[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/diy/ - Do It Yourself

[Advertise on 4chan]


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: 1604591024944.jpg (175 KB, 999x1000)
175 KB
175 KB JPG
Thread overvolted: >>2060313

>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

>Archive of Popular Electronics magazines (1954-2003):
https://worldradiohistory.com/Popular-Electronics-Guide.htm
>Some guy’s list of electronics resources:
https://github.com/kitspace/awesome-electronics
>Microchip Tips and Tricks PDF:
https://ww1.microchip.com/downloads/en/devicedoc/01146b.pdf
>Li+/LiPo batteries required reading:
http://www.elteconline.com/download/pdf/SAFT-RIC-LI-ION-Safety-Recommendations.pdf

>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

>I have junk, what do?
Shitcan it
>>
File: NCV8247.png (150 KB, 1072x1114)
150 KB
150 KB PNG
Thread brought to you by the NCV8247 series string pixel controller.

This is the first time I've seen LED shorting under a CC driver as a method of controlling them. Quite interesting.
>>
File: 1596116507822.png (321 KB, 1062x1375)
321 KB
321 KB PNG
this thread's digits brought to you in part by the MC56F8247 DSP controller
>>
THIS IS NOT WHAT WE AGREED ON
>>
>>2068260
weird... but follows the logic of the automotive world, where power saving is not as great a concern as luxury
>>
>>2068267
It doesn't waste power, the voltage output of the (switching) CC driver just drops.
>>
Can I use electric grill for BGA rework (CPU replacement on mobo)? It is 2kW, 230C max, but I think I can bypass it and make it higher.
Also, can I use rosin-alcohol flux for soldering BGA chips? Or I need proper BGA flux?
>>
>>2068345
I want to use grill as bottom heat, and hot air for actual soldering*
>>
>>2068345
Depends on what sort of temperature control it has. How close is the thermal sensor to the heating element? How quickly can it respond to changes in temperature? Does it have PID capabilities?
>>
File: modemExample.jpg (234 KB, 1543x655)
234 KB
234 KB JPG
Working with m.2 4G LTE cellular modem module.
On bottom of module there is a heat dissipation area used for bridging gap with a thermal pad to the main board IF the main board has an exposed heatsink ground plane.
Often the main boards available don't have an exposed plane.
>picrel is example mainboard that doesnt have ground plane and also has passives in the way
Is there a way to tether the m.2 module off of the main board (assuming I fabricate a way to mount the main board) so that I can add a large heatsink to the bottom area? I know there are ribbon cables used for this like with PC motherboards (IDE ribbon cable etc).
Is there anything available like that for this use case?
p.s. im ignorant as to whether or not using a cable like this would degrade performance if using an extension cable etc.
>>
>>2068379
I think it will have temperature all over the place, so I'm planning to use an arduino or similar shit with a relay and thermal couple. Maybe triac and implement PID, dunno.
>>
>>2068262
shit, sorry anon, it's been a couple of years since I baked the bread

>>2068393
the bottom heat is just there to avoid the powerful heat sinking action of the inner and bottom layers of the board. 230C is too much, you really want it to be well less than the solder's liquidus temperature so that any components on the bottom don't fall off, 150C max if you want to be kind to other components on the board
>>
File: Thermal-Pad.jpg (9 KB, 396x250)
9 KB
9 KB JPG
>>2068346
might be there for you to put a cutout on a custom board and install a thermal gel pad to connect the thermal pad to a chassis
>>
i got a 24v 1.75A switching adapter,
what do?
>>
File: DhIyXm.jpg (9 KB, 320x281)
9 KB
9 KB JPG
>>2068250
>>2068399
also with that low side topology and npn v_load can only be vin - 0.7 max. if you don't want the 0.7 drop put the led above the first transistor.
>>
>>2068430
stick up bum
>>
File: exampleDrawing2.jpg (114 KB, 1976x922)
114 KB
114 KB JPG
>>2068427
>might be there for you to put a cutout on a custom board and install a thermal gel pad to connect the thermal pad to a chassis
Aye, it's there for that but most of the boards dont have a cut out nor the exposed plane.
I'm wondering if there is some type of adapter/ribbon cable I could insert into m.2 connector on main board to extend the module away from the board to mount it and install a heatsink directly to the pad? Drew example pic
>>
File: ngff extender.jpg (49 KB, 1000x1000)
49 KB
49 KB JPG
>>2068443
there is
>>
>>2068417
Hm. Now I think about it, would simple triac dimmer be sufficient? After all I'm not a repair shop (yet).
>150C max if you want to be kind to other components on the board
Yeah
>>
>>2068493
the thermostat on the front would be more stable than trying to use a chopper
>>
Reminder to socket your JFETs
>>
File: heatDisArea.jpg (29 KB, 526x603)
29 KB
29 KB JPG
>>2068453
awesome, do you know if there are types of board that don't block the bottom of the module or could be cut so that the bottom of the module isn't blocked by the board? I see passives on that particular board so I don't think there'd be room to cut a 25x25mm hole in it to expose the area on the module. Trying to search but haven't found any alternatives.
>>
>>2068503
Seminder yo rocket tour EFJTs
>>
>>2068507
>do you know if there are types of board that don't block the bottom of the module or could be cut so that the bottom of the module isn't blocked by the board?
yes, the one you make yourself.
>>
File: nigometer.jpg (110 KB, 451x267)
110 KB
110 KB JPG
>>2068519
Makes sense, sorry for being so dense, kind of new to all of this. Will look into fabrication requirements to creating my own board that can handle an extension and all the components it may need on it. Thanks again anon, I should now be pointed in the right direction.
>>
>>2068519
>yes, the one you make yourself.
okay mr. balloon hands
>>
>>2068496
Hm. Is it worth it making my own thermostat, or built in thermostat would be sufficient (that is supposed to measure cooking surface temp)...
I've just have a concern that it overshoots and stuff. Plus it would measure cooking surface, not board temp
>>
File: Capture.png (13 KB, 190x216)
13 KB
13 KB PNG
So I'm working on this project and part of it involves measuring the current coming out of the photodiode inside of this laser that has the pinout of pic related. I'm not sure what the best way of doing this is, ideally I'd like to get a resolution of 1 microamp in the readings.
>>
is osha going to come to my house because i removed a ferrite bead for a 24volt 1.75a switching adapter.
EMI is the cause of concern
>>
>>2068537
>it would measure cooking surface
that's true. manual solutions including a dimmer and a thermocouple on the board and readout are probably fine. it's crazy but it just might work. gl anon
>>
>>2068519
Anon, I forgot to ask, maybe this question is impossible to answer without knowing more details on the module but in your opinion: do you think that the benefit to thermal performance would be worth the trouble of fabricating my own board? According to hardware sheet I can also just put a heatsink directly on top of the silver component shield on the module.
Just wondering if i'm being too autistic and over thinking things.
>>
File: 3500fig01.png (4 KB, 439x162)
4 KB
4 KB PNG
>>2068592
>do you think that the benefit to thermal performance would be worth the trouble of fabricating my own board?
that depends on the total power dissipation of the device and the temperature rise that it will create in different configurations.
you might not even need a heatsink if the power dissipation is low enough.
>>
>>2068542
Well you'll want pin 2 to be at Vcc, and pin 3 to be being pulled down into whatever constant current sink you have in mind. Hence you'll put some resistance from pin 1 to GND and measure the voltage across that resistance with an ADC, either directly, or with an amplifier in between. Use too low a resistance, and the voltage will be so small you need to amplify it, but use too high a resistance and the output impedance might start to mess with your ADC. Rail-to-rail amplification may well be required.

I'd also strongly consider making an analog feedback loop for the current measurement, pretty sure you could do so with an LM2596 and an op-amp and some passives. Same for almost any other common buck converter circuit. So long as you're not trying to do some sort of computer controlled laser brightness profile, that is.
>>
File: 1602853705568.png (39 KB, 514x522)
39 KB
39 KB PNG
>>2068542
picrel for random example off of image search. check the datasheet for your laser, they may specify a particular load condition for the feedback diode
>>
File: file.png (74 KB, 999x549)
74 KB
74 KB PNG
>>2068542
there are a bunch of packaged current measurement solutions available, you just have to dig around.
though 1uA/lsb might be hard to get.
pic related is an ina226.
>>
File: 1596995431798.png (200 KB, 418x315)
200 KB
200 KB PNG
If I don't have a proper crimping tool, is it OK if I take the skirt off and solder a wire in, then isolate with heat shrink tubing?
>>
>>2068772

leave the plastic sleeve in place
solder the wire where it peeks out the front
the solder will be sucked into the tube coz of capillary action
the plastic will survive but change color a lil bit
>>
>>2068782
>leave the plastic sleeve in place
>solder the wire where it peeks out the front
OK guess I could try that, thanks.
>>
>>2068744
Not him, but that's a nice simple circuit, using a DAC as a variable reference. Note that the comparator there HAS to be open-collector/open-drain, or just replace it with a single transistor in the first place. Can't say how the feedback network will react to the PWM either, either the op-amp goes into saturation sourcing current into the comparator, or C6 buffers the change by increasing the peak amplitude, and the PWM has no net effect. Both are subpar.

And seriously:
>470kΩ base resistor
>10Ω collector resistor
>without a darlington pair
Looks like a mistake if I've ever seen one. Those MMBT3904s are just normal hFE_min = 50-100 BJTs.
>>
>>2068772
if you plan on using those more often, buy a brand crimping tool used on ebay. They are plenty and cheap for those style crimps, like 1/5th new price.
The chink ones can suck hard, crimp looks good while wire pulls out easy.
>>
File: common-emitter.png (14 KB, 562x459)
14 KB
14 KB PNG
Being retarded and having a hard time figuring this out..

