[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/diy/ - Do It Yourself


View post   

File: 3.63 MB, 4032x3024, vintage memeboard.jpg [View same] [iqdb] [saucenao] [google]
1951716 No.1951716 [Reply] [Original]

Thread obsolete: >>1945940

>I'm new to electronics. Where to get started?
It is an art/science of applying principles to requirements.
Find problem, learn principles, design and verify solution, build, test, post results, repeat.

>Project ideas:
http://adafruit.com
http://instructables.com/tag/type-id/category-technology/
http://makezine.com/category/electronics/
Don't ask, roll:
https://github.com/Rocheez/4chan-electronics-challenges/blob/master/list-of-challenges.png.png

>Principles (by increasing skill level):
Mims III, Getting Started in Electronics
Geier, How to Diagnose & Fix Everything Electronic
Kybett & Boysen, All New Electronics Self-Teaching Guide
Scherz & Monk, Practical Electronics for Inventors
Horowitz and Hill, The Art of Electronics

>Design/verification tools:
LTSpice
MicroCap
falstad.com/circuit/circuitjs.html
NI Multisim
CircuitLab
iCircuit for Macs
KiCAD (PCB layout software, v5+ recommended)
Logisim (Evolution)

>Components/equipment:
Mouser, Digi-Key, Arrow, Newark, LCSC (global)
RS Components (Europe)
eBay/AliExpress sellers, for component assortments/sample kits (caveat emptor)
Local independent electronics distributors
ladyada.net/library/procure/hobbyist.html

>Related YouTube channels:
mjlorton
w2aew
jkgamm041
eevblog
EcProjects
greatscottlab
Photonvids
sdgelectronics
BigClive
Ben Eater

>Li+/LiPo batteries
Read this first:http://www.elteconline.com/download/pdf/SAFT-RIC-LI-ION-Safety-Recommendations.pdf

>I have junk, what do?
Shitcan it

>> No.1951719
File: 49 KB, 960x960, sisy.jpg [View same] [iqdb] [saucenao] [google]
1951719

>>1951716
>>1951711
>respirator projects
>How did that go in the end? IIRC they need pretty stringent safety interlocks of all sorts that would make getting a community-made respirator approved by some regulatory body exceedingly difficult.

It was a experimental setup made by a Brazilian university. Many were made and some quite competent. I lost touch with the project but I do not think it went past a prototype because the unfortunate stabbing survivor we call president cut the funding of all such initiatives (and most, if not all research desu).

I know that in Brazil a group made 3D printed consumables, used when you "tube" up a patient and had to be discarded after each use, free of charge for some hospitals. The regulatory agency must have looked the other way in that case I think.

After all investing in national tech is a very stupid thing to do since every student and researcher is a sex-addled, drink-and-drug sodden communist spies.

>> No.1951723

>>1951716
Oh and I can buy 2 of those boards for 5-15 dollars each. Same for 2 similar boards.

Anyone think it's worth doing? They look damn neat, so even if I don't scrap them for their parts they're possibly worth it to hang on my wall. Don't have hot air so desoldering would be a pain, not that I wouldn't try it with a solder sucker and some patience.

>>1951719
Sucks to hear, but it's not surprising. I wonder if any chinese university students tried the same thing, snuck it past incompetent regulatory bodies, accidentally killed a dozen people, and had the whole thing swept under the rug? Or are they not creative enough?

>> No.1951728

>>1951723
I cannot speak for the chinese, but the problem they were having was not in actually building the thing but controlling it since that requires accurate experimental data for the calibration and manufacturers did not provide that IIRC.

>> No.1951732
File: 140 KB, 1000x750, meh.jpg [View same] [iqdb] [saucenao] [google]
1951732

I had this shit on backburner for a year now but today i'm finally going to finish it!
Mainly because i have to hand it in tomorrow... man i wish i wasn't such a procrastinating slob.
I had to remove a layer of dust from everything.

>> No.1951747

Reposting because it got buried by a million replies to some post about an MPPT controller before anyone responded:

I'm planning on building a SMPS to boost a low voltage source to >200V at a fairly high current (for the voltage) of around 100mA. Due to this I need some pretty high peak primary currents of >10A. My issue is my input supply is only 3A capable.

What I'm wondering is, do I REALLY need a supply that can match the peak primary current I need, or, will the input storage capacitors provide that current? I'm switching at about 100kHz and calculated a switch on time of slightly under 5us. Here's the thing though, I found the A*s product during the switch on time to only be something like 125pA*s where as the total A*s product for a single fully charged 220uF input was like 2.6mA*s. Based on my calculations supplying 10A for like 4-5us should only drop the capacitor by a few mV. Basically fuck all. The cap charges back up in the switch off time and bam, I can get 10A peak current from a 3A supply.

Is this analysis correct or am I not understanding something?

>> No.1951749

I have a length of enamelled wire and want to completely strip it all of the insulation.
Is there an easy way to do this? Like soak it in acetone for a few days?

>> No.1951751
File: 8 KB, 400x201, ethernet cable.jpg [View same] [iqdb] [saucenao] [google]
1951751

>>1951749

easiest way to strip enamel is to not strip it, but replace it with easily-stripped ethernet cable, or easily-stripped wrapping wire.

>> No.1951754

>>1951751
I have some thick enamelled wire that I don't need for anything else.
Copper is expensive so I'd rather use that if possible instead of buying new wire that's easier to strip or already uninsulated.

>> No.1951767

>>1951747
>Pin = Pout+losses
You are wrong because the above must hold true under steady state

>> No.1951774

>>1951754

a guy here, jpanhalt, says acetone doesnt work, but he uses other corrosive liquids that do. do you like to play with acids?

https://forum.allaboutcircuits.com/threads/best-method-to-strip-enamel-off-magnet-wire.29861/

>> No.1951778

>>1951774
NaOH or NaCl wouldn't be much of a problem since they're cheap and available everywhere but don't they attack the copper too?

>> No.1951787

>>1951749
Crush the wire with a dough roller and throw it into a container of mineral spirits.

>> No.1951797

>>1951749
fire and sand paper

>> No.1951854
File: 78 KB, 800x800, INA3221.jpg [View same] [iqdb] [saucenao] [google]
1951854

>>1951350
>>1951478
So I'm going to build my own lab PSU from an pc ATX PUS, and have it driven by an Arduino and a LCD. The PSU already have 30A 3.3v, 30A 5v and 15A 12v out. I'm thinking of driving those directly out and have a variable buck booster attached to 3.3 and/or 5 to have a variable output driven by a digipot. Pic related and some shunts seems to be the solution for measuring the voltage and current. Anything else I should keep in mind before attempting this? (im 3d printing an enclosure and definitly putting in a nice 120mm fan or something, glass fuse for each channel etc)

>> No.1951862
File: 14 KB, 480x480, Eiechip LED Digital Voltmeter Ammeter DC Digital Multimeter 100V 10A Blue.jpg [View same] [iqdb] [saucenao] [google]
1951862

>>1951854

a PC supply is the worst thing you can use as a bench supply coz everytime you connect a transistor backwards, the supply shoots 50A thru it causing an explosion, and setting your wires on fire. it's like playing tennis with a hand-grenade: it never ends well.

as for using an arduino, that's equally retarded coz you can get panel meters for less than $5 on aliexpress, or amazon for a few bucks more.

as for digital control, you can use the digits on your hand to turn a knob.

>> No.1951865

>>1951862
I dont mind the work for controlling it with an mcu. Regarding the backwards transistor remark, wouldnt a fuse work?

>> No.1951874
File: 159 KB, 1488x705, step - down.jpg [View same] [iqdb] [saucenao] [google]
1951874

>>1951865

if you have to replace a fuse everytime you make an oopsie, you'll go broke. what you need is a controller that does current limiting. it acts instantaneously so fuses wont blow, and components are spared.

there is a series of front ends that do that, but they're all step-down. there may be one that's step up as well.

>> No.1951877
File: 53 KB, 1000x433, Overcurrent-Protection-using-Op-Amp-Circuit-Diagram.png [View same] [iqdb] [saucenao] [google]
1951877

>>1951874
Some quick google searching shows this op-amp based overcurrent solution, I guess having this in between the PSU and output could work?

https://circuitdigest.com/electronic-circuits/overcurrent-protection-circuit-using-op-amp

>> No.1951880

>>1951877

sure, but a 1 ohm resistor (R1) will screw your regulation. you'd need to find one with 0.1 ohm.

>> No.1951881

>>1951880
>sure, but a 1 ohm resistor (R1) will screw your regulation. you'd need to find one with 0.1 ohm.
alright, will experiment on a breadboard before I start designing all of this

>> No.1951893

>>1951881

another thing: it would be a lot better if it were variable. so, when you're playing with an LED, you'd set the limit at 20mA but with a motor, you'd set it at 250mA, etc.

>> No.1951927

seems like I can use something like LM4041 that are high precision variable voltage reference ICs rather than doing it by hand

>> No.1951929

>>1951877
uhh I just figured out googling overcurrent protection IC gives a lot of viable ready to use solutions

>> No.1951939

are oscope probe grounds connected? so if i say connect channel 1 probe ground to + and channel 2 probe ground to - on a battery, will it blow up?

>> No.1951945

>>1951939
yes

>> No.1951948
File: 146 KB, 1000x1057, IMG_20201115_180827.jpg [View same] [iqdb] [saucenao] [google]
1951948

>>1951854
Been there, done that.
It's not worth it. You can get a good lab power supply pretty cheap nowadays so dying your own from an old PSU is not worth it. The only advantage is that you get 3 clean, low ripple voltage outputs at the same time (3.3V, 5V, 12V) but thats it.
I am not going to go into detail, but there are many important features a lab power supply has which a PSU does not.
Pic related, it's a diy computer PSU lab power supply abortion at the bottom, and a beautiful baby proper lab power supply on top.
I still use the PSU one if i need to power some finished device like an arduino uno, but to actually develop new EE projects, a proper lab power supply is a must.

>> No.1951954
File: 13 KB, 964x507, file.png [View same] [iqdb] [saucenao] [google]
1951954

>that didly feelereeno when you finally find the bug, squash it, and then you see the correct waveform on the scope

fucking chip pinouts and registry bit fuckery will give me a coronary one day

>> No.1951958
File: 1.02 MB, 3900x2285, DS7014.jpg [View same] [iqdb] [saucenao] [google]
1951958

I want an oscilloscope but don't have any real use for it and would end up using it as a very expensive multimeter probably.

Pls help me justify the cost or convince me to buy something else.

>> No.1951961

>>1951958
Is that a fucking digital scope? I hope that is just a very bad joke attempt on your part.
Can that overpriced piece of shit do this?
https://www.youtube.com/watch?v=XziuEdpVUe0
Yeah, didn't think so.

>> No.1951967

>>1951929
Yes, and if you are using an arduino you can use it for current control as well.

>> No.1951974
File: 22 KB, 560x643, current.jpg [View same] [iqdb] [saucenao] [google]
1951974

>>1951967
Also I am not sure about this guy
>>1951948
But I made a 0-20V 1A linear power supply with a small xformer I found in the trash. It worked very well until I moved and had to throw it away. People do not realize how much current PC supplies can put out until they burn something expensive.
>>1951877
I use pick related when I need a quick and dirty current measurement thing. Guy asking could you that signal to trigger something to limit the current.
>>1951732
what is that?
>>1951854
If you are using low power stuff only you could try starting with a linear supply? They are very good until 20W. Also as it often said here, most times when people are thinking about using a digipot, there is a better and simpler alternative.

>> No.1951976

>>1951974
> Guy asking could you that signal to trigger something to limit the current.
> Guy asking could use that signal to trigger something to limit the current.

>> No.1951991
File: 23 KB, 787x612, current measurement.png [View same] [iqdb] [saucenao] [google]
1951991

>>1951974
Sorry that circuit is a cheap shit I got from the web and is wrong. This is the correct one.

>> No.1952022

>>1951747
Average power is what you need to care about, I think it might make more sense than average current. Either way, if the high-current peaks are short enough, then your caps will buffer that and keep the power source's output below 3A.
>Based on my calculations supplying 10A for like 4-5us should only drop the capacitor by a few mV
Sounds goof to me.

>>1951958
Best bang for your buck is a 2nd hand analog scope, which works just fine if you don't need 4 channels, single-shot, or math functions. Some analog scopes do have some of those though.
You could also get a USB scope if you're not afraid of killing your computer.

>>1951974
>>1951991
Should be using a proper output limiting circuit, with a darlington output transistor with both voltage and current feedback. Using instrumentation amplifiers as error amplifiers with reasonably high gains. Combining the voltage and current signals with diodes is kinda shit but it still works pretty well.

>> No.1952025

>>1952022
That is a current measurent circuit, not a limiting circuit. The Emitter gives a voltage proportional to:
>I*(1k/1m)*100 V

>> No.1952080
File: 57 KB, 883x450, 61Ql6wMTkAL._AC_SY450_.jpg [View same] [iqdb] [saucenao] [google]
1952080

is this style wire stripper worth buying? At least it says made in germany so it might be good?

>> No.1952093
File: 15 KB, 692x607, cade.jpg [View same] [iqdb] [saucenao] [google]
1952093

>no native LTspice for linux
I mean it runs ok in wine but it still makes me sad to not have a decent circuit sim in my main OS

>> No.1952126

>>1951749
razor blade/box cutter. don't cut into the wire, but angle it so you just scrape with the edge

>> No.1952128

>>1951732
Hi clive!

>> No.1952133

>>1951974
It's a 4 channel LED bulb dimmer which is able to process audio signal and adjust the brightness of said leds based on bands present in the audio signal such as bass, mids, etc..

>>1952128
h.. how the hell do you know my name?

>> No.1952175
File: 20 KB, 640x893, fanctrl2.png [View same] [iqdb] [saucenao] [google]
1952175

I am currently using the top circuit to turn a PWM output from a raspberry pi into a variable DC voltage, R1 just limits the current to the maximum my PNP transistor is rated for which is 6A, C1 stops my power supply being overloaded as the power supply is rated to 2.5A, but the average power flowing through the fans is max about 2A.

But the gain on my PNP transistor is only 15 at 6A, and it just occurred to me that probably the bottom circuit would work just as well with one of my 10A darlington NPN transistors... is my thinking correct? It's much less intuitively obvious (to me) that it would work... with the top circuit the capacitor is acting like a storage tank, and when the PWM cycle opens the top transistor it lets the tank fill slightly. But current doesn't actually flow from + to minus, it's just a potential difference, so supposedly it should work

>> No.1952180

>>1952175
Yes, using a Darlington is a perfectly valid decision. I’d use a MOSFET though. I’d also replace the capacitor C2 with an antiparallel diode, I think the capacitor will interfere with the dimming of the fan, but I’ve barely worked with those fans before.

