/ohm/ - electronics general

>>1239226
>outdated technique
For kilowatts of ac? Are you quite sure?

>>1239338
Where are you? Germany? Have you tried your local Premier Farnell company?

>>1239434
w-what is she planning on doing to that poor board?

>>1239456
Just adding an RF shield. Don't worry, it'll be fine, anon.

File: static-voltage-regulator-diagram-s.jpg (58KB, 626x963px) Image search: [Google]

58KB, 626x963px

>>1239436
>quite sure?
Yes, the Alibay servo model I posted does 7 kW. Here's a fully electronic version, 50 kVA. In the Radio-TV-tube era people had magnetic stabilizers, no electronics, no switching, fully continuous and totally robust. Tap switching was a clumsy amateurish approach that should be phased out.

File: 9999c840-f274-4477-9c1c-f92a8708cfee_300.jpg (6KB, 300x300px) Image search: [Google]

6KB, 300x300px

>>1239434
How could I hook up an rf remote control to one of these bad boys? I'd use one of the pre made ones but they don't sell ones that handle 15 amps

>>1239483
Big relay+pre-made one or some other kind of receiver.

File: 1496607074633.png (85KB, 679x936px) Image search: [Google]

85KB, 679x936px

>>1239466
Noice.

>>1239483
Did you check spamazon? You can get them for $50, 15A rated.
The simple, cheap(?) way is to build a 20A relay with 120VAC coil into a project box with two cords and an outlet. The power cord should be as heavy as needed for the load. The outlet neutral goes to the power cord neutral, the outlet hot goes to the common contact on the relay, and the power cord hot goes to the NO contact on the relay. The coil cord, which can be lamp cord, goes to the coil.
Plug the coil cord into one of the commercial switches and plug the contacts into the mains. It's not very pretty but it's functional.

>>1239502
impressive feature set for $50, with the safety and stripper. i wonder if the relay is latching.

>>1239510
It's whatever you want it to be, baby. Look up Z-wave.

File: 1342285389230.jpg (64KB, 600x334px) Image search: [Google]

64KB, 600x334px

>>1239466
>IGBT
Pic related. Fundamentally they have an energy-needed-for-switching, and it's on the order of milljoules. Doesn't sound like a lot, until I had an application I needed to be at least 1MHz. Their switching times are pretty bad too once you get above 100khz. They are amazing at high voltage high amperage, but damn. At least a kilowatt dissipated JUST IN THE IGBT

File: Screen Shot 2017-09-06 at 2.04.55 PM.png (168KB, 1572x1296px) Image search: [Google]

168KB, 1572x1296px

I need to switch a voltage going to a comparator from one voltage set by a potentiometer to another from a clock. It doesn't have to be using two potentiometers as the voltage dividers themselves, but I want the two voltages to be able to be set linearly and independently by turning two pots.

I thought since one pot will always give a lower voltage than the other, I could use a 500k pot and a 10k pot and simply switch the 10k voltage over the top of the 500k voltage with very little difference, but I found that in simulation the FET I used dropped too much voltage across itself and it didn't work. My backup idea is to have one voltage set by the pot and the other set by shorting the voltage to 0V through a rheostat, but this lower voltage isn't linearly controlled by the rheostat and is dependant on the voltage set by the other potentiometer. Any thoughts?

As seen in pic related, there's a slight voltage drop across the FET preventing the voltage from reaching the 3V. This isn't particularly a problem as the voltage will change fairly linearly, but if there's a better solution I'm all ears. The 500k pot might also interfere with the input impedance of my comparator, but I haven't messed about with that yet.

>>1239548
You could use a pfet and nfet tied to the same trigger. at <12V there's usually no risk of exceeding base to emitter or gate to source voltages. You could also use a not gate, where the + is the higher voltage, and the negative is the lower voltage, and the pulse swings between the two (AC couple with a DC biased around mid or whichever is safer in the pulse-off state.

>>1239555
And that would let me have the 10k pot be at a higher or lower voltage than the 500k? I don't need that at all, I was just wondering if there was a better method. Ideally I'd have a mechanical SPDT relay and simply switch between the two voltages, but I don't think I could bear my circuit clicking at me.

File: 1497774327624.png (7KB, 567x672px) Image search: [Google]

7KB, 567x672px

>>1239548
What've you got to work with, then? If linearity is not so important, you could use Pic related to switch a parallel resistor in and out of circuit, perhaps with an NPN instead of NMOS. The op amp should have a nice high input impedance anyway and won't care too much.

File: IMAG0786.jpg (4MB, 4000x3000px) Image search: [Google]

4MB, 4000x3000px

Repairing a dead right angle drill, what's going on with these wire connections? They're all firmly affixed, but what am I looking at? I've just never come across it before and it's piqued my curiosity.

File: IMAG0787.jpg (3MB, 4000x3000px) Image search: [Google]

3MB, 4000x3000px

>>1239573
Closer picture. With the exception of the two leads going to the LED, and the two going to the motor, they're all like this.

>>1239562
That's what I meant by:
>My backup idea is to have one voltage set by the pot and the other set by shorting the voltage to 0V through a rheostat,
>but this lower voltage isn't linearly controlled by the rheostat and is dependant on the voltage set by the other potentiometer.

I don't mind the non-linearity too much, my main problem with it is that this second voltage when the transistor is active is dependant on the primary voltage.

Why would you suggest an NPN instead, would they have a lower voltage drop or some other more ideal characteristic?

>>1239573
spot welding?

>>1239561
You're thinking way too much about this. A biased not gate coupled with a 50k pot between Vhigh and Vlow would let you swing between the two voltages. If your stuff is sink-safe it's pretty easy, otherwise you need some crossover rejection like an unbiased class B amp

>>1239577
I want to be able to change each of the two voltages with precision and independently from one another. If moving the potentiometer will significantly change both voltages, it makes adjusting both that much more complicated.

>>1239575
They may be easier to drive from a 5V supply, mostly. Depending on how much offset current the op amp supplies, the Rds(on) at Vgs might be insufficient to switch as you might like. I think you would do better with 1/4 CD4066 in there, or maybe a CD4053 triple 2-to-1 analog mux instead.

>>1239581
What is your actual goal here? You seem to be expecting a linear output despite being driven with sinusoidal waves. It's much easier to bias your point to change the trigger level than it is to change your voltages.

I even gave you TWO different ways of switching the voltages with and without crossover, and you seem to be missing it.

>>1239583
Running a test I can actually get better results with an NPN, though it depends highly on the value of the base resistor.

>>1239588
The sinusoid is just there to try and visualise the result, the aim is to switch between two PWM duty cycles, each one set by each voltage level. The transistor input voltage also won't be at 0V-5V, but rather at around 1.5V-3.5V as it's the output from an opamp oscillator.

It's to make bike lights with variable strobing, and stems from my design philosophy of "if you don't know what to set it at, add pots and figure it out later".

>>1239591
I'll mock it out on a breadboard later and see how responsive I can get it with the poor output voltages of my opamps, assuming I have nearly correct pot values lying about. For now the double pot circuit seems more promising than the 1 pot 1 rheostat circuit, so I'll give that a shot both with a FET and a BJT. I also need to make sure that the opamp output voltages won't give me different frequencies.

>>1239576
Maybe? It looks like it has something copperish melted between the wires, and there's not the usual telltale signs of spot welding, but I dunno.

>>1239594
>>1239576
Doesn't much matter either way I suppose, this things fucked for the time being. Turns out the problem was the last guy who repaired it; the gearbox is missing rollers. Into the doctor bag it goes!

>>1239596
I suppose if you collect enough broken tools you'll eventually be able to fix some with parts from others, but I don't know if it's worth doing. But I'm well on my way there in terms of broken soldering irons.

>>1239591
You can use a basic dual BJT oscillator or a 555 chip.

>>1239597
Probably not so much in the case of soldering irons. The $5 right angle drill needs a gearbox, the $15 impact needs a thorough cleaning and a $10 brush assembly after I sanded down the commutator, and the $130 still-under-warranty angle grinder I paid $25 for needs new brush holders and brushes. Gonna fix it all up and sell/swap the cordless so I can get some Milwaukee and Ryobi stuff.

>>1239605
I could never get a BJT oscillator to work, and just using a single quad op-amp IC takes up less space and lets me generate a triangle wave for PWM purposes with just one opamp. I also don't have a scope, so I have to simulate everything to make sure it will work as designed, and I don't know how to import 555 timer models into LTspice.

File: tp-link-wifi-powerline-extender-kit.jpg (41KB, 680x407px) Image search: [Google]

41KB, 680x407px

I have one of these TP-Link internet extenders and while it works alright, it does tend to lose connection requiring me to unplug then replug it into the wall. It happens often enough to be annoying but at the moment I have no other options for getting internet in my room.
I've got fuck all knowledge about electronics and I have been wanting to learn. I was thinking that as a project I might automate the process of resetting it so I don't have to get up every half hour. Perhaps have a relay controlled by remote, or have a Raspberry Pi ping Google every minute and trigger the relay if it doesn't get a connection.
Does this sound like a good starter project or does the fact that it would be connected to the mains mean it's a bad idea?
Bearing in mind I have the skills to create a script for a RPi, it's just the actual electronics that I'm new to.

>>1239634
why not get a wifi card (or usb wifi plug) with a big antenna on it?

>>1239636
It's a big house with thick stone walls, even if there were no walls between me and the router I'd still be more than twice the range where my phone would get no signal from it. Plus there's the fact that while there are probably many better alternatives to my idea, this one would would help me learn about this stuff.

>>1239544
Are you trying to be funny?
Why would you be banging IGBTs at 1MHz at an application like mains voltage regulator?

>>1239638
well I don't think you'll be satisfied with a solution that still involves losing internet access so many times during the day
but sure, the concept of checking for internet access and then switching it on/off using a relay would work

>>1239593
Alright, using a 10k trimpot and a 500k potentiometer I'm getting pretty nice results with the BJT+100k base resistor. With one set of pot values I'm getting the 10k pot's voltage fluctuating by around 200mV, which is somewhat high, but it's probably due to the 500k pot giving 1V by itself, and so having a lower resistance to ground. In the final version I'll only be moving the voltage around in (and a little outside) the 2.5V wide range in the middle since that's the size of my triangle wave, so I shouldn't get this fluctuation at a large scale in the final build, especially after tuning the base resistor.

I'm considering putting some pretty large resistors on either end of the pot (2* 470k if 1M pot, 2*220k if 500k pot), and was wondering if a collective resistance of 2MΩ could prove difficult to my comparator, the LM393. If I'm understanding the datasheet correctly, the maximum input bias current is 100nA, so a rough approximation with 5V gives me 50MΩ, which being significantly higher than my 2MΩ leaves me with no cause for concern?

>>1239658
>tfw your red LEDs have such a high forward voltage that your FET can't drive two of them from 5V
A shame, but I guess the extra 100mW is hardly going to pose a problem as far as power consumption goes.