Looking at common emitter amplifier like pic related. But having a hard time figuring out exactly what this resistor is supposed to do (circled in red), I know it's something to do with biasing and feedback for the transistor but I can't find anything that explains this clearly.
>>
Sup /ohm/ I need advice: I want to build an electrostatic precipitator, because fuck overpriced air purifiers, and the only thing I really need is the high voltage supply. From what I found in literature I need 50kV DC, current draw should be around 1mA, but probably even less than that. On amazon I found some brick that should step up the voltage from 4,8-6V to 50-800kV, with an input current draw of 2-5A, but I don't know how I can control the output voltage, and I kinda fear it would blow up in my face. What do you suggest?
>>
>>2068964
big clive has done like a zillion teardowns of ionizers, he has some kind of weird fetish for them. seems like they're usually a transformer plus a many stage voltage multiplier
>>
>>2068790
>Can't say how the feedback network will react to the PWM either
the time constant is about 50ms. it will maintain average power as seen at the feedback diode, which implies slight increases in instantaneous power during heavy channel usage. it will be received as ac at the receiver anyway
>Looks like a mistake if I've ever seen one
the 470k base resistor is a bit odd, I agree, unless the intent is to vary the current within a small factor (like 20% modulation/correction index?), and R1 is supposed to carry a significant part of the drive current. it would have to be like a 1mW laser diode for fiber or something

>>2068893
qualitatively, when c-e current flows, the emitter resistor voltage also rises, and the b-e diode passes less current (Shockley diode equation), which reduces the c-e current by beta times the difference
>>
>>2068990
I'd rather not buy a ionizer as they're expensive and I was hoping in something cheap that I could buy off the shelf or some relatively simple circuit. My idea was to step up the voltage from 5V or 12V DC so I could just use a spare wall wart to drive everything (the wires of the ESP and the fan). In the end a single stage dry electrostatic precipitator is just a pair of grounded plates and a few wires in the middle with a fan forcing air through, something I could easily strap together with spare junk I had lying around.
>>
>>2069067
Ionizers and electrostatic precipitators are just variations on the theme of applying high voltage to air to make the dust drop out, no? So you could copy the circuit from one or get a chinkshit ionizer and steal the high voltage supply from it. 12v-50kv converters are not exactly a common component.
>>
>>2069087
No, ionizer work on the assumed principle that free ions will sanitize the air by fucking with the chemistry of bacteria and viruses or something, but IIRC it's snake oil, so ionizers are literal electron guns. ESP instead work how you said, and actually work, also considering how they're mainly used in industrial applications for air pollution control (and I just want to scrub the dust from the air so it's the perfect use case).
>>
>>2068964
>50kV
what's the difference between this and an ozone generator? doesn't air ionize at only 3kV?
you can get ozone generators for like $30
>>
Lads, I'm looking for a single axis joystick that has both a potentiometer and (preferably multiple each direction) NO momentary switches. Does this exist? What kind of search terms should I use?
>>
>>2069123
As I said it's neither an ozone generator nor an ionizer. The working principle is complex but the idea is that dust particles are slammed against the grounded plates from the corona discharge of charged wires. I found this study that has a setup of similar dimensions to what I intended to build, and that's where the 50kV figure comes from: https://www.sciencedirect.com/science/article/abs/pii/S0304388699000443

I knew electronic components were pretty cheap, maybe even free if you scavenge them from junk, and that looked like an easy and fun DIY project, honestly I didn't know 50kV DC were such a big deal: I saw an electroboom video where he made a "magic wand" that's basically a high voltage electrons gun and that thing was pretty compact, but I figured I asked for better informations rather than just copying a meme youtuber.
>>
>>2069137
Doesn't an ionizer do that?
The particles in the air get charged, which attracts them to a grounded surface where they can lose the charge.
>>
>>2069087
>>2069137
either way, the high voltage supply is constructed in the same way, the rest is just what you plug it into.
>I didn't know 50kV DC were such a big deal: I saw an electroboom video where he made a "magic wand" that's basically a high voltage electrons gun
Please keep in mind that while he larps as a retard for those sweet sweet clicks, electroboom is an actual engineer who knows what he's doing. Just because he made it look easy doesn't mean you won't electrocute yourself.
>>
>>2069143
As far as I know an ionizer only shoots out charged particles and not much more, what happens after is not their concern. An ESP is a self contained device with a clearly defined design optimized to capture aerosol particles; dirty air comes in, clean(er) air comes out. I haven't looked more into it as that paper covers my use-case pretty well, but it's clear that there are lots of parameters that influence the efficiency of such a device, and there's been a number of studies to model their behavior.
>>2069146
Yeah I know that, it's pretty clear. To be fair I'm an engineer too, just not an electrical engineer, so I have just a basic knowledge of electronics, hence why I was looking for more informations.
>>
File: image.png (138 KB, 2250x1042)
138 KB
138 KB PNG
Would this work?

Based on this: http://www.mosaic-industries.com/embedded-systems/microcontroller-projects/electronic-circuits/push-button-switch-turn-on/latching-toggle-power-switch (fig. 6)

This is a TURN ON TURN OFF circuit, it is supposed to turn on after a tap on a button and turn off when the button is held for more than 3 seconds. The MCU should be able to turn ifself off by driving the latch pin down. The only thing I've changed instead of second MOSFET I've used an LDO to reduce voltage for MCU's needs.
>>
>>2069168
>a diode to ground
get a simulator instead of asking stoopid questions
>>
>>2069183
Whoops. I've tried PROTEUS, it's garbage. Other than this?
>>
>>2069194
MicroCap is good and free. might even have that LDO in the parts library
>>
File: file.png (82 KB, 602x446)
82 KB
82 KB PNG
Ohmbros, got a question on turn-ratio conversion of transformer from AC->DC before I set out to make it, I just want to know if my calculations are correct if that is ok with you guys. I posted a while back about this but then IRL shit got to me bad so now I'm back.

My goal is to find a turn-ratio of a transformer that converts a 120 Vrms -> peak voltage of 13.4V (not 12V because I need to use 2 diodes, so the 1.4V is just for the voltage drop.

So to do that, I need to first need to find the Vrms value of the 13.4V voltage peak, and I use the formula Vrms = Vpk / sqrt(2) where Vpk = 13.4V.

(13.4Vpk)/sqrt(2) -> 9.47524 Vrms

Then I think I need to do the transformer ratio formula which I think is pic related. So the secondary voltage, I'd want 9.47524 Vrms, and the primary voltage is just the mains so 120 Vrms. Based on the formula, it'd be V1/V2 so 120/9.47524 -> 12.66 which would be the number of turns on the transformer.

Is this correct? I feel like something is off, I want the output (V2) to be 9.47 Vrms but I feel like with this turn ratio, it'll be one of those step-up transformers rather than stepdown. Should it be the other way around, with 9.475/120? I'm doing the full-bridge rectifier btw. Had to delete and repost because my brain fucked up so I fixed some stuff, hopefully it makes more sense.
>>
Can someone explain to me what the pillow means?
>>
>>2069224
you probably flipped the meaning of n1/n2, which is the number of turns of the primary for every turn of the secondary
>>
>>2069233
It's a print of a lipo battery.
>>
>>2069233
it's a pillow printed with a picture of a dangerously inflated LiPo battery. just lay down, it's comfy~
>>
File: file.png (4 KB, 248x81)
4 KB
4 KB PNG
>>2069235
Ok yea anon I think you're right, I probably flipped the meaning. I tried 9.47524/120 on the transformer ratio and it showed pic related, this is definitely off, when I did the other way around it produced ~12V which is comfy. I get really fucked up by terminology like this, thanks for reminding me what it actually means anon, so every 120 turns on the primary, the secondary will turn 9.47 times?
>>
File: common emmiter gain.gif (80 KB, 1512x545)
80 KB
80 KB GIF
>>2068893
>what this resistor is supposed to do

main purpose is to determine gain of the stage, secondary purpose is to add temp stability.
>>
Sorry for the dumb noob question. I just bought this air pump for my inflatable paddle boards, it claims 6000 mah of battery, which I'm assuming is stored in 18650's. Assuming there's plenty of space in there, how easily can I add more 18650's? I have 3 that are just sitting on the shelf. Can I just solder them in willy nilly, or would I need a new controller?
>>
>>2069244
You could replace them with more cells in parallel but they should be matched otherwise the worst cell will drag all other cells down.
>>
I'm trying to wrap my head around what situations with electricity are genuinely dangerous
Currently I'm considering the case of being in the open ocean, when lightning strikes close to you. Presumably the salt water of the ocean has a significantly lower resistance/a shitload more ions floating around than your isotonic body, which is already "shielded" by skin with moderate electrical resistance.
Why then, would the lightning's current have any reason to pass through you?
If the lightning is acting to balance the charges between the clouds and the ocean, why are you even "considered" part of that system?
Is it a case of "the lightning's voltage is so high your resistance is negligible and it goes through you anyway"?

Then, would this be valid with lower voltages? Let's say I'm floating in a saltwater pool and (somehow) an extension chord laying across the pool is cut so that each end lies submerged at one end of the pool and the other. Now that my skin resistance of roughly 5000 ohms isn't negligible compared to 120 volt AC, would I still be in harm granted I don't touch both ends of the chord at the same time?
>>
>>2069275
>Why then, would the lightning's current have any reason to pass through you?
The current goes through everything.
It doesn't "choose the path of least resistance". It takes every path.
>>
>>2069276
is it safe to pass current through my testicles
>>
>>2069244
>Assuming there's plenty of space in there
I doubt there's plenty of space in there.
>>
File: image.png (36 KB, 1781x496)
36 KB
36 KB PNG
>>2069199
Okey can I come off like a retard for a second, how do I simulate this thing? There's "Transient" which I guess allows me to graph stuff. But how do I factor in the fucking button?
>>
>>2069295
clock signal+transistor
>>
>>2069241
correct

>>2069295
MC12 has some neat dynamic analysis but I'm not sure how to use it desu
I'd work in a MOSFET and a time-varying voltage to turn it on and off. a few ohms isn't going to bother anything in this part of the circuit
>>
>>2068893
bro add a cap across it

>>2069113
Ionisers are intended to just clean the air by applying high voltage to air to make the dust drop out. That's why you put a piece of paper underneath them that gets dirty. There's a demo with one being turned on in a room full of smoke. And it imparts a charge to all particle sizes via electrostatic induction, from dust to air molecules, so evidently it will pull microbes out of the air and onto the paper mat. These use very fine spikes (or carbon fibres) as electrodes that are nowhere near the ground/return path, and consume barely any current because of it.