>> No.1952186 [DELETED] 

>>1952180
If I was going to use a mosfet instead do you think I should be concerned about the risk of static damage (are normal BJTs more robust than mosfets?). I do have an antiparallel diode actually. I didn't want to make the circuit without a capacitor because I thought the constant stop-starting of the fan coils would cause thousands of instances of back emf per second, I thought that may be a problem. But I am fairly noob to analogue and digital circuits. The circuit works fine anyway... just I wasn't feeding the PNP transistor enough current because its gain at 2A is 100 which is good but I just looked at the spec sheet more closely after noticing it seemed to be needing quite a bit of current on its base and it drops to 15 at 6A (which is the rate it's actually fed during the pwm duty cycle)

>> No.1952188

>>1952180
If I was going to use a mosfet instead do you think I should be concerned about the risk of static damage (are normal BJTs more robust than mosfets?). I do have an antiparallel diode actually. I didn't want to make the circuit without a capacitor because I thought the constant stop-starting of the fan coils would cause thousands of instances of back emf per second, I thought that may be a problem. But I am fairly noob to analogue and digital circuits. The circuit works fine anyway... just I wasn't feeding the PNP transistor enough current because its gain at 2A is 100 would be fine but I just looked at the spec sheet more closely after noticing it seemed to be needing quite a bit of current on its base and it drops to 15 at 6A (which is the amount of current it's actually passing during the pwm duty cycle)

>> No.1952253

>>1952188
Once a MOSFET is inside a circuit, static ceases to be much of an issue. Add some 1M or 10M resistors if you’re worried.

You’re right, PWMing an inductive load does produce a lot of inductive voltage spikes if there’s no protection. The purpose of the reverse diode is to short all the back-emf spikes back through the motor, preventing them from getting above 0.7V. Using a diode here is a common solution to this called a “freewheel diode” or “flyback diode” (there are no flyback diodes in a flyback converter). Capacitors on the other hand aren’t commonly used for this. But caps (usually lower ESR caps like film caps) are sometimes used for EMF suppression from commutating brushes and the like.

Note that if it’s a computer fan with its own BLDC driver it might have its own freewheel diode internally.

>> No.1952286 [DELETED] 

>>1952175
>But the gain on my PNP transistor is only 15 at 6A

this is not a problem at all. it just means the NPN has to be able to pass about 200mA, which is easy for even a small 2N2222A transistor. assuming a gain of 50, it would draw 4mA from the arduino, which is only 10% of the 40mA available.

>> No.1952289

>>1952175
>But the gain on my PNP transistor is only 15 at 6A

this is not a problem at all. it just means the NPN has to be able to pass about 133mA, which is easy for even a small 2N2222A transistor. assuming a gain of 50, it would draw 2.6mA from the arduino, which is only 6% of the 40mA available.

>> No.1952292

>>1952253
Ok thx.

Ok now I've gone and collapsed part of my brain. I thought that a PWM circuit could be theoretically near 100% efficient. My fans run at 2A at 7V, that's the max voltage I want to run them from my 2.5A 12V power supply. So if my PWM circuit was powering my fans at 2A at some duty cycle with the capacitor C1 or an equivalent from my first pic taking care of the current spikes during the on part of the duty cycle for the power supply then the average current coming from the power supply would be 2A. Or, I could power the fans through a resistor to limit the current to 2A, except the resistor would be dissipating a lot of power. Except in either case the power supply would be supplying an average of 2A at 12V... the power supply would apparently be supplying the same amount of power. Where did I go wrong with this line of reasoning?

>>1952289
It would be 400mA, 6/15 is .4, which is just a tad higher than I wanted... I'm going to try a N channel mosfet tomorrow though as I just remembered I bought a whole bunch a while ago to experiment with but never used them, I have 30, 70, even 150A ones with logic level Vgs threshold

>> No.1952304

>>1952292
The only problem of the capacitor C2 is that, when the transistor stops supplying the motor with current, the capacitor will continue doing so. Hence a 50% duty cycle will end up giving it more power than it might otherwise. At the extreme, using a 100F super capacitor would result in the motor remaining at near 100% power.

As for your reasoning, it only works for linear systems as far as charge is concerned. Resistors and capacitors cannot destroy or create charges, just give or take energy from them. But inductive circuits can, you can put 12V at 0.5A into a switching power supply and get 6V at 1A out the other end, or at least almost. In other words, Kirchhoff’s current law can’t be applied generally to a switching topology. Purely capacitive charge pumps do this too, but we don’t talk about them. In this case, you can treat the inductive motor as the inductive element of a buck converter.

Though it’s just as valid to look at average current of a PWM circuit. In general, average current in and average current have to be equal here, but instantaneously they don’t, and that too can give you high-efficiency differing currents.

In other words, your 12V 2A PSU isn’t putting out 2A at all, but rather only as much as it needs to to keep C1 charged.

>> No.1952313

>>1952292
> I could power the fans through a resistor to limit the current to 2A, except the resistor would be dissipating a lot of power.
Add an inductor to the supply line to smooth the current draw. The capacitor needs to have higher energy storage than the inductor (CV^2 > LI^2) to keep the voltage stable.

>> No.1952319

>>1952292
>It would be 400mA, 6/15 is .4

not according to what you said...
- 6A is the current RATING of the transistor
- 2A is ACTUAL max current draw

so 2/15 is 0.133A

>> No.1952322

>>1952304
>The only problem of the capacitor C2 is that, when the transistor stops supplying the motor with current, the capacitor will continue doing so. Hence a 50% duty cycle will end up giving it more power than it might otherwise.
Yes I have figured that, 50% duty cycle no longer translates to 50% power... I figure when I try my 70A+ mosfets tomorrow I may be using duty cycles of less than 10% at all times (currently practically using 30-40%). But the raspberry pi PWM seems good, you set a period and duty cycle in nanoseconds, at the moment I'm using a period of .1ms, so 10,000Hz... Intel PWM fan spec suggests a PWM target of about 25kHz.
>>1952319
Right yeah, but because it's PWM average current is 2A but I'm aiming to max out the current going through the transistor during the duty cycle so I tried to get it near 6A, but then would have to supply 400mA through the first transistor. R1 was to set the max current through the transistor, tomorrow I'll switch the 6A transistor for a 60A mosfet and reduce R1 from 1R to .1R

>> No.1952333

>>1952175
what you set up is more like a constant current source. link related.

https://www.youtube.com/watch?v=iH4iWcGJVG4

gain is seen more on a npn as you have current running through base and collector but you would be using it on the high end above the load so resistance then determines your voltage. gain is listed on npn datasheets reliably and a common breaking point for noticeable increase is after 200 volts not current put in.

also if you want this to be a variable voltage power supply where is your inductor? and potentiometer

>> No.1952336

>>1952304
Ok this is strange, just playing with equations, suppose I power a 5 ohm light bulb and a 5 ohm resistor in series from 10V:
5ohm light bulb, 5 ohm resistor, P=V^2/R=100/10=10W, 5W through the light bulb, current=V/R=10/10=1A

Now remove the resistor:
I=V/R=10/5=2A
P=V^2/R=100/5=20W

Now run the circuit without the resistor at 25% duty cycle:
So if we run at a 25% duty cycle we'll have 5W in light bulb, same as the circuit with the resistor in series, and an average current of .5A. The power supply will be providing twice as much current when it is switched on (2A), so combined with the 1/4 duty cycle we have overall doubled efficiency compared to the circuit with the resistor (naturally).
Average current=.5A
P=5W=I^2*R=.5^2*R=.25*R
R is now apparently 20 ohms. The PWM duty cycle has affected the apparent resistance of the circuit.

>> No.1952340 [DELETED] 

>>1952333
I always thought of the capacitor C2 as a jug of water, and each time the transistor above it switches on it starts filling up quickly (at 6A) while all the time discharging at a slower rate into the fan, so the PWM cycle just controls the rate at which the jug is refilled, and thus the head of water (voltage) available to push current into the fan. It does actually work though, have been using it for a couple of weeks

>> No.1952343

>>1952333
I always thought of the capacitor C2 as a jug of water, and each time the transistor above it switches on it starts filling up quickly (at 6A) while all the time discharging at a slower rate into the fan, so the PWM cycle just controls the rate at which the jug is refilled, and thus the head of water (voltage) available to push current into the fan. It does actually work though, have been using it for a couple of weeks. C1 and C2 are 4700uF 16V caps (I had two spare so thought might as well use them, at that size I can't go wrong)

>> No.1952381

>>1951958
>>1951961
>>1952022
How much should I pay for a second hand analog sciope?

>> No.1952396

>>1952381
Anything from free to a hundred bucks, really depends what you can find that's either in your area or can ship for a reasonable price.

>> No.1952412
File: 52 KB, 537x768, cc2650-blockdiagram.png [View same] [iqdb] [saucenao] [google]
1952412

Reposting my question;
>>1949446
https://www.ti.com/lit/ds/symlink/cc2650.pdf
>The thing says it has an internal DC-DC converter as well as its own battery monitor.
>Am I correct to suspect this thing is idiot proof in that I can just plug a LiPo battery to it and have at it with no external power electronics whatsoever?
The datasheets really don't seem to go into any details. Some forum posts I've looked up seem to affirm that no external circuitry is necessary and it's all handled internally.

Followup question to the above;
I plan on charging my LiPo with a TP4056 (as was advised to me on here a while back).
I'm following this webpage for designing the circuit;
https://www.best-microcontroller-projects.com/tp4056.html
He suggests to use a P-MOS to flip between powering your circuit by battery, or charging via supply voltage.
My LiPo will vary between ~4.2V max and ~3.0V min overtime. My charging supply will be coming from a USB-C voltage (5V).
Will the internal CC2650 DC-DC converter be able to handle flipping between the LiPo and a 5V supply line?

>> No.1952416

>>1952343
thats interesting. so you get overlap between input and discharge through the load?

well if its working and you wanted to adjust the voltage just slap a potentiometer on it that can handle the current and adjust the resistance to get higher voltage . increases are minimal without a inductor or transformer earlier in the circuit to log a higher voltage. i think there is a push pull method that doesnt require a inductor or transformer that power packs for power tools uses to boost 12 volts to 18 volts. the back flow allows it but not every powertool company uses it

>> No.1952448

>>1952133
show us your bench pls

>> No.1952457

>>1952336
>PWM duty cycle has affected the apparent resistance of the circuit
Yes, under the assumption that the apparent input voltage is constant. But the voltage going to the load is actually a square wave, so saying it's changing the effective resistance is a bit of a stretch. Still, if it helps you understand what's going on.

>> No.1952477

is there any easy way to measure 230AC with a scope that can only handle 120AC peak to peak?

>> No.1952478

>>1952477
voltage divider

>> No.1952479

>>1952477
Get a 100x probe

>> No.1952483

>>1952477
Even with a voltage divider or 10x probe (add some zeners for safety) you'll still likely not be able to see the signal. My scope has an input voltage maximum of 400Vpk, but anything over 40V is outside the bounds of the CRT at maximum V/div. Though maybe that 400Vpk applies to when you're using a 10x probe too, though I can't see why it would.

You should be able to use a voltage divider either way, but note that it may interfere with the 1M input impedance of the scope if its resistance is comparatively sized. Dividing it then buffering it with an op-amp might be a good idea.

A common line step-down transformer is also a pretty good idea, though it will distort particularly low and high frequencies.

>> No.1952494

>>1952478
>>1952479
>>1952483
thanks, i don't really trust voltage dividers, i will try the transformer though, it would make me much less scared of burning something if the low voltage output is optically isolated by the transformer

I got one more question tho, i have a generic triac. I turn it on by suppling a 0.1ms pulse to its gate, how long should i wait when it hits the zero crossing to supply this pulse so that it's on for the biggest part of the wave? If hit it exactly at the moment it hits the crossing then it obviously doesn't work.

>> No.1952505

>>1951732
>>1952133
What uni course is that for? That's really nice.

>> No.1952509

>>1952483
What is the max V/div your scope has?

>> No.1952511

>>1952494
>i don't really trust voltage dividers
I'd probably only use a voltage divider with a couple of antiseries zeners for protection

>>1952509
5V/div, 8 divs from top to bottom, is that abnormally low?

>> No.1952515

Alright chums I need some circuit advice. I'm designing a power circuit for a raspberry pi. Right now I've got planned 2 18650s in series and then regulating that down to 5V with a voltage regulator, one of those cheap switching ones from ebay. I'd also like to charge these batteries and want to be able to charge the batteries while running the computer.

To accomplish this, I realize that I can notionally reconfigure the circuit so that the computer and batteries are in parallel, and I have a 8.4V, 2A DC power supply. I never took any electrical engineering classes and it shows, but I've read on forums where other people did this and just *magically* the batteries will draw whatever load is left over from the supply, they say.

As in, the computer only pulls ~0.5A from the supply, so there's 1.5A available. The batteries will only draw 1.5A? The charging board I have for my batteries limits current inflow to 8A, so what's to stop the batteries from taking the full 2A? I don't understand why it's guaranteed that the computer will be powered. It feels like I should add a current limiter to the battery circuit to guarantee enough current routes to the computer when I'm charging it, but I don't know. Please excuse that I'm an electrical idiot.

>> No.1952519

>>1952515
Draw circuit diagram.
>>1952511
well you should be able to see 230V ac in there with 10X.

>> No.1952521

>>1952519
we should start putting up again general recommendations for asking stuff again in the OP
I remember we had something about posting circuit diagrams, I wonder why is it gone

>> No.1952562

>>1952505
none, i'm making it for a fren, poor sod has been waiting over a year now
>>1952448
i don't really have one my bench is my computer desk (in case you haven't caught on i was kidding i'm obviously not big clive) but i did buy the cliff tester solely because of his videos because i have never seen a device like that before and had to have it and love it

>> No.1952571

Hello fathers i never had. Would it be okay to repair a frayed light switch cable with two connector blocks and a new bit of wire?

>> No.1952573
File: 77 KB, 634x703, 1585003915039.jpg [View same] [iqdb] [saucenao] [google]
1952573

Hiya lads,
Looking to get +- 12V out of a 12V LiPo battery with >500mA @ -12V and >1A @ +12V
Having a look around though I'm struggling to come up with an IC/Circuit to achieve this...? I'd rather avoid a separate battery and the majority of the circuits I can find are either AC->DC or the few DC->DC seem to be specialize audio/synth supplies at 5V or just 100mA and no diagram's supplied.
Is there some trivial technique I'm staring right past here or any other advice / suggested reading?

For reference I'm strapping a DIY show laser to a drone for a flying light show, galvanometer drivers require +- 12V and the only power's coming from the onboard LiPo

>> No.1952576
File: 66 KB, 703x703, 1603012719092.jpg [View same] [iqdb] [saucenao] [google]
1952576

>>1952571
Go for it champ, so long as just a light switch and nothing heavier duty. Just make sure your new wire / connector blocks are rated for the current and you'll be right, oh and make sure nothing's exposed contact wise and it's all secured in place

>> No.1952577

>>1952571
Yes.

Also I want you to know I wish I had a cute daughter instead.

>> No.1952583
File: 295 KB, 2000x1500, rsz_img_20201116_200120 (1).jpg [View same] [iqdb] [saucenao] [google]
1952583

This fan pulled from my induction stovetop stopped running. I am guessing that the copper coil overheated and melted itself. Good news: the part is modular plug and play. Bad news: I don't see any replacement parts.