File: 1496041355618.png (35KB, 400x417px) Image search: [Google]

35KB, 400x417px

>>1239691
Go bipolar. All the cool kids are doing it, and proudly.

>>1239634
Seems to be fairly common:
www.google.com/search?q=TL-WPA4220kit+dropping+connection

File: q.png (11KB, 1152x436px) Image search: [Google]

11KB, 1152x436px

Could someone please tell me the formula for the 647.1mV? I don't know how to calculate it, or what to google (didn't find anything useful). I've already spent too much time on this.

>>1239691
>such a high forward voltage
how much at which current?

>>1239658
100nA through 2Mohm develops 200mV. If you can deal with so much offset error, alright, but you might lose the top end of your adjustment range. Personally, I'd reduce the whole resistor string to no more than 500k.

>>1239702
Just a guess, but are you looking for Thévenin's theorem?

>>1239702
>don't know how to calculate it
Recipe using equivalent transformation:
Convert upper branch to current source
Ieq1=1V/2K=(1/2)=(5/10)mA
Convert lower branch to current source
Ieq2=3V/5K=(3/5)=(6/10)mA
Add currents to find Ieq
Ieq=Ieq1+Ieq2=(11/10)mA
Find equivalent parallel resistance
Req=2K||5K=(10/7)K
Find effective load resistance
Reff=Req||1K=(10/17)K
Find output voltage (across 1K)
Uout=Ieq*Reff=(11/10)mA*(10/17)K
Uout=(11/17)V=0.6470588235V

>>1239712
>Just a guess, but are you looking for Thévenin's theorem?
>>1239702

Or superposition, where you set the other voltage source as a short, which would be to ground in this case.

so the 1 v is on a voltage divider with 2k above 5k in parallel with 1 k which gives
2k above 833 ohms, or 0.294v

the 3 v is on a voltage divider with 5k above 1k parallel with 2k, or 5k above 666 ohms, or 0.353 v

superposition = add the two voltages to get 0.647 volts.

anon you should learn the basics. ohms low, parallel resistors, superposition, and more advanced topics like thevenin's theorem.

>>1239712
>>1239716
>>1239718

it's kinda like linux utilities: there's usually more than one way to do it. Use the one that makes sense to you, but try to understand all of them.

>>1239719
>>1239718
>>1239716
>>1239712
Thank you all! Now I gotta keep that stuff in my head!

>>1239706
2.3V each at 20mA, so I'm left with 0.4V across the FET if it could push 20mA at that, which it can't. Or at least not at the 4.5V I'm driving it at.

>>1239697
Will BJTs reduce that voltage drop? I might even be able to get away with using the BJT itself to limit the current if needed, but I don't enjoy relying on a transistor's unpredictable hFE.

>>1239712
The offset error I'm seeing at the moment hould be decreasable by changing he base resistor, 100k worked for a 2N2222 in simulation, but I'm using a BC547 here and I know they behave differently. This offset is mostly independent of the high resistance divider, and as I said I don't mind losing some of my adjustment range since the PWM triangle wave will only occupy half the space between the rails. The resistor string is at 500k at the moment and that's where those results came from, but the greater they are the less the low value potentiometer's voltage will be effected. I haven't tried to hook it up to the comparator yet, but if doing so causes the voltage to sag depending on the comparator's other input or output, then I'll have to go for a lower resistance divider. But if it's just a constant hundred mV or so then it won't bother me in the slightest.

>>1239736
>push 20mA .. which it can't
Sounds like the notorious 2N7000. Parallel a few, let them play 20Ω resistor.

>>1239720
>keep that stuff in my head!

Tips & Tricks

If you have small integer values with a simple common multiple/denominator the calculation becomes much easier using fractions. Example: 1K||2K||5K = (10/10)||(10/5)||(10/2) = 1/((10/10)+(5/10)+(2/10) = 1/(17/10) = (10/17)K. Exact value, no calculator required. If you have to parallel many weird values, RPN is your friend :)

Recipe using superposition:

Upper branch input shorted
U1=3V*(2K||1K)/(5K+(2K||1K))
2||1=2/3 because 1=2||2 and 2||2||2 = 2/3
U1=3V*(2/3)/(5+(2/3))=3V*(2/3)/(17/3))=(6/17)V

Lower branch input shorted
U2=1V*(5K||1K)/(2K+(5K||1K))
5||1=5/6, 2+(5/6)=17/6
U2=1V*(5/6)/(17/6)=(5/17)V

U=U1+U2=(6/17)V+(5/17)V=(11/17)V

File: wires.jpg (65KB, 800x600px) Image search: [Google]

65KB, 800x600px

how do you store your jumper wires?

>>1239759

sorted by color and length. how else could you.

File: IMG_4679.jpg (572KB, 3000x2000px) Image search: [Google]

572KB, 3000x2000px

>>1239759
all the short one always in range on my desk (left). Longer ones, not sorted by length, in the drawer (right, obviously duh)

>>1239780
Is it just me or this webm won't play? New laptop so idk if I'm missing some plugins or whatever

>>1239789
ok well that explains it...

>>1239594
>something copperish melted between the wires
spade connectors are plated with that silvery thing, and when you heat them and add flux, the plating tend to degrade and the copper core (or copper plating, depends on the connector) usually turns copper redish
I've seen that happens many times when soldering to a spade connector and i'm pretty sure the flux does that to the plating
no idea what happens here with the spot welds, but i guess heat + melting metal does that

>>1239702
Kirchhoff's current law (KCL): total current in/out of a node equals zero:
(v-3)/5000+(v-1)/2000+(v-0)/1000=0
=> (2v-6)+(5v-5)+(10v-0)=0
=> 17v-11=0
=> 17v=11
=> v=11/17 ~= 0.647

Coupled with Kirchhoff's voltage law (KVL, total voltage around a loop equals zero), you can solve any resistor network (and also networks involving capacitors, inductors and linear voltage/current sources).

Other approaches (e.g. series/parallel formulae) can only handle a subset of cases (although this does cover most of the cases which occur in real circuits). Also, the KCL/KVL approach is the easiest to automate (e.g. if you want to write circuit-simulation software).

>>1239736
Given that you're working in a 5V system with op-amps that can't even get to the upper rail, I would think carefully about using MOSFETs at all unless they are specifically designed for logic levels. Try a BJT in simulation. Vce is usually on the order of a couple hundred millivolts when the transistor is saturated.
Also, the overriding one voltage divider with another thing is just doomed. Look at your circuit again. What is your Vgs *in circuit*? What is your transistor's specified threshold voltage?
If you really want to use two dividers, you'll need to switch one of them off whenever the other is on. Consider one pullup resistor and two rheostats switched by out-of-phase FETs, with whatever inverter you like.

>>1239759
All mixed up. In a bucket.

>>1239767
>not keeping jumper wires joined together in EIA color order

>>1239746
>2N7000
You got it!

>>1239934
I'm using a BJT instead of a FET for switching between the voltage dividers and it seems to work mostly fine, and I should be able to tune it to work perfectly by changing the base resistor. I'll try using a BJT to switch the LEDs instead, but I have a sneaking suspicion that I wired the FET up on the positive side of the LEDs instead, so I'll check that first. I'll look into switching the two dividers with an inverter, but if there's any significant delay I don't want it because this thing will be switching at up to 10Hz.

>>1239759
I just have a bag with a spaghetti mess of solid core copper wire, who even uses those jumpers if you can't make custom lengths that lie perfectly flat along your breadboard?

>>1239634
Run a constant ping to your router. I have the same one, it's a little too eager to go into power saving mode so you need to constantly run data through it to keep it from going to sleep

>>1240029
>10Hz
What reactances would delay it for so long? You'll be fine up to hundreds of kilohertz, at least.
In general, low-side switches are easier to design with and use.
>spaghetti mess of solid core copper wire
Good-tier, but you probably haven't built up a collection of modules with pin headers yet.

File: IMG_1136.jpg (1MB, 3264x2448px) Image search: [Google]

1MB, 3264x2448px

>>1240096
I'm just a little concerned about what voltage will be sent to the comparator in between the switch, and whether it will effect the kHz PWM output of the comparator. I suppose it depends on the 0.5-10kHz strobe square-wave clock's waveform, and I don't really have a way of analysing it. But as long as it is close enough to an ideal opamp oscillator, I shouldn't have a problem.

I've got a few header pins lying about that I solder to things whenever I need them, pic related is a small PCB I etched to practice drilling holes 2.5mm apart, but not much else.

>>1240029
>I wired the FET up on the positive side of the LEDs
You want a switch, not a source follower. The 2N7000 has a typical Rds of less than 5Ω at Id=20mA, Vds=0.4V and Vgs=4V. You will either need an additional series resistor or you reduce Vgs to get 20mA.

>>1240040
That's a bug, it should wake up and reconnect. Can you disable the power save mode? Firmware update?

>>1240114
Well I did wire it up to the positive side of the LEDs, hence their low current, but now that I have wired the FET between the LEDs and ground, everything's working much better. I still can't get the LED current reliably past 15mA with the FET there.

With a BJT there (BC549) I can't get above 10mA, so it looks like I'l just have to put them in parallel instead.

In other news, the comparator doesn't change the voltage from the asymmetric potentiometers significantly, so I'm starting to wonder what the advantages of using extra components (two transistors + inverter) to switch between two pots of the same size are over the current setup. Is it only limited to the slight voltage droop and slight increase in efficiency?

>>1240110
>whether it will effect the kHz PWM output of the comparator
A missed pulse or two to LEDs at kilohertz won't matter much.

>>1240146
>With a BJT there (BC549) I can't get above 10mA
What's your base resistor, 33k or so? Try a third of whatever it is.
>the advantages of using extra components
As you said yesterday:
>I want to be able to change each of the two voltages with precision and independently from one another.
Switching one off while the other is on prevents interaction between the two controls and allows each to be set precisely. The design with a FET between the two voltage dividers will not be precise because of the low and setting-dependent Vgs well below Vgs(on).

>>1240146
Now by messing about with base resistors I've managed to get 400mV drop, so I'm starting to see it being a better idea to use the extra transistor+inverter. But that's when I use a 10k base resistor, which I found was ideal for stopping voltage fluctuations at a particular setting, chances are it was the base current giving me that extra voltage boost. When I use my largest resistor lying around, a 5.6M, the voltage sag is never more than 20mV, which I think I'm fine with. Except the voltage on the other side of the BJT is barely changed, which makes stupid sense. Time to mock out the new circuit then.

File: leds.png (2KB, 187x271px) Image search: [Google]

2KB, 187x271px

>>1240146
Do you trust your meter? Try do it like this.

>>1240172
>1Ω
My "bab's first resistor pack" only contained values from 10Ω to 1MΩ, so I'll have to get a few more sometime soon. But neither the 2N7000 nor the BC54X (with 10k base resistor) can give me 20mA, so I don't have a choice.

