/diy/ - Do It Yourself

qfn > qfp

this thread's digits brought to you in part by the 2SK3603 super FAP-G silicon power MOSFET

What's the smallest way to make a small delay in a signal? Like pic related. I've thought of using an SR/JK latch with a capacitive delay running from Q to R, but that doesn't have a delay before turning on, which I can't do capacitively since the input signal will be off before the latch toggles. Do I have to have 2 latches with different reset delays on them and a logic gate to compare their outputs in the right way (O = NOT(Q1) AND Q2, where delay1 < delay2)? I think you could do something similar with 2 comparators, but I doubt it's any simpler.

>>1343632
not to be an asshole but the smallest way is an 8 pin microcontroller and one decoupling capacitor.

>>1343632
There's the 74HC123 which is specifically designed for that, although several other options exist: a pair of 555 monostable circuits, a pair of transistor monostables, a pair of schmitt trigger monostables.... It's variations on the same aproach of using one monostable that goes high for X amount of time after being triggered, then using that to trigger a second monostable that gives the output pulse.

>>1343607
>FAP-G
heh

>>1343634
That's sadly true isn't it.

>>1343636
That 123 is a bit of a waste of pins, but probably still more compact than a 556, which would need a few more passives. The double Schmidt might be a good idea if I need 3 of these, which for now is yet to be seen. I'll see where the transistor method leads me since it's easier to simulate, get it working, and see from there how many I need.

>>1343661
see, this is what ATTINY10s are perfect for

>>1343632
Every gate has a delay. Two nand gates in series.

>>1343634
>not to be an asshole but the smallest way is an 8 pin microcontroller and one decoupling capacitor.
you can go smaller as there are 6 pin MCU's on the market

>>1343732
It needs to be controllable, consistent, and in the modest realm of 100µs to 100ns.

>>1343758
That's 100µs to 100ms.

>>1343758
well then, i was thinking maybe an IC delay line, but the first one i looked at has debouncing, so it would depend on the pulse width and whether that looks thin enough to be a switch bouncing.

http://cds.linear.com/docs/en/datasheet/699412fb.pdf

but there are other choices, so maybe.

https://www.digikey.com/products/en/integrated-circuits-ics/clock-timing-delay-lines/688

Hello, /diy/. Today is my first time writing here and I hope I get the help I am seeking for (I also hope it's the right place). I'm gonna write straight for what I'm searching - I need some kind of software, that will give me soil information (like what kind of soil, what kind of microorganisms, water and etc) of a particilar area selected by me in like 1 meter depth, or doesn't matter that much, but 1 meter would be great. Also the exact coordinates of the edges, sea level and things like that.

>>1343773
What are you asking for exactly? A piece of sensor equipment for taking your own readings; access to a database of measurements someone else has made; or software that just makes an educated guess based on your location?

>>1343774

Thanks for replying. I guess, you defined the things I am searching for way better than me, thanks. I am searching for a kind of program that will give me access to a database of measurements someone else has made; or software that just makes an educated guess based on my location. But if there isn't, the first one could also help me.

>>1343758
if you want small, go with the 6-pin micro
if you want smallish and analog-ish, go with the dual monostable
or, if a proper switch, use a diode, a cap, and a resistor

>>1343775
why don't you just send a sample off to your local extension office?

>>1343777
The thing is, I am searching this software, because the guy of mine couldn't find it and asked for a bit help. I told him I will try to find it(didn't ask why and for what, just told him I will give my best). Don't have any other information...

>>1343781
well this is absolutely the wrong place. try >>>/out/

>>1343784
Okay, will check there, and thanks again. Also, I will keep checking this thread just in case if someone finds something and decides to help.

What do you call these connectors? I have to buy some for a project of mine, but can't figure out what they're called.

>>1343803
"Dupont" connectors
you can find the exact thing in the pic by searching for "dupont male female leads"

>>1343803
jumpers?

>>1343803
If you go with the cheapest chink shit chances are you get copper clad steel wires and steel connectors.

I have some piece of shit media player which uses some obscure old chip from Actions Semiconductor - ACU7515. Is there any way I can dump/flash firmware, meaning, is there any general approach to doing that on non-brand-name chips like PIC, etc? I can remove the chip from the board. I have the pinout of it as well. It's a standard 80-pin chip so I could eventually try and get a logicaldevices programmer with an adapter for it, but I'd like to avoid that if possible. As a sidenote, there are GPIO pins and some pins marked as USB, but there is no USB port on the board. I'm assuming I can make some kind of adapter to connect a niggerrigged USB cable to it and try my luck. Is this correct?

That's one of a group of MP3 players collectively known as S1, there was a decent community around figuring out firmware modifications for them. s1fwx can probably dump it http://s1mp3.w1r3.de/tools.s1fwx.html

>>1343863
>>1343860

>>1343863
This isn't about the software part, it's about the hardware. As I said, it doesn't have an USB port, and it's not a regular mp3 player. It already has really custom firmware on it.

Newb questions here:

Rechargeable batteries with a higher mAh rating will mean longer battery life for the equivalent voltage/wattage than another battery with lesser mAh rating, correct?

Given that, where is a good source for batteries? I want to get some high-capacity ones for some projects (including some AAs on the side for use in small gadgets) and I don't want to make poor choices on brands or manufacturers. Like I've seen some 2800mAh and 3100mAh rated AAs but I don't know if they're reputable/won't blow up/etc.

>>1343894
>correct?
Yes, but like you suspected, the stated values are often exaggerated and your 3100mAh battery might actually last longer than 2800mAh.
I have no recommendations for brands.

>>1344051
Hurr. Was supposed to be the other way around.

>>1343607
>Super FAP-G
>High speed switching

>>1343894
For NiMH batteries like most AA ones, if it says capacity above 2700 it's almost definitely bullshit. Also if you're using it in a relatively low power device that you'd expect to be charging after more than a few hours use you probably want a low self discharge type battery like the Sanyo Eneloop cells, don't take as much capacity as some other models, but waste less just sitting around with self discharge so can be more usable energy anyway.

>>1342684
>>1342689

I could not get the original circuit (>>1342675) to work. The circuit acted in strange ways, leaving some part of the leds glowing at all times.

This is the version that I'm working on at the moment. It works, but still lacks the PWM drive input pin.

How can make a bike brake light that turns on when breaking? I have mercury switches, but I think the mercury drop will be bouncing all the time.
Any popular accelerometer suited for this application that can sense deceleration? Can I just average the readings to filter peaks? Any better method? I'd rather not mod by brake handles and install switches there.

>>1344658
You'd rather not install an easy switch under your handle to detect when it's pressed and wire it down the frame, you'd rather have an unreliable solution that'll break in no time?

>>1344266
>low on-resistance
the animated gifs almost author themselves

>>1344654
ffs number your parts and add values

>>1344661
it's always fucking Borat isn't it

>>1344654
why don't you put the static strings on O6/O7 of the transistor array, put the diode array on I6/I7, and drive E with pure PWM?

>>1344665
I was going to suggest this, a logic level MOSFET on E. 2003's outputs are 500mA, 300mA recommended, so it should be fine, you can also share the load between 6 and 7 with inputs connected to VCC. Less components to put on the board.
The circuit seems fine, the 2003 has pulldown resistors, so your code might be the problem.
>>1344661
Why should it "break in no time"? It's a circuit with no mechanical components. A switch or my plastic brake handles will probably break even sooner if I don't mount it correctly, the wire can be ripped accidentally... I prefer an all-in-one solution, it doesn't has to be unreliable necessarily if done correctly, an epoxied PCB will last forever. I'm just looking for a generic accelerometer that fits my purposes.

>>1344690
>if I don't mount it correctly
then mount it correctly
>the wire can be ripped accidentally
then work out a sensor at the caliper end, right below where you would most likely want a brake light
KISS don'tcha know
if you insist on an accelerometer, then get a chinky breakout board of your choice and data-log it to a reasonably large storage device while you ride, coast, etc. then, deciding between brakey/no-brakey will purely concern software

>>1343603
>Mfw I spent 3 hours finding and salvaging parts for a voltage divider needed for my project
>Mfw doesn't work

>>1344703
>not using a pot

>>1344706
>Assuming by the part that I'm salvaging resistors from old pcbs that I have a pot with the right value

>>1344395
Is there a clear cut line for what is low and high energy requirement? Like would 800mA be considered high, or is it something more substantial?

My amplifier got lost, so I was thinking I should just make myself a tube amplifier and a nice wooden case for it.
Trying to find some kits online, I find only either amplifier kits or pre-amps for phono steps. Anyone know one with a combination set, or can I just get a set of each and jam the in the same box?

And, those amplifiers kits found on aliibaba or eBay that say amplifiers and not just pre-amps, will they be enough to drive some speakers, or are they just intermediates between a normal stereo receiver and the speakers?

Just want to drive my record player and an AUX-input, no need for anything fancy.
Semi-adept at soldering and basic electronics, but no skills or inclination at this point to design anything myself.

>>1343603
>greatscottlab
if i hear "let's get started" one more time i'm going to put my head through my desk

>>1344714
Alibay has tube amp kits? I must have missed something.
>drive my record player
Amps don't drive record players, they drive loudspeakers.
>no skills or inclination
Then a kit is what you want. Look for phono and Aux input and a watt or three of output power.

>>1344797
Congrats, you misread the question, answered none of it and managed to be a pedantic ass.
Thanks anon.