This looks like it could be saved or replaced--any advice on whether or not similar fans are available for purchase?

>> No.1952589

shieet turns out that skipping say every 10th cycle in a dimmable led bulb and instead using that cycle for calculations will fuck with the led and make it flash randomly, i had hoped it would simply case that led to be off once every 10 half sines but nope
So i will have to rewrite my code somehow and tickle the triac even during the logic cycle

>> No.1952591

>>1952583
Can you power it? If so, can you try starting it manually? (literally just slapping the fan in any direction while it is powered but not spinning). That is a single phase shaded pole motor, with age the mechanical stuff might get stuck and it will have problem starting, but not so much running. Sometimes it also happen that it is touching something inside the enclosure and cannot start properly, so check wherever it was for dirty, kinks in the casing and other things that might cause trouble.

>> No.1952593
File: 32 KB, 599x529, burger.jpg [View same] [iqdb] [saucenao] [google]
1952593

>>1952589
I had a stroke reading this

>> No.1952595
File: 19 KB, 501x192, file.png [View same] [iqdb] [saucenao] [google]
1952595

>>1952593
Just imagine this, except every 10th half sine is missing, because the triac is never triggered during it, because the micro controlling the triac has no time to trigger it, because it is running some calculations.
And as it turns out, dimmable 230V LED bulbs don't like that at all and will act out and get violent. I had hoped the buld would simply be off during that missing half sine, but no, it just messes up everything.

>> No.1952598

>>1952589
>instead using that cycle for calculations will fuck with the led and make it flash randomly,

interrupts anon, interrupts.

>> No.1952600

>>1952598
I am already using interrupts to detect the AC zero crossing, but i will have to delve deeper and figure out how to recode that shit so it somehow triggers during that data cycle as well.
It's not like i can stuff the whole thing into an iterrupt, if interrupts take too long they mess up things like timers and so on

>> No.1952616

>>1952600
You cannot make an interrupt that fits in 1/120 seconds? you literally only need to send a pulse.

>> No.1952618

>>1952616
>zero is crossed
>wait the amount of time for the phase angle you want
>trigger = on
>wait the time the datasheet recommends
>trigger = off
>return to main program

>> No.1952628
File: 130 KB, 1567x677, dc-dc converter, 500V isolation.jpg [View same] [iqdb] [saucenao] [google]
1952628

>>1952573

how does $36.42 CAD sound to you?

>> No.1952629

>>1952595
you are basically adding a signal which frequency is 1/5 (because its every 10th half sine) of the original wave frequency.
so if you are going with 50hz sine wave, you are adding a 10 hz pulse signal which is very noticeable for humans.
use interrupt as someone said, a 50 hz interrupt is not so demanding

>> No.1952630 [DELETED] 

>>1952583
>This fan pulled from my induction stovetop stopped running.

odds are better than 90% that some oil in the bearings will make it work a-ok.

>> No.1952632 [DELETED] 

>>1952583
>This fan pulled from my induction stovetop stopped running.

odds are better than 90% that adding oil to the two bearings will make it work like new.

>> No.1952633

>>1952629
>>1952616
It's not just a dimmer, it has to do other calculations which take about 5ms so it misses about half of the 10ms window in which it has to trigger the triac during that half sine
If i was only trying to make a led dimmer and nothing else it would be trivial

>> No.1952636

>>1952583
>This fan pulled from my induction stovetop stopped running.

odds are better than 90% that adding oil to the two bearings will make it work like new.

as for replacements, there are tons of online and bricks-and-mortar stores selling appliance parts. just google.

>> No.1952640

>>1952583
It's just a fan motor. You could replace it with a bathroom ceiling fan.

>> No.1952656

>>1952633
are those calculations used in the interrupt?

>> No.1952691

>>1952573
>Looking to get +- 12V out of a 12V LiPo battery with >500mA @ -12V and >1A @ +12V
>>1952628
>Posts converter with one +12@500ma
Try again.

>> No.1952693

>>1952573
Traco THN 30-1222
https://www.mouser.com/ProductDetail/TRACO-Power/THN-30-1222?qs=ckJk83FOD0Uwgp%252BPXcGZgw%3D%3D

>> No.1952694

>>1952691

you're such a dumbass.
1) the output is isolated, meaning you can use it as either a positive or negative source. just reverse the wires.
2) you dont need a converter to get +12 from +12

>> No.1952700
File: 23 KB, 480x522, soyjack glasses.jpg [View same] [iqdb] [saucenao] [google]
1952700

>>1952691
>Try again.

>> No.1952707

I'm this guy >>1947730
I finally got the Brit guys' board working. Here's what was wrong with it, if anyone's curious.

>used an underrated buck-boost converter (0.8A max current, even when the peak current for the ESP itself is 0.8A excluding other components)
>traces for power were thin af 0.15mm
>output from buck-boost converter was being fed into the power plane through a thin ass 0.15mm wide track and 0.8mm via. The guy basically had to bypass the entire power supply by soldering a 3.3V power bank to a test point to supply power.
>transistors being used to flash ESP32 (connected to RTS and DTR) had a gate threshold voltage of 0.5V, meanwhile the RTS and DTR lines outputs "low" voltage was 0.7V. Basically the transistors were on all the time and grounding BOOT0 and CH_PD/EN, and ESP32 was held in an undefined state. Guy had to cut traces to the transistors gate with a hobby knife
>accidentally put 470k resistors in place of 470 ohm to current limit manual reset and boot mode. Buttons didn't reset the board. The guy had to solder wires across the resistors to bypass the 470k ones.
>accidentally routed the I2C lines too not connected pads of the ESP32. Guy had to solder bodge wires to the correct pads from a test point.

What a shit show man, I am embarrassed

>> No.1952722

>>1952707
You did all that or he did it?

>> No.1952752

>>1952595
I don't understand what you're doing.
Why not just use PWM for the dimmer? Don't those work as a background process?

>> No.1952755

>>1952752
He said they are 230Vac led bulbs.

>> No.1952762

>>1952722
He did all the soldering and bridging stuff and probed test points with a multimeter, I looked at the schematics and figured out what wasn't working, and told him where to probe.
If the guy didn't own a multimeter or soldering iron, he wouldn't have been able to get the board working.

>> No.1952769

>>1952762
So those were your errors? I'd give the money back if I were you. Or order a correct border on your expense

>> No.1952799

>>1952769
>unemployed
>having any integrity

>>1952707
Reading your bullshit I can't stop wondering where you fucked up
>used an underrated buck-boost converter (0.8A max current, even when the peak current for the ESP itself is 0.8A excluding other components)
Literally non-issue. I've run for years on tiny piss SOT-25 with max output current of about 150mA for ESP32 without any issues. If you have any capacitors on PCB connected to same power net you'll be fine.

>transistors being used to flash ESP32 (connected to RTS and DTR) had a gate threshold voltage of 0.5V, meanwhile the RTS and DTR lines outputs "low" voltage was 0.7V. Basically the transistors were on all the time and grounding BOOT0 and CH_PD/EN, and ESP32 was held in an undefined state. Guy had to cut traces to the transistors gate with a hobby knife
>gate
Did you use mosfets?

>accidentally routed the I2C lines too not connected pads of the ESP32. Guy had to solder bodge wires to the correct pads from a test point.
>no erc
>no drc
>taking money

>> No.1952802
File: 20 KB, 255x247, 1598975204433.jpg [View same] [iqdb] [saucenao] [google]
1952802

>>1952691
lol, retard

>> No.1952809

>>1952755
I didn't know it took 230v to trigger a triac

>> No.1952845
File: 1.53 MB, 4160x3120, basic fan contraption.jpg [View same] [iqdb] [saucenao] [google]
1952845

I've made a simple fan contraption that I'm later planning to use as a solder fume extractor. I bought a 24V 0.16A computer fan, a PWM module, a plastic case, DC plugs m/f and mistakenly bought a 12V 2A DC adapter plug instead of a 24V one.

So now I have two choices, buy another power supply rated at 24V for around 4$ and wait months or I boost the voltage from 12V to 24V if I can. I'm thinking about doing the latter as I may have the components at home to make a simple buck-boost converter. But, because I have the PWM module there, I'm not quite sure how to proceed. I'm thinking that I would have to probably smooth the output of the buck-boost before it goes into the PWM module.

From what I understand, the way buck-boost converters work is that I would need an inductor (how would I measure the right value for it?) and short it to ground periodically to boost the voltage of the circuit. I'm thinking that it's not going to be possible with a simple 555 timer because the voltage spikes will be higher than the 555 is rated for and on average it will be 24V. Thoughts?

The PWM module is this one if it matters:
https://www.aliexpress.com/item/32836419349.html?spm=a2g0s.9042311.0.0.27424c4ds494uy
Image related is my super simplistic contraption.

>> No.1952867

>>1952845
Order the 24V PSU and use three 9V batteries in series with a power switch until it arrives.

>> No.1952869

>>1952809
>PWM
>triac

>> No.1952877

What finger do you (almost) habitually burn after unknowingly touching a component? For me, it's my thumbs.

>> No.1952970
File: 3.24 MB, 3120x4160, IMG_20201116_195509.jpg [View same] [iqdb] [saucenao] [google]
1952970

I designed this board with one of the potentiometers flipped so now 0 is 10 and 10 is 0... Is there any conventional way to fix this without ordering boards again?
Maybe inverted potentimeters?

>> No.1952976

>>1952519
>well you should be able to see 230V ac in there with 10X
Yes I know, but with 10X mode, the voltage at the end of the BNC will be only 23VAC. I'm wondering if I actually can put up to 400Vpk into the BNC, or only 40Vpk.

>>1952573
Two 3S lipos in series, with the centre-split used to give you ±12V. If you need those to be regulated, use two 4S lipos with buck or linear regulators on each.
While a bit of a pain, it's going to be simpler than getting a power rail inverter of some kind. If you've already got a battery and don't want to get another one or two, lmao have fun with that

>>1952618
Pretty sure onboard timers can work behind the scenes while you do other shit. I'd add some logic to call the math function either before or after the "wait the amount of time for the phase angle you want" section, depending on which time period is longer.

Or you could just use a normal analog TRIAC dimmer, with the pot replaced with a vactrol. No need for zero-crossing timing bullshit.

>>1952970
snip legs, solder bodge wires
bend legs, solder facing different direction
just put labels 10 and 1 on the opposite sides

>> No.1953002

>>1952976
I think snip and solder is the best thing I can do here. Thanks anon! I just wish rev-log ment what I thought it ment...

>> No.1953029

>>1953002
You could flip all the components to the opposite side of the board, though this would mean bending the DIP8 upside-down and having nonmatching silkscreen

>> No.1953041

anyone ever repaired a plasma TV before? I have one with a power to ground short that I suspect is blown MOSFETs on the Y sustain board. I bought a repair parts kit but I'm not sure I need to replace them all, and some are sma enough that it would be risky to do so I think. Also since the MOSFETs tend to be in parallel they might not all be bad. Do you think it's worth checking for the short after removing each part so I don't take out any good ones or is it likely that they're weak and partially damaged and I should replace them anyway?

>> No.1953074
File: 19 KB, 665x350, diag.png [View same] [iqdb] [saucenao] [google]
1953074

>>1952521
>>1952519
I have returned with a circuit diagram. The switch on the voltage supply is just to represent dis-/connecting the supply plug. The computer load will vary based on usage, but most power estimates show an upper end of around 1Ah. Sorry for the shitty diagram, I'm still learning, so if I've done something wrong please tell me

>> No.1953090

>>1952769
Yeah they were my errors. I told him how to fix them (bridging what to where etc). I took 70$ for the job, he even tipped me 10£ (12$). I'm highly considering returning him the money, but he's already spent more than 300$ to get 5 boards manufactured and assembled. Of course I can't refund 300$, so 82$ will have to do.

>>1952799
>Literally non-issue. I've run for years on tiny piss SOT-25 with max output current of about 150mA for ESP32 without any issues. If you have any capacitors on PCB connected to same power net you'll be fine.
It's most likely something else, the ESP kept boot looping when he tried to power it from the buck boost converter. Also should mention both pads of the inductor are connected to the buck boost converter using similar thin 0.15mm traces without any kind of large pads or plane like the design guide suggested.

>Did you use mosfets?
No they were NPN BJTs, tiny SOT-23 ones. BQ something

>> No.1953092

>>1953090
>the ESP kept boot looping
Use a beefier PSU.

>> No.1953110

>>1953074
Oh no I'm obviously retarded and the power supply positive is supposed to go to the top lead.

>> No.1953193

>>1953074
Use a 12V PSU. Put the buck converter set to 5V on the Pi and the regulator on your BMS. Use a switched DC barrel jack so when it's plugged in the battery charges and the Pi is powered, but when it's unplugged the battery discharges into the buck converter.

>> No.1953195
File: 2.74 MB, 4032x3024, 9877F426-66AB-4A72-AE28-33B285720777.jpg [View same] [iqdb] [saucenao] [google]
1953195

Can anyone identify these DB158 parts? Pretty sure they’re not bridge rectifiers. The PCB had a bunch of these daughter boards and analog switches, which looks really interesting.

>> No.1953198

>>1953195
>Part : DB158
>Category : Power Management
>Description : Dual-in-line Glass Passivated Single-phase Bridge Rectifier Voltage-50 to 1000 Volts
>Company : Surge Components Inc.

>> No.1953211
File: 209 KB, 1080x1084, rice.jpg [View same] [iqdb] [saucenao] [google]
1953211

>>1952694
Ahh that was next question about wiring up a converter backwards, thanks this looks great

>>1952691
Not me but I appreciate it

>>1952693
Also looks good I'll weigh up shipping times between these

>>1952976
Might consider another battery too in the mean time...


Last night I tried connecting up a boost converter to a 12V supply, boosting to 24V then connecting 12V to the GND input on my circuit, PSU GND to -12V and 24V to +12V on the circuit. Probing with my multimeter this all looked good, had the right potentials relative to my "12V Ground" but my circuit refused to work, the galvanometer mirrors just froze in place instead of spinning like on my +-12V AC adapter. The only difference I noticed was that the AC power adapter was reading 0V between +12V and -12V instead of the 24V I was expecting.

Does anyone know what the difference is here?

>> No.1953246

>>1953193
That's more or less the design I had in mind, but what current will the circuit draw? Like I said in my first post, how can I be sure that when plugged in the pi will get the 0.5A at 5V it needs? Am I relying on a current limiter built into the BMS to cap it at e.g. 1A so that the power supply always has enough for the pi?

>> No.1953253

>>1953246
The BMS shouldn't draw more than 1A per battery when charging, and the Pi no more than 2A at full blast, so ~4A is the total maximum current draw when all circuits are powered. Make sure your buck converter can supply at least 2A constant current and your main PSU can supply at least 4A constant current.

>> No.1953265

>>1953246
Short circuit & thermal protection is part of the BMS. A sinlge 18650 cell alone can discharge a continuous current of 20A, so the main limiting factor is the buck converter.