But they're different from HV ozone generating devices, which both actively kill microbes through chemical oxidation. Often marketed as "air fresheners", as opposed to "ionisers", but there's definitely some overlap. These can use a variety of different topologies of electrodes for ozone generation, but all of them have the positive and negative electrodes that are reasonably close to one another, in order to produce a lot of corona discharge.
A UVC lamp will usually work better than these for sterilisation, aside from ozone's ability to seep through fabrics and go around blind corners.
>>
>>2069236
>>2069238
Ahhh ok. I see it now. Thank you.
>>
File: 1584151589340.png (171 KB, 596x622)
171 KB
171 KB PNG
>>2069352
>>
>>2069328
>correct
Thanks bro, it's always nice to have a platform to talk on/ask questions to, especially when I make stupid errors left and right
>>
File: donkeywearinghat.png (598 KB, 612x952)
598 KB
598 KB PNG
Kinda new to this so please excuse possible retardation.

So it is not possible to either step up from between 3.2v-2.8v to something between 4.4-3.2v _or_ stepping down from 6.4v without shit getting wasted as stupid heat that also needs to get dissipated to prevent shit from literally cooking?
My device can only operate at around 4.4v down to ~3.2v while the battery I want to use has a nominal voltage of 3.2, holds this pretty well but will eventually be around 3.0 halfway through and I want to get the most out of it without also having to taking care of stupid ass heat that I want to avoid with efficiency in the first place.
>>
>>2069497
Forgot to mention that the device should get provided with at 2amps, better 2.5 to not make shit crash if it spikes for a moment.
>>
>>2069497
>either step up from between 3.2v-2.8v to something between 4.4-3.2v _or_ stepping down from 6.4v
yes, a buck-boost converter
>without shit getting wasted as stupid heat
no such thing as a 100% efficient conversion unless you can change the laws of physics.
>>2069500
2A is feasible but you might need extra bulk to handle fast transients if the change from 2-2.5 happens quickly.
>>
>>2069504
Alright, many thanks
>>
>>2069411
sure thing. in a practical transformer, you choose primary turn counts and wire gauges according to the temperature rise, size, and flux density limitations of the core and wire. in particular, size the wire to pass the inductive current (E/2*pi*f*L), and try to minimize the resistive part of the total power due to inductance, because it'll just become waste heat

>>2069500
buck is more efficient than boost. better to step down from 6.4. there are plenty of cheap buck converter options on aliexpress that are well under $1 each
>>
>>2069497
Buck converter from two batteries in series would be my method. Pretty sure an eBay LM2596 set to 4-4.4V will work fine at up to 3A, but you’d have to check the datasheet.

Bucks are more efficient and simpler than buck-boosts.
>>
>get a chink LCD
>it's dim as fuck
>can't tell whether it's my init code or just chink quality control
thank you xi
>>
>>2069559
ur welcome, thanks for purchasing our product, please leave a good review

I purchased their uni-t recently, pretty comfy
>>
>>2069559
>want to save money, buy the absolute cheapest thing available
>doesn't work, all money spent wasted
the electronics jew strikes again.
>>
>>2069580
buy the second cheapest. you get screwed a lot less often
>>
>>2069612
Buy all from one seller, and choose a seller with plenty of orders. The orders for the item doesn’t matter as much as the seller’s total number.

Never been chinked hard at all. And yes my LM358s work just fine, thank you.
>>
>>2069639
>Buy all from one seller, and choose a seller with plenty of orders
This is pretty much the golden rule, if enough people agree we should probably put it in the sticky if it fits
>>
>>2069673
We have alibay in the OP with:
>for component assortments/sample kits
next to it. I think it’s good enough advice to include, but in general I’d probably recommend ordering from LCSC instead. Presumably only the consumer/hobbyist-oriented assortments and kits are worth getting on alibay, though I can’t say I’ve done a comparison. Or bought anything from LCSC for that matter. If anyone with more experience wants to chime in to provide an informed opinion on buying from either service I can include it in my paste next time I make a thread (OP for >80% of the last dozen threads).
>>
>>2069673
From my experience, amount of orders doesn't mean anything in slightest. Like 0 order dude can scam you just as good as 9000 order dude. If anything, 0 order dudes tend to be better.
>>
>>2069685
what does the caveat emptor mean senpai
>but in general I’d probably recommend ordering from LCSC instead.
is LCSC a good option for the states? I remember someone saying to use it only if you're in Europe or something
>(OP for >80% of the last dozen threads)
thanks for picking up after us slowpokes...
>>2069686
>Like 0 order dude can scam you just as good as 9000 order dude. If anything, 0 order dudes tend to be better.
also a good point, I think its cuz they're more desperate to get positive reviews. I've used aliexpress for clothes/limited edition mugs/components and they've been pretty reliable, at least for the ones over 200 orders
>>
>>2069639
>Buy all from one seller, and choose a seller with plenty of orders
or buy from authorized distributers that have verifiable supply chains.
>And yes my LM358s work just fine
still could be fake and work, just not meet performance specifications listed in the real datasheet. its rare that a fake outright doesn't work.
>>
>>2069690
>is LCSC a good option for the states?
it was fine in 2019. far preferable to ali, and I imagine ali's only gotten worse. otoh, that was before COVID and the bitchy trade stuff
>caveat emptor
buyer beware
t.the anon who put it there long ago

>>2069710
this is the way. you may be surprised how close the pricing is between LCSC (verified good parts) and ali (probably stored in a coffee can)
>>
>>2069686
My anecdotal evidence disagrees with your anecdotal evidence lmao. Seriously though, using aliprice or whatever, ensure that the seller has high ratings, and sells a bunch of related stuff. If they've got a huge splotchy assortment of products from different categories, then they're probably just riding the money train on particularly popular items, and may not even have stock of their own. You know, the sort of seller that's selling arduinos with a name like "home and beauty" or something.

Also if you buy from one seller constantly and never get chinked, then that's a good seller. Never going to find a good seller with that level of certainty if you constantly hop about.

Not to mention the shipping price usually gets consolidated.

>>2069690
>what does the caveat emptor mean senpai
"Caveat emptor" is a latin phrase meaning "let the buyer beware" that's made its way into the english lexicon, alongside "et cetera" and "ad infinitum" and such.

>is LCSC a good option for the states
If you live somewhere like the USA or Western Europe, you've probably got a Digikey or Farnell or Mouser warehouse nearby, making it less worth buying from China directly. But check the shipping prices and such anyhow, all those bulk services have a free shipping threshold anyhow. I think Arrow and LCSC have relatively low free shipping thesholds.
>thanks for picking up after us slowpokes...
Well I'm a slowpoke this time around I'm afraid. Was gonna use the same pic too.

FYI, I'm also an ali main.

>>2069710
>or buy from authorized distributers that have verifiable supply chains
Depends on how unusual the part you're ordering is, and how much you value QC over getting your parts relatively cheap and in small batches. For me, far away from any warehouses, buying 5 ICs with less than $20 of shipping costs is worth it.
>still could be fake and work
It was a joke. LM358s are such old basic op-amps that they're probably what fakes of higher-end op-amps are under the hood.
>>
>>2069744
Idk. Some stuff is impossible to get from verified dudes (for ex, go find CPU for laptop in digikey or rs components).
Some stuff is better bought from verified dudes (for ex, LEDs if you need to be sure they are decent, power MOSFETs (for motor drives for ex), driver ICs, and other shit that might result in frustration and costly damage if it is shit)
Some shit like Arduino, modules for Arduino, microcontrollers, programmers and shit is just so cheap in Aliexpress, you can buy 10 fakes and still save money.
>>
>>2069751
Idk, I got scammed only once on ali, when I ordered FR4 and received fucking nothing.
> You know, the sort of seller that's selling arduinos with a name like "home and beauty" or something.
Well. This is not a deal breaker, I got tweezers, soldering tips from such sellers.
Actually, I don't think anyone stocks anything, they all are resellers, they get order, go to their Shenzhen market, ask bros if they have shit, buy it for half price and sell to u.
>>
>>2069757
>Some stuff is impossible to get from verified dudes (for ex, go find CPU for laptop in digikey or rs components).
oh, yes, ali is superior for gray market replacement parts, breakout boards, and tooling. there are even a few custom transformer shops on ali, which is amazing given what you had to go through 25 years ago. but often, you are just paying for another middleman

>>2069673
there's another argument against ordering all from one ali store: ali had a rationalization push a few years ago, penalizing stores if their products were too mixed. some vendors may own two or more stores, each covering a particular category (ICs, modules, switches). furthermore, when stocking a lab, you will probably not find one vendor who sells everything you will want. e.g. nobody has ALL the SMT adapter plates, you'd have to order from 2 or 3 for a full collection, and one of the adapter plates is only available from 2 shops
>>
>>2069762
>they get order, go to their Shenzhen market, ask bros if they have shit, buy it for half price and sell to u
Well considering it took like a week for them to put together my 25 item list, maybe that's true.