Using a 74LS245 as piezo driver. It works, but prpbably isn't the most elegant way to do it. Suggestions to alternatives are welcome.
How worried do I have to be about the radio nazis comming for me for running this?

>>1344816
That happens here a lot. Here I go not answering anything either.

I'm learning programming and thought this thread might be the correct place to ask. I'm currently starting a small programming project related NTC (and possibly Ptc) thermistors. It's currently just a command-line program to keep things simple. What kind of features would you like to see in such a program?

>>1344658
This is called a limit switch, and it's exactly what you need.
Have it pushed by the brake handle, or by the brake itself. You don't need to mod anything, you can screw the switch to a piece of wood or plastic that you can attach with hose clamps or something similar.

>>1344821
How about an explanation of what the fuck it is,what its for,how to use it,what input it needs,what output it provides,what it does,who would use it,when would it be used etc.
In fact you could even post that here because wtf are you talking about

>>1344817
slightly, since any radiation would be crapping all over WWV
maybe choose a slightly different frequency like 11.0592 that nobody gives a shit about
>alternatives
literally any gain element that can handle 10MHz ffs

>>1344714
if they specify watts of output, it's probably meant to drive a speaker. if not, it's probably a pre
the "aux" is probably line level, doesn't need a pre. you'd probably want a pre specifically designed for phonographs with a de-emphasis circuit

>when the datasheet doesn't specify the LSB or MSB

>>1345082
Fuck I think I just got it around the wrong way, it's still not working after I swapped the leads. But who the hell does this?

>>1345073
Thanks. Your post confirmed my impressions from trying to research this on my own. Most of the kits I see that say amplifier and not just pre-amp list the effect at around 8W, which sounds little, but I guess it would be okay. My googling also suggest AUX-inputs don't need pre-amplification.
But, I have not yet found a tube kit that includes amp and pre-amp, but I'm still trying to see if I can find one before giving up.
Worst case Ontario, I guess I can make an amp and pre-amp fit in the same chassis, and just make it a bit bigger, but of course a board specifically designed with both implemented would be more space efficient.

>>1345087
i've never seen zero not be lsb in bcd

>>1345095
ATM I'm building a 555 oscillator because it's easier than a debouncer so I can test each pulse and deduce what pins are wrong from there, but I can't seem to get the thing working. I'll be back after eating.

>>1344711
If you're using it continuously I'd definitely call 800mA high.

>>1345088
>more space-efficient
it's highly unlikely that you'll find your exact specs in a single board. just go with the two boards, anon. if you find the inclination, you might be able to run the pre off the power amp's transformer or something

>>1345100
you could make a debouncer with the 555
to reverse output polarity, swap ground and +V on the R/C/switch input

why did none of you tell me this happened??

>>1345107
>you could make a debouncer with the 555
Well considering my "switch" was just 2 pieces of solid-core being touched together I think the oscillator is a better idea. I've got a big enough cap for it.

>>1345107
I'm starting to fear the same.
It does seem strange to me that no one would be making these kits tho, as I suspect a huge number of tube nerds make them specifically to run record players on.
Oh well. I'm in Russia for work now, but will start work on a 2-in-1 solution when I get home if I haven't found anything by then.

I have a Raspberry Pi. It runs on 5V DC, and draws 2.5A at full load. I also have a 7,2V DC NiMh battery for RC cars.

Question 1:
Will this circuit work?

Question 2:
The LM338T voltage regulator outputs maximum 5A. Does this mean that it'll only output 1.5A if that's what the Raspberry Pi needs at the moment, or will this fry my Pi?

>>1345131
Fuck, forgot image.

>>1345100
Aha, two segments are wrong, which was complicating my process of getting the bits the right way around. It works now. Messing about with a breadboard again makes me remember how handy those "digital logic trainers" were; a bunch of switches and a 1k/1M clock built in to a console with a breadboard in the middle.

>>1345131
1. probably
2. sigh, A

>>1345132
I'd add some filter caps (par for the course when it comes to voltage regulators) and check that your battery can handle the current.

>>1343603
I went up to see the links in the OP but holy shit that's quite a soldering job.

>>1345134
hah, it never hurts to have a few blocks like that built up on little breadboard-friendly boards for when you need them: switch debouncers, crystal oscillators, level shifters, voltage regulators, etc.

>>1345136
image search it, there's quite a story behind it too

>>1345135
Thx.

>>1345136
Thanks for the advice!

>>1345139
>quite a story
You mean the one on EEVblog, or the one on HamSexy? If the HamSexy one isn't just them picking a random image (the kid's bad soldering was mentioned only once) it could make sense, but I do like the idea that the same kid is behind both.

>>1345142
yep, same story
>"an electrician"
>"professionally install"
an electrician professionally installed my sides

>>1345095
Me neither, and there's also the option of checking the datasheet's truth table, internal circuit or whatever.

What is a modern alternative to the 28C16 EEPROM?

>>1345109

>>1345171
In addition, does anyone know of a good modern alternative to something like a 74HC189? It doesn't seem to be in production anymore.

>>1345171
NOR flash ig
except you now often have to deal with rewriting whole sectors yourself instead of just bytes. the 28C64 is still in production if you need a more direct replacement with better byte-handling capabilities
>>1345175
uh, 62256 with a bunch of address lines tied low?
obviously the correct replacement depends on what exactly you intend to do with them

>>1345171
>>1345175
If your device needs those, it's obsolete in many ways to begin with. "Good modern" alternative would be something quite different, like an I2C or SPI EEPROM or a microcontroller with enough built-in EEPROM.

Assuming you don't want to modernize the circuit itself, there are bigger parallel EEPROMs and flash memories like the other anon said. Battery backed SRAM might also be a valid option.
If separate inputs and outputs are the key feature of 74189, you could try replacing it with a (much bigger) dual port RAM. IDT has those. Or, if you need something resembling 74189 more closely, you could use a small CPLD. Or a bunch of 74670s with some glue logic, if you're desperate. If separate IO isn't needed, the other anon's suggestion would be an option.

But why not just buy what you need from eBay?

>>1345073
>11.0592MHz
The peoblem is that I need to hit the resonant frequency of the piezo, and according to my testing thats 10MHz. I'll research one other piezo that might work but if it doesn't I'm out of options.

>>1345250
>obsolete in many ways to begin with
well fug

>>1345266
on second thought, 10MHz at less than a watt won't be much of a problem. worst case is the FCC sends you a letter telling you to stob dat :DD and then you cross that bridge when you come to it, but again, not really something to worry over
as to better designs, maybe investigate the patent literature such as these auto-resonating designs https://patents.google.com/patent/US3989042

>>1344726
More like
>hieh we heff ay modah, en we food like to duaif eet

Is there a name for these flip over terminals? I cant find any how-to info on Google because I dont know what to search for.

The 10# wire doesnt fit around the screw, so it appears it goes under the terminals instead which flip over it? Does the connection on the left look right?

>>1345292
If the conductor is sandwiched between that terminal plate under the screw then yes, that is the correct way to connect it, just tug on it and make sure you always have a secure connection

>>207260443
Thanks for the advice. My problem is that none of my 8+ attack medals are guilted or have skills as Im still relatively new (only got this far because I first made the account awhile ago and came back to a ton of jewels I used to get a KEX).

So really Im just fucked for 160 until I get a few handfulls of M&Bs and/or some more attack skills come up.

>>1345297
I apologize, its not giving me a delete option.

>>1345296
Thanks, that seemed too simple I figured I was doing something wrong. Im used to always wrapping around the screw

>>1343803
I saw your thread on /g/ before, faggot. They're called wires.

>>1345250
>>1345176
Thanks, these are good alternatives.

>>1344701
>>1344890
>piece of wood
>hose clamps
I sperged out and I'm going to use a hall effect sensor on the rear wheel and a uC to automatically turn on the lights and off when the bike hasn't moved for 5 minutes. It'll count the rpms and light up when decelerating. I even made a case for it >>1345380

Apparently it's hard to get 433MHz modules that actually work. I bought a set of FS1000A transmitters and XY-MK-5V receivers. The transmitters were missing an inductor and a capacitor and the receivers just didn't work. The transmitter worked fine after I added an inductor and capacitor to it.
I also bought a RXB6/KXD10036 set. The RXB6 receiver works great, but the transmitters don't work at all.
I don't know if I have shit luck or if it's so common for these things to be DOA.

who are you calling sloppy, dirty kuffar?

>>1345487
You need to use them properly. They're AC coupled, so your data must have enough transitions in it for the reception to work. Also, the receiver does not wake up immediately on receiving something, so whatever data you send, you must take this into account. And finally, since it's a radio (pretty bad one, too), you have to handle the reception errors and missing data one way or other.

>>1345487
too much chink crap, and of course, we always buy the cheapest we can find

have you tried the NRF24L01 transceiver?

>>1345472
I had hoped you'd curb your autism, but too late. Do inform us about how it turns out.

>>1343603
I have a ti-84, and a entire row of buttons does not work, but everything else works fine. How would I go about fixing it?

>>1345487
>I don't know if I have shit luck or if it's so common for these things to be DOA.
It's common for chinese stuff. Also remember, you get you what you pay far. Keep shopping and buying the cheapest

Am I retarded for wanting to have a USB Type-C connector on my babby's first PCB? Is picrelated even solderable by hand (of the steady variety)?

Of course I only need the standard 4 USB pins, but Type-C is just so comfy.