>> No.1953269

>>1953211
Flipping a converter only works if it’s an isolated converter. These have integrated transformers and use an optocoupler to talk across the gap, and are more expensive as a result. They’re usually rated to a few hundred or thousand volts of isolation. Common DC-DC converters are non-isolated since they just use a single inductor, and have a direct connection from their negative input and negative output (give or take a current sense resistor). Trying to invert a non-isolated converter means shorting it out completely.

Note that there are some non isolated circuit topologies that invert a rail (see: some rs232 transceivers) but these usually use capacitive charge pumps and hence aren’t suitable to high currents.

>> No.1953273

>>1953265
Right, that's clear to me for the case without the power supply connected. These balancing protection "BMS" boards on ebay mostly cap over current at around 8 to 10A, which is gonna max out whatever supply because it's mostly for short circuits. So I either need to add a current limiter to the batteries inline with the supply or get a more sophisticated BMS for the batteries, right?

>> No.1953283

>>1953273
The 8-10A overcurrent protection is for the discharge cycle. The BMS will limit the charge current from your PSU to each 18650, usually at 1A per cell.

>> No.1953286

>>1953273
Misread your post. If you're really worried about protecting the Pi from too much current, power it through the micro USB/C jack which is protected with a PTC fuse and TVS diode.

>> No.1953291

>>1953286
No I don't think you misunderstood it. The worry is that the Pi won't get enough current when the batteries are being charged. The BMS is still in the mail so I don't know what ICs it has on it, but we'll assume it does limit the cells to 1A each. A 12V 2A supply gives 24W of power, leaving 6.6W for the pi, which will work out to only 1.3A available. So either higher voltage or amperage on the supply to compensate.

>> No.1953303

>>1953291
>So either higher voltage or amperage on the supply to compensate.
This, although the BMS should still charge each cell at 500mA if there is enough power.

>> No.1953336

>>1953198
Wrong package mate, these only have 4 legs and are in a DIP package, this is a SOIC8 or something.
http://pdf.datasheetcatalog.com/datasheet/surge/DB104.pdf

>> No.1953355
File: 67 KB, 800x800, photo_2020-11-16_21-27-40.jpg [View same] [iqdb] [saucenao] [google]
1953355

>>1953303
I won't be able to tell for my board specifically until I get it and can examine the components, but I want to understand how the circuit limits the current to the battery. The HY2213 seems like a pretty common IC on these boards and there's one for each cell. That datasheet says Vdd will be around 3.9V and the limiting resistor is 100Ohm, so that means the current provided to the battery is only ~40mA?

Pic related is a different board, but let's say I'm using this one. It looks like the two small components are the charge ICs, so those are 101Ohm resistors. Then, the batteries will only receive 38mA there?

>> No.1953427

If I have green and blue cables which should be positive and which should be negative?

>> No.1953432

>>1953427
Well, with mains wiring green is ground and blue is neutral. Neutral is less safe to touch than ground. Also I think blue is one of the lives in some 3-phase wiring color schemes. So I'd make green your negative and blue your positive. Unless you've got a positive ground plane and centre-negative barrel connectors and such, in which case do it the other way around.

I'd want to buy more wire though.

>> No.1953447

>>1953432
It's for putting a second level of charge on my breadboard, red for 3.3v to drive the chip/logic, and 7v to drive a display
not using it for anything high voltage or permenant
Just want to have a consistent approach for future projects

>> No.1953499 [DELETED] 

Does anybody know the name of this circuit? At first glance it's a common emitter with an active load and active bias, however I am told it has a special name. It has somethig to do with load curves, that's all I know.
I'd appreciate any help. The gain is around 0.6 with a 2k2 load and around 0.2 with an infinite load (that is, open circuited).
Thanks!

>> No.1953500
File: 29 KB, 574x380, Circuit1.png [View same] [iqdb] [saucenao] [google]
1953500

Does anybody know the name of this circuit? At first glance it's a common emitter with an active load and active bias, however I am told it has a special name. It has somethig to do with load curves, that's all I know.
I'd appreciate any help. The gain is around 0.6 with a 2k2 load and around 0.2 with an infinite load (that is, open circuited).
Thanks!

>> No.1953509

>>1953447
>Just want to have a consistent approach for future projects

this is autism talk. just do ''whatever'' like normal decent people do.

>> No.1953513

>>1953500
Those gains are with R1=R2. At first my theory was that it was some sort of programable current source, controling the R1 and R2 ratio to determine how much of the current sourced from the up mirror is sinked in the lower one. But I can't find any info online about it. Thanks for any help

>> No.1953631

>>1953500
class A amp

>> No.1953633

>>1953500
why do you keep posting this? it's unity gain, low output impedence, low power consumption.

>> No.1953639

If a battery doesn't have an internal NTC, could I just slap a thermistor to the casing of it and monitor the temperature that way?

>> No.1953641

>>1953639
Also; does anyone have recommended websites for buying low quantity Lithium Poly or Lithium Ion batteries?

>> No.1953648

>>1953641
There are name brands, which is what I'd aim for. I bought some name-brand CNHL lipos off an official aliexpress seller, they work well. 18650s are much more prone to being faked, but you can probably ask about to find a reputable seller.
>>>/diy/rcg has a fair bit of knowledge in this area, otherwise check reddit or wherever

>> No.1953652

>>1953648
>CNHL lipo
>18650s
Just looked these up. These are pretty beefy, I assume used in RC applications.
I need small form factor stuff. The kind used in cellphones and other similar devices.

>> No.1953658

>>1951716
if i have a analog voltmeter and no oscilloscope to measure a dc pulse signal do i do dc or ac mode? i did both and i get different readings. its like the ac one is 2 times as large. is this due to the radioshack voltmeter thinking dc is supposed to be continuous so this pulse set doesnt feed enough or is the ac one needed for any frequency measurement?

i can accept that the dc pulse is genrating ac when its at the low because the current wants to flow backwards in a circuit and its not that important because i was going to throw it on a transformer anyways but its a noticeable difference and the volt meter reads batteries properly

>> No.1953663

>>1953652
>I assume used in RC applications
Yes. At a few hundred amps, really slick shit.
There are some very small RC quads and cars that might have decent lipos for, but in general small electronics lipos have very different requirements compared to RC ones. So I'm not really sure where to get the standard lower current ones, so look about the place.

>>1953658
No clue how an analog voltmeter works on AC mode, but it should be relatively simple to calculate.

>> No.1953669

>>1953658

you have the wrong tool, so neither AC or DC modes will tell you the truth. in DC, the signal is integrated by the physical movement of the meter, so you'll get a lower or higher voltage depending on the duty cycle. but it's very very approximate.

on AC, the input capacitor will differentiate the signal, causing spikes on both rising and falling edges, leading to inflated results. anything that's not a sine will be wrong, and a pulse will be the wrongest.

there's no substitute for a scope in a case like this.

>> No.1953676
File: 347 KB, 1020x570, file.png [View same] [iqdb] [saucenao] [google]
1953676

What sort of probe do i use that is small enough to allow attachment to legs on soldered dip chips or smd chips?
Probing multiple legs on chips that are soldered on is a fucking nightmare.
Like for example let's say i have pic related, it's SOIC8 and i want to probe say 4 legs at the same time, what kind of probe is small enough where i can attach 4 of them in a line to all the 4 legs?
Even those small hooks on normal oscope probes are MASSIVE and stand no chance here

>> No.1953680

>>1953676
you put test points/vias in your pcbs to probe or solder something in

>> No.1953684

>>1953680
None of those things are an option, i often don't design the pcbs i need to probe so i can't put any test pads and soldering small wires to attach probes to is an absolutely LAST option, it's so tedious to do, i would rather castrate myself than do it. I need need some tiny probe advice.

>> No.1953685 [DELETED] 
File: 312 KB, 1045x784, 20201111_130837.jpg [View same] [iqdb] [saucenao] [google]
1953685

>> No.1953687

>>1953669
damn. well thanks i guess i just have to wing it

>> No.1953690

>>1951854
https://www.youtube.com/watch?v=CIGjActDeoM

>> No.1953702

>>1953658
if it's below 200khz, a dso138 off ebay is a cheapish o-scope. I would seriously consider buying a big boy scope though, multimeters just aren't good enough
>the ac one is 2 times as large
what's the duty cycle of the pulse?

>> No.1953704

>>1953684
steady hand, press up against the pin. You aren't actually using the hook, are you?

>> No.1953706

>>1953704
enjoy your shorts and dead ICs fren

>> No.1953707

>>1953690
>aussie.
I run the audio through pulse and boost the bass and tame the treble to make it listenable....
human asset tho.

>> No.1953708

>>1953706
git gud faggot

>> No.1953713

>>1953676
What you need is a probe clip.
https://www.aliexpress.com/item/4001159503853.html
https://www.aliexpress.com/item/4000838223639.html
This kinda shit.

>> No.1953714

>>1953355
Are you talking about the balance circuit? 40ma could be accurate. It doesn't need to be much because it only activates when the cells are out of balance, which should happen rarely if ever.
That's not the same as current limiting. Normal BMS don't do current limiting, they just shut off when the current is exceeded. If you want to charge at a particular current you need a separate current limiting cirucit.

>> No.1953716

>>1953713
The trouble with those are that these are made for specific use only, you can ONLY use them with say SOIC8 chips and nothing else for example. So you would need to have 50 different clips if you wanted to be somewhat universal.

I actually already have that exact clamp, but often i need to probe things that aren't that exact package form factor..

I'm looking at this now, it seems like it could be pretty good.

https://www.thingiverse.com/thing:2111631

Stabby probe arms like this are the perfect solution to probing absolutely anything, but the biggest problem always is the arm material.. i tried wires, CNC mill coolant nozzles and other things, but they simply never hold position well enough for the tiny probe tip to stay in place

>> No.1953718

>>1953716
That's some interesting item you linked.

>> No.1953725

>>1953716
why not just scrape the solder mask off a more accessible point? If you're precise, you won't uncover neighboring traces. Even a monkey could safely probe it.

>> No.1953727

>>1953725
>scrape solder mask
jesus christ...
this would be acceptable if you are trying to fix a walkie talkie on a deserted island to call for help, but no EE with any class would ever do something so abhorrent

>> No.1953729
File: 347 KB, 1541x1920, soy based chickem.jpg [View same] [iqdb] [saucenao] [google]
1953729

>>1953727

>> No.1953730
File: 108 KB, 1080x1080, 10pcs-High-Efficiency-Test-Hook-Clip-Logic-font-b-Analyzer-b-font-Cable-Gripper-Probe-Test.jpg.jpg [View same] [iqdb] [saucenao] [google]
1953730

>>1953716
Keysight sells something similar to that for like $600
Have you tried these style clips? They work pretty good for around SOIC size and bigger. There are also even smaller ones that look like just a straight stick but have little grabbers that come out when you push the back, haven't had a chance to try those much.

>> No.1953734

>>1953730
yeah, i actually have these ordered from chinks, but it always takes over a month for ali orders to arrive, so i will find out in about two weeks how good they are

>> No.1953738
File: 828 KB, 2160x1080, e2c4cd69-8061-49e6-bb52-d08688e821a6..jpg [View same] [iqdb] [saucenao] [google]
1953738

got this Epson (tw3200) projector for free and I want to try and fix it.

it seems the colours have shifted slightly, as well as the focusing for those colours (the crosshair in the image)

anyone know where I should start looking?
I'm going to open it up tomorrow and see what I can do

>> No.1953739

>>1953730
Sigrok/Pulseview's site has a good comparison for different probes:
https://sigrok.org/wiki/Probe_comparison

>> No.1953740
File: 36 KB, 1919x936, mppt.png [View same] [iqdb] [saucenao] [google]
1953740

hmmmmmm... mppt

>> No.1953755

>>1953739
thats great, thanks

>> No.1953756 [DELETED] 

>>1953676
>What sort of probe do i use that is small enough to allow attachment to legs on soldered dip chips or smd chips?

i solder 1/8-watt resistors to the pins, pointing at the sky, and use small hook clips to grab the resistor.

>> No.1953759

>>1953676
>What sort of probe do i use that is small enough to allow attachment to legs on soldered dip chips or smd chips?

i solder 1/8-watt resistors to the pins, pointing at the sky, and use small hook clips to grab the resistor.
if i feel i might need that test point in the future, i'll cut the leg and leave like 3mm of it sticking in the air.

>> No.1953764

>>1953738
Maybe a new bulb? Mine started acting fucky and that fixed it. Pro tip (maybe idunno it's been years) bulbs are expensive but for some reason sears had them online for almost nothing.

>> No.1953766
File: 2.04 MB, 3456x4608, fd72ec07-1d40-4abd-9a92-ba9963bbdf7c..jpg [View same] [iqdb] [saucenao] [google]
1953766

>>1953764
I think its going to be this cracked lens that I can't find a part number for

feels bad men

>> No.1953771

>>1953766
That's a compound Fresnel lens or "fly's eye" lens.

>> No.1953798
File: 1.10 MB, 1080x2160, 6898aa4f-98da-4838-87d4-7a3bdf3b83c4..jpg [View same] [iqdb] [saucenao] [google]
1953798

>>1953771
cool, I didn't know what they were called.

anyway after pulling that out, the image looks exactly the same, maybe it's the polariser?

this is my first attempt at projector repair so I'm not really sure

>> No.1953805

>>1953798
Seems like the projector was dropped and some optical elements fell out of alignment. It also probably needs to be cleaned.

https://www.projectorjunkies.com/epson-tw3000-%CE%B5%CE%BA-%CF%84%CF%89%CE%BD-%CE%AD%CF%83%CF%89-2/

>> No.1953809

>>1953805
yeah that's what I was thinking, I will have another look at it tomorrow and see what happens.

it was free anyway, so I won't be too sad if it doesn't work, but it would be comfy if I get it working

>> No.1953812

>>1953809
Just pick a color and only use that one color and disable the other two

Perfect for stuff like spread sheets or tetris

>> No.1953820
File: 85 KB, 385x305, dip.png [View same] [iqdb] [saucenao] [google]
1953820

if you need to pull the chip out alot, are these expensive round sockets better than those regular cheap ones with flat legs?

>> No.1953826
File: 21 KB, 598x574, booba2,.jpg [View same] [iqdb] [saucenao] [google]
1953826

>>1953820
In my personal experience the round ones last longer.
>>1953805
>projector junkies

>> No.1953829

>>1953820
>are these expensive round sockets better
Yes. Machined sockets are better.

>> No.1953863
File: 623 KB, 647x548, Screenshot_2020-11-17_10-16-05.png [View same] [iqdb] [saucenao] [google]
1953863

>>1953820
>if you need to pull the chip out alot,
Consider a zif socket.

>> No.1953864
File: 19 KB, 800x800, 16-Pin IC ZIF Test Socket.jpg [View same] [iqdb] [saucenao] [google]
1953864

>>1953820

if by ''a lot'' you means dozens of times over its lifetime, then these are worth the extra 2 bucks.

>> No.1953879

>pretty basic 4 layer board
>pcbway $50
>jlcpcb $7
Is pcbway just not geared for one-offs or does jlcpcb cut corners somehow? This is my first time ordering one.