I bet the forward voltage is so high because the LEDs are a red/orange in colour, not a proper red. Though maybe that was the warm lighting, as now it looks normal. I wonder how hard it would be to diy some photometry equipment, like a spectrometer, something that tells me intensity, and whatever else would be needed to analyse LEDs?

whats a good multimeter for a beginner under $20?

>>1240231
I'd actually recommend just getting two shitty <$10 meters instead, it lets you measure two things at once and second guess one meter's reading. Getting one of those LCR meters/component testers that EEVblog looked at recently is also a great idea. Only get a better meter if it has a brand name behind it or if it has a feature you need, like frequency or something else. I only buy something more expensive if I'm getting a really good meter, in which case it's better buying a $200+ model, because $50 and around that don't appear to be worth getting. Sub-$10 meters work just fine for 80% of hobbyist stuff, and one of those LCR meters fills in another 10% of that. The remaining 10% is stuff that requires a better CAT rating, true-RMS, auto ranging, frequency, etc. and you should be able to get some combination of those for under $20 excluding a legitimate CAT rating.

But that's just my two cents, I'd listen to a few other guys.

Can noise from the alternator of a car be magnetically coupled into a 3.5mm cable carrying audio and if so would a shielded 3.5mm cable (if such a thing exists) stop that from happening?

I have my car audio connected by aux to my phone and I get some weird noise that increases in frequency as I accelerate. Moving the cable can reduce the noise significantly though I haven't played around with it much to know the best placement and even if I did I don't really have a way to secure it in place.

I figure a shielded cable would be best since if there is magnetically coupled noise it'll couple to the grounded shield on the cable and not the signal line.

>>1240231
dt830 series

>>1240236
Yes, and to some extent, yes. But you could try putting ferrite beads on the cable going to the alternator to see if that helps, since if it isn't the alternator it will help you to find where the noise is really coming from. If it is the alternator, then turning on the AC will probably also give you an increase in audio noise, which you can test for. But shielding your audio cable in the first place is a good idea, so do that anyway.

File: Screen Shot 2017-09-07 at 8.37.35 PM.png (206KB, 1534x1338px) Image search: [Google]

206KB, 1534x1338px

>>1240206
I hate this transistor's threshold voltage.

>>1240261
Do you really need the diodes?
Also you can divide the inputs to the opamp down closer to ground by the same factor so that you have more headroom to work with on the FETs.
Also, dude, a buck fifty. https://www.jaycar.co.nz/4066-quad-bi-lateral-switch-cmos-ic/p/ZC4066

File: Screen Shot 2017-09-07 at 9.06.14 PM.png (170KB, 1024x1352px) Image search: [Google]

170KB, 1024x1352px

>>1240261
Wrong pic; it freaks out when I select 2N7000s as my transistors and won't go above 3.5V. Trying it with BJTs and I get a much better result, and I don't have to use extra diodes. I just hope my BC55Xs and BC54Xs are up to the reverse voltage. It does feel a bit daft using current switches like voltage switches though, I hope this result works in practice. 50MΩ resistor to simulate maximum current drain from comparator.

>>1240265
The FETs have built-in reverse diodes, so when one FET is on current can go back through the other FET and to the other voltage divider, and messes everything up. Dividing closer to ground is definitely a way to avoid this solution, assuming I do the same to the triangle wave, but it looks like the BJT solution is probably much better. So that IC is pretty much 4 SPST switches? Pretty neat looking, but I really don't want to make my circuit bigger than it already is with 1 DIP8 and 1 DIP14. But if this BJT method falls through then I'll check it out. More ICs should be available in DIP8 or other more compact formats, I hate installing 1 IC and not using half of it.

File: 1493263008093.png (286KB, 1275x1650px) Image search: [Google]

286KB, 1275x1650px

>>1240267
Oh, derp.
That IC is, indeed, pretty much 4 SPST switches.
>More ICs should be available in DIP8 or other more compact formats
They have been for years. Just not at jaycar.

>>1240269
I could always just buy a single solid state SPDT relay, but I would rather not simply because of their size and it's hard to get a SPDT one. BJTs aren't letting me get above 3.7V, or below 1.1V. Actually since my triangle wave will oscillate between 2V and 3.3V thanks to the shitty range of the opamp, this won't be a problem. Well that worked out nicely.

>>1240276
>buy a single solid state SPDT relay
If you can find one that's meant for conveying voltages faithfully, anyway. Usually they're higher-power devices.
You could use the third of four 4066s as your inverter, and maybe the fourth for something to do with audio switching.
There's also the 74LVC2G53 which is a single 2:1 mux, exactly what you need for this, if only you could source it locally.
>worked out nicely
The old engineering joke is that happiness is a straight line on a log-log graph. If you wanted more margin, you could reduce the triangle to 200mVp-p or less and still get useful comparator action.

>>1240278
I guess so, but that's just the default with the voltage divider on the non-inverting input being a 50/50 split. I could decrease the amplitude to get a more triangular wave, but for PWM I don't think it matters since the capacitor-triangle wave should have its width vary linearly with respect to time. I might do so anyway just to make sure I'm getting a reliable waveform, maybe a 1:2 divider like I've got in my square wave generator, it seems to give me a more predictable frequency when dealing with very low frequencies, but it's still a little off.

>>1240278
Also I've dumped the speaker idea because ALL 5 of the tiny speakers I bought died when I soldered on them, even for as short a time as I could manage. I'll put a 5V buzzer on the handlebars and see how fellow cyclists like being buzzed at. Worst case scenario I get mad and mount my 10mW laser diodes to the handlebars and try to set fire to their e-bike's battery as they glide effortlessly up the hill past me.

I realised that all this circuitry will have to be seperate from the main battery pack because it will take up too much space, unless the remounting of one of the cells to fit them all inside the sardine can+wood enclosure means I have some extra room, of course. Maybe I'll cut the can in half and add a small wooden extension in the middle instead, which would look really good if I wasn't throwing the thing together out of sardine cans and radiata pine. I trust that sandpaper and danish oil will make it not look like miniature pallet furniture.

>>1240236
do you also have the phone plugged into a power outlet?

>>1240206
>My "bab's first resistor pack" only contained values from 10Ω to 1MΩ
10 10Ω resistors in parallel = 1 1Ω resistor

File: DSC_0177.jpg (652KB, 2064x1272px) Image search: [Google]

652KB, 2064x1272px

The USB Port on my Blue Yeti microphone broke, and I want to rewire it directly with a usb cable, so I won't be using a usb port anymore (since I've got USB cable with a broken end and I don't want this to happen again).
But I don't know how to solder the five USB cables, I've labeled the needed ports on the left (1 to 5).

>>1240371
solder pins to them first and clip the wires to the pins
if it doesn't work, clip the wires to different pins until it does work

>>1240206
Pretty common because of how the 3 bands are

>>1240231
As long as you aren't going over ~50v it doesn't matter

>>1240377
Thanks, I got it!

File: led-red-clear-10mm.png (42KB, 591x772px) Image search: [Google]

42KB, 591x772px

>>1240206
>no 1Ω
Get 1Ω 1% and 0.1Ω 1%, always measure the voltage drop directly at the resistor, 4 leads, Kelvin style. Meanwhile measure the real resistance of your mA meter, leads, fuse, switch (if rotary wiggle a bit). Post picture showing your m( (facepalm).
>LEDs are a red/orange in colour
Proper red would be 680 nm. Cheap LEDs are only ranked for brightness, typical Vf is not even specified (added it in pic). I tested three of mine (10mm, clear, ~630 nm, 35¢) and found 2.05..2.07 V at 20 mA.

>>1240371

there's a standard color code for USB cables that most manufacturer's stick to: black red green white. but nothing is guaranteed. see http://pinoutguide.com/PortableDevices/mini-usb-cable_pinout.shtml

the important thing is to get the 5V polarity correct. as you can see in your pic, 1 and 5 have thick lines to they're for power. you can find which one is ground coz it's connected to the USB jack case. you can confirm this coz ground is also on the phone jack, and possibly the case of the pot.

that should determine ground on the mic. to find it on the cable, also test that it's connected to the case of the jack. then use voltmeter to measure 5V across red/black when you plug the cable into a computer to confirm.

as for white/green, if it doesnt work one way, reverse them. no harm done.

Kind of electronics related

I'm planning on making a portable 100M shortwave radio antenna and I don't know what size wire that I should use. I've got a sangean one which gave me this idea

>>1240422
i got bored and googled. link says 14awg is code compliant. it doesn't actually matter from an rf perspective though.

http://www.eham.net/ehamforum/smf/index.php?topic=38785.0;wap2

>>1240425
The sangean uses 16 or 15 awg which is what threw me off.
OK sounds good I'll just pick up a big spool of 14.
Thanks for the help!

beginner here.
i want to have fun with brain EEG.
you know those inductors plugs you can use to play videogames with your brain?

i want to build a brain alarm clock that wakes me exactly when REM phase is over.

problem is that the boards i've seen so far, either purchased or "have fun soldering your components" cost 200$ or more, but they are overkilling for what i need: i merely want to detect the stages of sleep and not actually fetch high quality usable input.
is there any cheaper lower quality alternative?
is there any chinese product i can cannibalize?

>>1240422
>a portable 100M shortwave radio antenna
100 metres? That's a lot. You will need a good ground and a tuner to really make use of it. In case of lightning don't forget to ground it before it grounds you.

Shortwave is more or less dead currently because of solar minimum. 41m and 31m has some activity, mostly middle east and of course CRI. What's that funny thing on 9995 kHz, time signal? Waiting for cw id..

>>1240471
I mistyped
I meant 100 ft but was thinking about getting a better receiver and my mind was stuck on meters

>>1240459
Either Thinkgear AM or you can make your own amplifier with opamps like this http://www.prointellekt.ru/EEG2.php Be careful, chineese like to sell fake opamps on aliexpress.

>>1240459

nope, EEG is complicated to do, and not many units are made, so expect it to be expensive if you want it to be good.

one company created a dedicated EEG chip, and is selling it in home gadgets like the NeuroSky MindWave Mobile BrainWave Starter Kit, for about $100. i expect it to be shit, but check the reviews.

in any case, i cant see myself being able to sleep with wires attached to my head, or wearing a headset, so maybe some other avenue is worth exploring, like movement detectors (piezos) or breathing rate detectors.

I want to replace some old rechargeble batteries that won't hold a charge. I know they won't hold a charge because I tried for 12 hours (the recommended time to charge) and agree that they did not produce sufficient output for my needs. Based on this I know I need to replace the cells inside, however I worry about shock. Is there a way to ensure that the old battery cells are fully discharged before I go and shock myself? I'd rather not turn on and run the item they go with for however long a discharge would take.