>>1345607
>disassemble it
>reassemble it
I once fixed a numpad with that problem doing exactly that

>>1345609
>apply flux
>apply solder
>remove excess solder with wick

>>1345610
I disassembled it, but what do I do next? I hear I cannot solder tracers or some shit, but I am not entirely sure how to troubleshoot the exact key that is fragged

>>1345613
I know of the basics of SMD soldering, but for some reason that connector looks more intimidating than a QFP chip or something.

>>1345614
do you have a tester?
anyway, put it back together and see if it works now.

>>1345616
make sure to attach it firmly in place.
do those connectors require holes on the PCB?

>>1345618
I do ...... kinda... one the probes fell off , but i can still get a reading, opening it up didnt work, I think it a connection, not sure how to check which one in the entire row, i think it reads it as stuck

>>1345616
Are you getting it made for you? Just thinking that soldering would be the least of your problems if you had to drill those holes and etch the thing (assuming you don't have a CNC router).

>>1345585
this anon gets it

>>1345595
>I had hoped you'd curb your autism
hi redd!t

>>1345613
>GOTO 10

>>1345609
>0.50mm
sure, use a fine tip. personally, I'm a huge fan of bevel tips but might consider smaller conicals for this after a cerain amount of cursing
I haven't read up on the electrical joy involved in USB-C but one of these days when my head's out of RF mode I will probably do so

How do I even start working with mains power shit without dying? I study electrical engineering (second semester) and I`m just getting in touch with electronics, so far nothing that would kill me. I know the math and stuff but I don`t know the trade. (Like lack of experience I guess). I`m planning on doing a power supply for a amplifier or a 5A supply for grow leds

I am interested in doing current control for a brushed DC motor. I'll use an old maxxon motor rated for 24V and 2.4 Amps for continuous duty.
Do you have any recommendations for a motor driver?
I borrowed a Japanese one (servotechno) that does the job but it costs around $60 converted from JPY.

>>1345628
>mains>fuse>transformer>rectifier>filter

even books for children can explain this

>>1345632
>current control
Can't you just use a PWM solution if it's for unregulated speed control? Or does constant-current give you the sort of behaviour from the motor that you're looking for?

>>1345633
oh, and add a regulator too

>>1345635
>>1345633
thanks. As I so distracted I migth think I have brain damage I`m usually afraid things will blow up because I fucked up.

>>1345637
just dont touch things powered by mains
add a switch before the transformer so it doesnt get power as soon as you plug it

Can I just buy any old multi-meter, or is there something I should look for?

for general use, not just small electronics

>>1345634
I want to do torque control. I can do it manually with a PI controller but this require knowledge of the armature resistance and inductance. But then I have to worry about model mismatch.
Or I can use a current sensor, but it would be less troublesome to directly buy a motor driver with current capability directly from the shelf.

>>1345628
two tips on working with hazardous voltages or non-isolated circuits, as given by the late Bob Pease, analog IC pioneer (who, incidentally, lived a long and healthy life with electronics, becoming an hero by running his car into a tree):
>add a neon lamp and resistor into the HV part of the circuit while you're working on it as a warning lamp not to go sticking your fingers into it
>keep one hand in your pocket (literally) while working with HV to avoid current passing across the heart
third common-sense tip:
>use an isolation transformer if you have one and there is no isolation already in the circuit

>>1345622
Oh yeah, of course. I'm considering PcbWay or JLCPCB or something, one of those cheap Chinese printing shops that do 10x10cm PCBs for 5 bucks + shipping.

>>1345625
I don't think the tip would matter -- the SMD pins are just not exposed. So it's like a QFN chip but with a gap between pins and the outside casing which is also made of metal. Sounds like a pain.
There are also "straddle" connectors with SMD pins on both sides of the PCB, but they don't look too stable.
I dunno. I guess I'll keep looking for the most hand solderable kind of Type-C jack.

>>1345663
>the SMD pins are just not exposed
hand-solderability is as much or more a property of the footprint than of the device. if you can't get to the terminals, extend the pads so that you have room for the tip of an iron and some solder, which flows back and conducts heat (and draws more solder) over the pad and terminal. thin (0.5mm or less) solder wire and extra flux helps a lot

>>1345668
This makes a lot of sense. Thank you for the insight, anon.

>>1345577
This I understand. I'm using some encoder/decoder ICs which send a data burst four times and verify that at least three are correct, that is all handled for me.

All proper use aside, shipping units that are clearly missing components is disgraceful. It actually does work a little bit, from a distance of a couple of inches (with antennas), and not reliably even at that.

>NRF24L01
have not tried it. for now I only need one-way communication.

>>1345677
>clearly missing components
SHAMEFUR DISPRAY
it is true that chink QC is generally weak, and it's not usually in one's interest to shop too hard on price
care to name the vendor you bought from? I suspect some have better or worse QC, just as some are better or worse on shipping promptly
>nRF24
highly recommended, also modules containing it or its clones can be gotten for about the same price as the cheapie 433s. the only disadvantage is that you need a microcontroller with SPI port to set it up and talk to it

>What software should I use to layout boards?
>Altium
>EAGLE

I'm getting real tired of you double niggers telling people to use these two.

I work in industry and we don't even use either of these two because of their shitty practices and overpriced bullshit.

If you're going to be tremendous faggots and put a bunch of expensive software there then add Easy-PC as a cheap alternative. It takes the spirit airlines approach to schematics+layout, only pay for what you need, no recurring costs but you have to pay a small fee everytime you want to upgrade.

The best newbies can do is use Kicad, they can do anything they need to and then some for the low price of learning the software. Don't get memed into the faggy 10k+ boards per year mfg packages.

>>1345695
this is fourchan, nobody here actually buys software

>>1345695
>implying that 4chins shalt know their station as hobbyists and never, ever work with microwave or high-speed logic
your indignance and your ignance are adorable, triple nigger
next time I OP, I'll be sure to arrange them in decreasing order of general preference, instead of alphabetically

>>1345713
>Implying Kicad isn't beyond capable for RF or any other perceived voodoo magic

Now it's pretty clear you're a fucking exanigger that doesn't know what they are talking about.

>>1345646
Thanks my man. A transformer here would be 5 times the price of a power source. I think I`ll have to plan ahead, make so I have minimal time with mains on for testing and go with it. Has anyone here made a power supply?

>>1345741
that+dumpster diving for PCs to harvest their internal organs.

>>1345628
>a power supply for a amplifier
You need no power supply.

>>1345741
>Has anyone here made a power supply?
Yes, I used a transformer salvaged from a clothes dryer. Before putting it inside my supply I used it as a weight lol.

>>1345739
yes, dear, I have one of those sitting about a meter in front of me
I stand niggeringly corrected, microwave tools have been part of KiCAD since about 2010
that said, it still looks like a bitch and a quarter to output files for pick-and-place

>>1345741
you still gotta watch out for capacitors
and if you're designing any significant circuitry directly off the mains, seriously consider using a variac when starting it up
>made a power supply
I presume you mean of the off-line variety... not yet. as easy and tempting as all the instant off-line SMPS chips make that endeavor, flybacks are not something I have taken much time to study yet

If I have a Wirless sensor network in the configuration of pic related, and each node has its own clock, how do I synchronize them all to have the same or as similar clocks as possible?

The user can only set the clock on gateway node A, and the yellow lines between nodes represent 2 way communication of data(including clock data), and each node's clock can be read or written to by that node.

I asked this question in QTDDTOT, too but I don't think I'll get a response soon enough, so here I am.
Do simple buttons (not switches) have a signal bounce, too? And if so, should I debounce on a hard- or software level?

>>1345832
>(not switches)
Huh? The rest of the text makes it sound like you mean push-button switches. Assuming that's correct: yes, they have contact bounce, and the shittiest ones have plenty of contact noise too, because the contact is dependent on how strong the user presses the button.
Everyone debounces with software, unless there's some decent reason to not do so.

>>1343603
Alright lads.

I want to learn a programming language so I can crank out embedded electronics projects faster. So far I've been programming in python, should I move to C for microcontroller programming or verilog/vhdl and do FPGAs?

Which is more in demand if I were to get a job in that field? (Systems engineer atm, no programming required)

>>1345832
Yes, any mechnical switch, pushbutton, latch etc. Requires a debounce. If you have the software cycle time to spare, it can be a way of saving yourself space or wiring (I know some PLC programmers do software debounces).

Honestly thoiugh, I disagree with the other guy, I think it's best to debounce at a hardware level. Add a resistor-capacitor low pass filter and you know that when your input is 1, it's a damn 1. If you have software timers etc, it can go wrong depending on how fast a person presses and withdraws a button and your values, and may require some annoying tuning.

>>1345836
You're right, I meant push button switches, I guess.. I'm kinda retarded when it comes to electronics. I'm literally holding a soldering iron for the first time with just Ohm's-Law in my theoretical toolkit.. On the other hand I thought about saving some clock-cycles time thus having a longer battery life.

Plus I have not much space to spare on my board since it should be a handheld. Like >>1345847 said.

Some many questions.. But thanks for the quick response, that kinda helped me clear things.

>>1345837
Go for C. Everything after that will just be learning the syntax.

>>1345837
Those are very different paths you're considering.
I'd recommend going with C and C++ so you could then transition smoothly to Verilog/SystemVerilog.