>> No.1953889

>>1953879
>4 layer
>basic
oi m8

>> No.1953892

>>1953879
A big reason JLPCB can be so cheap is that they put different PCBs from different customers on one big single panel

>> No.1953896

>>1953892
>being this retarded
Every similar service does same shit
Nobody is going to prepare all the fucking tooling just for you

JLCPCB literally is just cheaper and not sure where they cut corners

>> No.1953898

>>1953896
no suicide nets in the factory

>> No.1953906

>>1953896
Look at how much "Panel by customer" option increases the price in JLPCB, while in PCBWay it stays the same.
JLPCB seems to able to reuse the same setup while PCBWay always seems to need extra setup.

>> No.1953911
File: 893 KB, 4160x3120, IMG_20201117_165624.jpg [View same] [iqdb] [saucenao] [google]
1953911

>here's your soldering iron tip bro
Wtf happened? I've been soldering with this cheap 5$ china iron quite long using leaded solder and everything was fine. I wanted to try lead free soldering and it looks like it ate into the iron and maybe even fused with the solder. I've soldered before at "350°C" (doubt it's accurate) with leaded solder and this did not happen.

This is supposed to be a chisel tip.

>> No.1953917

>>1953911
Call the customer service hot line and tell them. In most cases they will give you a code so you can ship it back to be refurbished for free.

>> No.1953923

>>1953906
check to allow on x-out when calculating panel on PCBWay (also update panel size)
Overall JLCPCB is still using uighur labour

>> No.1953929

>>1953917
>call customer support
what a stellar idea!

>RING RING.... RING RING
>TIAN PONG CHONG ONG CHONG?
>uhhh....h-hello, is this customer support?
>PONG CHONG TIONG ANG???
>um....

>> No.1953937

>>1953911
The plating wore off and your iron tip started dissolving into the solder. It happens. Swap to a spare tip and order another. Tips are consumable items, and you get what you pay for.

>> No.1953948

>>1953911
tips are iron covered but have a copper interior, it will dissolve into the solder pretty quickly once breached.

>> No.1953953

Very stupid question. I need to replace capacitors on some old electronics. A friend tells me there are "permanent" caps so I won't have to yet again replace them some 20-30 years later. Trying to research that, I'm told Ceramic caps are basically what he meant but aren't "eternal" and are superior to the old leaky venty electrolitic ones that I'm replacing.

What's true and what's not? What do I get?

>> No.1953959

>>1953953
Electrolytics have a far higher capacitance rating than ceramics and are used for different purposes, and nothing is permanent.

>> No.1953960

>>1953953
film caps, ceramic caps, polymer aluminum caps
Depending on what type of device are you talking about, capacitance/voltage range, operating temperatures. They might be suited or not.

>> No.1953969

>>1953959
>>1953960
Thanks... Is it that I have options for each device's capacitors (these are old computers from the 80s) or is it likely in almost all cases I'll have to replace them like for like?

>> No.1953974

>>1953953
>20-30 years later.
That is hardly an issue, you will be pooping in a diaper by then, so replacing leaky caps will be the least of your worries

>> No.1953975
File: 100 KB, 1281x666, Schematic.png [View same] [iqdb] [saucenao] [google]
1953975

sup
I'm following this guide to make a packet radio interface, except the fag who made this schematic didn't put the units on C4 and C5.
This might be obvious to some what it should be, but I don't make assumptions.

He says the following in the guide:
The amplified signal passes through D1 and charges C4. D1 prevents the cap from discharging back through the transistor, so you get an accumulated voltage up to around 4.4V DC during a full volume signal. This gives an excellent range on the input to the Arduino, and allows for sensitivity adjustment to be done in software.
R6 bleeds the capacitor down when the audio stops.
Q2 takes the PTT signal from the Arduino and grounds the radios PTT line to key it into transmit.

BUILD NOTES:

C1 and C2, I used non polarized caps since I had them on hand, but regular electrolytics will work fine. If you're concerned about isolation, you could use two 600:600 Ohm transformers instead.
C4 is not critical. If you don't have a .22, put two .1 in parallel
R7 and R8 don't have to be 10K, can be anything from 10K up, don't have to match either.
C3 can be anything from 1uf up to 10uf

TLDR: what is a .22 or .001 cap? can I assume µFor is it usually something else? (they're all ceramic caps)

>> No.1953982

>>1953969
Replace them like for like.
>>1953975
uF = micro Farad
1 uF = 1000 nF = 1000000 pF

>> No.1953983

>>1953974
I'm not that fucking old

>> No.1953997

>>1953982

That doesn't answer my question though. What does the author mean by .22 or .001, especially since the other caps have a rating and unit listed

>> No.1953998

>>1953975
C5 is definitely uF, C4 could be either uF or nF, it might be obvious based on the time constant C4*R6, which is then either 0.22 seconds or 0.22 milliseconds

>> No.1954001

>>1953997
0.22uF and 0.001uF
Since you mentioned that all caps are ceramic

>> No.1954002

>>1953975
>what is a .22 or .001 cap?

if there's no unit, assume uF.

>> No.1954029

>>1954001
Excellent thank you

>> No.1954034

>>1954002
>>1954029
also keep in mind this
uF is in general most used range
If it's anything else - then generally there is a reason to put it there and extra effort will be spent to write other letters

>> No.1954046

>>1952093
>Freeware
>Doesn't run on Linux
LMAO Linux is a joke

>> No.1954047

>>1954046
It runs fine, but using wine.

>> No.1954049

>>1954046
It's basically adware for LT.

>>1954047
Learn to use KiCAD's integrated simulator instead.

>> No.1954071

>>1954049
>>1954046
low iq posting hours

>> No.1954095

How do I test an Op amp? I am having kind of a brainfart.
I have an OP27, I hook up supply voltage and Signal+ On the op amp to the + side of my function generator and – to signal generator ground? Or do I need to generate a 180 degree shifted signal on Channel 2 and connect that because I think a differential signal is expected at the input.

Sorry if that’s a dumb question

>> No.1954122
File: 54 KB, 440x826, 1605641254310.jpg [View same] [iqdb] [saucenao] [google]
1954122

>>1953917
They would tell me to fuck off in Chinese. I don't think anyone can expect quality for this little price and they know that you know you're getting what you pay for. Besides, I bought this ages ago.
>>1953937
>>1953948
I switched to a new tip when I tried soldering lead free and the first thing I had to solder was wires and I'm thinking the juices of the insulation may have mixed in with the flux and possibly got on the tip, disintegrating the thin coating.
I tried a different tip and soldered a capacitor without the thing disintegrating so that may have been it.

Soldering lead free seems fine, not as bad as all the people complaining about it online. It's not as nice as leaded but it's not horrible either.

Thanks all.

>> No.1954127

>>1954095
test in what regard? there are several parameters in a opamp. If you want to know if it is working as it should I usually do two sanity checks
>open loop comparator
>ground +
>put vcc in -
>Vout should be low
>ground -
>put vcc in +
>signal should he high

and
>unity gain buffer
>>1954122
good soldering fren

>> No.1954200

>>1954095
Op-amps are meant to be used with negative feedback. They can be reasonably well estimated with the following equation:
Vout = (V+ - V-)*G, where G is a very large number. Naturally, the output voltage cannot be above or below the power supply rails Vcc and Vee. Using them without feedback (e.g. with signal into V+ and GND into V-) will result in the thing acting like a kinda shitty comparator.

To use it practically you need negative feedback, which is almost always done by feeding some of the output signal back into the inverting input V-. They do not need a differential signal. Common circuits that do this are the inverting amplifier, non-inverting amplifier, and unity-gain buffer. Then there’s a few interesting ones like an all-pass filter. Test a buffer and an inverting amp, that’s almost certainly sufficient to know if it works.

See the Wikipedia page “list of op amp circuits” for more.

>> No.1954219

>>1951939
They're usually connected to earth too so careful if you're probing anything with a floating ground

>> No.1954246

>>1953195
Can't find much but I think the manufacturer is thomson which became STMicro in the 90s

>> No.1954336
File: 5 KB, 1073x316, aaaaa.png [View same] [iqdb] [saucenao] [google]
1954336

Is there a way to make the outputs not an LED but a pin that shows me 1 or 0 rather than light on or light off? I've tried connecting a pin to the output(s) but it doesn't seem to do much. This is on logisim btw

>> No.1954341

>>1954246
Thanks! I didn’t recognise the manufacturer logo, but even knowing it I’m still not getting any search results.

On a side note, is there a list of IC logos somewhere? Would be really handy.

>> No.1954350

>>1954200
>Op-amps are meant to be used with negative feedback
Only true for linear circuits. There's plenty of op-amp based RC oscillators which have to be used with positive feedback by definition.

>> No.1954373

>>1954350
Those are Schmitt-trigger oscillators, that are better made with a comparator due to the higher slew rate. You can use an op-amp, but because the two inputs are barely ever equal, the higher linearity of the IC is wasted. If you need a part to function with positive or no feedback, it’s almost always better to use a dedicated comparator. Some op-amps even have clamping diodes to prevent the inputs from being too different to one another. Op-amps should only be used for linear circuits.

>inb4 but i used my 741 for everything and it worked fine

>> No.1954462
File: 178 KB, 900x463, timing trace.png [View same] [iqdb] [saucenao] [google]
1954462

Can someone explain to me why the frequency from Q to clock of the 2nd D-flipflop is half of the frequency (5Hz) coming into the first flipflop (10Hz)? The professor said that we can see this in example doing a timing trace but I don't really understand what he wants us to show on it to be able to understand it.

>> No.1954476

>>1954462
The point is, by wiring a D-FF’s inverted output up to its input, you can get a T-FF. T-FFs divide a clock by 2. Chain a bunch together and you can get a 1Hz signal for a clock from a 32768Hz crystal oscillator, among other uses.

>> No.1954480

>>1954462
>>1954476
If you’re wondering why it actually works on the other hand, it’s pretty basic. A D-FF operates by latching its output to be equal to its input at the instant of the clock transition. Say Q = low, so D = high due to the inversion. On the pulse, Q will turn from low to high, but one propagation delay after the clock pulse. Hence at the moment of the clock pulse D is still high and Q is still low, but shortly afterwards Q is high and hence D is low again.

>> No.1954481

>>1954462
Q toggles on a clock edge. Toggles again on the next clock edge.
So it takes 2 full cycles on the clock to get one full cycle on Q

>> No.1954503

>>1954476
>>1954481
>>1954480
>The point is, by wiring a D-FF’s inverted output up to its input, you can get a T-FF. T-FFs divide a clock by 2. Chain a bunch together and you can get a 1Hz signal for a clock from a 32768Hz crystal oscillator, among other uses.
>Q toggles on a clock edge. Toggles again on the next clock edge.
Oh, we didn't go over the T-FF yet or even what toggle means but that sounds really useful rather than chaining up multiple D-FF. I will definitely write this down so that when I reach T-FF I'll be able to read your explanations again.
>On the pulse, Q will turn from low to high, but one propagation delay after the clock pulse. Hence at the moment of the clock pulse D is still high and Q is still low, but shortly afterward Q is high and hence D is low again.
>So it takes 2 full cycles on the clock to get one full cycle on Q
Okay so if I'm getting what you're saying, the process goes:
(1) assuming that D starts low, then Q will be low, the inversion of Q will then be high
(2) Since the inversion of Q is connected to D, the next D input will be high so that the output Q will be high. The inversion for this Q will now be low. Because the inversion of Q is now low, D will be low.
Based on this, it will take an extra step/clock cycle, (2), to get a full cycle on Q where it goes from low->high->low which is why it becomes half the frequency as it takes a longer time to get a full cycle?
Thank you three so much for taking the time to answer!

>> No.1954543
File: 15 KB, 308x218, 71xPUsEOFDL._AC_UY218_.jpg [View same] [iqdb] [saucenao] [google]
1954543

Looking for a silicon soldering mat. What do you all use, if anything, and if you do what are some handy features i should get?
Thanks

>> No.1954639

>>1954503
Yes that's right. Though it's less "one extra clock cycle" as it is "twice the number of clock cycles". Same result in this case though.
You probably haven't covered shift registers and such, but you can do the same thing with more than one D-FF chained together like Q>D Q>D Q>D, with the last FF's inverted Q hooked up to the first FF's D. Connecting all the clocks together. Here each clock cycle the signal passes from one FF to the next, until it loops back to the beginning with an inversion. This is called a twisted ring counter, and will divide the clock by 2N. Say you have three of them, you'll end up with f/6.
Once you learn shift registers it will make more sense. Shift registers are fun.

You can test all of this in Logisim (perhaps try the more recent evolution fork, the original one stopped development a decade ago and doesn't necessarily run on new OSs). But note that this form of cascaded synchronous logic relies on propagation delays to work properly, and by default I think logisim doesn't include those. Maybe. I was doing logic testing with LTSpice, which has that issue.

>> No.1954641

Ok, consider an amplifier topology that only uses JFETs and resistors and capacitors, no MOSFETs or BJTs or diodes at all. Only requirements are that gain is like 10 or more and that the output impedance is at least 10 times lower than the input impedance. Would it be possible to construct such a circuit such that inverting the power rails would have no effect on operation?

Using MOSFETs wouldn't work because of the body diode, and BJTs wouldn't turn on and diode biasing and such wouldn't work at all. JFETs also have bias requirements due to being a PN junction, but I'm hoping that it's relatively easy to solve that issue.

>> No.1954642

>>1954462
Draw a timeline in a paper. Remember the flip-flop only changes its held/output value on rising edge.
It should be obvious when you draw it.
Also, if you don't have a FPGA yet, grab iCESugar and nanoDLA from aliexpress.

>> No.1954643
File: 655 KB, 715x1200, Electronic component manufacturer logos 2015-01-06_002.png [View same] [iqdb] [saucenao] [google]
1954643

>>1954341
>is there a list of IC logos somewhere?

dont think so.

>> No.1954661
File: 27 KB, 1479x819, nicewave.png [View same] [iqdb] [saucenao] [google]
1954661

you guys like my wave? quite difficult to get it to work at least for me. this off and onboard dac connected to an onboard opamp. the whole thing loads samples from an eeprom by spi saves it on a buffer and plays the buffer. i can program the eeprom with this little C program on wrote for a raspi that used syscalls. just needs to be a raw file i create from audacity. sounds like complete trash and the way i did after it gets to the end of the buffer there is some delay in the samples so I need to figure out a better way to do it.

>> No.1954664

>>1954661
>nice wave
I see your wave goes from 0 to 1, so it kinda looks more like a DC pulse.

>> No.1954668

>>1954664
It goes 0 to 1 just because I set the internal reference shit for the dac to 1v I can set it to anything I want.

>> No.1954671

>>1954668
The expectation is for the centre to be at 0, not 0.5.
This can easily be why it sounds like shit.

>> No.1954674

>>1954671
I can just make the reference of the speaker via a pwm to .5 volts if I cared enough. I’m working on making it when it gets to the end of the buffer it doesn’t have a small delay in the samples right now.

>> No.1954677

>>1954674
0 to 0.5 is still not centered in 0.
That's centered in 0.25.

>> No.1954678

>>1954661
That's not many total samples, can't you increase the temporal resolution? Surely there's enough room in the EEPROM for more samples. The voltage resolution looks fine for now though.