>>2069820
>breakout boards
This is a big plus with ali or ebay. For rapid prototyping, it's often worth designing and etching a PCB that a bunch of modules (eg MCU, DAC, ADC, audio amp, PIR sensor, etc.) just plug into via 0.1" socket headers.
>>
File: 3742019907.jpg (361 KB, 1000x682)
361 KB
361 KB JPG
TV suddenly turned red, pic should be all black/white text.
hasn't been hit, didn't hear a pop when it changed, just watching it and suddenly went red.

Is it reasonably fixable?
I thought it might have been a blown/leaky capacitor but I opened it up and didn't see anything that looked bulged, leaky or charred.
tried re-seating the ribbon cables too.
>>
File: 391012084.jpg (728 KB, 1200x800)
728 KB
728 KB JPG
>>2069854
Sony 40CX520
>>
>>2069854
The red channel is getting more than it bargained for, and that's the signal to the LCD going wrong somewhere. The red is presumably staying in the same place, and not flickering about? If it's flickering then you've got a digital signal gone to crap somewhere, and troubleshooting it is gonna be night impossible. If it isn't flickering, it could possibly still be a digital problem, but unlikely since 99% of all data going to the LCD will be serial, little chance for one corner of the screen to stay black and not the rest. Thankfully, if the problem is digital, this means that the problem is either down after the LCD driver, or all the way back at the RF receiver. The latter of which would be easy to troubleshoot by plugging an HDMI in, but I doubt it's the case.
More likely, if there's no flickering, it would mean damage to the LCD itself. Is the LCD hot at all?

>>2069858
That's the power board, not going to be too useful. But judging from all those magnetics, it's possibly be a HV CCFL backlight and not an LED backlight. The HV from a CCFL backlight is arguably more likely to damage the LCD panel than anything else.

Either way, I suspect unless it's an intermittent issue that might go away without doing anything or by delivering some percussive maintenance, it will be a broken LCD. Might be able to get a cheap enough replacement on eBay, maybe someone with a similar model TV on craigslist burnt out the logic board something, otherwise you're looking at a new television.

Or you could just keep it that way for as long as it lasts for the sake of having an edgy red TV. I tried to do that by buying a 2nd hand monitor for $1 that was stuck displaying a magenta image (dead green channel, would look cool) but when I plugged it in via HDMI-DVI connector, it looked normal. Must have been in the VGA port, which I've no idea why people still use considering the monitor and computer on either end of the cable are both fully digital.
>>
File: IMG_6887.jpg (3.51 MB, 2520x1903)
3.51 MB
3.51 MB JPG
>>2069866
>percussive maintenance
gave that a go

It's red on HDMI and broadcast TV, the red doesn't flicker but it does fade in and out in patches as it turns on.
>>
i have chip with pin for 868mhz antenna, if run a 3 cm long trace on my pcb to ipx connector where the antenna will be plugged in will that fuck something up? like for example won't the trace act as part of the antenna and change the frequency?
>>
>>2069961
>868mhz
>3 cm long trace
That's ~1/12th the wavelength, which arguably means you should care about transmission-line effects like impedance matching. Poor impedance matching can cause reflections of the signal, resulting in a lower SNR and lower effective signal power. Though at 1/12 wavelength it's not that big of an issue, people usually mark the threshold at 1/16 wavelengths or 1/10 wavelengths, so you're on the cusp one way or another. If you're worried, use KiCAD's microwave PCB design feature to make a single/double-layer coaxial line. Or just solder a short length of thin coax instead of routing on a PCB.

If the trace has grounding next to it / around it then it shouldn't change the resonant frequency of the antenna, not that a quarter-wave monopole probably has a particularly peaky response anyhow. The frequency of the wave being transmitted will not change under any common circumstance.
>>
>>2069854
>>2069858
>>2069910
Check/clean all flat flex cables and LVDS cable. Other than that, you could still have bad caps or t-con board.
>>
File: IMG_0447.jpg (2.02 MB, 3264x2448)
2.02 MB
2.02 MB JPG
I have this old egg scrambler thing that is basically a DC motor attached to 6 C batteries. C batteries are expensive and wear down because this thing gets used alot. I was thinking about converting this to a wall wart set up.
I know its 9V but how do you determine how many mA you need so I can find a suitable plug?
>>
>>2070068
12V 1A should be fine.
>>
>>2070068
Measure the power draw?
>>
>>2070068
the capacity of the average C battery (4000mAh is a good estimate for motor loads) / the number of hours the thing runs on a single set of batteries = the current you need to supply (add 33% or more margin to the wall wart rating)

>>2070077
133% overvolting a motor in a consumer product that is probably already overvolted wouldn't be my first choice of strategy
buit if OP had a buck converter board on hand, that'd be bretty good
>>
>>2070079
It's more likely a 12V motor undervolted at 9V.
>>
File: file.png (323 KB, 612x612)
323 KB
323 KB PNG
>>2070079
>>2070077
>>2070078
problem with this thing is that its a cheesy 70s product so everything is just plastic frames with melted plastic rivits so I'd have to destroy it to get the motor out
i tried reading the label and it says ----buchi motor
also looks like pic related which i guess could be a 9 or 12v motor
>>
I'm currently reading through "Make: More electronics" and I've came across the LM386 power amplifier. It looks like a regular op amp but it can output more power. Am I right in believing that it's just another opamp configuration? (similar to how you can get low noise, transconductance etc.)

Also, the datasheets for my regular opamps (lm358p, tlo72 etc.) don't have power output but have power dissipation in their characteristic tables. That's the amount of power lost through heat, right? Why would they include that value but now power output?
>>
File: 1597137513744.png (48 KB, 416x283)
48 KB
48 KB PNG
>>2070109
>Am I right in believing that it's just another opamp configuration
not exactly, it has internal feedback/gain resistors and other shit, which can make it behave unlike an op amp. see the schematic
>That's the amount of power lost through heat, right?
yes
>Why would they include that value
because power dissipation is important for thermal design and load planning. but you need to figure out actual dissipation yourself based on the intended load impedance, the load average power, and the supply voltages using the traditions of Ohm and Kirchhoff
>but now power output?
because it isn't rated for power output, because that's not a relevant figure of merit in the intended applications

>>2070105
ah, Mabuchi
I'd stick to the intended voltage then. you don't want to
>getting this mad about eggs
flying everywhere
https://www.amazon.com/PHC-Power-Adapter-Regulated-Supply/dp/B006QZ7RG8 should be more than enough, if you don't have anything in the wall wart box you can use
>>
Are these legit? Are they actually zero crossing?
>>
>>2070166
yeah im pretty sure thoes have a little light in them and a photosensor on the other side. unless current can jump between photons they should be fine
>>
>>2070155
neat thanks. its stupid gadget but an 8 pack of C batteries is like $20 here so a wall plug would save me some money
https://www.youtube.com/watch?v=GdonmCgg3lE
>>
>>2070166
>Are they actually zero crossing?
It would be a world first to find an honest ali listing.
>>
Let's say I have 310VDC resistive load. And I want to adjust power of it.
So I get IGBT to control it. But. IGBT creates 30W of heat at 100%. Can I use 220VAC relay in parallel with IGBT?

Like, when I need 100% power, turn IGBT on, and then turn relay on? And when I need to turn it off, just turn off relay first, and then IGBT?
>>
>>2070182
Pic related.
>>
>>2070182
>310VDC
ok, how much current does it draw? you're trying to tell me it pulls 50A?
>>
>>2070219
No more than 16A.
>>
>>2070170
I think you misunderstood. Zero crossing means they turn on after the mains voltage crosses 0 volts. That means a softer current peak, since it starts from 0 and ramps up to 230v, instead of just connecting it randomly and potentially landing on the peaks of the sine wave, instantly connecting the load to said voltage and creating a brief but large current peak.
https://hackaday.com/2017/09/26/an-introduction-to-solid-state-relays/
>>
>>2070222
30W loss = 99%+ efficiency, that's unironically not too bad
have you considered a MOSFET instead? FCP190N60 is suitable for your application, and dissipates only about 3W under the application conditions. costs about as much as a relay, but requires zero fucking around
>>
>>2070272
>30W loss = 99%+ efficiency, that's unironically not too bad
30W = heatsink too big.
>have you considered a MOSFET instead?
Yes. First, they are expensive as fuck, second, high voltage ones have high resistance resulting even in more heat
> only about 3W under the application conditions
>Rds 199 mOhm
>0,2 ohms
>P = R* I^2 = 0.2 * 16^2 = too much
>>
>>2070279
fug, I made a rookie error on the math
I mean, you could put the relay in parallel with the IGBT, but then there's no real reason to have the IGBT at all when a simple capacitor would do as much to protect the contacts
>>
>>2070289
Capacitor will kill contacts, because if relay is open circuit, it charges to 310VDC, and once you close relay, it welds contacts.
And IGBT is needed for power control, not just to turn shit on and off, but at full blast, 30W are being generated, and I would rather have those 30W in heater, not some heatsink.
And it is not like I can mount IGBT to heater, because temperatures I aim are too high.
>>
>>2070068
9V wall warts aren't uncommon. I'd go for a 1.5A one just to be on the safer side. You can probably find exactly that on amazon. Note that originally, the ESR of the batteries would probably mean it's running below 9V, so you may get a bit of extra power anyhow when powering it with a lower impedance switcher.

Also if it's particularly old, you may find that the brushes in the motor are almost worn to their limit anyhow.

>>2070166
Read the datasheet

>>2070182
>I want to adjust power of it
Are you PWMing the IGBT for the intermediate power ratings? What kind of speed do you need? If you've got the IGBT PWMing at 90% power, it's still going to be dissipating almost 30W, and can't be replaced with a mechanical relay. Unless you're fine with <1Hz switching times, that is.