>>1240502

WTF dude, you dont have to worry about shocks until you get to 50V or higher. as for your question, ''how can i make sure batteries are discharged", you short them if they're small, or wire them to a light bulb if they're big, like ''coffee cup'' big.

>>1240508
>you short them if they're s
Bearing lion in mind this is a dangerously stupid comment.

>>1240508
>50V
depends how wet your skin is though. 50V is for perfectly dry hands, but having a bit of moisture (eg sweat) brings the limit down to 24V, and if you're by a pool the max safety voltage can be as low as 12V.
For AC it's a bit lower.

>>1240531
This, I just got done reading about voltage safety, it's not really wise to say that an electric current is ever truly safe unless you're adequately protected from shock.

>>1240480
>better receiver
Traditional or SDR?

>Waiting for cw id..
got .-. .-- --
RWM, a Russian time beacon on 9996

>>1240325
Sardine cans are stackable, if you add little flat hook-and-pin latches between them.
>make it not look like miniature pallet furniture.
Nope, you're fucked.

>>1240356
It only has 10 of each, and I'm not cramming all those on a PCB.

>>1240394
I'm not doing a 4-lead measurement that relies on my $6 faultimeter, neither am I going to plot a V-I graph of my LEDs experimentally. But making a device that can give you the Shockley diode equation for any diode you plug into it would be a pretty neat idea.

>>1240537
No way in hell am I going to buy another can of those bastards.

>>1240600
You could just cobble together a current source and measure once, at the current you're actually supplying them in the application, then work out Rseries from there. You don't need Kelvin connections for that, as the few milliohms of component lead won't matter for selecting 5% values.
Read the FET datasheets, look at the curves.
>No way in hell am I going to buy another can of those bastards.
Fair enough.

>>1240459
>>1240496
EEG is actually stupidly easy. You just need an analog source with >100hz sampling rate and an FFT. The harder problem is noise (electrical signals need to be amplified a lot, so the closer the amps are to the scalp the better) and the amount of channels for an interesting picture. A simple way is they make 4-channel audio input devices. If you could time-division multiplex it, you could get easily get 5 100hz channels if the bandwidth was only 1000hz and you left a full cycle width between samples. The waves are slow enough you probably wouldn't even have to worry about rise/fall times, and if you "overdrove" the signals (i.e. like 60% signal 40% timeoff) you could have time for the signal to settle (in regards to switching overshoots)

Having the cycle being directly driven from computer and only sampling after the settling time would guarantee that you'd only sample during the intended cycle period. What would be more annoying is ensuring clock accuracy, but if you used real-time clock recalibration it wouldn't be difficult.

>>1240609
Forgot to add:
http://openeeg.sourceforge.net/doc/

>>1240533
the only case a "12v" line is ever dangerous is in an inductive system where it's not really 12v at all during spikes anyway. if you try hard enough you can kill yourself with anything, and pedantry is the enemy of safety when it comes to good rules of thumb.

>>1240606
Or I could just put a trimpot in series and adjust it until I have 20mA and write down that R value. If I do it in the same arrangement with a FET powering it from a comparator's voltage, there shouldn't be any error.

>>1240650
Also true. It looks like trimpots can handle a lot more power than I thought they could.

File: nes.png (296KB, 800x558px) Image search: [Google]

296KB, 800x558px

I'm not sure if this goes here, or /v/, or somewhere else. Who else has had experience restoring old vidya consoles? I tried the boiling trick on the 72pin connector today and it worked great. Then I put it all back together and it was back to being a stubborn little shit.

I know that there's completely different style replacement parts out there, but I don't really feel like pouring money into these things unless I go in for the RGB upgrade. Which I don't intend to do just yet. I want this NES reliably working before I proceed any further, and I want it to stay the front-loading style.

File: Screen Shot 2017-09-08 at 4.48.46 PM.png (71KB, 1040x1124px) Image search: [Google]

71KB, 1040x1124px

>>1240669
A 1/2W is good for maybe 10 of these LEDs, which is tempting enough that I've replaced the resistor with a 25-turn sealed trimpot permanently, in case I want to change the LEDs or add more. It also gives me the freedom to change between a BJT and FET to switch the LEDs, but either way I don't think it makes much difference. Pic related is the probable final circuit, and it only took 2 unneeded jumper leads, excluding the off-board pots and LEDs.
>tfw it all fits within a 3.5x3.5cm PCB
I'm wondering whether to do it on dot-board or to etch it, I'm thinking of etching solely because it looks better.

>>1240687
Replace the cartridge connector first. That's the most common failure point, even the originals went after a few years. I don't know why you'd even try boiling when a replacement is like $10 shipped or less.

>>1240724
Should I leave the unused comparator inputs floating, or tied to a rail? They're made of BJTs so it shouldn't matter if I leave them floating, right?

>>1240810
Better to ground them in case they should decide to oscillate.

>>1240815
I don't like the idea of tying the output to anything without a large resistor, something my circuit doesn't have too much room for. Would tying both inputs to one another stop any oscillation? Or is tying them to the negative rail fine? My "ground" is ~2.5V higher than the negative rail, hence the slight confusion. Would their oscillation cause any issues with the other comparator or use up a particular amount of current?

>>1240818
-V is fine for the inputs. Leave the output open. It's really just best practice. Oscillation could increase power consumption slightly, especially if you find yourself riding near a powerful radio transmitter of some sort.

>>1240825
>radio station
Ah that makes sense.

File: DSC_0021.jpg (2MB, 3840x2160px) Image search: [Google]

2MB, 3840x2160px

Used the wrong tool for the job and burnt the boards contact points to a crisp.
It's a crossfader and as far as I can tell it's digital, it has 4 pins on one side and 2 on the other.
Any idea the diagram this follows to a 4 pin connector?
There no other electronics on the board.

>>1240834
Nope, sorry.

File: DSC_0022.jpg (2MB, 3840x2160px) Image search: [Google]

2MB, 3840x2160px

>>1240834
Pins are labeled 1-2-1-2 on the right and 3-3 on the left. The connectors pins go from 1 to 4.

>>1240840
Looks like a double-gang slide potentiometer, which is analogue. But I'd test it.

>>1240840
Aanlog stereo fader, by the looks of it.
You can try removing your extra wiring, scraping bits of the solder mask off, and using a continuity tester to map out the connections. Other than that, who knows.

>>1240842
I wish there was some exposed contact but the soldering iron I used stripped everything away.
Next time I'll definitely take my time and look for my heat gut.

>>1240845
I think you need more temperature control, not less. The pins on the back should be connected directly to the slide pot itself, just measure from those. What exactly did you break, judging by the look of that PCB there weren't any traces on the front of it, so what happened?

>>1240845

because you're dealing with an a simple analog pot, you can just trial-and-error the external connections until it works the way it's supposed to. there should be no concern about something blowing up coz you reversed some wires.

this assumes you wired the PCB correctly. find an ohmmeter or continuity tester to make sure all you amateur-hour connections are good, and none are shorted that shouldn't be.

>>1240857
I've trial and error'ed this for 2 nights now and no luck.
It's from a vestax VCI 380.
The controller doesn't power on fully without the pcb plugged in, the closes't I've got with this is have the controller power on and reset as soon as I hit one on the two extremes.

>>1240854
Yeah all I had at hand was a red hot glowing soldering iron and not my tiny prcesision one.
What I ruined was on the other side now all that's left on the soldering points are brown pcb.

>>1240868
Then you're probably shorting power to ground. Why were you soldering on this anyway?

>>1240874
I was replacing the crossfader for a new and improved model

>>1240875

you can get a schematic for the VESTAX PCV002 mixer which uses a similar cross-fader pot. it's possible they used the same design, and you might learn something from it. however my PDF programs are too old and dont display the text correctly, so i cant find the crossfade in the schematic. file is here: https://elektrotanya.com/vestax_pcv002_mixer.pdf/download.html

>>1240616
Fair point.

Either way, is the best method for discharging the battery really to just build a connection between the two poles and let it zap out? For reference on size, it's an old game system battery, rated to 7.2v and 1600mah. Not huge I know, but again, I want to get into electronics and I figure doing it the right way consistently will be better than haphazardly doing shit until I fry myself some day.

>>1240926
You will destroy lithium batteries and possibly cause a lithium fire which you probably don't have equipment to extinguish. Don't try to discharge them.
When replacing battery packs, remove one lead at a time if possible, and put insulating tape on each lead as you remove it.

>>1240531
>For AC it's a bit lower.
it's higher?

File: pinout.png (14KB, 930x220px) Image search: [Google]

14KB, 930x220px

>>1240906
Hey, you motivated me to really search hard for the service manual for my equipment and I actually found it!

I was stumped by the fact the fader had 6 pins and the connector 4 but only 3 pins are actually used!
It now works like a charm! Thanks! I'll be able to perform this sunday!

>>1240959
The part marked in green really confused me because it by looking at the PCB you don't see the last bend that leads to the second pin and it looks like it goes straight to pin 1

File: safety.png (795KB, 1368x1788px) Image search: [Google]

795KB, 1368x1788px

>>1240616
>>1240947
These are the diagrams they showed us in my installations course. They used the electrical installation guide "according to IEC international standards" that you can find on schneider electric's website.

File: bodily_effect.jpg (27KB, 480x218px) Image search: [Google]

27KB, 480x218px

>>1240926
>doing shit until I fry myself some day
Modern reporting has claimed that it sometimes takes several shocks to be lethal, and that the condemned person may actually catch fire before the process is complete.[citation needed]

what software you use to model circuits? Matlab has a bunch of shit for power systems and maple is useless in this regard

>>1241020
check the OP

anybody with experience building guitar pedals? Do I want mono or stereo jacks? I can do mono with a stereo jack, but not the other way around, right?

>>1241039
an easy source of income. Just make sure to make them rugged, because they literally are beaten

>>1240807
Tried the boiling because the replacement was every but as unreliable. I can't get consistent results between methods and cartridges. Everything has been getting cleaned with alcohol until gray stops showing on the q-tips.

>>1241039
>Do I want mono or stereo jacks?

guitars are inherently mono. a guitar player occupies one spot. mono. two guitar players can be separated, thus stereo.

there's no such thing as a stereo mike for a singer; same principal.

now if you are bringing out more than one signal for other purposes, fine. but the idea of a stereo guitar simply makes no sense.

>>1240978
I see that 10kHz is less effective than 60Hz because of the skin effect, and if ElectroBOOM is to be believed the most painful shocks happen at 1kHz. Relationship between frequency and heart fibrillation would be something to find an infographic on.

>>1241065
i read something about using stereo for better grounding or to disconnect the battery when the pedal is not in use...

File: psychtar.png (8KB, 462x427px) Image search: [Google]

8KB, 462x427px

>>1241081
>disconnect the battery
very common

how do I use the graph things in LTspice? This thing is cancer. I mean how to I adjust it like a osciloscope, to fit the wave in the screen without getting eye aids

>>1239573
I'd say that looks like a silver solder join - it's better mechanically than a regular solder join and it means you don't have to use shitty lead free solder to tick RoS boxes. Depending on the materials being silver soldered, you may not need to use flux.