>>1345815
i believe NTP has solved this problem quite nicely, but in short, compare the passage of time periodically to the next server up, adjust the speed of passage of local time accordingly, account for round-trip time when querying the next server up
>or spring for TCXOs and get 5ppm precision for just $2, then naively ask the next level up for time every hour or so

>>1345847
you don't need precise timers, silly. you just need to do nothing unless the last 50-100ms of samples, which you can take at relative leisure, all agree. no tuning to it if you're not in a big hurry
>it's a damn 1
only if you're lucky enough to have Schmitt trigger inputs

>>1345837
tbqh FPGAs are pretty niche and aren't as broadly applicable as the huge variety of microcontrollers that can be targeted by C compilers
otoh if one really must work on complex signalling faster than 50ns or so, FPGAs are the way to go for those who can't afford custom chips
of the two, go for C
>not learning Forth
>2011 + 7

>>1345641

>>1345641
buy the shitty $2 special from harbor freight. when you know you need more, you'll know what more you need
if you're feeling lucky, buy the $20 digital chink-o-meters from the home improvement stores on alibay

>>1345849
>>1345853
>>1345848

Thanks lads, will take your advice.

Following question, what's a decent microcontroller (not too expensive with development board) that you'd recommend?

I used to have arduinos which I've heard is pseudo C but I'd rather go with the real ANSI deal, as it were. Thoughts?

>>1345861
STM32. Very cheap, very good.
You can get an Arduino Nano-like board for 2 bucks from AliExpress. It's like a more advanced version of Arduino in most ways, but you can program them with Arduino IDE if you wish as well and then it'll just be a better Arduino without the added difficulty.

Check out this video for more info: https://www.youtube.com/watch?v=EaZuKRSvwdo

You might need a programmer for it if you don't have a device for that. Costs about 2 bucks from Ali as well (but original ones with fully featured debugging and stuff are 8-20$).
Also, STM just bought the best pro IDE for STM32 and made it free. TrueSTUDIO + STM32CubeMX (powerful project generator/chip configurator) is a combination with great potential.

>>1345861
everything
>>1345864
said. take the blue pill

>>1345149
underrated kek

>>1345695
>>What software should I use to layout boards?
>>Altium
>>EAGLE

Eaglefag here. I've used this POS for well over a decade and I'm jumping ship for KiCad.
The licensing updates have been so ridiculous for the past 3ish years.
I just don't want to have to learn a completely new suite, and all my current designs are already built in Eagle ;_;

>>1345882
fortunately for you, KiCAD has some EAGLE import capability, or at least I've seen a lot of development activity with names suggesting as much in the nightlies

I think I have some harmonic problems out of a transformer. I want to put a simple RC circuit on the output to test this.
For a low pass filter, is the Resistance value the internal resistance of the transformer?

>>1343603
Thanks for making me feel good about my soldering skills.

>>1345893
my god that would be amazing. I'm a KiCad n00b, so didn't know that.
individual designs that still are being supported I can go in and revise in Eagle. what I'm most worried about losing are all the parts I've built

Im trying to build a roadmap for embedded engineering, im gonna post it here an see if anyone has any feedback

F>>1345937
> design patterns
This is a joke right?
> python/ruby
For embedded? Seriously?
Seeing as there is absolutely NO electronics in there whatsoever I would say this is seriously software rather than embedded. If you can't build it and get it certified you are literally worthless as an embedded engineer. At a minimum you should understand the physics of laying out high frequency traces, proper grounding without creating antennas, etc. Embedded is getting hard-core now and the faster it goes the more effort you need to keep it happy and clean. Got to keep up!

>>1345948
Python and ruby are somehow used extensively in the automotive sector dont ask me why

So what do I need to learn/know if I want to build some LED fixtures? Like really high power aquarium/vivarium type shit, preferably controllable so I can vary intensity and color to simulate natural lighting and weather and shit...

On the one hand it seems simple, but on the other I've seen build threads on that sort of stuff and it looks incredibly complicated and expensive.

>>1345643
In which case you'll want a buck (or buck-boost) converter with constant-current capability. The LM2596-based constant-current boards (have 3 trimpots not 2) will work for this since they're 30V and 3A maximum (4A with a heat sink), but if your motor has a higher stall current you might want to look for something else. But if this does work you can remove the current-limiting trimpot and replace it with a panel-mount potentiometer and maybe some resistors to set the min and max range. Just set the maximum voltage trimpot to 30V and be sure to power it with a few volts more than the motor requires.

>>1345909
no, it's the impedance of the transformer, which is frequency dependent and not equal to the dc resistance

>>1345954
it can be as simple as pre-fab constant current controllers with a panel control for individual brightness of a few LED emitters attached to an aluminum heat sink with a possible fan, or it can get as complex as a small computer fetching actual weather conditions in the Galapagos and simulating weather conditions in the sky over the course of the day in the Galapagos including passing clouds, lightning and moonlight
Where Do You Want To Go Today?™

>>1345695
>I work in industry and we don't even use either of these two because of their shitty practices and overpriced bullshit.
I'm using Eagle for years, not paid anything, you know, it's ok for hobby to use a cracked software
If anything I'd suggest to use OrCad/Allegro or Mentorgraphics suite. These are also available "for free" for hobby use, just don't tell anyone and keep it installed on an external eSATA/USB3.0 HDD

>>1345909
You can calculate it instead, f_cutoff = 1/(2π*R*C). Even if you do put the filter on the end to see what frequencies are getting out it will be hard to measure them with only a true-rms multimeter, let alone anything worse. You essentially need an oscilloscope, though hooking a speaker up to it and getting the fft of it from Audacity also works. But unless it's a damn big transformer (or high frequency) you're not going to be getting any funny resonances with whatever stray capacitance you've got lying about the circuit. If you do have a capacitor in an AC part of the circuit I'd redesign it.

Can you post the circuit and/or what frequency it's meant to run at?

>>1345966
>impedance
Well it's easy enough to measure the complex impedance of the thing by measuring its DC resistance then measuring the amplitude across it with a known circuit element as a load.

do you have to remove flux after soldering electronic stuff?

>>1345989
with flux remover

>>1345989
no but isopropyl alcohol is cheap and tasty

>>1345968
except that EAGLE isn't all that great anymore. if KiCAD is good enough for CERN maybe it's good enough
also not cracking is one less installation step to fuck with

>>1345989
depends on the flux. assuming a not particularly sensitive circuit: no-clean = nah, RMA = an alcohol wash is not strictly necessary but is a good idea, anything more active than that = yes

>>1346010
>>1346007
what percentage do i buy and say some alcohol doesnt get wiped, this wont do anything to the circuit and will evaporate anyway right?

bit of a perfectionist in fixing stuff, sorry

>>1346011
i use 99%, which is recommended. i assume the filler in 70% is water which will probably evaporate poorly. 99% evaporates immediately if you blow on it.

>google decides you're a bot and you get to fill out 10 fucking captchas before you give up and f5

>>1346011
the very point of isopropyl vs regular alcohol is that iso is 100% pure
regular alcohol can get up to 98% I think

>>1346013
>>1346012

thanks guys! ill buy the 99% isopropyl

What would cause a 7432 OR gate's output to remain at 1.5v after going from high to low?

>>1346020
>excessive loading
>an input between ~0.8 and ~2 volts
>oscillation manifesting as 1.5v on a dc meter reading
>undervoltage fuckery
>fried chip
measure supply voltage, input voltages both ac and dc, and measure output ac again.

>>1346015
1 last thing, 99% isopropyl wont dissolve any plastic and i can use it to remove dust from my pc parts right?

>>1346027
i can't say that it won't damage *any* plastic but i've cleaned plenty of plastic headers without issue and obviously it works on ic encapsulants.

>>1346026
Thanks anon

>>1346032
ok thanks again!

>>1346011
Low percentage alcohols are just for medical. Get the highest you can.
Part of what makes it so useful is that it evaporates so quickly and leaves behind no moisture.
It's safe on most paints and plastics too. Almost all of them.

do i check the faulty row with power on ? or off

>>1346074
you check connections

>>1346077
Do i test ohms of inner and outer?

>>1345972
I'm having some harmonic spikes at that show up on my oscilloscopes FFT function at 180, 240, and 300Hz. They are coming out of my current transformer I have on a high voltage line.

That is then going into a 30:115 transformer and becomes saturated. That gives a constant 115+10% voltage out at the rated 5A.

However, since there are higher frequency harmonics It's saturating at a much much higher current than I expected.

So I figured I would just put an RC circuit on the output of the CT that can handle the ~45v coming out at saturation.
If you have a better Idea I'd love to hear it. I only know about low pass filters in theory

>>1346079
diode test, make sure things that should be connected are connected and things that shouldnt arent

Whats the best way to save as much as possible of this corroded connection? (Small red circle) I was thinking multiple white vinegar treatments, then typical flux & de-solder.

Schematic wise, the corroded connection is on an AC Adapter port. You can see that the red wire with corrosion is going to a battery terminal (larger red oval), but the black wire going to the other pole is disconnected. Do you think that both the red & the black were connected at that same (corroded) point? there is no other points that it would be connected to at that ac port.

batteries connected to AC?
wouldnt the black wire be connected with the other black wire?

>>1346089
Actual harmonics? How strange. I don't have much experience when it comes to HV or current transformers, but that's certainly something I've never heard of. What is the relative amplitude of the harmonics? If it's less than whatever error threshold of current measurement you're after then it should be fine to filter them out, but I'd use a fairly harsh active filter to get only the 60Hz, instead of a single passive RC filter.

Since you've essentially got a current source, you could either buffer the measured signal from that across a shunt, or try to push that current through the filter itself. The first one is going to be easier, but if you have to go for an analog method due to the voltage/current being too high then you'll have to treat the CT + shunt resistor as an imperfect voltage source (using the Thévenin/Norton equivalence V = I_CT/R_shunt). If your filter impedance is much greater than the shunt impedance you'll be fine.