Consider looking for an IGBT that will dissipate less power at the same current, or looking for a MOSFET that suits your purpose. Parametric search on digikey should work.
>>
File: 20210329_195545.jpg (2.96 MB, 4032x3024)
2.96 MB
2.96 MB JPG
I've an op amp in a non inverting configuration with gain like in the top picture.
I need to be able to drive a bit more current so I want to use BJTs like in the bottom picture but should I connect the feedback resistor directly to the op amp's output or to the output after the BJTs?
>>
>>2070320
>Are you PWMing the IGBT for the intermediate power ratings?
Yes.
>What kind of speed do you need?
Well. Faster than relay.
>If you've got the IGBT PWMing at 90% power, it's still going to be dissipating almost 30W, and can't be replaced with a mechanical relay
Yes. But at lower power (lets say 20%) this won't produce much heat (switching losses would be minimal due to relatively low frequency of couple Hz)
But at 100%, 30W is just too much, and I'm afraid I'd need active cooling, and I don't want active cooling
>>
>>2069275
>don't touch both ends of the chord at the same time?
what difference would that make since you specifically mentioned a saltwater pool so it is presumably a case of water resistance being extremely low? then this whole setup is as safe as touching the chord with both hands, right? probably even less so since is not resistance of the skin, it is resistance of every part of your body that you need to consider since currently would be flowing through your entire body. you measure resistance between two points, right. so in this case you would have to measure resistance between every point where current flows in and where it flows out of your body. for example your balls would be floating by themselves separate from your body, so you would have to consider the ball to ball resistance, etc.
>>
File: 1610842336675.png (20 KB, 374x286)
20 KB
20 KB PNG
>>2070334
which output do you want to be a well-behaved function of the input?
op amps try to solve continuous equations by driving a difference function to 0. the opamp will try to correct for any (monotonic) distortions you put in its way. a simple class B driver stage is no problem but will probably show bad crossover distortion. you probably should add some little bias to your power stage like the resistor-diode string in pic (but without the cap)
note, non-inverting opamps are not necessarily stable at low programmed gains (<10 or so). extra gain stages only expose and worsen the problem. if that opamp is much faster than an LM358 you might want to cut high-frequency gain by placing a few tens of pF from opamp output to opamp - input, and/or increase the programmed loop gain and attenuating the signal input
>>
>>2070334
Connect to after the BJTs. As the other anon says, you'll get some crossover distortion, but it won't actually be that much since it's operating as a closed feedback loop. The output voltage of the op-amp will just be jumping suddenly to keep up. Naturally, a higher slew-rate op-amp would be better for eliminating the crossover distortion.
Can't speak for stability issues though.

>>2070337
Well either you put a hard-limit on duty-cycle with the IGBT to keep it below 10W or something, or you have multiple IGBTs, switching in unison or alternately, to spread the load. Unison might have thermal runaway issues.

Have you done the thermal calculations yet? 30W isn't too too much. With a fan the whole thing will be smaller, but you may have reasons for avoiding active cooling.
>>
Why do we use 358's when class AB's are so easy to make, and have less constraints on input range?
>>
>>2070398
>when class AB's are so easy to make, and have less constraints on input range?
because, no matter how easy it is to make a class AB, it's easier to follow an opamp recipe. also, literally every reason you didn't mention:
lots of gain
excellent linearity (result of lots of gain)
easier-to-use than a discrete design
unfussy
easily reproducible
versatile (filtering, dynamic range compression, etc.)
power-efficient
cheap as chips
>>
>>2070407
I meant 386, fuck. 8ohm audio drivers
>>
>>2070420

one reason to not use the 386 is that it's rated for 0.7W but in reality, you have to limit the output power to 0.125W to get a proper distortion figure. at 0.7W its fucking 10% which is unbearable. 1/8W is just too low for most purposes.
>>
File: 1598496968620.png (220 KB, 1062x1375)
220 KB
220 KB PNG
>>2070420
ah
because LM386 are even easier to apply than an op amp
>preconfigured with a voltage gain of 20
>inputs are referenced to V- for easy usage
but there are better, just not for 6.5c each

>>2070479
>fucking 10% which is unbearable
oh well, I guess I'll have to use something else. TPA620x looks really easy to apply
>>
File: s-l1600.jpg (231 KB, 1600x1200)
231 KB
231 KB JPG
Who's the manufacturer of these transistors?
>>
>>2070567
some sweatshop in asia. children as young as ten work 12 hour days doping silicon.
>>
>>2069559
it was my init code after all. ilitek drivers have a ton of confusing settings. i take back my racism against based xi.
>>
>>2070580
I was here when you said Xi had a microscopic purple wiener with seamonkey size nuts.
>>
>>2070567
8/10 nice mold flash
>>
>>2068391
Better question would be: are you having thermal issues?
>>
>>2070420
Yeah the 386 is ass. Unironically use a PAM8403 monolithic filterless class-D amplifier module instead.

>>2070492
Looks like those TPA620X are flexible with regards to inputs being both differential or single-ended. Neat.
>>
File: pin assignments problem?.jpg (207 KB, 1062x1375)
207 KB
207 KB JPG
>>2070578
does /b/ know this?

>>2070625
>PAM8403
intredasting, wants large and many caps
>When the PAM8403 works with LC filters, it should be connected with the speaker before it's powered on, otherwise it will be damaged easily.
>When selecting a ferrite bead, choose one with high impedance at high frequencies, and low impedance at low frequencies.
the datasheet had to have been written by an /ohm/boi
>>
>>2070728
>wants large and many caps
Lmao the $1 ebay modules don't bother with the 470µ. All those 1µs are just going to be ceramics.
>filter
From what I've seen, LC filters aren't terribly necessary. The ebay modules don't use any, but I imagine when they do there's a risk of inductive spikes if the load is disconnected. Note that they're only rated up to 5V or so, for higher voltages there's a PAM8610 that's also rated up to 10W.

I want to design a prototyping PCB where the 8403 module just slots into socket headers, but I can't find the dimensions of the pin lineouts, so I'll just have to wait for it to arrive and measure it.
>>
File: wg1608board22.jpg (546 KB, 2460x1634)
546 KB
546 KB JPG
Anyone know the name of the type of heat sinking going on for these three chip areas on my router? Looks like a shield that has tabs you bend under the the board. I see exposed perimeter of copper where the board sits and assume heat is transferred through the copper and into the shield?
I notice the shields are able to be wiggled (the finned blocks arent loose, the shields are), trying to see if this is normal as I assumed the interior of the shield/heatsink would have a thermal pad to mesh it with the board. Are those finned alloy sinks useless if the shield isn't thermally interfaced to the board?
>>2070622
been away a couple days. I see the modem says it reaches 68C during use and the max is 70C according to hardware sheet. I removed the cover and put small fan on it for now keeping it around 50ish. It's operating in a hot uninsulated cuck shed with 80-90F+ weather daily. Wanted to address it sooner than later. Will be placing a large heatsink on it and a thermal pad below it. also checked.
>>
>>2070824
engineering kludge
>Looks like a shield that has tabs you bend under the the board.
it's a "can". it's primarily for RF shielding, not thermal management, but may have thermal functions good or bad
the finned parts look almost like a standard raspberry pi heat sink kit lol
>useless if the shield isn't
they aren't entirely useless. but I bet the effects were well characterized in the lab and carefully approved by lifecycle engineering. or the engineer gave zero fucks
>around 50ish
that's not bad
airflow might be the best win for that installation. adding heatsink only decreases the thermal resistance of one component of the thermal circuit, airflow reinforces many at once
>>
>>2070824
>Are those finned alloy sinks useless if the shield isn't thermally interfaced to the board?

they're thermally interfaced to the air, which is what 99% of heat sinks do. the chips already send heat into the PCB, which is not where you want it.
>>
>>2070824
It's kinda fucky, looks like there might be a couple mm between the top of the can and the top of the IC. Are there thermal pads on the inside of the cans? Even if they are, those are some big heat-sinks that probably aren't the limiting thermal resistance and hence are overly large.
As the other anon said, those look like RF shielding cans, though usually those would be soldered for the purpose of getting better electrical connectivity. In the case that they need both RF shielding and heat-sinking, this technique makes some sense, but I do wonder if there are better solutions.

Also what they've done here can mean they can support a heavier heat-sink than just gluing atop the IC would provide, but A: you can buy proper PCB-mount heat-sinks with mounting holes specifically for this, and B: the aforementioned high thermal resistance from IC to can would be a significant bottleneck.
>>
my MAX7219 LED driver is outputting ~4V on a ~4.5V supply. Since the output pins are supposed to be ~0.3V, is it likely that the IC is bust? Currently rechecking my soldering but knowing that it's a broken chip would speed up troubleshooting.
>>
File: 61BDwvvCBWL._AC_SL1096_.jpg (57 KB, 874x1043)
57 KB
57 KB JPG
Why the fuck are these things over $100? Am I really going to have to waste a weekend setting up a raspberry pi just to avoid getting jewed by the GPIB cabal?
>>
>>2070937
I mean it's an adaptor to a very old bus. I'm not really surprised, the market is probably pretty small so they have to recoupe costs somehow.
>>
>>2070937
>GPIB
Looks to be 8 bidirectional data pins, a bunch of grounds, and some signalling pins like you get in extended RS232. Almost certainly a similarly simple interface as RS232 too, so very easy to program. Assuming you can find the connector, that is.
What software does it need to interface with? It's gotta either use a custom driver, or use something standard like a USB-to-serial class. If it's the former, you might have some trouble, but it looks like there might be some somewhat standard drivers for it out there, hopefully with the technical documentation required to match it on the MCU's end.

>raspberry pi
A microcontroller is the best option, unless you meant using the raspi as a programming computer directly with its GPIOs. Use an MCU with native USB, so probably a lower-end STM32, maybe an ATMega32U4 or one of those USB PICs if you can't into STM.