>>1954664
>>1954671
He can just add a high-pass filter before his amp, or a capacitor in series with his speaker.

>>1954674
If that delay is significant with respect to the duration of each sample, then you're probably doing something wrong. How fast is your SPI running? How are you timing the circuit?

>> No.1954679

>>1954677
No I mean connect the black wire of the speaker ie the reference to a pwm output with an average voltage of .5 volts.

>> No.1954680

>>1954679
Sure, it'd work but... horrible :/

>> No.1954683

>>1954679
>>1954680
With enough low-pass filtering and a high enough PWM frequency it would be fine. Though I'd prefer just buffering a voltage divider. Or using a cap like a sensible person.

You could also do differential signalling by sending an inverted audio waveform to the other side of the speaker, that way you get twice the effective voltage.

If it's a piezo speaker it probably won't sound any different at all.

>> No.1954689

>>1954683
might as well use a pwm and have it run at 1mhz or whatever the fastest speed is going to be so no additional hardware required.
>>1954678
right now im abusing interrupts. The spi is running at 1mhz so thats not the issue. the issue is once the spi interrupt fills the buffer it disables itself and then I have dac timer, once the dac gets to the end of the buffer, start up the spi again in the interrupt. the spi start up time is the delay since it has to fuck around with the chip select send the command and address. I need to set a flag or something so it starts up the spi again at some point in the main loop. Maybe once the dac timer gets half way through the buffer and when the dac timer gets to the end half the spi fill up to that point.

>> No.1954698

>>1954671
the DC sound level? what the fuck retard

>> No.1954701

>>1954698
DC is the problem.
Negative voltage is needed for a proper audio signal.

>> No.1954711

>>1954689
>might as well use a pwm and have it run at 1mhz or whatever the fastest speed is going to be so no additional hardware required.
It still requires the filter hardware, which is literally the exact same filter hardware required by just putting a filter before the speaker. Actually the filter before the speaker can be simpler since it doesn't need the resistor.

Also an onboard op-amp will be shit at providing enough current for a speaker, you need a proper output stage. Check the datasheet to see its output impedance and/or maximum output current.

>> No.1954713

>>1954701
No it does not.
>>1954689
it sounds like shit because it is unfiltered. You have the main sine wave component and a big mess of high frequencies mirrored around the sample frequency and its harmonics. You need to filter them out. If you are messing with DAC/ADCs you should learn about anti-aliasing filters for sampling and "reconstruction" filters for the output (it has a billion names so it does not really matter what they are called),

>> No.1954720

>>1954701
It isn't required, but it helps. And inductive speaker running with a DC bias added to its AC audio signal will sound distorted because the diaphragm of the driver will have an off-centre equilibrium position. Technically, you could do this on purpose for a certain sound if you'd characterised how your speakers react, or equivalently use that characterisation to send your speaker a modified audio signal such that the DC bias makes it sound normal again. I wouldn't exactly advise doing either of those though, as the DC current will heat the driver up more than necessary.
If your signal level is low enough (like sending a 0.35Vrms signal to a large speaker) then the distortion from the DC will be minimal.

As I said earlier, some technologies like piezo (or other capacitive/eletrostatic) speakers will be less effected by this.

>> No.1954725

>>1954720
You're right in that a piezo will not care much for this.
The assumption was that, if there is a piezo, is it just there for testing.

>> No.1954740

if a power supply that coverts 230V AC to 5V DC silently buzzes during operation is that a problem?
i don't mind the sound but i don't wait it to melt or something
The current draw is 30ma and the source can supply up to 500ma so it's not overloaded or anything

>> No.1954742

>>1954740
how it can buzz and be silent? And no, it is ok if it buzzes. The transformers inside it and capacitors vibrate due to several things and can cause some amount of noise.

>> No.1954908

>>1954740
It depends on the topology of the power supply. Buzzing while only pulling low current like you said would concern me.
Is it an smps, or an old school bulky step-down transformer?
I've seen transformers with dual primary coils so you can connect them in series for 240Vac or parallel for 120Vac. However, if the primaries are connected opposite of each other, the magnetic fields will fight each other because they are basically 180* phase shifted from each other. That can cause humming, and is a bad thing.

>> No.1955000
File: 23 KB, 800x480, SDS00020.png [View same] [iqdb] [saucenao] [google]
1955000

I need to test over many runs that channels 1 and 2 of pic related remain without any time offset. This used to be a frequent issue that I now seem to have finally fixed. The thing is, my scope only allows to use a mask on one channel at a time (are there even scopes that don't?) and the signals are not in-phase so math function doesn't help me either. How would I go about this? I don't wanna sit here all night watching the damn screen myself.

>> No.1955004

>>1955000
>not in phase
>no time delay
what

>> No.1955014

>>1955004
Well you can see it's an RF transmission. The carriers are not in phase. The messages, however, need to arrive simultaneously, save for a nanosecond or two.
>yellow go big when purple go big

>> No.1955015

>>1955000
>>1955014
invest in dual time-base scope
If time difference is identical in all cases - might be possible to adjust trigger delay

>> No.1955023

>>1955014
well in that case I would just run the cases, save the date to a usb drive and write a script to calculate delta_t

>> No.1955025

>>1955000
Capture traces into labview, do envelope detection, get delay between envelopes with cross correlation.

>> No.1955044
File: 17 KB, 640x640, SSW-101-x1-G-S.jpg [View same] [iqdb] [saucenao] [google]
1955044

why the heck these 1pin female headers seem such unobtanium on fleabay? I don't want to bother making order from digikey to just get handful of these, but I can't seem to find them anywhere sane? Even male versions seem pretty rare?

Also why there are round and square pinned/holed headers??? Idiot me bought some fancy ones with round holes and of course now square pins don't actually fit into those so that was a waste

>> No.1955046

>>1955015
>>1955023
>>1955025
Thanks for suggestions, it really does look to be a task for the PC. Gonna go find a USB-B cable.

>> No.1955048

>>1955044
Dupont connectors?

>> No.1955058

>>1955048
for pcb mounting, not for wires

>> No.1955072

>>1955058
>>1955044
you buy them in bars and break them as needed.

>> No.1955076

>>1955072
The female ones don't break very well at all. Best to buy them at the correct size like he's trying to find.

Not that I particularly understand why he isn't just grouping all connectors on his board into a single 2x3 or whatever, as it makes it more difficult to insert something incorrectly.

>> No.1955115

>>1954543
I got this one recently: https://www.amazon.com/gp/product/B07MB2DFG9
I like the little scooped out bits for corraling components.

>> No.1955133

>>1955072
>>1955076
yeah, trying to break out single female receptacles is pain in the ass

it would have been just convenient in this case to have individual sockets and I thought they would be a dime in a dozen on ebay, never thought that it would be an issue. but I guess not.

>> No.1955134
File: 38 KB, 727x598, timing trace 2.png [View same] [iqdb] [saucenao] [google]
1955134

>>1954642
Is it something like this? I wanted to do a timing trace/timeline but I didn't know what I wanted to check the timing with, but it seems like they want the output frequency which is Qnot? Or is it Q instead?
>>1954639
Okay, I'll go test it out on logisim, thank you anon! I didn't know they have a newer version, that's cool...
>But note that this form of cascaded synchronous logic relies on propagation delays to work properly, and by default I think logisim doesn't include those. Maybe. I was doing logic testing with LTSpice, which has that issue.
Yeah I think it was called ripple counter or something...

>> No.1955182

>>1955134
>but it seems like they want the output frequency which is Qnot? Or is it Q instead?
Anon, the frequency of Q and Qnot is obv the same. They both flip at the same time, as Qnot is Q inverted.

>> No.1955183

>>1955182
Sorry, bro..... I'm starting to get dumber...

>> No.1955242

Have had a hand tremor all my life
How do I get good at soldering

>> No.1955278

>>1955242
Have really good convenient work holding. Have the workpiece in a swivelling vice with plenty of rigidity, and something equally rigid nearby to rest your hand against. Using integrated clamps or spring-loaded pins to hold components against the board instead of using tweezers should also help you. Naturally, being able to do your work with hot-air (or a reflow oven) will reduce your reliance on a steady hand. Some fab-houses do part soldering for you, if you’re that desperate.

Also some amount of physical training should help, the less stress your muscles are feeling then the less they’ll be shaking.

>> No.1955284

>>1955278
I'll look into those
Thanks anon

>> No.1955338
File: 2.51 MB, 4160x3120, what are these.jpg [View same] [iqdb] [saucenao] [google]
1955338

What are these? I thought they were diodes like the others, but these have a voltage drop across them in both directions.
0.017V across the marked one on the bottom left and 0.023V across the marked one on the top right.
Backstory is that I'm trying to figure out what's wrong with this pcb. I replaced a melted/bubbly cap and a fuse and I'm now trying to figure out what else could be defect.

>> No.1955352

>>1955338
Did you remove them from the circuit to measure, or are you poking at it in the circuit like a dingus? 0.023 is close enough to 0v for me, but whatever voltage they have would be dependent on the rest of the circuit.
look like diodes to me by the package and the XDn labels. Could be some special diode like a zener, but I don't know. The color bands should mean something.

>> No.1955356

>>1955338
They almost certainly are diodes, but by measuring them in-circuit you’re exposing the surrounding components to the same current, and it might be finding a path of least resistance through an unpowered IC. Or a dead IC. You’d be much better off probing for voltages wrt ground while it’s turned on. For more nuanced troubleshooting, a scope or logic analyser would be required.
Looking via thermal IR (or feeling) for hotspots is also a good method of troubleshooting.

>> No.1955399

>>1955242
anchor with your pinky, use clamps, exhale as you dive in. only solder on good days.

>> No.1955516
File: 1.77 MB, 1920x1080, AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA.png [View same] [iqdb] [saucenao] [google]
1955516

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

>> No.1955518

>>1955516
>software
you have only yourself to blame

>> No.1955520

>>1955518
I didn't realize until after college that hardware is more rewarding
I'm trying to make up for past mistakes through embedded development

>> No.1955527

>>1955518
>>1955520
Explain, I plan on taking an Electrical/Electronics Certificate program in college and want to know what you two mean by this. Should I focus on the hardware? the aspect of the study rather than the software aspect of it?

>> No.1955534
File: 163 KB, 1080x1687, TS100Tips.jpg [View same] [iqdb] [saucenao] [google]
1955534

>>1951716
Gentlemen, I was one of the guys who bought the TS100 chink meme iron. Anyone familiar with this tool? I have 2 questions:
>Tip
I bought the BC2 tip which is a chisel, it feels a bit big on some of the TH pads on this 70s stereo reciever im working on. Trying to find a better option but unsure. I think the C1 tip is a smaller BC2 but I may be wrong. Any of the other tips a better choice for work like this? I read chisel is the best but maybe i'm wrong?
>Ungrounded tip
Found out after buying it that the tip isnt grounded, supposedly risks damaging delicate things with ESD. There is a ground screw on iron to tether it to ground with a wire. This a big deal? Working with TH transistors like you'd find in old radios etc.

pic of pads im working with. Kinda shit pic, not much for scale. Will try to get pad measurements incase pic is useless
https://images2.imgbox.com/6f/d9/CJm4oqQn_o.jpg

>inb4 shouldve saved 100bux and bought pace ads200, rip

>> No.1955535

>>1955516
Turns out I'm a dumbass and wired my chip wrong
Fucking hell

>> No.1955541

>>1955527
gtfo back to /sci/ unicuck, we actually make shit here

>> No.1955543

>>1955541
>gtfo back
>go the fuck out back
Retard detected lol

>> No.1955549

>>1955543
Gtfo is Get The Fuck Out

>> No.1955552

>>1955549
yup

>> No.1955553

>>1955534
I think the Bs are bevel tips, not chisel tips. The Ds or Ks are more chiselly to me. The D24 or similar makes for a pretty damn good tip, not too fond of bevel tips myself.

Personally I have a D12 and a K for my T12 station (similar to TS100 tips but not identical).

>> No.1955561

>>1951716
any 1 ever found a working sub 100hrz oscilator circuit that doesnt need a 55 timer? shits annoying all the ones i use can only go as low as 10 khz then crap out and refuse

i undid a damn steel core for this. it was like a hour of my fucking life and the transformer cant give it back its in pieces right now. all but 1 lamination are fine. whole thing was like a god damn miracle and now i find that not of the oscillators i use do 60 hrz. boost to 15kv bleed sure boost to 300v with usable current nope

>> No.1955568

>>1955561
an arduino

>> No.1955576

>>1955561
any oscilator that does not use crystals?

>> No.1955577
File: 118 KB, 1920x1080, Screenshot_2020-11-19-01-53-51_compress76.jpg [View same] [iqdb] [saucenao] [google]
1955577

>>1955553
Thanks for pointing that out. Youre right, my BC2 is a bevel tip, wrongly thought it was chisel. Is it correct that a properly sized chisel should fit across the pad's width and not hang over the pad?If that is the case, it looks like D24 is similar width to BC2.
Thanks for mentioning T12, I searched and was reminded that hakko t12&t15 tips fit. Theyre alot longer but it looks like there is a printable adapter to couple with a hakko rubber grip to make it less shitty. Probably going to try and track down an adapter piece, grip and hakko tip to buy.

>> No.1955586
File: 13 KB, 332x404, circuit-20201108-2159.png [View same] [iqdb] [saucenao] [google]
1955586

>>1955576
you say that but pic related if i scale up the caps it craps out. idk y

>> No.1955590

>>1955586
pic related can be fet or npn as well but if the caps go above 1uf it doesnt seem to do anything

>> No.1955592

>>1955586
Do not connect the oscilator to whatever you are testing. Make the oscilator circuit, put a unity buffer in the output and then connect that to whatever.

>> No.1955593

>>1955592
And if you are in falstad, go to the circuits tab, opamp, relaxation oscilator. I use that for audio mixing sometimes for 0-10hz and it works fine

>> No.1955600

>>1955592
>>1955593
thank you for the responses i will try both of those

>> No.1955602

>>1955586
supposedly S&H stages with mica, or poly-something caps can last weeks. You might need to buffer your states with those

>> No.1955633

>>1952633
can you precompute,outside of the interrupt, part of this calculatons?
>>1952600
> how to recode that shit so it somehow triggers during that data cycle as well.
not sure what kind of micro you are using, does it have nested interrupts? if yes, just give the zero crossing interrupt an higher priority so its ISR can be called even while the data cycle ISR is going on

>> No.1955664

>>1955352
>Did you remove them from the circuit to measure, or are you poking at it in the circuit like a dingus?
>>1955356
>They almost certainly are diodes, but by measuring them in-circuit you’re exposing the surrounding components to the same current, and it might be finding a path of least resistance through an unpowered IC.
Ok I was poking at them in the circuit like a dingus, makes sense. I'll try measuring wrt ground or desoldering the larger components and testing them individually. Thanks.

>> No.1955704

I'm going to use one of those cheap $5 ebay BMS boards for a project. I'm not exactly sure how to use it. Can I connect cells in parallel or does the BMS only allow one cell per "channel"? can they discharge and also charge at the same time?