Choosing what order filter you need is important, a normal RC low-pass attenuates at a rate of -20dB/decade (that is, per order of magnitude of frequency), especially since you have harmonics so close to your original 60Hz. See if you can replicate your original signal (multiple sources of different amplitudes) in Spice or some similar program and try out a few different filters.

>>1346179
For example. I've found that an AC analysis is more useful than a bunch of sources.

>>1345567

>>1346089
oh hey fiber tester anon, glad to have you back
I'd probably consider a "tuning" capacitor between the inductors, but maybe I've just been too deep into radio lately. worth a try

>>1346152
looks like that would be shorting the battery. I'd look closely at the other terminals with a magnifier to see if there's anything like a little stub of wire that came out of them, especially the one with the other black wire

>>1346179
I think this guy's building a power supply, actually

I'm going to use an attiny85 with a lithium battery (3.5 to 4.2 volts), will the attiny detect a voltage as low as 3.5 volts in one of its inputs as a HIGH?

(The attiny's VCC and input signal voltage will be the same since they come from the same battery)

>>1346219
you should be fine if you keep the input voltage out of the 25-75% of Vcc range

Tell me if I understand this correctly:

If you’re trying to use an SRAM Chip that has 2^19 memory locations and 8-bit word lengths, you would need an address bus 19 bits wide but you would only need your data bus to be 8 bits wide due to the size of the words?

>>1346259
correct, assuming you do want to use all 2^19 memory locations and that the device is in fact a 512kx8 device, not a 64kx8 device

>>1346219
You can find this info from the "DC characteristics" section of the datasheet.
It says the threshold for "high" is 60% of VCC, if VCC > 2.4V. That is, it varies with supply voltage and will be 2.1V with 3.5V supply voltage. On the other hand, the maximum input voltage at 3.5V supply is 4V.

>>1346278
In other words, whether it is a 512kbit device or a 512k byte device. Memory manufacturers like to sell their stuff in bit sizes.

>>1345966
>it can be as simple as pre-fab constant current controllers with a panel control for individual brightness of a few LED emitters attached to an aluminum heat sink with a possible fan, or it can get as complex as a small computer fetching actual weather conditions in the Galapagos and simulating weather conditions in the sky over the course of the day in the Galapagos including passing clouds, lightning and moonlight
>Where Do You Want To Go Today?™
The latter, absolutely the latter. Also want to DIY it all because I'm unemployed and bored and not in school anymore so I miss learning shit and taking classes.

>>1346336
>galapagos
you must really love your fishies
>first, find weather server for sea location
>then, turn weather server info into one or more bundles of weather data over some period of time
>then, calculate moon phase and sun/moon rise/set times, and incorporate their visibility and position into weather bundles according to time
>then, translate one bundle into animation directions and color observations
>then, use animation directions to perturb color observations
>then, translate each perturbed color observation into a numerical control factor for a particular wavelength of LED
>then, convert numerical control factor into electrical control signal
>then, convert control signal into high-current power for the LED
>then, power all your LEDs
iow I suggest you work your way up to it

>>1343603
Hey guys, can I get a quick MCU recommendation?

I'm implementing a WSN and 4 of my current nodes are using Arduino pro mini (pic related).

I have to order a few more soon, and was thinking if there are better options than this.

Eventually I will adapt the WSN to be part of a robotic swarm of small mobile robots.

>>1346179
The RC filter was a quick test fix to see if it would work.
The harmonics are pretty bad, I would say on the FFT analysis the first odd harmonic(180hz) it was 2/3 the height.

I'm trying to model it in Pspice atm, I'm having some issue with how to generate harmonics for an input wave.
Its a little difficult to choose the right filter, considering it going to have to handle high voltages.
I'm thinking an LC circuit might be a good first step
>>1346203
I'm always here, just usually just reading, and sometimes answering.

>>1346353
>>then, translate one bundle into animation directions and color observations
>>then, use animation directions to perturb color observations
>>then, translate each perturbed color observation into a numerical control factor for a particular wavelength of LED
>>then, convert numerical control factor into electrical control signal
>>then, convert control signal into high-current power for the LED
>>then, power all your LEDs
I guess this part is really more what I'm interested in, not sure how to do... The origin for weather patterns is simple enough, I may even have a go at creating completely sythentic/simulated weather or environment to drive it just for kicks.

But I guess my main issue is I don't really know jack shit about electronics beyond a general physics course back in undergrad. And I don't really know anything about micro controllers and like GPIO type stuff.

I'd love to find some books or other resources on that stuff but not sure what exactly to look for.

>>1346365
Maybe this will interest you: >>1345864

Blue Pills (the most popular STM32 boards) are a bit bigger that a Pro Mini, but they are worth it if you're willing to sacrifice some space.

>>1346374
What do you expect from a saturated transformer other than harmonics. Why is that a problem for a battery charger?

>>1346171
>>1346203

You guys are probably right. the black terminal is close enough to the disconnected wire.


Any idea on how to handle the corrosion at the point connections

>>1346399
Corrosion is only at the surface. If it aint broke dont fix it. Pump it off and solder it new if you must.

How are IC's turned off in the electronics industry? Like, if I have an OP amp which uses 1mA of current constantly and I want to turn it off, what is the standard way of doing it? A logic level P-channel MOSFET on the OP's VCC? N-channel on ground?

>>1346374
I doubt you'll need anything as big as the broadcast industry uses, but they're out there. polyester film caps with 630V ratings are easy to find in reasonable capacitances

>>1346375
sure, if you can draw the simulated weather in a window, you probably can do all the calculations necessary to drive LEDs
>general physics
you remember Ohm's Law, E=IR, and Kirchhoff's current and voltage laws? that's most of what you need for something like this
you've got a few options. RGB LEDs come in many shapes and sizes. RGB may not have all the color accuracy your fishies want though. so you could supplement with other LEDs for specific wavelengths as they prefer.
the most commonly used RGB LED for makerstuff is the WS2812 addressable LED. they take power and a data line, driven by a simple self-clocking code that has some generous, but not loose timing requirements, and are sorta bright. there are also LED driver chips like the PCA9685 that take a slightly more complex but less stringent, industry-standard serial data stream with separate clock, but then you get to/have to hook the LEDs up for yourself as you want them, possibly amplifying the output current with MOSFETs
maybe start with something arduino. among its good points: the community documentation you can find on the interweb is very accommodating toward total electronics noobs, the environment is well integrated and not super fussy, the arduino board communicates easily with your own programming on a host computer (so you won't have to bring in a whole TCP/IP stack to fetch marine weather or a whole graphics/touchscreen library for a UI), many of these chips/LEDs have easy to use driver and example code openly available
another consideration is that aliexpress/ebay have some of the more common chips mounted on boards and some types of LEDs mounted on aluminum or tape substrates, and you need only wire them together and maybe stick them to a heat sink to get the functionality you want

>>1346412
a lot depends on the circuit around it. a switch on the + rail is one way to do it, but probably not wise to do to a device with voltage on the inputs due to latch-up and leakage currents
some opamps and other devices have purpose-designed shutdown inputs with maximum off-current guarantees
some opamps also consume a hell of a lot less than 1mA in full operation

>>1345861
>used to have arduinos which I've heard is pseudo C
It uses 'Wiring' which is basically an implementation of Processing. You can think of this as a dumbed down version of c++. 99.9% of Arduino programming is done functionally, for some reason you don't use oop outside of libraries.

>>1346412
1. Use chips which have a power down pin. One example is AD8030. Too bad it won't turn "totally" off. Instead, the current consumption drops only about 90%. This is quite common behavior, but there are better ones, like LMV710.
2. Put the power hungry chips behind their own regulator. Use a regulator which has a power down pin.
3. Use high side switching. P-FET can be used like you suggested, but if the current consumption is low enough, you might be able to use processor pins or something similar for powering the chips.
4. Like >>1346419 said, there are also opamps which consume only microamps or even less. They suck in pretty much every other way, though.

Low side switching of IC power supply isn't exactly common, but might be a valid option for stuff like pressure sensor bridges.

I have a 5v 550ma rated phone charger. Is there a quick and easy way to cut its voltage and ma to a 1/4 just using resistors?

I am electroplating stuff, and instead of using batteries I want to use this charger.
The closer to say 1.0v and 100ma I can get the better, its slow but the plating ends up better. I have resistors and ceramic capacitors laying around.

>>1346451
(+charger) -- 39Ω -- two diodes in series -- (-charger)
output across the two diodes

>>1346470
that's more like 1.2Vdc, depending greatly on diode, but that's about as close as an I-have-junk-what-do-poster deserves

>>1346392
>Blue Pills
Just read up a bit about it. Is there a version that doesn't have the wrong resistor that I can buy?

>>1346470
>>1346471
i have a bag of in5404s laying around, im gonna put it together and see what happens

>>1346473
on many USB ports, they work just fine with the "wrong" resistor
on one copy I had, I tried swapping it out for the "right" one and the board failed to connect to USB, so take those cautions with salt

>>1346475
hmm, looks like you'll get about 1.1V with two, if the curve keeps going in that direction

>>1346480
>fails to pic related

>>1345861
I'm a PIC man, their PICKit 3 can program most of their microcontrollers (you program them on your breadboard or in circuit on a PCB, very flexible), they also have their own free IDE for project development. Documentation on all their chips and software is good too.

so i shorted the arduino 2560 attached to my ramps 1.4 board when the power supply touched my heated bed on a wanhao i3, the 5v regulator is working still (usually what fries when i fuck up something). The board works fine except i can't connect to my pc now.