>>1240534
Traditional
Probably don't want to go over $70 for it. I go camping a lot and bring my old portable with me which is why I need the wire to help receive signals better

>>1241144
>I go camping a lot
Small baggage or room for a larger radio?

>>1241157
I've got some room but I would prefer it being portable with the option to use battery power

>>1240966
>>1240978
These graphs are gold for my TENS project. Thanks anon.

>>1241072
I heard that 7Hz is the natural frequency of something or other and runs the most risk of inducing fibrillation.

>>1241065
>there's no such thing as a stereo mike for a singer; same principal.
But there is such a thing as a balanced signal or phantom power over TRS cable/jacks. TRS is used inside the booth all the time, in part because it's sort of compatible with isolated unbalanced equipment.
>stereo guitar makes no sense
It's not going to stereo, it's going to be 2-channel. There have been many times I have wanted to automate the pickup fader, both for timbral repeatability and effect.

>>1241174
Baofeng?

File: RP2100-3.jpg (36KB, 469x404px) Image search: [Google]

36KB, 469x404px

>>1241174
Not many small SW receivers can stomach a long wire antenna, some can. Best solution is a selective antenna tuner. You will at least need a passive attenuator (Local/DX) plus adjustable RF gain. I only have a 10m antenna and another 10m as ground wire (no tuner) on a 2100, more didn't improve reception. More often than not switching from external to internal (110 cm telescopic) doesn't make much of a difference but disconnecting the ground wire always does.

File: trace.png (653KB, 640x480px) Image search: [Google]

653KB, 640x480px

what's the best way to fix this?
I'm guessing remove component, cover the trace with something (what?), and then try to resolder the component

>>1241195
is that trace on a lower layer? otherwise it seems like its intent is to connect the two pads. if so then i'd just remelt those shitty joints and then cover the site with a bit of superglue (or anything) to prevent oxidation. if it's on a lower layer you'll have to explain how you managed that.

>>1241196
I'll just assume the pads are supposed to be connected then and just solder over it
not sure how it happened, didn't notice it till today

>>1241130
Change your ".tran" function to get the correct timescale, and/or click+drag a box around the area you want to zoom in on. You can also customise the graph equations to change their Y scale, if you want.

>>1241195
If it's the component that's broken or if the solder is particularly bad then you should reflow the pads and wipe them up with solder wick and remove the component in whatever order is easier. If not, just reflow the solder (maybe with a little extra solder or flux) and those pads should connect up again. Looks like a cold joint to me, though the trace of solder mask on that central copper is slightly worrying.

File: IMG_1137.jpg (3MB, 3264x2448px) Image search: [Google]

3MB, 3264x2448px

Mirrored circuit on the left, correct one above it, and PCB below. I drew a pencil grid to help me align the pen, and I hope the pencil lines don't mess with the etching. Wish me luck!

>>1241207
If they do, you can scribe them out.
Luck!

File: SR.gif (13KB, 470x250px) Image search: [Google]

13KB, 470x250px

>>1241185
>7Hz is the natural frequency
7.83 Hz is the fundamental natural resonance of the surface - ionosphere cavity.

File: aaa.jpg (986KB, 1920x1155px) Image search: [Google]

986KB, 1920x1155px

i made a 12->5v/2a switcher with shit i had sitting around so i can charge my phone off a car battery after the hurricane. unfortunately i soldered the deadtime circuit wrong and i think that resulted in the tl494 getting fried somehow. not really sure how because it's 41v rated and in a buck configuration with a 24v supply used to test it. might just be a fake chinese ic.

>>1241213
Diodes might matter here. 1N400x are too slow for switchers and 1N4148 is weak.

>>1241213
Post schematic, not breadboard

>>1241208
I'm more worried that the ink drawn on top of the pencil will fall off.

>>1241213
Is that a broken wire I see under/above the "5" written on the IC?

File: Capture.png (45KB, 1231x686px) Image search: [Google]

45KB, 1231x686px

>>1241220
that's just the gate driver. i used some schottky for power. i'm not looking for a diagnosis, just sharing, so i'm too lazy to label values here.

>>1241226
it's the end of the yellow wire bent over to bridge two rows

File: IMG_1138.jpg (1MB, 3264x2448px) Image search: [Google]

1MB, 3264x2448px

>>1241223
Shit, it's happening. I guess I'll redraw it and hope for the best.

>>1241190
I've been taking my old Tecsun R-9012 with me and I've got a 20 ft external antenna that clips onto the telescopic one that improves the quality by a lot
I've been eyeballing the Degen de1103 but I'm still looking

>>1241213
I've never seen someone solder resistors together like that

>>1241237
that's what happens when you wing it with perfboards

File: XMerm.jpg (169KB, 1024x768px) Image search: [Google]

169KB, 1024x768px

>>1239434
Working on making a solar light for my doggy in the backyard. Just need some advice since I'm still very new to all this.

I've noticed that the voltage of my solar panels increase or decrease depending on how much light is available, would it be alright to wire them in series (0.5v panels) to roughly 3.5 or 4v and then step that up to 4.2v for my 18650 battery pack.

I'll be using the mt3608 because I bought a pack of 10 of them for cheap and it'll be within spec, along with your typical 18650 protection PCB to keep the batteries within a safe voltage range.

Also could I connect the mt3608 in parallel safely? I found pic related in a forum from some guy saying he did it and it worked. Just want to decrease the current passing through each PCB when stepping up from 3.7v to 10 or 12v for the LED lights.

I can link all the parts I'm using if you need it just in case I missed some details you'll need to help me with.

>>1241232
Don't worry too much about pinholes. Just patch it up as best you can and tank it again. Your board would be fine with 0.6mm traces all around. As long as you have that much cross section in toto, you're not putting enough power through it for a few holes to matter much.

>>1241240
DO NOT attempt to charge Li batteries without a proper charge controller AND protection board. Frankly, if you're new to all this, you probably shouldn't be fooling around with Li batteries at all.

>>1241245
I've got the protections boards, I was just wondering if I could feed them a constant 4.2v from a boost converter instead of using a charger.
Big no?

Stepping up to 5v and then using a charger would make the whole thing a lot less efficient which is why I was wondering if this alternative was possible or too dangerous?

>>1241247
Don't the charging boards typically take more than the cell voltage to power their built-in buck converter? And I think the protection boards usually just power themselves off the battery.

>>1241247
lipos are low impedance so constant voltage charging will cause them to draw more current than they can handle and thus overheat. the other anon is right, don't fuck with them until you understand them. at least read a charger datasheet.

if you insist though lipos can be safely constant current charged up to 3.6V. they'll be mostly charged at that point so it's not justified to add the constant voltage phase in your case. it'll only decrease their cycle life. lipos have a listed charge/discharge rating in "C"s, where a 4C rating means it can handle however much current it would take to charge it in 1/4th of an hour. stay well below this limit.

File: Screenshot_1.png (59KB, 402x332px) Image search: [Google]

59KB, 402x332px

evening, guys. where do I put an external crystal on this PIC? just starting to learn to read datasheets

http://ww1.microchip.com/downloads/en/DeviceDoc/40001919A.pdf

File: image.png (279KB, 1701x949px) Image search: [Google]

279KB, 1701x949px

>>1241256
microcontroller datasheets all have a table that calls out the special functionality of each pin. to find these i ctrl+f'd "crystal", found the section on the oscillator, and then referenced the pin function names in the table (OSC1, OSC2). be sure to read up on configuring it.

>>1241260
Many Thanks!

>>1241251
Not sure about the charging boards. The ones I have (tp4056) take a 5v power source though and output 4.2v at roughly 0.85A.

I wasn't sure if there'd be more to it, I know they don't charge constant current though and wouldn't have any features that a charger would need but I was hoping the protection board would help compensate to at least protect me from exploding or combusting cells.

>>1241254
>Safely constant current charged up to 3.6v
Thing is though my project is mostly with old recycled laptop 18650s.

I guess I'll just use the mt3068 to step the solar panels up to 5v and then use the tp4056 to charge the batteries?

Also what's a safe method of wiring the mt3068 up in parallel to split the current going to the LEDs? If not then any idea on a cheap boost converter that can provide roughly 1A 12V from 3.7v power source?

The limits of the solar panel will effectively give a constant current anyway.

>>1241251
The TP4056 is actually a linear regulating charger. Buck/boost chargers that run direct from a solar cell are available but they're not very designer-friendly.
>I guess I'll just use the mt3068 to step the solar panels up to 5v and then use the tp4056 to charge the batteries?
That will definitely work. As you mentioned, you may lose a bit of power.
>Also what's a safe method of wiring the mt3068 up in parallel to split the current going to the LEDs?
It doesn't look like you'll get more than 300mA when stepping up from Li voltages. Why not separate the load into 300mA or so sections?
Also, what about capacities? When building Li packs, you need to match their capacities. The weakest one will drag your whole pack down. A pack with 144Wh of Li will be hard to build out of recycled batteries.

>>1241235
>1103
Plays in a totally different league than the 9012 because of the high first IF (55.845MHz). http://ok1ike.c-a-v.com/soubory/degen_modif.pdf

File: IMG_1140.jpg (1MB, 3264x2448px) Image search: [Google]

1MB, 3264x2448px

>>1241232
I SEE THE LIGHT!

>>1241278
Well by connecting them in parallel I was hoping that would split the current up. I know it's not an even split, but 3 in parallel should keep the current low enough to provide "good enough" efficiency, I hope.

This is not turning out as I'd hoped, but at least it's shaping up into something...

>>1241292
I don't know that I'd trust them to be well-matched enough to share current halfway equally.

>>1241295
They wouldn't share equally regardless, they won't even be operating at an absolutely identical voltage to begin with either. There will always be a seemingly insignificant voltage difference that'll fuck with load sharing.
It will help though.

>>1241279
would you recommend upgrading to the 1103 or should I spend a little more to get something better?

File: 36c71d9d-edb6-402a-abf7-419251b4a5c6[1].jpg (24KB, 361x361px) Image search: [Google]

24KB, 361x361px

>>1239434
this is on sale for lowest it's been and I have one and it's a really good meter :3 I use it at work! I paid like 17 dollars for mine after shipping and it's only 11 right now...
https://www.banggood.com/ANENG-AN8002-Digital-Ture-RMS-Multimeter-ACDC-Current-Voltage-Frequency-Resistance-Temp-Tester-p-1145700.html?p=2T111812202419201707

File: 17245217922056234323.jpg (180KB, 600x600px) Image search: [Google]

180KB, 600x600px

>>1241317
>$17
>$11
Expensive. Been using pic related for a few months now and it's really good. Does everything I need from a multimeter with good accuracy.

Well would you look at that? I too can share a link and shill a product, heh.