>> No.1955750

can someone explain to a retard what JFETs are for/when you would use one over a BJT/MOSFET?

like, BJT:
>current applied to base controls current through collector to emitter
>comparatively higher power draw due to the necessary gate current
>faster switching speeds
>not optimal for VLSI due to size and heat

MOSFET:
>voltage applied to gate controls current through drain and source
>significantly reduced power draw and significantly increased signal sensitivity since the gate doesn't allow current flow
>gate is essentially just a capacitor that can turn on/off the drain/source path
>significantly reduced switching speed capability due to the need to undo capacitive hysteresis before turn-off/turn-on can be achieved

JFET
>is a transistor
>????
>profit

>> No.1955776

>>1955750
>when you would use one over a BJT/MOSFET?
Well, you probably wouldn't. They're sort of an obscure specialty component from a bygone era.
I think nowadays they are mainly used for audio signal processing. They don't have a body diode so can be used easier for AC signals like audio filtering. And the logic is inverted (normally open) unlike BJTs and most MOSFETs, though you can buy depletion mode MOSFETs.

>> No.1955781

>>1955750
infinite input impedance for weakass signals, mostly.
also switching and variable resistors. Check out the MXR Phase 90 pedal

>> No.1955810 [DELETED] 

>>1955781
>variable resistors

yep, widely used to make automatic gain control circuits (AGCs) because of this characteristic. also, audio on/off switches because of same.

>> No.1955813 [DELETED] 
File: 8 KB, 261x193, agc circuit.jpg [View same] [iqdb] [saucenao] [google]
1955813

>>1955781
>variable resistors

yep, widely used to make automatic gain control circuits (AGCs) because of this characteristic.

>> No.1955815
File: 14 KB, 500x370, mic-preamp-agc.gif [View same] [iqdb] [saucenao] [google]
1955815

>>1955781
>variable resistors

yep, widely used to make automatic gain control circuits (AGCs) because of this characteristic.

>> No.1955892

Hi, retard here. I wanted to know if I can use an amplifier power supply as a bench power supply. Basically putting these two together, and expect to get 48v 10a adjustable (or at least close to that)

https://www.aliexpress.com/item/1005001407304522.html
https://www.aliexpress.com/item/4001231282413.html

Or should I just invest a bit more and buy a generic switching power supply of 480w?

>> No.1955895

>>1955892
the power limit for retards is 20W. You need to keep bellow that until you can answer your own question

>> No.1955900

>>1955895
Are you afraid I might get hurt?, how considerate UwU

>> No.1955955

Is there anything wrong with putting a jumper across the travelers on a 3 way switch? I have a circuit that needs to stay on regardless of switch state and I can't think of any reason this is bad or dangerous. Noted of course that there is a brief (~100ms) interruption in power when a switch IS changed due to neither traveler being in circuit, but this is just to prevent anyone from accidentally turning off the outlet for our deep freezer.

>> No.1956040

>>1955561
What do you need it for? What kind of waveform do you want? For a simple square wave, then a comparator-based relaxation oscillator is my go-to, it's marginally better than an NE555. For a triangle wave, I'll use one schmitt-trigger comparator and one integrating op-amp. Sawtooth waves are the same but with a diode across the integrator's resistor. For a sine wave, I think the emitter-coupled-oscillator topology was a decent one, not that I've ever tried it. Shaping a triangle wave with diodes and/or filters is also a pretty reasonable way to go.

Anything with inductors or especially transformers should be reserved for when you need the oscillator itself to put out significant current. Like for radio circuits or resonant power circuits. If only because op-amp virtual inductors work better than real ones.

If you want to go full shitlord, there's a circuit that uses a single reverse-biased NPN transistor to form a square wave.

>>1955664
Desoldering is going to be a massive pain on any SMD board. Unless you have hot air or some sort of desoldering shim set, don't bother trying. Even if you do find a faulty part it's going to be a pain to replace it, though destructive removal of ICs is more feasible. Repair of SMD boards is a niche area because of this.

>>1955892
Looks ok to me, but as a switching supply it will be somewhat noisy compared to a trafo+linear supply, and it doesn't have split/multiple rails. Probably no push-pull either. But if that's all fine by you then go for it.
IIRC there are more tried and tested variable power supply front-ends though, EEVblog reviewed one.

>> No.1956056

>>1956040
I already have a linear one, only gives 2a tho, and I can't give it more without having a massive trafo, that's why i wanted a switching one, the noise doesn't bother me that much.
>power supply front-ends
Yeah, can't buy those. They'd rape me with the taxes and shipping, it's better to buy it for parts. Thanks m8.

>> No.1956060

>>1956056
>Yeah, can't buy those
I'm literally talking about the sort of thing you're connecting to the amp PSU.

>> No.1956063

>>1956060
Oh sorry, I misunderstood you. I'll look into them.

>> No.1956074

>>1955776
> And the logic is inverted
Not inverted, but offset. There's partial conduction at Vgs=0, which increases as Vgs increases and decreases as Vgs decreases. The fact that they're in the linear region at Vgs=0 makes them useful for circuits which need to operate with very low supply voltages.

>> No.1956088

>>1956074
>Not inverted, but offset
It's both. JFETs are depletion mode devices, so while increasing the gate voltage on an N-ch MOSFET would increase the conductivity, doing so to an N-ch JFET would decrease the conductivity.
I think.

>> No.1956096

If any EE can help me with this consulting problem I'd greatly appreciate it.

I'm a first year eng student. I contracted for a company over the summer and quit in Sept. I made hardware for them and they keep bugging me to do them free favors because the project is failing since they don't have a competent hardware employee anymore. I did a bunch of free stuff for them but they took it apart and broke it.

I made a new device with similar functionality but I had to re-do it the way I inisisted on from the start and it works great. Got it done in Altium from scratch and am about to send the PCB for printing (first time doing this).

I haven't said anything to them about the rebuild. I don't have a contract anymore and nothing that says I can't do other work after I quit in the last one. How can I keep some ownership rights if I tell them? What would you do in my situation??

>> No.1956137
File: 23 KB, 404x294, circuit i have found.jpg [View same] [iqdb] [saucenao] [google]
1956137

>>1956040
its just for a transformer but if its way above 60 hrz it probably wont have enough current. hex inverters would have to be made and add to the total parts on board since i dont have any on hand and i would rather not implement a ring oscillator

worst case scenario i can use the led blinker circuit but the current draw from the transformer drains the flow into the oscillator and then it rebonds and causes current drop across the transformer primary.

i found pic related but i dont think its 60 hrz. some have claimed 1 ma from it with a 9v which is believable ish if you used a small transformer but when i look at it it seems like its atleast 10 khz. i can slow it down some with a 10 uf cap non polarized across the primaries but i dont know if it will be 60 hrz

>> No.1956157

>>1956096
You have to make them believe that they need you. Then you need to know the value of your services and negotiate from twice that amount downward. Don't tell them you have a finished product until they agree to a contract and timetable, then when you're 3 weeks from the deadline ask for more money to complete the project.

>> No.1956163

>>1956137
>its just for a transformer
I'm gonna go out on a limb and assume that you haven't done enough research into inverters and switching power supplies.
Explain what you're going to do with the transformer, so I can explain why A: your circuit can't source enough power for anything meaningful, B: your transistors are going to explode under any load, and C: using a common boost converter is almost certainly a better idea.

>> No.1956173

>>1956157
well I know they are overextended and desperate on this deadline. I led them to believe making the PCB would be easy then I got bogged down in coursework. Not sure I have the charisma to pull off asking for another contract empty handed. They need what I have but I don't want to give it to them for free and they have legal resources. I'd be happy with some of the rights to the project for some re-occuring income while I'm in school.

>> No.1956179

>>1956173
Wait until the last second and squeeze them for as much as you can or walk away with nothing.

>> No.1956190

>>1956179
Okay, I'll just start talking again like I have nothing and see what kind of state they're in. All I can think is they're pissed at me so we'll see how this goes.

>> No.1956230

I just got a rasberry pi zero and I kind of want to do some kind of electronics thing
is it worth doing in your opinion?

>> No.1956255

>>1956230
A raspberry pi is a good way of learning low-level programming for interfacing with peripheral electronics. But so that you don't destroy your raspberry pi by connecting it to the wrong voltage or trying to pull too much current or giving it an inductive spike, I'd advise that you learn some amount of basic electronics theory beforehand. Probably best to mess about with electronic components on a breadboard with a multimeter too, you'll want all that stuff when using your raspi for it anyhow.

>> No.1956320

Redpill me on current transformers. Why are they different from regular ones?

>> No.1956366
File: 369 KB, 800x692, stupid fucking dumbass idiot.png [View same] [iqdb] [saucenao] [google]
1956366

>can't get what should be a dead simple circuit to work
>now your oscilloscope is dying too

>> No.1956380

>>1956366
all I'm trying to do is use the PWM output of an STM32 to drive a BJT, which drives a bunch of N-FETs, which are switching a bunch of solenoids/fuel injectors.

but nothing happens
The STM32 is working just fine, the only thing I could say is weird is that the PWM signal seems to be at max roughly 1-1.5 volts instead of the 3.3 I'd expect
is there something fundamentally wrong with my circuit?

>> No.1956381
File: 134 KB, 1538x692, 2020-11-20-011006_1538x692_scrot.png [View same] [iqdb] [saucenao] [google]
1956381

>>1956380

>> No.1956384

>>1956381
Is stm32 ground the same ground as the solemnoids?

>> No.1956387

>>1956384
it's not actually, I'm powering the STM32 via a 7805 on a separate breadboard

>> No.1956388

>>1956387
Well, there's your problem. You need to connect those grounds together.

>> No.1956393

>>1956380
>>1956381
You'd better be using freewheeling diodes on those coils, the inductive kickback is gonna hurt otherwise.

>>1956384
>solemnoids
nice

>> No.1956395

>>1956393
>nice
No, solemn :-)

>> No.1956396

>>1956393
yeah I forgot to draw them in, but I've got diodes on the solenoids.

>>1956388
I joined the grounds and still nothing. I swear I wired everything correctly but I guess I'm just going to have to start from scratch tomorrow

>> No.1956400

Any suggestions for getting into RF design?
Books? Online courses?
Affordable spectrum analyzers?

>> No.1956402

>>1956381
disconnect the mosfets. is it still only 1V at the STM's pin? you might need a proper FET driver.
What frequency and duty cycle is your PWM?

>> No.1956407

>>1956402
I'll try disconnecting them tomorrow, but the duty cycle is supposed to be 15%, 50%, and 80%. The frequency is 110 Hz. The voltage didn't really vary regardless of the duty cycle (at least, it was relatively constant at the base of the BJT)

>> No.1956425

>>1956407
at 100hz you would not need anything. Try only turning the BJT on and off, if that does not work you are doing something wrong. If you do not have acess to a scope put a LED in series with it to see.

>> No.1956429

>>1956396
>>1956407
Check if the voltage rails are sagging. What kind of caps do you have?

>> No.1956433

Do I need an oscilloscope/pulse generator/variable power supply?

>> No.1956450
File: 97 KB, 1150x505, all you need other than dildos.jpg [View same] [iqdb] [saucenao] [google]
1956450

>>1956433

yes, less than $20 total.

>> No.1956452

>>1956450
So, it is not a good idea to buy the large types they use in labs at the university?

>> No.1956460 [DELETED] 

>>1956452

not if you're gonna quit the hobby after 2 months coz it's a lot of work for very little fun.

>> No.1956465

>>1956452

not if you're gonna quit the hobby after 2 months coz it's a lot of work for very little fun, and 95% of your efforts will be failures.

>> No.1956491

>>1956452
learn the basics. You will notice when the lack of scope becomes an issue

>> No.1956546

>>1956465
>>1956491
I think I know the basics and I have used scopes a few times before. I mainly do prototyping that requires simple automated or controllable mechanical projects and microprocessors. I need to be able to make the prototypes on my own, but not any final products.
I noticed the univerities sell a lot of scopes, but I have no idea about buying scopes.

>> No.1956550

>>1956387
>I'm powering the STM32 via a 7805 on a separate breadboard
Are you raping your STM32 with 5V or is there another voltage regulator?

>> No.1956557

how fast is an ordinary optocoupler? i want to use it as a driver for a mosfet and i wonder if it can keep up with the mosfet since i want to pwm it

>> No.1956562

>>1956557
as fast as the datasheet tells you it is. You can use this circuit to increase the speed too.
https://electronics.stackexchange.com/questions/136928/under-what-conditions-does-an-optocoupler-work-fastest

>> No.1956563

>>1956562
>datasheet
those things are fucking hieroglyphics to me

>> No.1956572

>>1956550
Whatever-pills have a voltage regulator onboard. Vin pin likely takes a range, including 5v .

>> No.1956579

>>1956572
then instead of mosfets connect simple led, change code to just toggle pin and see if it works at all and work from there.
If everyhing is wired correctly on breadboard it should work so either code is not working or something is wired wrong

>> No.1956580

>>1956579
Solenoid Anon went to sleep AIUI. I'm just a lurker.

>> No.1956587

>>1956579
solenoid anon here
yeah it's running on 5 volts but it can handle it according to the data sheet.

The PWM does work with the LED. That entire side of the board works fine, it's the mosfet part that's shitting the bed. I'm wondering if I just got a bunch of bad 2n2222s. I haven't been able to re-wire anything yet, that'll have to wait until after work.

>> No.1956662

>>1956546
If you can get any oscilloscope for under $50, it's probably worth it.

>>1956563
You'd better get practiced reading them, you can't practice the hobby without it.

>> No.1956674

is a "pure" sine inverter just a power mosfet H-bridge driven by a microcontroller via SPWM?

>> No.1956813
File: 97 KB, 1918x1056, laserEngraver.jpg [View same] [iqdb] [saucenao] [google]
1956813

I'm making myself a simple laser driver for my engraver.
But this time instead of wasting my own time by making the pcb myself, i'm gonna try out on of them chink fab houses that makes pcbs for like 5 bucks, i bet it will look great, i will have it made in black solder mask with gold plated through holes and white silk screen, oh yea i'm getting hard just thinking about it
>but anon if the circuit is this simple why don't you just use perf bo-
shut up

>> No.1956820

solenoid anon here
guess who just learned what a pulldown resistor is for

>> No.1956837

>>1956820
Ahem. That's some progress, considering your grounds weren't even common yday.

>> No.1956842

>>1956813
hm... should i add some solder mark anime girls to the pcb? I wonder if the fab houses have any graphic rules, so if they would reject a nude big tiddy solder mask waifu design

>> No.1956900

I wonder what's the best and cheapest method for insulating a home electronics workbench. I just want to protect a few equipments like a 30V 5A PSU, a 2in1 soldering and smd rework station, a low end laptop, a Hantek 6022be oscilloscope and a function generator from my initial project.
I'll be working with automotive ECU's. I'm just at the beginning of my studies but I would like to build the workbench and setup the electrical installation correctly already. Thank you, anons.