Any likely parts to check?

If not how would i go about fixing this, because i have like 6 fried boards, and i'd like to bring them all back if possible. :3

>>1346503
the USB-serial chip, maybe, or the 2560 itself
you should really stop doing that though

I designed a Chebyshev-1 LPF, and it came out to be:
H(s) = 0.2756 / (0.9528s^3 + 1.45392s^2 + 0.742679s + 0.275633)

Which confuses me because, this function has no zeroes? Shouldn't it have zeroes. I verified in Matlab, and the frequency response is what I want, but I'm still hesitant about not having zeroes.

>>1346515
Actually it was 4th order:
H(s) = 0.2756 / (s^4 + 0.9528s^3 + 1.45392s^2 + 0.742679s + 0.275633)

>>1346310
Thanks for answering, and yep the datasheet specifies 512k x 8

>>1346515
cheb-1 lpf got all poles and they are located on an ellipse inside the unit circle.

>>1346553
Thanks, man.

>>1343803
jumper wires.

Holy shit. I just found this thread and it's everything I wanted /g/ to be. I never liked that board.

>>1346515
>linear frequency axis

>>1346630
Well I never liked pancakes so what are you gonna do about it

>>1345695
>>1345882
Circuit Studio. I'm telling you.

>>1346630
Glad to have you. Heed that we are a slow board, and a reply might not occur for many hours.

>>1345646
>the late Bob Pease, analog IC pioneer (who, incidentally, lived a long and healthy life with electronics, becoming an hero by running his car into a tree):

awwww, you gave me a sad. i used to love reading his articles in a trade mag, back when i did electronics test for a living.

that said, also just butch up. i've been shocked a few times myself from mains power. not as dangerous as we have to make it out to be.

>>1346074
the problem is probably the soft plastic conductive pads (not pictured) have worn out.

>>1343803
>>1343808
is correct, these often are copper-clad steel, and in like 30 gauge too. Total shit.
This is something you should buy on digi-key and not amazon. The amazon ones you never know what you're going to get.
Bud Industries makes the good ones.

>>1346720
yeah, I had his Troubleshooting Analog Electronics book. there were some really good street smarts in there. he was leaving the funeral of one of his lifelong friends, and from what I can tell about Bob, either had a heart attack of his own or just said fuck it
>what's all this Widlarizing stuff, anyway

>>1346768
>being this mad about copper-clad steel eggs
personally I just buy solid-core 22AWG or 24AWG wire by the 30m spool. it's somewhat more expensive, but that way I have a better idea of what I'm getting, don't have to waste Dupont pins, and can cut wires to length or discard wires as required. multiple colors would be helpful though

>>1346794
>funeral of one of his lifelong friends
Yeah, Jim Williams'. Pic related.

>>1346801
>pic
what

Does anyone know of an industrial rotary knob with 4-5 positions, that i can mount in a 22mm hole?
Can only find 3 positions

>>1346831
there are standard items from ABB and Schneider Electric

>>1346720
>not as dangerous as we have to make it out to be.
lets be honest, if you dont make it out to be dangerous then tards and normies will feel safer around it and be willing to do stupider shit.

>>1345301
Cut it shorter tho, you have a mile long gap

I bought this just past a year ago, I think I've used it with 2 or 3 small projects and it's dead. Going through the manual, the issue is that there's no voltage going out. Should I open it to find what component failed?

>>1346893
>Should I open it to find what component failed?
As opposed to what? Throwing it away and buying a new one? Crack the bitch open.

>>1346893
There's a fuse in series with the output. Check that. They're 4A fuses IIRC.

>>1346893

hmm, you can't operate a camera, so I'd say toss it and buy a new one.

>>1346898
>>1346902
I checked the output of the transformers secondary and it's reading 27VAC which is a little above the spec of 24+-10%. The next step was to check the wiring to the PCB, and it looks like that's fine. Step 3 is to replace the pcb.
There's no fuse that I can find.
What's next you think?
>>1346908
4chan erases exif data of the images you upload to it, and orientation is included in that.

>>1346915
>orientation is included
Then act accordingly. Learn to crop, scale and rotate before posting. No one needs megabytes of tilted blur.

>>1346930
>tilted blur
I think I'm gonna name my band that

>>1346930
Holy shit can you control yourself?

>>1346908
>>1346930
lmao imagine being this guy

>>1346930
i don't know if you've ever used a phone camera but they're designed to go in and out of focus randomly with no manual option. idk how you took this picture but it's probably a $20k setup.

>>1346893
>>1346915
I think I'm shit out of luck, the transformer is obsolete and I can't find it listed anywhere, and +15% voltage may have broken an IC on the pcb so replacing the transformer may not even fix it.
>>1346957
Some phones do have options for manual focus, but it's not that convenient to use on my phone. I did it on this one though.

>>1346915
The transformer secondary voltage is 27V because it is unloaded, you aren't drawing current on that circuit so the voltage is higher than you expect.

That board looks pretty simple, take some good, clear, photos.

If you have main power form the transformer and no output I would look at the device doing the output switching/driving. From the soldering iron work back from the plug and determine which traces are connected to the thermocouple and which are connected to the heating element.

The other thing could be the thermocouple is open which the circuit will most likely interpret it as an "over-temperature" and so the logic circuitry may be working but will keep the output off.

>>1347017
Just in gener it helps to thuink about how the thing has to generally work in order to debug.

You have some power supply that provides juice to a heating element based on a switch that driven by some logic. The logic is based on the heat setting, and the thermocouple feedback. Typically these circuits have electrical fuses that break on overcurrent OR thermal fuses that break on overtemp.

>>1347017
Just to be clear: That 27VAC is something I would expect to see. If you saw less than 24VAC it means the transformer is overloaded. And if you saw something fucky then probably the transformer's internal thermal fuse has gone, or a winding has shorted or some other fucky shit happened. Basically you are good from the transformer end.

Work your way back from the side that is not working.

good night and goodluck.

I recently bought a cordless WEN rotary tool kit. The charging port for the wireless battery uses a US wall socket but I'm in the EU. I had a US-EU socket converter and stuck it in to charge it. After a few moments I noticed a strong burning smell coming from the socket and smoke plumes rising. I quickly pulled it out but sparks shot out as I did that. Now the charging port is dead but the battery is still functional.

What do you think caused the problem? What exactly is the problem and can I repair it?

>>1347064
I presume the charger was made to run off 120V only. Stick that into 240V and bad things happen. A socket converter just makes it physically connect, you still need to make sure the device supports 240V 50Hz.

>>1347075
This. A universal phone charger has a switched-mode flyback converter inside it that will step down any voltage from 60VAC to 300VAC or so, since it has voltage feedback on the output and generous part tolerances. A US/EU converter, or any other cross-voltage converter is made to work with these devices. You'll need a transformer if you want to run a generic US appliance off EU power, but if it just has a switched-mode converter within it that happens to not be up to the tolerances of 240V, there may be something less expensive you can do.

I'd say that if the battery is working there's definitely something you can to to try repairing, the most drastic of which is removing the converter (part of the) circuit board and replacing it with a $5-$10 prebuilt 240V PSU board. So take it apart and see what you can see. Trying to trace the circuit itself is also helpful, as is knowing what the names of any chips are.

>Q for kicad users
if I have a design with a lot of hierarchical sheets can I make a PCB only from those sheets I select so I have design separated in multiple PCBs ?
e.g. I have analog and digital sections and want them on 2 separate PCBs but they are all interconnected in one project with hierarchical sheets

>>1347097
Doesn't have that functionality built in, the ways I've seen suggested are either export netlists from each sheet you want individually then combine them for each board. Or the kinda hacky solution, just draw both PCBs in the same document.

>>1347100
well I guess it isnt a problem if I get all the PCBs from the same manafacturer, Im a relatively new kicad user so I didnt know what is the preffered way to deal with this
thx anon

>>1346392

NodeMCU will be best for you. it is the almost the same zise as the mini but cheaper and has WiFi and Bluetooth

>>1343803

In my country they call it male-to-female jumper wires

I'm looking into resurrecting my favourite vintage amp, a couple of years ago a speaker cable shorted and let out the magic smoke. These two 100 ohm resistors in parallel right before the speaker connection burnt out. I just have a few questions I was hoping someone more knowledgeable could answer.

What's the chance that these will be the only things I have to replace? I'm concerned about the nearby transistor.

How to I determine what wattage replacement I need? Local store only has 100ohm up to 1W in stock.

I am 99% sure the cause of this was external from a dead short.

>>1347158
>How to I determine what wattage replacement I need?
you would need to know the voltage drop at those resistors, follow the traces (shouldnt be that hard on an ancient board) and figure it out

>>1347158
>What's the chance that these will be the only things I have to replace?
there's likely other components that are busted if you pulled enough current to melt those resistors like that, especially at audio signal levels

1W should be more than enough though, those just look like 1/4 resistors to me

>>1347182
>>1347171
Thanks boys, think I need to do a bit of reading but this is a good start

>>1347158
You can desolder the transistor and test it with ease. Those resistors will nominally take less power to fry than one of those transistors, but will probably take more time to fry, because thermal runaway. If you can trace the circuit and look at where that current was coming from you can deduce other parts that would be under stress. You only need to trace back the parts that deal with the amplified signal or the power supply itself, or from the output to the transistors. If the transistors are fine then the small-signal area of the board will be too. But check R613 and R614 and the one beneath them, they looks burnt, and check that the 100µF electrolytic is fine and that the yellow liquid is intentional. The big ceramic power resistors to the side are also possible points of failure, though if some 100Ωs were the initial points of failure I can't imagine what they're for.