Brand New 1Pc LCD Digital DT-830B Electric Voltmeter Ammeter Ohm Tester Digital Multimeter
http://s.aliexpress.com/jMRFrMb6

>>1241324
i had one and gave it away and bought a $50 extech instead. its autoranging takes millenia and the forward voltage meter has never worked. imo if you're going to buy a meter you should either stick to your pic or get a fluke.

>>1241326
I second this.

File: IMG_1141.jpg (1MB, 3264x2448px) Image search: [Google]

1MB, 3264x2448px

>>1241287
I might have been a tad aggressive with the bradawl, the 1mm bit didn't bore out all the peened copper. Preparing for delamination!

>>1241353
Also please ignore the extra hole at the top right.

>>1241232
Is that CuCl etchant? Cheers to you if it is.

>>1241326
I have never liked autoranging meters. Would pay extra for manual ranging buttons/stops.

>>1241371
Nah, ammonium persulfate.

>>1241201
How would I go on doing that? Thanks

>>1241384
Yeah, persulfate's kinda slow. HCl+H2O2 takes 1/3 the time, which might have spared you some resist dissolution.

has anyone here tried with hydrogen peroxide, vinegar and salt solution? How did that go?

>>1241419
But why tho? You're still generating toxic waste, and possibly even more toxic due to that organometals generated by organic acids might be even more bioavailable.

>>1241424
Idk it was just something I read. Haven't done any etching myself. So what's the best solution to use? And how to dispose?

What are the chances of getting arrested if I order a bunch of electronics and weird toxic chemicals, given that I live in Belgium?

File: Screen Shot 2017-09-09 at 11.02.03 PM.png (213KB, 2132x714px) Image search: [Google]

213KB, 2132x714px

>>1241403
>>1241419
I swear, one of these days I'll just dip my PCB in bromine water and y'all be jelly. Then I put an electric current through it with a copper electrode and pull all that copper out of solution again for 100% reusability. Alternatively, iron(III) sulfate would probably work fine too, provided the iron comes out of solution onto an electrode if an iron electrode is present during etching. I would like to see if anyone has tested this. Technically it would work with ferric chloride as well as ferric sulfate, but I like to avoid electrolysis with chloride ions in the mix.

>>1241399
A transient analysis is the simplest and most useful. Your ".tran" SPICE directive has a few parameters that I know of. If you type ".tran 0 3m 1m 10u startup" your simulation will start recording the simulation at 1ms, end it at 3ms for a total of 2ms graphed, the maximum timestep will be 10µs, and any DC supplies will start at 0V when t=0 thanks to the startup command. I ignore the first 0, the max timestep, and usually the startup command and just write ".tran 0 50m 40m" or even just ".tran 0 10m" which plot from 40ms to 50ms and from 0 to 10ms respectively. When you've got a graph running you can drag a box around a section of the curve to zoom and enhance, but it is limited by the graph's resolution.

Otherwise just make sure you have a ground node somewhere on your circuit for all voltages to be in comparison to, and make sure all your components have values and don't require any extra information, which some do. In the end, you can get some damn informative graphs out of it.
Try http://www.editorialdigitaltecdemonterrey.com/materialadicional/id212/cap1/LTSpiceGuide.pdf for a handy guide.

>>1241430
>I swear, one of these days I'll just dip my PCB in bromine water and y'all be jelly.
With your current rate of progress, we all die of old age before that happens.

>>1241430
>jelly
>bromine
It's your nose m8. If it weren't for needing to electrolyse outdoors, the CuCl process can supposedly do much the same.

>>1241425
I doubt you'll have serious trouble in Belgium, except for the usual customs duties if any. That said, all you really need is drugstore antiseptic 3% H2O2 and hardware store muriatic acid ~31% HCl. To two parts 3% H2O2 solution, add one part HCl in a shallow glass container. Add board. Etch in 3-4 minutes. Save mixture for reuse and regeneration. See http://www.instructables.com/id/Stop-using-Ferric-Chloride-etchant!--A-better-etc/

>>1241436
>CuCl
Why not just use copper(II) sulfate? The idea is that the Cu2+ oxidises the Cu(s) and they both turn into Cu1+, right? I don't see why you'd fuck up the perfect opportunity for simple electrolysis with chloride ions.

First of, thanks for the sticky. The first book's 100 circuits at the end are great. The 2nd book is top notch.
Now, how do I send GPS data from my drone to my phone?

>>1241439
Admittedly I don't know much about electrolysis chemistry, but the intent of the CuCl system is reducing waste via easy renegeration, just by bubbling air through it, since the CuCl is also a copper etchant. I'm a bit too tired to chemistry right now but maybe there are other ways to pull the copper back out of the Cu2Cl when the bath grows too large.

>>1241440
This would be a /g/ question if anyone on /g/ knew what the fuck they were talking about.

>>1240724
What am I looking at?

Have any of you ever ordered something from mouser and had to fill out a form that asks you if you're on some government blacklists?

The form asks me if I'm on any of these lists:
https://www.bis.doc.gov/index.php/the-denied-persons-list
http://www.pmddtc.state.gov/compliance/debar.html
https://www.treasury.gov/resource-center/sanctions/SDN-List/Pages/default.aspx

They also ask me what I'm going to use the product for, if I work for any foreign government or my own government, and some other stuff like that.

This just seems a little bit extreme for a switch that costs $1.75.

>tfw no oscilloscope
what's even the point?

File: Digital-Multimeter.jpg (11KB, 300x225px) Image search: [Google]

11KB, 300x225px

>>1241326
It's always good to have a throwaway meter in your collection. Mine is yellow. The other one is dark gray, much smaller and has 10 MΩ.
>autoranging takes millenia
The [R-H] range hold button is your friend. Let it seek and lock it. Some autorange meters are damn fast which is rarely mentioned. Also: Do not measure current unless you have a spare fuse..

>>1241531
>>tfw no oscilloscope
>what's even the point?

you can do a lot with a multimeter. scopes are like 4x4 pickups and guns, basically compensation for inadequacy

>>1241539
Depends on terrain. Anything µs-related is off limits for a multimeter and even a toy scope is better than no scope at all.

>>1241539
can't decide if this is low effort or good bait

File: relay.jpg (67KB, 879x879px) Image search: [Google]

67KB, 879x879px

I need to power a small 12 v circuit to control some lights, it would consume less than 100 mA
what do I power it with other than a transformer?

>>1241353
burred holes. did you drill from the component side? not good. first use 0.8mm instead of one, ream only if need be.

>>1241060
Did you try bending the pins back into place? You can get to one half of the pins and bend them upward using a toothpick. I don't remember it that well but I did it once in the past. Also, never actually press the cartridge down when you play, it's not necessary and is actually what bends the pins out of whack in the first place.

>>1241557
there's a large variety of nonisolated mains supply circuits. they're dangerous and should not be used if your project has any human interfaces. there's a dozen simple, cheap circuits you can find online and other than those you can buy small board mounted (nonisolated) supplies for $3-5 that are almost standalone:
https://www.digikey.com/products/en/ac-dc-converters/923

the obvious choice is a wall wart though. there's no reason not to use one unless you really want everything to fit in one package.

File: ACDC.png (124KB, 265x284px) Image search: [Google]

124KB, 265x284px

>>1241557
Transformer if you want AC, small AC/DC adapter/power supply if DC.

>>1241324
why would you even buy that one when harbor freight gives those away for free? Those things are never accurate compared to my fluke, they're really only good for a few basic measurements.

Does anyone have a known working LTSpice model for the MC34063A or equivalent? The two I've tried so far haven't worked with the example circuit in the datasheet; I'm not entirely sure if the models I've been using are bad or I'm fucking something up, I'd at least like to narrow that down to one or the other.

Almost total noob here: I'm trying to reverse-engineer a PCB from images and picture schematics (no actual files) and there are a few components which have pins that are supposed to go to ground, and on the schematics the pins are marked with the signal ground symbol. Does this warrant adding a ground plane inside the PCB? If yes, when I add it, since those are pins designed to be soldered, can I just leave the pads I already placed in and they will ground themselves since there are no traces connecting to the pads on the top or the bottom layer?

>>1241635
>Does this warrant adding a ground plane inside the PCB?
Depends entirely on the chip and your requirements.
>can I just leave the pads I already placed in and they will ground themselves
Depends on your PCB editor and how you use it. For example, the software I'm using allows me to use a dedicated ground plane. In that case, I have to configure the relevant ground pins to be connected to that plane. After that, the software makes the actual connections automatically. Another option would be to do a (ground) copper pour over the ground pins.

did anyone buy this bundle?
https://www.humblebundle.com/books/electronics-programming-make-books

if so, can someone share it? I am a poorfag, but i'd love to read this shit

>>1241653
pretty sure most of these can be found on the net, check out libgen.io, a lot of good electronics ebooks there if you have a title to search with

>>1241659
thanks, this site has plenty of interesting stuff

>>1241436
>I doubt you'll have serious trouble in Belgium
i was just joking about the terrorists.
Thanks for the info though, so I buy the stuff once and then it basically lasts forever?

>>1241639
It's not a surface mount chip. It's an issue of multiple through-hole components, like a DC jack, capacitors and a voltage regulator, all of which have one pin for ground. I'm using Altium and I made a GND net and it warns me that the pins are not connected to each other on that net, which is what made me think I needed a ground plane. That said, if I only need a ground plane, for me to use OSH Park, will I need to add another completely redundant plane to use the 4 layer board option, since they don't have a 3 layer option?

>>1241697
You know, you could post a pic of your PCB...
I'd guess your other routing prevents Altium from connecting those pins to the ground polygon. Change other routing or wire the grounds manually. If neither is possible, then move components, use bigger board, use jumpers or use more layers.
Very few PCB houses make 3-layer boards or any odd-layer boards besides 1-layer.

>>1241522
Where are you ordering from/shipping to, Ahmed? I don't recall that I ever had to fill such a thing out.

>>1241555
It's just meh-tier bait.

>>1241681
You got it. You may have to add acid from time to time, as it is consumed by the regeneration process. A litre of acid will be more than you will ever need.

>>1241707
I would but I'm a dumb phoneposter and the pics are on my PC. Ground polygon? What do you mean by that?

>>1241518
PCB layout on GIMP. Each component is its own layer with its own name which corresponds to values on a spice file, so you don't have all the information needed to understand it.

I've connected a pair of tp4056 boards to a battery and charging current has almost doubled.

Is there any significant risk to doing this?

>>1241740
blowing up your battery

>>1241747
How?

If I have a protection board between the charger and the battery would it make a difference?

>>1241756
Probably not. The battery is designed to charge at a specific maximum rate. If you exceed that, you will degrade the battery.