>> No.1956913

>>1956842
Do it. But do it tastefully, or go all out on the /e/ cunnyposting. Or put the Tiananmen Square copypasta in there if it’s a Chinese fabhouse.
>big tiddy
Shit taste detected

>> No.1956933

>>1952080
We have them in the lab, they are really good but once someone manages to fuck up the blade they are useless

>> No.1956943

>>1956842
Yes, you should. Particularly if you want to sell them.
waifuduinos are selling like pancakes in /csg/

>> No.1956989

>>1956943
>search “anime pcb” and “anime circuit board” on aliexpress
>no relevant results
this need to change asap

>> No.1956990

>>1956989
Try rgbduino 1.2.

>> No.1956992

i have bought a compass module for a robotic project and it isn't linear AT ALL (90° reads 80° and 180° reads 210°.)
I need to linearize it and since i'm way too lazy to do it degree by degree, i'm making a machine that turn the module 0.1° and sends the data to an excel sheet to that i can make the linearization here. here's the issue : how can i prevent the magnetic field from a small hobbyist servo motor from interfering with the compass module ?
would wrapping the servo in tin foil be enough ?
if that glorified conspiracy hat doesn't work, what would ?

>> No.1957001

>>1956992
You need to wrap in in something ferro-magnetic. TIn foil does not work for low frequencies. Put it in a mint/tea box or something. Also the datasheet should have something about linearity.

>> No.1957004

>>1957001
and I have no idea about how the metal bawks will affect the earth magnetic reading.

>> No.1957046

>>1957004
The datasheet has nothing about it sadly.
The servo will turn a 3d printed gear that will turn another gear with the module or my phone stuck on it, that way it doesnt have to come into direct contact. Thanks for the help.

>> No.1957051

>>1956990
I already ordered one, for the meme. But they’re hardly terribly good. Pink hair and big diddy and skimpy clothing just screams chinese knockoff. I want a lainboard. I’d settle for a tuhu or yotsuba, but not a botegirl or similar.

Good thing I bought my own solder mask for my etch kit, I’d see about testing myself if I actually had a printer where I’m living.
Would rather use the CNC laser ablation of resist/mask though, looks speedy and fun.

>> No.1957057

>>1957046
Put the servo on a long belt/shaft to keep it as far as possible away from the sensor.
Or just turn it off with a MOSFET on its PSU once it’s in place before the measurement.

Also do it in both directions for not just linearity but repeatability. Take it somewhere else to ensure you’re not in a magnetic anomaly created by a large price of iron nearby.

>> No.1957107

I've rewired the circuit, added pulldown resistors like I should have, and made sure everything is wired right and actually functioning

Even with the BJT driving it, or directly connecting PWM to the gate, the solenoid won't trigger by the signal alone. So I've bitten the bullet and just bought some mosfet drivers

>> No.1957111

>>1957107
Have you checked your pinouts? Draw a diagram of the voltages at the different sections (BJT base, BJT collector/MOSFET gate, MOSFET drain). Replace the STM32 with a switch to 5V or 3.3V for simplicity.

>> No.1957165
File: 22 KB, 408x385, inductor.jpg [View same] [iqdb] [saucenao] [google]
1957165

>grab pic related out of a switching PSU
>hook one side up to lap PSU and the other side to a multimeter
>the input side draws a lot of amps but less than a volt
>0 volts on the multimeter
Intredasting...

>> No.1957175

>>1957111
Yeah, I've checked everything and it's all correct. The PWM can switch an LED, the BJT connected to the PWM can switch the LED, the MOSFET on its own can switch the solenoid when the gate is connected to VCC

it's just the BJT can't drive the gate to switch the mosfet

>> No.1957203
File: 35 KB, 902x87, IRFZ46N gate threshold.jpg [View same] [iqdb] [saucenao] [google]
1957203

>>1957165

you have no idea what you're doing. transformers dont do DC.

>>1957175

as a rule, i stay away from FETs, so i might be ignorant. but this line in the datasheet seems odd.
for threshold voltage, there's both a min and a max. why a max?

>> No.1957209

>>1957203
So it's a transformer and not an inductor? Makes sense because I saw it on the AC side of the PSU

>> No.1957211

>>1957165
Learn Faraday's (or Lenz's) law.

>>1957175
It can take a fair bit of current to get them on and off to minimise switching losses, which is what dedicated MOSFET drivers are meant for. The ones I bought can push ~2.5A instantaneously on a good day, they're half-bridge drivers too so do the low-side and high-side. SMD tho.
But if you don't need to minimise switching losses (relays don't switch quickly anyhow, so it's not like you'll be slamming the FETs at hundreds of Hz in the first place) anything should do. A BJT driver like you're using will be able to get the voltage up to 12V and below 2V, provided it's working properly. I had some issues where my 2N2222A just wasn't turning off properly once. Replaced it with an audio amp.
In other words, it's very much doubtful that getting a MOSFET driver will help you.

Can you double-check for me with this table:
>______| OFF | ON
>Vbase | 0.7V | 0.0V
>Vcolle | 0.7V | 12V
>Vdrain | 12V | 0V
All voltages are measured with respect to common ground.

I'm assuming Vdrain is not behaving like that, correct? If Vcolle doesn't go low enough or high enough, then it's a problem with the BJT. If it does, then it's a problem with the MOSFET. If Vbase isn't behaving properly then it's a problem with your driver.

Also check your breadboard(?) connections as they can be a bit shitty, and test your MOSFETs out of circuit. It's possible that you destroyed your MOSFETs by using a diode improperly or just got counterfeit ones in the first place.

>>1957203
Because it isn't an assured spec, and can vary part-to-part and on temperature. If you want to use a circuit to turn the MOSFET on and off, then it needs to be at least 4V when on, and no more than 2V when off. Anything in between could turn it on, could leave it off, could leave it in a semi-on position where it eats a fuckton of heat and dies. Then again, I wouldn't drive it at only 4V either, those transistors usually function best when their gate is up at 10-15V.

>> No.1957214

>>1957165
>>1957209
Pretty sure it's a common-mode suppression choke. Technically it's a type of transformer, but chokes aren't designed to be efficient at high frequencies like inductors and transformers, hence the magnetic core isn't low-loss / high-Q.
Common-mode suppression chokes are installed before rectification on the HV side of SMPSs for noise suppression purposes, not for any kind of power converting.

For some info on cheap and simple power supplies, take a look at some of big-clive's teardowns. Often LED light bulbs and the like have a cheap AC-DC power supply in them that uses minimal parts. Sometimes isolated, sometimes not.

>> No.1957217

>>1957107
try making a class B amplifier in the meantime. Just a PNP, NPN, and a couple resistors

>> No.1957220

>>1957217
A class-B will only go to a voltage as high as the input signal, no higher. He needs something that uses a pullup to 12V, if not a smarter level shifter.

>> No.1957225

>>1957220
doesn't his CE amp pull up to 12?

>> No.1957233

>>1957225
It does. Though I'd be of mind to call it an inverter or level shifter, as opposed to an amplifier.

A class B / totem pole on the other hand won't pull up to 12V.

>> No.1957237
File: 83 KB, 1578x854, cm-choke-example.png [View same] [iqdb] [saucenao] [google]
1957237

>>1957165
It's a common mode choke
It has no "input" side. Since it's magnetic component - doesn't really work with DC other than letting it through.

What you want in common mode choke is common mode attenuation to be as high as possible while differential mode to be as low as possible

Pic related shows difference between common and differential modes. In common mode essentially created magnetic field is in same direction and causes impedance to increase and attenuate signal. In differential mode magnetic fields by each coil are supposed to cancel each other so impedance is very low (diff mode setup shows that currents should flow in opposite directions)

>> No.1957240

What is the purpose of using groundfill on pcbs?

>> No.1957241

>>1957240
reducing noise radiation, noise susceptibility, reducing crosstalk, that sort of thing.

>> No.1957242

>>1957240
By that i mean should i remove all copper except traces or keep all. Copprr and only remove say 1mm around traces to isolate them

>> No.1957243

>>1957046
Also I have assumed you were already doing it but put ferrite beads on the motor cables. You can buy clip-on ones or scavange them from old electronics (mouses, keyboards, printer motors)

>> No.1957250

>>1957242
It's good practice to keep it for noise reasons. As far as PCB manufacture goes, it means more etch resist remains on the board (less sodium carbonate required after exposure) and less etching to be done (your etchant lasts longer). If you're not etching it yourself then those don't matter as much.

In the case that you're using higher voltages, it's somewhat common to have no ground plane to dissuade jumping/creepage. Like on an AC-DC power supply, the HV side would have no ground plane but the LV side would.

>> No.1957256

Do I still need a BMS if I got an active balancer?

>> No.1957258

>>1956992
What's your module chip?

>> No.1957282

Is there some universal smd part that is like legendary sop223 lm317, but is a buck converter and not a linear regulator, so it allows me to save more power? i need it to operate with voltages up to about 12v and be able to bring them down to 3.3 or 5V with 900ma draw tops

I am using sop223 lm317 but i am just not happy with all the precious precious energy it's wasting as heat since it takes 1.5V and just throws it out the window as heat which my tiny lipo battery frowns upon

>> No.1957283

>>1957282
it is called "search the part catalog"

>> No.1957286

>>1957283
that will only give me some random results, i am looking for some widely used famous parts , like for example every child knows what lm317 is, but nobody has ever heard of say LM2941CT which does basically the same thing but is not as famous.
Same situation with for example attiny85 vs some other micro nobody ever heard about

>> No.1957299

>>1957286
LM2596

>> No.1957305

>>1957299
>LM2596
That requires a bunch of external components, i am looking for something that is standalone like lm317 where it just werks without having to add billions of caps and coils and shit. But as i am looking at it i don't think anything like that exists and i will probably have to keep existing with the good old smd lm317

>> No.1957312

>>1957305
switched capacitor regulators

>> No.1957313

>>1956163
there isnt any real load. and i have read tons on the subject. the transformer will be rectified and charge a cap for smoothing then go through a homemade relay. a 9v can power it but a 15 kv bleed cannot so a hundred ma should be fine since the 9v can do it at 50 ma. now that said as it powers the relay its charging a cap with low esr but to adjust it to make sure it gets the right cvoltage a minimal resistor in series with it connect the circuit to the full bridge but when the cap is charged enough it cannot sustain the relay as it restricts the flow before it fully cuts off and reconnects and that dumps charge into the tank circuit which is only connected to the full bridge circuit with the relay and the process begins again. the homemade relay only moves about the thickness of a piece of paper and has a single normally closed connection. it does not use a copper leaf spring . it moves a plastic arm with permanent magnet on it (small)

so you see there is no real load and nothing blows up. passive draw is all that happens. no motor or tv. the tank circuit should erode whats put in and the limited window that its connected to the fullbridge through the relay isnt a big deal. i have done this with a fet but the fets have a damn body diode and shits back flowing... also the voltage was to high and the current to low so i had to wind a steel core and find a lower frequency oscillator. but nothing blew up. nothing died. anything that hits 800v triggers avanlanche diodes and goes to the negative of the power source and fullbridge which can handle 900.

since the process unfolds over time and caps are charging and discharging the current cant all go back to the fullbridge and then transformer. it only puts so much out and only takes so much in. from what i can tell most people who do this stuff for motors dont get it. much like you they say that doesnt work . it does. at no point does the transformer ever think i need to take all of that back

>> No.1957314

>>1957282
There is no such "one solution fits all" option for buck converters. Depending on load, input, output voltages, currents, noise and efficiency etc there are many options and easiest and most versatile options are for buck controllers with some minimal external components like AP62200
Seeing you being such cheapfuck there really is no solution for you

>> No.1957318

>>1957313
the transformer itself is a inductive load. That is why everyone so far has told you to use a buffer between your oscilator and whatever you are driving.

>> No.1957320
File: 10 KB, 400x400, tegaki.png [View same] [iqdb] [saucenao] [google]
1957320

>>1957313
>the transformer will be rectified and charge a cap for smoothing
Then you don't want an inverter style circuit at all, those are specialised for delivering AC. You need a proper switching power supply, be it a boost or a flyback. No need for a high-power oscillator, especially not a sinusoidal one. Just use a 10k-100kHz relaxation oscillator of some kind (comparator schmitt or NE555) connected to a N-ch MOSFET, which pulls on an inductor or transformer. I'd use a transformer if you need it to be isolated, or otherwise need a high step-up ratio (over 10x or so). Transformers are a pain to obtain/wind though, so you might want to do as I did for my geiger tube PSU and just boost using a 220µH inductor, a UF4007 diode, and an IRF840 500V MOSFET, even though it will be less efficient. Add an extra comparator for feedback of course, to prevent anything from going too close to your breakdown voltages.
How high do you need your output voltage to be, and why?

>homemade relay
>15 kv bleed
>passive load
>capacitor
Gonna need a block diagram here chief

>the thickness of a piece of paper
>paper = 100µm thick
>dielectric strength of air = 3MV/m
>breakdown voltage across 100µm = 300V
watch yourself

>> No.1957322

i can't seem to figure out if those two capacitors are mandatory or not..?
https://static.chipdip.ru/lib/552/DOC001552809.pdf

>> No.1957327

>>1957322

you could build 20 of those and 19 will work fine without the caps but 1 wont. so, yes, 100% necessary coz you never take chances with missing condoms or caps.

>> No.1957329

>>1957327
i feel like if the power source is a liion battery then there is no ripple and the source is perfectly clean so the caps should make no difference at all... but i suppose i will include them in the PCB design and then simply not solder them on and if there any problems later i can easily add them

>> No.1957362

>>1957305
They all need external components, that's just how switching converters work. You can get modules where they mount the inductor and stuff on a small PCB and then you can solder that on, but you pay through the nose.

>> No.1957366

>>1957362
it's 2020, we were supposed to have hoverboars and flying cars for like 10 years now, i can't believe we still have to solder coils to shit to get less wasteful voltage reduction

>> No.1957370

>>1957366
You got a problem with coils? My little brother is a coil.

>> No.1957376

>>1957366
You can get all-in-one smd packages but they do not come cheap
For example, Wurth 171031801 is great

For cheap versions you can get assembled module boards. Just google "TO220 buck replacement" and you will find answers

>> No.1957380

>>1957366
Stop complaining brainlet.

>> No.1957383

>>1957380
>>1957366
And as I said before you can get charge pumps which have no inductors.
https://www.ti.com/power-management/non-isolated-dc-dc-switching-regulators/charge-pump-inductorless/products.html#p20min=30;1150

>> No.1957387

>>1957383
>low input voltage
>low output current
>useful in any way

>> No.1957389

>>1957383
charge pumps suck due to the two capacitor paradox, you can't power anything significant with them.

>> No.1957390

>>1957387
>>1957389
So go make your own ICs faggot

>> No.1957406

what is something good to use for programming and debugging outputs on a pcb?
normally i simply use regular 2.54 pin headers but on a small project i am making from smd parts regular pin headers are simply too huge and take up too much space

>> No.1957409

lazy bastards, i have to do everything around here
>>1957408
>>1957408
>>1957408

>> No.1957411

>>1957406
JST connectors/headers are commonly used. You could also use card-edge connectors, FPC/FFC, pogo pins, DIP sockets, etc.