>FR2
>no support for the transistors
Looks a bit shit tbqh

>>1347190
Thanks anon this helps a lot.

Those resistors that look burnt are only covered in soot from the two output ones that burnt, it's a vertical board so it made a mess everywhere, just missed them when I was cleaning it.
The yellow glue/jizz is intentional, some sort of shitty glue.

Here's the right channel board, both have perfect symmetry.

I'll give those transistors and caps a test and see how we go.

>Looks a bit shit
Eh, maybe, but the rest of the guts in this thing are very solid looking. It's one of the first quadrophonic amps so it was well regarded in its time. The main transformer and the heatsinks for the big boy transistors look pretty mean.

>>1347207
>>1347190
Here's a full view in case you're interested

>>1346967
Before you do anything crazy, check all the pins on that connector. I've had one or more fail on one of those. Did you measure the voltage as +15%?

>>1346967
Before you do anything crazy, check all the pins in that connector. I've seen them fail before, slip down inside where they can't make contact.
Always try the easiest thing first.

>>1347105
if the only differences are in your top-level schematic, you could make one copy of the .sch and .kicad_pcb files for each board, edit the top-levels as desired (excluding irrelevant sheets, including any interconnects), annotate and assign footprints to any new components (keep old annotation), export netlist from eeschema, and import netlist into pcbnew. thus you preserve your work on the layouts. the main disadvantage is that F8 to update board no longer works, you have to update the board by way of the netlist as above
btw, thank you for having this problem anon, I've been kinda wanting to do this to one of my projects in progress

>>1347107
NodeMCU is a software stack. the ESP8266 only has WiFi, no Bluetooth

>>1347207
>right channel board
if that still works, then you've got some luck on your side. you can compare dc voltages between the two boards and work your way toward toasted parts, if the board is not actively blowing smoke when powered up. sadly I think that with a shorted output your power transistors have very likely been toasted, if not also the drivers behind them

Arduino Mini? Micro? ESP32?
RPi? Tinker Board? Beagle Bone?
How does one choose between all the different controller boards available (both at a hobby and professional level)?

>>1347274
At a hobby level it's about what's easiest to use.

At a professional level it's about the support the manufacturer offers, the toolchain, the libraries available, security, speed, peripherals, and cost.

None of those products you mentioned will be used in industry outside some one-off test apparatus for company internal use only.

The best thing you can do is pick a micro, learn C and learn the micro inside and out. If you can do that then you can pretty easily transition to other platforms.

If you pick something like RPI with python you're going to make learning other platforms really hard for yourself in the future. But if you pick some ATMEL part without the arduino bootloader you are are going to be able to easily pick anything else you get your hands on up.

The tricky shit is multicore or other crazy architectures, but again if you have a solid foundation in C and can read application notes+datasheets you will be OK.

I'm not an elitist, you should do whatever you want, I am just speaking from what little industry experience I have.

>>1347283
yeah r e d d i t s p a c i n g I know.

fite me faggots

>>1347283
Cheers. Are the controllers I listed not used for final products because they're overpriced and too generalist (like, jack of all trades) for industrial applications? But besides that, I don't really understand why the arduino company would make all those boards when to me there are only very marginal differences...
Thanks for the info though. I'm currently teaching myself c++ and will move on to python after i'm comfortable enough. Hoping to get into robotics though i still don't know which aspect of it exactly (vision, control, AI, etc).

>>1347297
>industrial applications
That's too broad.
If you're making an advertisement display for a store should you just stick a RPi on a regular monitor? Yes.
Should you use a RPi as a dashboard computer for a car? No.

For MCUs it's a bit different. It's much easier to design and print a PCB with chips you need for a quadcopter or something than making a solid Linux computer when one is already available for 5-50$.

Those different boards aren't made for industrial applications, they are made for tinkerers who don't want to make a custom PCB, so they need a variety of packages to choose from for various use cases.


>currently teaching myself c++ and will move on to python after i'm comfortable enough
I still find it weird that in this day and age people are still learning low level first and high level afterwards. I feel like starting with Python is the way to go, and then you can go lower if you ever need to (which might never happen).

>>1347297
For industry, it depends entirely on how many you're building.
I build test equipment, with short development cycles, short lifetimes, and small production runs. We mostly optimize for quickness of development. Where I might use an arduino (even including the whole Uno board in a final product) a product designer might use an FPGA or a smaller microcontroller, because when you're making thousands it's worth additional development time. It's about minimum cost per unit for the level of function required. You'll never see an arduino (though maybe an atmega328) in a product to be sold.
Arduinos are aimed at the hobbyist market. You'd use a mini or micro instead of an uno when you need the compactness. you'd use a mega or a due where you need more pins or faster clock. There's always tradeoffs in terms of size, cost, power consumption, and capability. There are many ways to balance all those and that's why there are so many of these things available.

Has anyone used these sort of springy bent wire clips to mount PCBs? I got the idea that they might work for a project I'm doing, but the problem is that so far I haven't been able to find any good resources about designing them. In my case specifically the idea is kinda to hang a pcb on back of laptop (well, not exactly but anyways).

What I'd like to learn is basic stuff like what sort of wire to use (diameter? grade?), how big I need to dimension the holes for that, how to ensure that I can manage to thread the wire through the hole with all the bends etc. Also what sort of hook design makes sense to the end, I've drawn one variation, but idk, maybe more "J" shaped design would work better?

>>1345937
There is a lot wrong in that graph but the fact there isn't any signal processing in it, while there is the notion of OOP makes the graph is pretty much worthless. EE is signal processing first, all the other bullshit you've tried to organize into irrelevant categories comes after that.

>>1347283
>>1347318
Agree with these guys, if you're going professional, then its best to pick a microcontroller and really get to grounds with it instead of in the nice simple environment RPi/Arduino etc uses. However, if you're doing it for a hobby, the amount of time and effort you put into doing things "the pro way" is exponentially higher than making some nice DIY stuff.

Basically, DIY hobbies = RPi/Arduino, because of the inordinate amount of time saving.
Pro = Pick a microcontroller and C, learn both.

That being said, if you have literally 0 knowledge, starting out high level (RPi/Python) is good, because you're faced with a staircase rather than a sheer cliff, and you will save yourself time and effort in the long run by ascending a staircase rather than climbing a cliff. It's not idiotic to start out easy, it is moronic to start out at the hardest difficulty (I learned this the hard way).

>>1345864
Going back a few days now but I decided to go with the STM32. Hope you (or someone else) can help - can you point me in the direction of a tutorial or basic guide on using truestudio + STM32CubeMX?

Since you suggested them, you have experience in this directly. I'm still starting with C so I could use the help. I think it's gonna be rough for me to even get an LED to flash on and off but if I can do that I'd be dead happy.

even if you do hobby stuff I recommend going the low level way and learning simple circuits and MCUs without any abstraction layer first, that way you will create a foundation which will help you develop your skills for more complex projects and you will be able to design stuff from ground up without lookin up google tutorials

>>1347349
I'm tempted to write a quick guide on getting started with CubeMX because it looks intimidating at first but it's not really complicated. LED flasher can be made in a minute and I can get you there, but for everything else you'll have to go straight to the docs.

TrueSTUDIO is just an IDE like any other, but it's the best one for STM32 specifically. You don't really have to use it since CubeMX can export projects for various IDEs, as well as just Makefiles.

>>1347355
If you could do a quick guide for someone new to C I'd really appreciate it. I don't know how much time it would take though, if its too long just pointing me in the right direction to whatever you used to get started would be great.

>>1347343
>starting out high level (RPi/Python) is good
Have to disagree, you are going to be spending a huge amount of time relearning the same thing different ways and being crippled by other peoples limited choices of implementation.
If you are genuinely interested in embedded as a hobby then there is no harm in getting used to being confused by datasheets as early as possible. It's not just for picking and sticking, the better you know a chip the better placed you are to compare it with others to suit a job

>>1347403
this.
Its best to start /hard/ and work your way to assembly, then c/c++, then hll

>>1347403
>>1347433
>this is what /ohm/ actually believes
i shiggy my di/g/gy at you

>>1347325
Most home-machinists who need to make springs just get commercial "Music wire" like this
https://smile.amazon.com/Music-Wire-020-Diameter-Long/dp/B002WXNLTK/ref=sr_1_6?ie=UTF8&qid=1520803117&sr=8-6&keywords=music+wire

And just use a little bit of heat and some tiny needle-nose pliers to get the shape you want. 90's might be easier with a small vise and a hammer.

>>1347441
Not him. I think that starting with assembly if you just want to program well-known microcontrollers or even simply write higher level programs is a bit too roundabout, but a lot of people suggest starting with relatively low level languages like c and c++ and then if interested moving on to python.

9-pin D-sub that only has 7 pins from the factory, to a 15-pin.

The 15-pin doesn't have solder cups.

Any tips and or advice?

Where should I seperate the 9-7 pin side?

I don't know what's inside that rubber housing, is there anything I could use in there?

>>1347509
That's interesting.

>>1347480
>>1347433
>>1347403

I used to be in the "start with assembly" crowd until I was teaching other people. You need some fundamental understanding of what programming is, what you are doing, and how to break tasks down to describe them in a programming language.