Also there are cheaper ones than that particular model on eBay, though I can't say what the difference between them is.
>>
>>2070937
>>2070948
oh and there's at least one instructables article of someone making their own
>>
>>2070933
0.3V when set low? or do you mean 4.5-0.3?
>>
>>2070933
all the outputs are multiplexed, no? If you just stick a meter on it you'll just get some average-ish value, you need to use a scope or something to see if it's pulsing.
>>
File: 1614524126590.jpg (45 KB, 1024x622)
45 KB
45 KB JPG
>>2068247
How I can make a DC power supply with the following output 100kv 10mA
>>
>>2071039
>100kV
>10mA
That's 1kW of power. Are you trying to kill yourself by any chance? Anyway, take a microwave oven transformer, rewire it to output 10kV p-p, then use a 10-stage Cockroft Walton voltage multiplier. Use components rated for at least 12kV: you can find rectifier diodes for microwave ovens for that value and ceramic caps rated for 20kV easily, they're not too expensive either. Put everything in an acrylic tube or a plastic box and fill it with oil (vegetable oil or mineral oil, haven't looked into this yet), but be mindful to support it along its length so it stays in the middle of the tube. That's my plan anyway for a 50kV sub 1mA supply anyway, I will use a much smaller transformer but the rest should be pretty much the same, caps and diodes are on their way, but I now realized that I need to make a HV probe to test all this so it will be a while before I report back as I'm still researching the subject.
>>
>>2071067
no I want to make my EBL system
>>
>>2071069
Good luck anyway. I'm the anon that wants to make an electrostatic precipitator, I'm not knowledgeable in any way shape or form and this is what I could gather from a bit of research online, so take my post with a grain of salt. If anyone says anything, trust them rather than me.
>>
File: 1563214592762.gif (493 KB, 480x336)
493 KB
493 KB GIF
>>2071067
>Cockroft
>>
>>2071067
>mineral oil
You want mineral oil for submerged electronics.
>>
>>2071082
Not me. No way.
>>
>>2070948
>Almost certainly a similarly simple interface as RS232
lol no. it's a multi-master network with multicast capabilities. nested state machines everywhere. read a fuckin' book
>>
Can't find a more appropriate general for my question, so here it goes:
I want to put my 3D printer in a (heated) enclosure and thus have to move the PSU and control board away from the frame. I want to use connectors to go through the enclosure.

For the signals and stepmotors I can just use a D-SUB or something, but what do I use for the heated bed which draws a lot of current, something automotive perhaps? Who's got suggestions? I don't want to use something that can be confused with AC.
>>
Why is youtube so filled with unwatchable indian videos on anything electronics related? It didn't used to be this bad.
>>
>>2071175
That's how jewgle hires h1-b poos now.
>>
>>2071175
And why do they always have bad (dangerous) practices?
>>
>>2071160
RC connectors like the XT60 are designed for low voltage, high current and used often for heated beds.
>>
>>2071175
If you're talking theroetical stuff, then it's been that way for many years.
>>
>>2071192
We should talk about the SCAT-1 supersatellite more often.
>>
>>2071192
Mostly that, I was looking up Ebers-moll and I couldn't find a single video that wasn't indian. Maybe these are the only videos that exist on the topic, but I doubt that.
>>
>>2071183
>XT60
Heyy that looks alright, can find chassis versions of it too. Thanks.
>>
Super noob here, i have a 12v 0.3a brushless dc fan and i power it with 5v usb source. The fan runs quite slow because of the low voltage. Is there any cheap and simple way to increase the voltage to make the fan spin faster?
>>
>>2071160
the word you're looking for is "bulkhead connectors"
XT60 is good for the heated bed. for the rest, Creality and some others use plain old dual row IDCs and ribbon cable, and a breakout board on the frame to individual switches/motors
DB isn't bad but you will want a few pins in parallel for the nozzle heater

>>2071224
look up XL6009 modules on the online flea market of your choice. but beware, USB ports are only designed for 2.5W output, and may not have enough power to run a 3.5W fan. you will need to select your power source carefully
>>
>>2071235
>bulkhead connectors
Ah yes that gives me a some examples of things I could use. Thanks.
>>
>>2071098
Oh I see, it uses the bus management pins to tell it when it's sending an address
>>
>>2071183
>>2071210
>rated to 60A
It's likely pretty overkill. I'd recommend an XT30 instead. Or one of those T connectors.
If you do get XT60s or XT30s, get name-brand Amass ones.
>>
>>2071268
If I go with XT I think I will use a XT60 for the bed and a XT30 for the nozzle, just so I can't swap them by mistake. Since I can't find these with 4 poles.
>>
>>2071265
yes. pretty nicely done for the area in which it was intended to operate. too bad Commodore's serial implementation was beset by buggy shift registers and unimaginative workarounds
>>
>>2071271
There are ones with 3 poles, in case you want to share a common rail.
>>
>>2071286
Nah I don't want to make assumptions, for all I know one of them is switched on the negative end with a transistor.
>>
>>2071288
>for all I know one of them is switched on the negative end with a transistor
If they both are, then you can have the positive rail common instead. If one is high-side and the other is low-side, you don't have a choice but use 4 wires, of course.

Though the 3-way connectors can only be plugged in one way, it still might be possible to short the wrong leads together momentarily if you plug it in wrong, so it's probably a good thing to use seperate connectors.
>>
>>2071286
depending on the printer they might not have a common rail. also, heated beds are among the things that typically gets upgraded when refitting a machine for ABS, and some of the upgrades use mains ac to power the heater. better to keep that wiring entirely separate
>>
>>
>>2071330
>>
>>2071341
Hm... Do I really need hot air?
>>
>>2071343
Ok, reballing with iron is impossible...
>>
>>2071351
you really did not think this through, did you
>>
>>2071069
>electron beam lithography
I’m assuming you have the ability to make a vacuum chamber with an insulated heat-proof holder for your thermionic element? Could probably break open a vacuum tube or incandescent bulb and just use JB weld where it isn’t as hot, assuming JB weld doesn’t offgas enough to drop the mean-free-path of your electrons below the length of the chamber. You’ll also need some high-voltage insulators for your shaping electrodes, though external shaping coils may be a better option.
Ideally you’d have the ability to make custom glass structures.

What kind of wavelengths do you want to get down to? I forget the equation, but the wavelength will be inversely proportional to momentum, and momentum will probably be proportional to V (maybe V^2). Then, from what I remember about Gaussian optics, the waist width of a focused beam depends on the angle of convergence, and that’s what will enable you to get focus spots smaller than the wavelength. IIRC they currently use UV at 200nm or something and focus the beam down to tens of nm.

Imagine if we had an optics general. DIY laser tables would be quite the sight.

>>2071351
How did you manage to get it into your head that you could?
>>
File: 1612571184347.png (17 KB, 220x324)
17 KB
17 KB PNG
>>2071466
>and that’s what will enable you to get focus spots smaller than the wavelength
that, and clever mask distortion
>>
>>2071354
>>2071466
Not really. I thought this would be easy:
https://youtu.be/ZP_upYsuwmc?t=224 (3:44)
But in reality this dude is just a fucking wizard. I mean, compare pads on Sandy Bridge CPU and that whatever mobile chip thingy.
I'd need definitely leaded balls and definitely a stencil. Luckily I can get all stencils for CPUs I want to swap.

And hot air. Because I can't risk damaging non-garbage boards with flames.
>>
>>2071496
are you german? i read your posts like im watching marcoreps
>>
>>2071521
No. But I watch germans, and deal with them every day
>>
Let’s say that for any electronics project, there’s a variety of different ways to approach it. If input and/or output are analogue, then you could have either digital in the middle or analog. All analogue segments can be either transistors, discrete op-amps and comparators, or dedicated ICs. All digital segments can be made from transistors, discrete logic, dedicated ICs, or an MCU/FPGA/etc. Then there’s the option of using “obsolete” tech like vacuum tubes or delay line memory.

Wouldn’t it be neat if we had a project each thread, with a bunch of different ways to tackle it, and we’d all provide our own solutions? Would be a good way to learn from one another.
>>
>>2071604

sounds like communism, comrade.
>>
>>2071469
I assume that's some sort of diffractive optical effect?

>>2071496
What kind of tip were you using? What about flux? I think it might be feasible to just flush the whole thing with flux, add some solder to your tip, and just run it over a bunch of balls at once.
>>
>>2071496
I've been soldering for multiple decades and I am poggers

>>2071658
in that other image, blue is original mask. green is the mask with resolution enhancement applied to compensate for edge effects, penetration effects, malfocus, etc. red is actual. they also make slits alongside tracks to increase the effective depth of field and do some other fun tricks

>>2071635
just think of the amazing codex of science we could be reading right now on the chinese dime if we knew mandarin
>>
>>2071658
>What kind of tip were you using?
Comparable with cpu pad size. At 400c since cpu is quite heavy mass.
>Flux
Pure rosin. When desoldering - rosin alcohol, because this is what I had.
When I will do it for real, I will get chipquick no clean or mg chemical no clean or topnik zel. Or nordson flux plus.
>add some solder to your tip, and just run it over a bunch of balls at once
This will collect solder from cpu. Balls have excess solder. Which is why all stencil madness exists
>>
>>2071758
>Balls have excess solder
Relative to the tip, that is. The surface tension forces of a large ball of solder make them have a higher affinity to the flat surface of your iron. If your tip has a large surplus, and has a geometry to suit, you might have some luck. Might just be impractical to achieve with any real tip, however.
>>
>>2071767
this is literally the only use for the conical tip
>>
>>2071758
>I will get chipquick no clean or mg chemical no clean or topnik zel. Or nordson flux plus.
or Kingbo RMA-218
>>
>>2068247
When I am soldering and I touch the solder (or any other conductive thing) with my hands while I touch it with the soldering iron, I get small electrical shocks. This shouldn't be happening, right?

