts.

>>1566385
kek, it turned relay 2 off, didn't it?