Once you have that then you can explain how simple operations like arithmetic and memory manipulation are handled architecturally.

I still think exposure to assembly is a MUST for any embedded programming, but I think most of us forget we were introduced to programming in some form or at least some formal logic prior to getting into assembly.

>>1347509
Started cutting into pic related, was a bit difficult, am just going to sever at a lower point, now.

how do you all
1. stay motivated
2. get anything done
when it takes weeks, if not months for components to ship from china?

>>1347552
buy from mouser or digikey, problem solved, also you avoid counterfeit shit

>>1347552
most projects i abandon are in the design phase. once i've ordered a pcb and parts they just sit in a corner until they arrive and i feel like assembling them.

>>1347557
I don't have that much money and digikey confuses the shit out of me though

>>1347561
what can't you afford? integrated circuits are cheaper on digikey than ebay, and not that much more expensive than ali in the rare case where ali has your parts. discretes and passives are pretty generic and can be ordered on ebay or ali in bulk once and used whenever, except for when you need a specific part in which case you can often only get it on authorized distributors like digi anyway.

mechanical parts are comically overpriced though.

>>1347552
it helps to have a few things in the pipeline at once
also, express mail from China is cheaper than domestic mail in some cases

>>1347558
also this

>>1347561
not much to be confused about
>enter search term
>choose proper category
>enter quantity needed
>sort by price
>find cheapest item in stock with suitable specs
>click buy
>repeat

quick qttdtdtdtot.

On a wall wort (power adapter, w/e they're called) how can you tell what wire is the positive? I've been using a car charger for electrolysis but want to try something with a bit less amps. and this 12v 2a plug is perfect (though i may want to use even smaller if i can find a one).

Both wires are black, one has a broken white line going down the length.
I assume a multimeter could tell me, but I'm not sure how to use one properly because I am electrically retarded.

>>1347616
learn to use the fucking multimeter.
go on youtube, search "how to use a multimeter" and you'll get plenty of tutorials.
You can learn anything on youtube these days no matter how retarded you are.

>>1347616
the dashed wire is *supposed to be* positive. it was on the last wall wart i repurposed.

anyway with a multimeter just turn the knob to "V" and put the red probe on one exposed wire (while it's on) and the black probe on the other. if the resulting number on the screen doesn't have a negative then the one the red probe is on is the positive wire.

>>1347616

>>1347625
i think he intends to chop the jack off and just use the bare wires

>>1347625
hmm this might actually help a bit. I never realized thats what that symbol meant. Thanks anon.

>>1347623
Thanks anon. makes sense.

>>1347620
I would, i just never have a reason to use one. I will eventually as i keep needing it for specific things.

>came across old boombox while traveling
>speakers have clipped wires
>no toolson me
>stick a nail into the 6.3mm jack
>move it around while touching it with 3.5mm headphones
>get only slight background noise

Does that mean the thing has busted output? Or is what I did just stupid and it just wouldnt pick anything up either way?

(its SANYO C20, sorry for the shitty photo)

>>1347626
>jack off
mild zoz

>>1347726
>stick a nail into the 6.3mm jack
no, it means you're a fucking idiot
get a proper 1/4"-1/8" headphone adapter and try again

>>1347726
what you did is just stupid and it just wouldn't pick anything up either way

>>1347726
>stick a nail
My sides

>>1347384
Try the authoritative source

>>1347728
he probably shorted it fucking the amps

>>1347728
>>1347759
>>1347760
Well, I literally have nothing useful on me, I didnt even have the nail, had to wiggle it out of a wall in a bathroom at the nearby gas station!
Im in a smallass hotelroom in a smallas town, no idea where I would get proper jack adapter.

How do I test the thing to determine whether ist worth taking it? I mean it lights up, the tape drive spins, the record deck spins. So the only thing that could be fully dickered is something on the amp side would be my guess.
Man I wish it didnt have the speaker wires clipped...

Do I just leave this on the side of the road?

>>1347765
Shouldnt the headphone output be too low power to mess up stuff even if you short it?

How hard is it to install all the pick-ups and pots for an electric guitar?

I have a background with electric RC cars, doing all the assembling and soldering. Is this any harder? I know someone who has an guitar that needs its components installed and soldered and he wants me to do it.

>>1347770
You are well prepared. Identify components and look for instructions, there are many.

>>1347770
Electric guitars aren't SMD, you'll be fine as long as you can make attractive, clean, strong joints. That means not OP's pic.

Knowing the laws of electricity and circuits will help for troubleshooting or modifying a circuit, but if you're following well laid-out instructions you should be fine. Remember that electronic components are usually within fairly loose tolerances, so ±10% isn't uncommon. Also don't short voltage sources across low resistances.

Does anyone know whats the part number of the regulator used in this:
https://www.ebay.com/itm/4X-plus-or-minus-12V-Power-module-2-8V-5-5V-input-output-5V-DC-DC-converter-MA/121990297977?epid=508177620&hash=item1c67303979:g:n5MAAOSwf05XPA5e

Schematics with values perhaps?

>>1347890
ask a vendor maybe? all I know is it's "definitely imported chip"
in fact, the general principle can be applied to any boost converter if you don't give a fuck about cross-regulation, Pic related

>>1343603
I am trying to detect the caps and num lock leds on my computer over usb with a teensy++ 2.0 with the arduino ide, does anybody know how this could be done or point me to some library?

>>1347763
That's more for implementing C rather than programming C though.

Any advice on getting started on programming PICs?
I'm using a PIC18F2520 that I had lying about.
Basically when it boots, I want it to get time information over UART from a GPS module, write this to a DS1307 RTC over I2C.
Then I want it to keep reading the time from the RTC and shifting it out to 3 shift register.

Should I get into programming using assembly or with C first?
I've a fair amount of experience with Java so may find the C a bit easier

>>1347923
>>1347890

Chink sellers, they dont know shit about what they are selling.

Looks like this matches the circuit, just by looking at the top layer.

>>1347976
probably C
PIC assembly is pretty shit tbqh

>>1347970
>lolduino
nah
how about you actually learn what you're actually doing http://www.usb.org/developers/hidpage/HID1_11.pdf
>hint: it's all in the output report, and keyboards don't share modifier lock state under linux

>>1347979
kek, point taken
looks right

>>1347760
>facebook

>finish mechatronic/robotic degree
>Feel like I know nothing still
I'm starting my first real job in two weeks, a graduate position at a large company. I'm stressing that they'll let me go because of my lack of knowledge.

>>1348225
Chances are they expect that, and will be training you how to do something significantly different from what you've studied anyways. But that's just a guess.

>>1348225
If its a large company, they will 100% know that. Just be prepared to learn more, and if you're struggling, first few months you may need to look stuff up in your spare time.

Don't be afraid to ask questions, just be sure to phrase them in a non-ashamed but non-arrogant tone. You shouldn't feel bad for not knowing something, but don't act you know it all because you have a degree. Because you don't. Most of what I learned on my degree has been useless tbqh.

>t. 1 year out of graduation "engineer"

>>1347769
it's relative. transistors on ICs are very smol indeed

anyone wanna make a new bread?

>>1347976
do not buy a pickit 2, even though you can get them to work sometimes it's not worth the hassle. chinese knockoff pickit 3 works fine. download mplabx but don't use the cancerous IDE. it's not as bad as TI's CCS but still easier to compile using their command line tools and program with the IPE (or pk3cmd)

>>1348006
i have to counter this guy. i find pic asm to be fun and stupid simple. i wanted to learn it to begin with though and there's nothing wrong with C.

>>1348225
>>1348261
as a marginally more experienced engineer that gets to mentor new hires, my tips for not being hated are:
>take notes when you ask questions so you don't repeat them
>actually read (skim) the documentation so you don't need to ask trivial questions like "what section is this guideline defined in"
not knowing a rule exists or where the documentation can be found is of course fine. in my experience it's fairly hard to get fired once you've been hired, but negative perceptions of you as a lazy millenial will spread if you don't show signs of trying to stand on your own after a few months. frankly even that perception doesn't matter so much depending on your ambition.

>>1343603
I am currently looking at using an Atmega 328p into a project, i intend to use the dip for my prototype and an smd version for the final (if all goes well) one thing i noticed was that it is a dip 28 pin but qfp is 32 pin, can i get some help on why this is and what the diffrence is?

>>1348424
DIP Version

>>1348424


>>1348424
QFP version

>>1348424
it must be hidden in your pics

>be me
>third year CpE
>decide to open of Art of Electronics
>look through chapter 1
>no nodal analysis
>not a matrix in sight
>throws out half a semester by saying capacitors and inductors are useless in DC cicuits

this book makes my circuits textbook look like regular math but with funny shapes

>>1348454
also, I just now noticed they have an appendix section for ICAP/4. my uni has this software but i've literally never heard of it outside of that.

>>1348451
lel

>>1348424
>why more pins
leadframe cavity vs. die dimensions, cross-sectional area of leads/bond wires, less resistance/inductance for power supply pins, other thermal/mechanical considerations

>>1348454
>C/L useless in dc circuits
except circuits that are purely dc aren't usually very useful

>>1348451
yo, two more

>>1348469
Thank you, that makes sense

HX711 has a +-20 to 80mV differential input, and a 24bit ADC, its designed for loadcells, what could prevent me from slapping a thermocouple on the inputs instead?

NEW BREAD

>>1348497
>>1348497
>>1348497

NEW BREAD

>>1348488
The somewhat annoying common mode voltage range?
Not that you can't bias your thermocouples suitably, but still.