Anyone has any suggestions on what to do?
>>
>>2071890
>or Kingbo RMA-218
Nah. No chink fluxes are allowed under BGA. I trust rosin more in this application.
Who knows, what if it corrodes and conducts?
>>2071767
Idk. Doing what that dude did is kinda impossible. I tried conical pointy tip (which I don't like in general) and I should try using knife tip, but idk... And no flux at all.
But in general, I think I would be better off buying overpriced piece of stainless and overpriced tin-lead balls... But I still hope I would not have to reball a CPU, and I hope it installs normally on stock lead-free balls.
>>
>>2071890
>activated flux under BGA
lurk moar. a lot moar

>>2071958
>leaving extra solder on the lands is kinda impossible
you don't solder SMT much. I do it all the time even when I don't want to
>>
>>2071975
>you don't solder SMT much. I do it all the time even when I don't want to
Yes, but CPU has a bare tiny ass pad with no components. Leaving extra solder so you can later heat and form balls is kinda hard.
>lurk moar. a lot moar
I'm pretty sure all no-cleans have activators, but manufacturers of those probably made sure they decompose to non-conductive non-corrosive shit.
Which is why I will never trust chink flux. Hell, I'd rather take chunk of rosin and put it under BGA. Will double as compound XD
>>
>>2071985
>Leaving extra solder so you can later heat and form balls is kinda hard.
that I'll agree with. but sheeit, solder bridges are normal when hand soldering. the trick is in how you pull away from the pad
>I'm pretty sure all no-cleans have activators, but manufacturers of those probably made sure they decompose to non-conductive non-corrosive shit.
in other words, they deactivate upon heating. that's fine. plain rosin wasn't activated with strong acid in the first place, therefore pretty safe (but since rosin activity increases with heat, I wouldn't leave it on)
>Which is why I will never trust chink flux.
eh, fake american flux from aliexpress is usually a bad buy, and it's kinda silly to buy chemicals at a boot sale anyway, or really, to expect to buy anything first quality at a boot sale
>will double as compound
kek
citric acid is a highly aggressive water-soluble flux. just clean it very thoroughly from the work when done, which can be tricky under low-clearance components
>>
I have a shop downtown in my city. We have a garage. Faggots keep stealing stuff out of our garage. Last week a couple of punk teens came in and stole some spray paint. I saw them do it so I chased them down and got it back. But yesterday, some bum came in and raided our ash tray. I don't care that they basically took my trash but I do care that they trespassed.
We have cameras but it seems like people ignore them. So I want to make some sort of thing that goes off if someone enters our garage.
I'm not exactly sure what to do. I was thinking of using the same type of sensor that triggers a garage door from going down (I think it's a laser) and then have that hooked up to a flashing red light. I want it to go off for maybe 30 seconds.
Is there an easy or straightforward way of achieving this?
>>
>>2072136
no, there isn't, because bums and delinquents are creative animals that learn
if you want easy and straightforward, use a PIR sensor and a relay to turn on the indicator of your choice, and maybe add a key-operated switch to disarm the whole apparatus while you're in there
>>
>>2072136
Just have the same tripwire rigged to a shotgun
>>
>>2072136
why do you leave it open when it's not being supervised?
>>
File: 20210331_192116.jpg (3.32 MB, 4032x3024)
3.32 MB
3.32 MB JPG
I've a unit I'm trying to test that's supposed to do power factor correction for three phase systems.
You connect the three phase voltages into it and then there's three current transformers.
It's supposed to then switch banks of capacitors on and off in order to correct the power factor.
In order to not have to deal with actual three phase mains and have to get high power components, I've built a low voltage three phase circuit that uses a wien bridge oscillator at 50Hz that then gets buffered and split into three phases 120 degrees apart. I then have them put through class B transistor amplifiers so I can get enough current out of them (There's a small bit of distortion from the B class amps but not too bad).
Anyway, I have a load set up as shown, the three phases (they're about 12V RMS phase to phase) are passed through the current transformers (The transformers have a ratio of 1 to 360 so I get readings of hundreds of amps even with small current) and then they're connected to a resistive load in a delta shape.
When I do this and check the unit, it gives me a power factor of 1 which is fine.
However, if I change the switches so that the inductors are in series with the resistors, I still get a power factor of one.
Would I be better switching to a purely inductive circuit to make the power factor change?
Given that my frequency is only 50Hz, I assume I need a fairly large inductance to have a decent effect?
I'm currently using 150 ohm resistors and 1 miliHenry inductors
>>
>>2072216
>X_L = s*L = 2πf*L*j
>100πj * 1E-3 = 0.3142jΩ
>0.12° angle
>cos(0.12°) = 0.999998
Looks normal to me.

Try with capacitors for now to confirm that your power factor measurements are correct. 100µF and 10µF give 32Ω and 320Ω respectively. Make sure they're non-polar!

Note that reactive loads might cause distortion out of the amplifiers, so keep an eye on those waveforms to ensure they're not interfering with your measurement of power factor. Is the "power factor correction unit" also being used by you to calculate power factor?
>>
>>2072241
I actually have a delta arrangement of caps set up that the unit should switch in to correct the power factor, I've 3 banks of 10 22uF ones in parallel so about 220uF on each bank.
Yeah, I suppose it's much easier to change the power factor with capacitors at the low frequency, I just had a look and any inductors over 100mH with more than a couple of miliamps rating are hard to come by
>>
>>2072254
Best shot at inductors is probably to find some mains transformers.
>>
>>2072271
Yeah, I'll see what I can get my hands on in work tomorrow.
>>
>debugging wonky phone line
>isolate affected line
>clamp ohmmeter on
>watch reading swing wildly between open, dead short, and everything between
>account for both ends of the wire, visually inspect its whole length, find no obvious faults.
>resistance still fluctuating as if somebody was scrubbing a wire brush over the patch panel
I need an old priest and a young priest.
Why do phone lines always have the weirdest failure modes?
>>
File: 1596156145147.jpg (53 KB, 600x591)
53 KB
53 KB JPG
What is /ohm/ planning for April Fools?
>>
>>2072254
aren't you worried about contact arcing damage?

>>2072413
>I need an old priest and a young priest.
did you look in the confessional, you may find them both
>Why do phone lines always have the weirdest failure modes?
that's a very large antenna you've got there, fren. are you sure you're reading the cable and not the local AM radio station?

>>2072418
I knew someone who got the brilliant idea of putting several hundred ms of digital lag on a light switch to simulate tens of thousands of km worth of cable propagation delay. this was 20 years ago, before IoT lighting made that normal
>>
>>2072445
>did you look in the confessional, you may find them both
kek
>>2072445
>that's a very large antenna you've got there, fren. are you sure you're reading the cable and not the local AM radio station?
Completely sure. The short circuits are very much real. This phone line wasn't actually being used as a phone line, I was using one pair as a convenient way to distribute a low voltage.
The other pair in the cable was just fine.

I have actually seen a similar failure mode before (on an actual operating phone line), in a house that was struck by lightning. I opened and checked every junction box in that place, the shorts just kept coming and going.
Some kind of damage along the cable meeting a very specific coincidence.
>>
>>2072464
Copper oxide can act like a semiconductor, they used to make diodes out of it with a bunch of blued copper shavings in a small tin. Chances are you’ve got a bit of copper oxide making a pressure-dependant resistor that fucks all over the place when you move the cable by 1mm.
>>
Hm.
AMD has two sockets/BGA pinout/whatever.
FP5 and FP6. FP5 is older, FP6 is newer... But. They have same amount of pins and it looks like pins are in the same place... How do I know if they are electronically compatible with each other without buying and testing and potentially damaging nice cpu and motherboard?
>>
>>2072600
The mobo should have it listed in its service/assembly manual. Assuming you can get it. The CPU will have it in its datasheet, but this only works if you know exactly what CPU was on it in the first place.
>>
File: Untitled.png (1.57 MB, 1500x3924)
1.57 MB
1.57 MB PNG
>>2072625
>The mobo should have it listed in its service/assembly manual. The CPU will have it in its datasheet, but this only works if you know exactly what CPU was on it in the first place.
Well, good luck getting board view or/and schematics for newer (2021 and 2020) laptops that use FP6 chips. Also I don't think AMD wants people to know their pinout much, because all I've found were those websites: https://en.wikichip.org/wiki/amd/packages/fp6 and https://en.wikichip.org/wiki/amd/packages/fp5 .

I also don't understand compatibility of older CPUs. Like let's say 2xxx and 3xxx series. They are same ""socket"" but some motherboards have specific code name printed on them and I checked BIOS, and there is no older CPU CPUID in there to be found. Which confused me, because I didn't find spec difference between 2700U and 3700U that sufficient to pay chinks twice as much for CPU, that might be even faulty since nobody knows where do they come from.
Worst part is that I can't find schematics on those boards to check and find the difference. Not even for shekels from some place like vinafix.

But then again... Those are SoCs, nothing smart is left on motherboard, no north bridge, no south, this CPU is like cloaca, does nearly everything.

Also, how often do companies use same package, but different pinouts? Because, unlike Intel, AMD seems to keep sockets on desktops long enough, so what prevents them from doing the same thing for laptops? Idk.
>>
>>2072665
This is block diagram of one of AMD laptops I have. Like, I don't see here anything that would really care much about new CPU. Maybe except GPU, but hehe, i dont have one.

Idk, I just have 10/10 /b/ tier business idea, open laptop CPU update shop in my shithole and swim in money, because me as a poorfag student would definitely swap shit dual core CPU or quad core with meh integrated graphics to something 8 core... with decent internal graphics... and same TDP and thus battery life.
Idk. Really confused. Should probably buy broken motherboards and check stuff.
>>
>>2072668
oh, and did I mention it be cheaper buying some shit amd athlon laptop, and then soldering $200 CPU than buying laptop with said CPU? Seems too good to be true, but idk, if shit works, it is based... And yes, imagine still gettin 1366x768... JUST fuck marketing dept.
>>
>>2072672
>And yes, imagine still gettin 1366x768...
that's what the external displayS are for
>>
>>2072673
This is where 1080p LCD update comes into place. On most laptops u don't even need to disassemble much... If you're OK with getting 19V on CPU, because eDP isn't isolated with IC like HDMI from CPU...