Considering adding some FAQs to the pasta. Comments?
>Q1. Lithium batteries
>Will explode if you insult them. Know exactly what you're doing. Read and understand all relevant datasheets and have proper safety equipment.
batteryuniversity.com
>Q2. I don't know what this is, what can I do with it?
>Probably nothing. Search the web for part numbers, maybe you'll find something.
>Q3. My circuit doesn't work. Halp?
>Check wiring, part pinouts, and soldering, then post schematic.
>Q4. Consumer electronics
>>>/g/

>>1241762
Most lithium ion batteries can charge at or over 0.7c comfortably. As long as his batteries are 2400mAh or over then it most definitely won't be an issue.

Got a cheap headtorch from Amazon but it's spazzing out on me. Checked the batteries and they are fine. Opened it up but it doesn't look like there are any loose wires but I know very little about electronics. How do I go about diagnosing the problem?

https://sendvid.com/qzahegxi

>>1241771
There's usually not much to cheap LED torches but LED and batteries. Sometimes, rechargeable batteries give them fits. Try fresh non-rechargeables.

File: file.png (69KB, 421x352px) Image search: [Google]

69KB, 421x352px

Here's one for the DSP guys:

I'm doing to calculate the spectral kernels of a Constant Q Transform by calculating the temporal kernals, then doing an FFT as in this paper:
http://academics.wellesley.edu/Physics/brown/pubs/effalgV92P2698-P2701.pdf

My MATLAB code is at https://pastebin.com/kqG69fYJ (embed)

I'm stuck on the "calculating the temporal kernels" part. I'm trying to do it with the centered Hamming function and complex exponential as shown in Eq. 7, but my temporal kernels don't look anything like the authors, which increase in frequency and become less spread out in the time domain with increasing kernel number.

I noticed that with my current setup, every other complex exponential is simply 1 across the time domain. Is this part of the problem, or is there something else I've overlooked?

Thanks in advance.

File: 38[1].jpg (16KB, 235x235px) Image search: [Google]

16KB, 235x235px

I'm sorry for gearfagging it up in here but the TS100 "open source" soldering iron is on sale today for only $40 and it's the lowest i've ever seen it. If any anons need a decent soldering iron this one is apparently really good and takes hakko tips.
https://www.banggood.com/MINI-TS100-Digital-OLED-Programable-Interface-DC-5525-Soldering-Iron-Station-Built-in-STM32-Chip-p-984214.html?p=2T111812202419201707

>>1241810
>meme iron that gets too hot to use after an hour
>out of stock for three weeks
>affiliate link
Into the trash

>>1241810
just get a 936 clone mr shill

>>1241804
Why are you using for loops in MATLAB. The whole point of MATLAB is so you can avoid for loops by using element wise operations.

>>1241813
yeah i know it's on backorder right now, kinda annoying. if you're soldering for an hour without letting your tips cool down they aren't going to last very long anyways if they're cartridge tips, even the weller quickchange tips burn out from that.
>>1241815
this is also an option, but it is not digital.

>>1241817
936D then

>>1241818
I don't think the 936D takes cartridge tips either does it? also this is tiny in comparison. also it can run off my cordless drill batteries.

>>1241816
Because I'm comfortable with them

>>1241317
>>1241810
>on sale today for only $40
11. Advertising (all forms) is not welcome
4chan.org/rules#global

>>1241834
Well you should suck it up and get comfortable with matrix operations.

You forgot the summation. You should also learn the difference between * and .*

File: IMG_1144.jpg (2MB, 3264x2448px) Image search: [Google]

2MB, 3264x2448px

>>1241353
It's alive! Oddly enough I can have either of the two voltage divider pots low or high and it still works. I'm certainly glad I didn't use three 1" pots. The strobing duty cycle isn't 50/50, it's more like 60/40, so I can choose whether it's bright on the 60 or bright on the 40. Yes I did underestimate the size of the capacitors.

>>1241863
>LM348
Amazed that it even works since your supply is 3V lower than recommended minimum.

>>1241804
Stop changing your matrix size each iteration. Just initialize it to the samples X bands and input repmatted arrays of 1:bands and 1:samples to get rid of your retarded indexing.

>>1241869
>+4 to -4
Huh, I am equally surprised. I have an LM249, but I can't find the minimum Vcc on the datasheet.

>>1241861
What summation are you talking about? The one that you see in Eq. 7 because I mentioned that in my post? Because if that's what you mean, that's a discrete Fourier transform of the function I care about, and you need to read the article or my posts more carefully. I'm asking about the temporal kernals, not the spectral.

>>1241885
A whaaat? The only datasheet I can find is on the Internet Archive from back when AMD made op amps. It looks like the only difference is in the internal compensation cap.
Still, awesome.

>>1241953
I tested their voltage output ranges and found the 348 had a slightly wider range, so I used that instead. They're both quad 741s, I guess there really isn't much difference. My LM324 might work instead, but it has different output specs so I'd have to change the resistors on the potentiometers, and it might not work with my dual BJT setup.

>>1241959
If it works and won't blow up, who am I to argue.

>>1241800
Just tried fresh batteries, same issue.

>>1241973
My guess is there's a bad connection in there somewhere, maybe a dry solder joint, or an equally overzealous solder overflow. See if you can non-destructively get into the circuit part of it, chances are the processor is just an anonymous blob of epoxy but you can still check for bad connections and broken passives. No point in breaking the thing to take it apart if you can send it back for replacement. Can you post a picture(s) of the electronics?

>>1241985
The 2 outer LEDs work fine, it seams to only bbe the centre one that had this issue. I'm not very knowledgeable about this but the solder between the PCB and the circle in the middle looks low quality?

File: IMG_20170910_113525-01.jpg (2MB, 2134x1599px) Image search: [Google]

2MB, 2134x1599px

>>1241990
Forgot image

>>1241992
Is there a number or name on that 8-pin chip? If it's a no-name this might become a little more difficult, but still doable. How about what's on the other side of the board? That solder joint at LED- is pretty bad looking, but it certainly has a lot of solder so I don't think it's the problem. HY-DR16A isn't turning up anything relevant, unless that headlamp is made from a salvaged circuit breaker.

File: IMG_20170910_120832.jpg (2MB, 3968x2976px) Image search: [Google]

2MB, 3968x2976px

>>1242000
No writing on the chip.
This is the model:
https://www.amazon.co.uk/Flashlight-Noza-Tec-Waterproof-Backpacking/dp/B01GNUIJJO/ref=cm_cr_arp_d_product_top?ie=UTF8

>>1242009
that star heatsink looks like its shorting out a bunch of stuff

>>1242014
Should I desolder it and see what happens?

>>1242014
The solder mask should stop that, but it is a possibility. There are marks on the board that resemble possible water damage, but it's probably no-clean flux. I assume the chip is a microcontroller since it has too many unused pins to be a specialised headlamp IC, but I could be wrong. I'll see about tracing the circuit tomorrow to see where that gets me, but chances are the microcontroller is busted, though there doesn't really appear to be a cause, so maybe it is the bad soldering.

>>1242033
Well I was assuming that you didn't have acces to that kind of equipment, but go for it. First I'd reflow the existing bad-looking connections to see if you can get the thing to work better just like that. Also see if putting pressure on the PCB makes the LED run differently, which would indicate a bad connection between the LED module and the PCB, which is the most plausible explanation since the other two LEDs work fine.

File: arrow.jpg (25KB, 489x441px) Image search: [Google]

25KB, 489x441px

>>1241992
I would resolder this.

>>1241860
>>Where do I get components and lab equipment from?
>digikey.com
>jameco.com
>sparkfun.com
>ramseyelectronics.com
>allelectronics.com
>futurlec.com
>ladyada.net/library/procure/hobbyist.html
>mouser.com
>alliedelec.com
>newark.com
>ebay.com
>aliexpress.com

you can't ban me without banning the op he advertised 12 websites

>>1241317
Why would you advertise the old version the AN8008 replaces this.

>>1242092
it was on sale for like 11 bucks. The AN8008 is also very good for a bit more money, but doesn't come with a thermocouple.

Circuits l test tomorrow. With an autistic teacher who only gives as autistic didatic circuits with 3040214 dependant sources and short circuits to solve. send help

>>1242104

understand that a ''circuits'' class has nothing to do with actual real-world circuits, and thus nothing to do with /diy/, and everything to do with proving you're intelligent enough to handle complex abstract concepts and math you way through them so you can get that diploma. if you cant do that you dont deserve the diploma.

>>1242049
>I would resolder this.

i would resolder all 4 of the big hand-soldered connections, coz they all look pathetic. using leaded (Sn63/Pb37) solder, of course.

>>1242127
I can do normal circuits, I just suck at solving system of equations by hand. I usually find the right circuit equations and then it goes downhill from there. I know, circuits class is just a "topology for retards", without the math and with repetition. Anyway, wish me luck.

Im making a product that would be connected to cars 12V.

What is needed on my supply rail to conform to the automotive requirements?

>>1242157
Buddy, You are fucked. Your mind is broken.

>>1242160
Care to elaborate?

>>1242161
I'll bite.
Let us pretend you have a product.

In what way could you have a supply rail that didn't "conform to automotive requirements"? Drawing too much current could be one way. It really depends on how this product is installed in the vehicle.
There is no way to help you without some information about the vehicle or the product.

If this is a troll. Good job.

>>1242157
if you need 12v out, find one of those automotive passthrough regulators that will regulate for spikes but otherwise is just open. load dump conditions can put ~80v on the rail. use a local fuse and an extra tvs (with a higher breaking energy than the fuse) if the regulator fails.

i don't actually know aftermarket automotive electronics standards i've just read up on it so take this with a grain of salt.

>>1242167
Thanks, thats what i needed to hear.

File: 20170910_194833.jpg (1MB, 2048x1536px) Image search: [Google]

1MB, 2048x1536px

I bought these nixie tubes and they have a small disk on the side, about 4mm by 1mm. I'm curious about the purpose of that small disk.

I thought it was getter, but in that case I think it should've been flashed, idk.

>>1242165

there are, in fact, specific requirements for car-powered gadgets: the fact that voltage is not really 12V but can vary from like 10-14. the fact that the supply is very dirty, with spikes from the generator, and whatever else. the fact that when you use the starter, the battery voltage drops significantly. the fact that, if it's constantly powered, it must use a small current or risk killing the battery.

>>1239434
Can somebody please translate the following into retard? I mean english so basic I couldnt possibly fuck this up. Translate was no help and anglesk technical terms are not covered at tech school. Feel free to explain the terms in the last part.

"To derestrict, the blue wire coming from the CDI unit is disconnected at the bullet connector. The wire that comes from the speed sensor in the housing is simply tied back into the loom. Best to tape the connector.

The blue wire from the CDI unit now needs a switched live. It is reccomended to splice into the black wire with a grey trace on the CDI unit to get the switched live."

File: heaterntc.png (11KB, 779x560px) Image search: [Google]

11KB, 779x560px

I have this heater module with a built in NTC thermistor, circled part in pic related.

I have used 555 for temperature regulation before, right of pic unrelated.

The problem is i dont have access to both of the NTC pins in the heater module. How would i go to make this work?

I can swap the heaters polarity if needed. I dont see how could this work like the circuit i used before.

File: heaterntc.png (9KB, 507x523px) Image search: [Google]

9KB, 507x523px

>>1242205
I figured it out.

>>1242181
>there are, in fact, specific requirements for car-powered gadgets:
What regulatory body is responsible for overseeing these 'specific requirements'?

>>1242234
and what's the penalty for non-compliance??

>>1242237
>>1242234
probably the ISO thing in your country+the transit thing in your country, which is usually state (like nation, state, city, not like nation state) centered

Anyone have experience with arrow.com? I wanna buy some stuff but I can't register since their fucking system is broken. I've called their US number like 5 times now and it always redirects me to the voicemail box. I've accounted for the timezone difference, before you ask, and also tried calling the number for Euro countries but it's not even working. I've also tried mailing them and doing the live chat thing, but no operators are ever available. They seem like a really decent company, cheaper than mouser and free shipping, but their customer service is terrible from what I've seen so far. I've yet to try contacting them on Facebook or Twitter. Could the non-responsiveness have anything to do with Irma?

>>1242268
they aren't gonna be cheaper than aliexpress
just buy from aliexpress

File: leddimmer-1_schema.png (66KB, 1092x888px) Image search: [Google]

66KB, 1092x888px

Trying to implement an analogue global dimmer for my 12V LED strip which is based on environmental lighting intensity.

The LED performs some light shows but when it gets dark, the intensity of the LEDs should go down but the light show should continue.

Thought it would be easier to do in an analogue method than trying to code each colour (they are RGB but for simplicity the schematic only has one). Perhaps reducing the 12V supply or the 5V input in the FET.

The higher the light the lower is the resistance in the opto-transistor.

Any tips?

>>1239759
Small utility drawers sorted in 4.
1. short male/female
2. long male/female

Rarely use the female ones so I'll probably put them in a bag in a less-prime location.

who makes good general purpose snips?
there was a company whose name started with X that I forgot

>>1242321
Xcelite?

Yeah got some many many years ago.

>>1242324
thanks

File: headlamp.png (5KB, 500x500px) Image search: [Google]

5KB, 500x500px

>>1241992
>>1242009
I think I've figured out the schematic, it's pretty simple and since it uses every pin on the IC it probably is a specialised headlamp chip. But there aren't any obvious electrical failure modes that make sense, aside from potentially the capacitor. I assume that the capacitor is used for timing purposes because a crystal is too pricy, so it's used to time the strobing light and/or its PWM. If the capacitor had the wrong value or was faulty then I can easily imagine the main light flashing strangely while everything else worked fine. The problem might also be with the transistor by the 4 parallel resistors or the 103 resistor above it, so see if these components are getting hot. If you could measure the voltage across LED+ and LED- while it's running that would also help tell whether the voltage powering the LED is fluctuating, or if the LED itself or its connections are faulty. Again, see if putting pressure on the heat sink or PCB near the LED have any effect on the LED's performance.

>>1242321
Those are good for small wires and close cuts.

Buy a crappy pair and an xcelite. The crappy pair works for 98% of the time. The Xcelites work better for the finer work.

>design circuit
>plot the paths
>draw them manuallly on the board
>etch it
>wait for it to be done
>remove the ink
>realize I forgot to mirror the design

this shit has an IC, maybe I can solder everything on the copper side, or just the IC

>>1242358
do you guys not realize you can have your pcbs made for $5 with 6/6 rules and delivered in a week or do you just like making them yourselves

>>1242360
I live in south america, shill-kun

>>1242360
Still possibly cheaper to do them ourselves, and it means any trial and error doesn't take at least a week. I'm not going all "diy spirit" on you, but I find it to be another fun process along with designing and soldering, though if you're making high density SMD boards you don't have much of a choice unless you have a really HD (laser?) printer and photosensitive film.

Anyone know if that UV-backlight LCD trick works? It looked like a neat way to go if you have a high pixel-density cellphone or tablet screen that you could easily drive.

>>1242360
Do you realize i can turn out pcbs 168 times faster than they do for a fraction of a price?

>>1242347
>specialized headlamp chip
Maybe not. I've designed microcontroller circuits where I've used all 6 GPIOs available. If the supplies are on pins 1 and 8, that's strong evidence for a PIC12.
Cap could be an oscillator, or just MCLR.

>>1242358
Yes, you can. It'll be a bit of good practice for when you into SMD anyway.

>>1242234
Customers who don't like your shit product blowing up or draining the battery.

>>1242365
Photosensitized blanks are almost as cheap as regular ones, fwtw. I don't happen to have any handy, or I might try it. 6mil is about 4 pixels on a 600x600dpi printer, which is about standard these days for LED "laser" printers.
>Anyone know if that UV-backlight LCD trick works?
Here's someone using a projector which presumably has had the UV filter removed. http://www.gnexlab.com/making-circuit-boards-in-20-seconds-with-dlp-projector/

>>1242358
You can bend the leads back and mount it as it was a regular ic.

>>1242378
That url is just click bait, pcbs in 20 seconds? really? Id love to see them try tho.

>>1242360
Sometimes you just don't need 6/6 rules tho.

>>1242387
Sure it is. That's Turks for you. The point is he's exposing boards with a video projector that are apparently sound. If you have pre-sensitized stock it won't take that much more than 20 seconds.

>>1242358
>>1242360
I managed to surface solder the IC, it was an insertion component, kek

>>1242358
>just the IC
Easy solution but difficult to remove if need be. Alternative is pin header and 'dead bug' style on the component side.

Is there a way to learn basic electronics as taught in beginner books from OP using just one's PC? What circuit emulator would be the best for this purpose?

>>1242414
You can get a good grasp of a lot if it, but you'll be missing out on a lot too. A breadboard, jumpers, and components will run you less than $20, and LTspice is free software, I use both in conjunction.

>>1242414
Do you expect a spell checker to teach you creative writing, freelunch-kun?

Got basic-tier Altera MAX II dev board and basic-tier USB Blaster. Installed Quartus Lite. Recalled enough Verilog to make a flashing LED. Installed, ran. Surprisingly gratifying.

>>1242414
More than half the fun of it, for me anyway, is making things happen in the physical realm. Also, simulators are slow and simulated components don't blow up to remind you to be more thoughtful.

A build I'm trying to clone includes a 25k trimpot, but my shop of choice doesn't have that value. What are my options? Buy 20k and 5k and put them in series? Or should I just order it elsewhere (don't really want to)?

>>1242171
Anyone?
>>1242314
Nigga U have an attiny85 with an ADC, what's the point of going "analog" when you're already using a microcontroller that can do the light sensing for you?
If you want to dim LEDs going analog is a bad idea, you should be using PWM to change the brightness of the LEDs, using transistors in the ohmic region will dissipate a ton of heat.
Make a small program that varies the PWM depending on the ADC input.
>>1242557
Depends on the circuit you're trying to make, post pictures of the application if possible. A 22k pot is fairly standard and a 3k difference won't make too much of a difference, specially when it comes to voltage dividers and similar applications.

File: fuzz stompbox.png (24KB, 351x540px) Image search: [Google]

24KB, 351x540px

>>1242561
It's a guitar effect pedal. I believe the trimpot would be "VR3", top right -ish.

>>1242562
aaand i'm not sure why the image got rotated.

>>1242562
It seems the pot's function is to limit the current that can go though the transistor and acts like a voltage divider to the output capacitor, I'm not versed in guitar amp circuitry, but I think that by varying the pot you vary the speed at which the output coupling capacitor drains out when the transistor is on, providing a certain audio effect, probably filtering low frequencies in the process.
Using a pot 3k lower (a 22k pot that is) means that capacitor will drain just a tad faster when the pot is set at its limit compared to the 25k pot, nevertheless the difference should be minimal.

>>1242574
if I get a 30k trimpot I can set it to any value the 25k one could do, right? Would that be a solution too?

>>1242576
Yeah, a 30k pot would allow yo to cover and extend the range of the 25k pot. I think that's a better idea.

>>1242595
Alright, I'll go with that then. Thanks for the help

File: weird.png (5KB, 120x221px) Image search: [Google]

5KB, 120x221px

>>1242603
This doesn't make sense. Why would one want to connect the (-) of the adapter to the input rather than to GND?

File: led thing.jpg (1MB, 466x258px) Image search: [Google]

1MB, 466x258px

A big advertising led panel was removed near my home. As I'm in a third world shit hole the shitheads simply left it there to rot on a plaza, I managed to get 6 power supplies (all bust) and a couple of these led square things. But I have no idea of how they connect to the power or anything. Thre are 4 cables coming out
>blue, green,yellow and red
I imagine 3 of them must be to control which are ON and the other a ground ? There is also the hole on the back. There is no writing or part number anywhere.

Anyone know how to use them?

>>1242619
hmm I hadn't caught that. I'm honestly not sure as i'm just starting out. I'm doing this with a friend and he was the one who found the schematic on http://www.byocelectronics.com/board/.

Is the (-) actually connected to the input? If I understand correctly the connector is stereo but I think a guitar's signal is mono (?), so maybe the (-) is actually connected to the unused connector of the stereo jack for... reasons? Idk

I'm not very familiar with microwave RF design so I need some pointers.

I want to roll my own microwave frequency counter that can work up to about 40GHz. What things do I need to consider regarding component selection, PCB design, and connectors when designing for such high frequency?

I don't have any instruments with enough bandwidth enough to make measurements at over 150MHz or so nor do I have signal generators capable of outputting more than 20MHz though I can always roll my own oscillators for test signals up to a few gigahertz when I'd start needing tunnel diodes and Gunn diodes. At that high frequency I'd settle for an accuracy down to 10MHz but I'd prefer 1MHz or even 100kHz if possible.

File: fuzz-box.png (24KB, 540x351px) Image search: [Google]

24KB, 540x351px

>>1242629
It should look like this, following the tradition to activate the box by inserting the input (mono) plug which connects sleeve (S) and ring (R) while the audio signal comes in through the tip (T). This was done to prevent empty batteries (note bat option included in pic) and required no ON/OFF switch.

>>1242664
I see, that definitely makes more sense than what I said. Thanks for catching the mistake!

https://www.amazon.de/Regelbare-L%C3%B6tstation-ZD-99Kit-48W-Zubeh%C3%B6r-blau/dp/B017DN8VUC/ref=sr_1_6?ie=UTF8&qid=1505154245&sr=8-6&keywords=soldering+station&language=en_GB
Is this a good buy?

File: electronics-wirecutter.jpg (5KB, 314x161px) Image search: [Google]

5KB, 314x161px

>>1242676
Low end but usable. Doesn't regulate temperature but you can adjust the heater power. Shearcutter missing. I prefer pencil-type handles with a closer distance to the tip.