Hm... I found cheapo motherboards with dual core shits. Do they have new gen dual core shit?
Anyway, why did they stop socketing CPUs on laptops?
>>
>>2072677
I don't recall ever seeing sockets on a laptop, not in the past 20 years anyway
>Why
"slimness", cost, effort, design-for-manufacture, the same reasons they stopped socketing everything else
>>
>>2072672
>it be cheaper buying some shit amd athlon laptop, and then soldering $200 CPU than buying laptop with said CPU
only if you succeed. if you fail due to obvious lack of experience and knowledge you've wasted the cost of both the laptop and CPU.
>>
>>2072687
Yes. This is not gonna stop me.
Something suggests me I should get a trashed laptop first. Fix it, and then try this.
Idk, this is more of a challenge than anything practical.

What are most common bga fuck ups?
Because so far with my cave man tools (soddering iron and blowtorch) I've only managed to bubble one PCH on some mobo and then heat warped the mobo... Because bottom heat is needed... But then I think I successfully removed cpu. I wish I could test it somehow. With a fucking hot glue instead of vacuum tweezers.....
>>
>>2072694
And yes, I won't be soldering real board with blowtorch. Obviously.
>>
>>2072145
That's a good idea. Maybe I could use one of those motion sensor switches for bathroom lights. I'll look in to it his, doesn't sound too hard.

>>2072148
I considered it. The next best thing is booby trapping some cigarettes. Nothing too harmful, just one of those round balls from the inside of a green crackling ball firework stuffed down into the cigarette. Bought a pack of Newports specifically for this. They used to sell prank exploding pieces ade specifically for cigarettes but the crackling ball pieces should work just as well.

>>2072203
The teens came in at the end of the day when we were packing up to leave. I opened the garage and was topping off my coffee before leaving. My wife said there was someone in the garage. I saw three teens walking by the front of our shop. I went in the garage to see what they had done and noticed two cans of spray paint gone. I went out and saw them holding the cans, so I ran down to them and took the cans back. Fucking faggot teens.
The ashtray got raided early in the morning. I was out smoking and a custie interrupted me, so I went in the shop for a few minutes to cut some keys for them. When I went back, the ashtray had been moved to near the garage door. Bastards must have been scoping it out because the tray is out of view. The door was also only halfway open. I got both the teens and the bum on camera but I don't think there is anything I can do about it. I really don't think the cops will give two shits about missing trash and I got my spray paint back.
>>
Top one is a DIN but what is the bottom connector? Something proprietary?
>>
>>2072867
What's it connected to, mains? If so it may be a connector used before the IEC connectors were used. Though judging by the colour of that FR4 the PCB isn't that old at all.
Not sure where is the best place to look for connectors like that, besides filtering by number of pins and current and voltage rating (maybe also dimensions) in digikey or wherever and hoping they have them.
>>
>>2068893
Put a cap across it, or the gain varies with your analog signals
>>
>>2072871
It's a DC 40v out, sorta like a molex. Reminds me of the connectors on a PSU but 2.5 times bigger
>>
Anyone have any resources or recommendations on a heater plate for smd soldering? I've seen Marco Reps using an MHP30 but I'm not sure if there are better ones for similar or cheaper price/if it's cheaper to build one myself.
Any tips/advice?
>>
I've put together a basic audio amplifier circuit that pre-amplifies an audio signal from an electret microphone and then amplifies the signal's power. However, the only speakers I have to test it is a small ~50 ohm speaker and my headphones. I've always used these so I don't notice the difference that the higher wattage provides; What's the point of increasing the signal's power? Does it increase amplitude / reduce signal clipping?
>>
File: 1617286797861.jpg (1.23 MB, 2573x1707)
1.23 MB
1.23 MB JPG
I want to hook up two TMC2130 V1.0 stepper drivers to arduino. How do I do that?
>>
>>2073002
with jumper wires
>>
>>2073002
depends on the library you're using to control it. Look up TMC2130 arduino. The library will probably have an example program.
>>
>>2068247
I made a battery pack for a portable drill with three 18650 batteries and one of these Chinese modules.
There are no specifications in how much voltage should I use to charge them, but some places recommend me to not go above 12.8 volts as input.
However, when I use that voltage, my battery pack don't go any higher than 11.8V.
Should I crank up the charging voltage, check my batteries or I'm just using a shitty charging/discharging module?
>>
>>2072418
An AT-TINY plus a 433 MHz module wireless doorbell that controls a 12V car siren.
Do printed PCB antennae work for these things?
>>
>>2073027
So long as its a big conductive loop.
>>
>>2073027
>>
>>2072826
>Bought a pack of Newports specifically for this
Fidel Castro's thread was two editions ago

>>2072997
yes. solenoid excursion is proportional(ish) to force which is proportional to current which is proportional to voltage (neglecting reactive phenomena) which is proportional to the square root of power. more watts = more displacement = more loud

>>2073025
>but some places recommend me to not go above 12.8 volts as input
bruh, that board doesn't have a charging controller on it. it is a safety device only. you MUST use a CC-CV algorithm and controller to charge the battery

>>2073027
433MHz is a bit big. try a 1/4 wave spiral around the board perimeter if you're short on space
>>
>>2073025
You should also be aware that that BMS doesn't have balancing circuits, either. It will protect from overcharge, overdischarge, and overcurrent on any cell, but it won't control the charge, and it won't balance your cells.
>>
>>2073099
>I∝V∝power
Ah, I see. So if power (loudness) is P=IV then the pre-amp focuses on bumping up the V and the main amp focuses on bumping up the I, both of which serve to bump up the P. Got it.
>>
>>2073161
>the pre-amp focuses on bumping up the V and the main amp focuses on bumping up the I
Not intrinsically, a main amp can have some amount of gain also. Like a gain of 20 as seen in the LM386.
When I design audio circuits like this, I'll want to work at a ~1V level since it's well within the ranges of any op-amps even at 5V, and also about right for sending into headphones. So I'll have my preamp amplify to about that, possibly with an AGC if needed. Then the output amp will have a variable gain to tune the volume, getting much closer to the voltage rails of the circuit (say, ±12V).

Look into converting voltage levels to dB, and converting dB into sound pressure. I think dB out of the amplifier and dB into your ears are roughly equivalent, minus some efficiency factor. They're logarithmic, remember.
>>
How does this shit work? The transistors are just replacing resistors in a normal low pass filter right?
>>
>>2073239
moogs are black magic to me, sim it and see
>>
>>2073239
the ladder filter on its side is similar to this thing. It resembles a basic RC multi-pass filter reflected horizontally. Apparently the two lines will produce two synchronized copies of the filtered signal which helps reduce noise. I've never used transistors in my filters though so i don't know how that would work. Something about Vin needing to make Vbase > Vcollector?
>>
>>2073251
>Apparently the two lines will produce two synchronized copies of the filtered signal which helps reduce noise
looks more like a long tailed pair to me, with the two copies being opposite of one another and used for negative feedback
>never used transistors in my filters
same
someone make a moog with op-amps, then i can understand it
>>
>>2073251
So it's similar to π-attenuator but it takes a balanced input and uses reactive components rather than purely resistive ones.
>>
>>2073239
it was replaced by the 303's diode ladder filter, the resister ladder turned out to be pointless. some guy on yt has a good video on it
>>
>>2073259
A filter is pretty much an attenuator anyway. While the attenuator would as act as a voltage divider to reduce the signal amplitude, the capacitive reactance of the RC filter actually filters out the harmonics (i.e. phase shifting)
>>
>>2068772
>is it OK if I take the skirt off and solder a wire in, then isolate with heat shrink tubing?
Only for temporary use.
The purpose of crimping is to give a reliable and robust connection that doesnt let go under stress or vibration.
Soldered joints tend to fail relatively quickly under vibration or load, especially unsupported. The heatshrink would provide some mechanical support, but not much, I wouldnt rely on it.
>>
>>2072874
>Put a cap across it, or the gain varies with your analog signals
That is not the reason to put a capacitor across it.
The reason to put the capacitor across it is because the gain of the circuit is determined by the collector resistor, and the emitter resistor + intrinsic emitter resistance.
However, in the absence of the emitter resistor you would leave you with A) thermal instability (it adds local feedback which stabilizes this), and B) a hard time biasing the circuit.
The negative consequence of adding the emitter resistor is a reduction in gain as mentioned prior. However, by adding a capacitor in parallel you still get the advantage of biasing and temperature stability, but at the frequencies of interest the total emitter resistance gets reduced to the intrinsic emitter resistance, meaning you get the maximum gain possible out of the circuit (it is not reduced by the added emitter resistor).
Note that this is of course only applicable for an AC amplifier, not a DC coupled one, but that is the case in most situations anyways.
>>
File: 543h7k.jpg (82 KB, 500x749)
82 KB
82 KB JPG
Why is Mouser such shit?
>>
>>2073351
Your order got stuck in the Suez canal.
>>
>>2073351
I've never had an issue with them. My experience was that they were fast but expensive.
>>
>>2073351
Waiting for parts is such shit I agree. Just do what I do and re-read Boku Girl.
>>
File: no waitiing.jpg (164 KB, 1065x708)
164 KB
164 KB JPG
>>2073438
>Waiting for parts is such shit I agree

i never wait for anything, coz waiting or not waiting takes the same amount of time.
>>
How much do mask ROM chips cost to get made and what's a typical MOQ? Parallel interface, 2kbit x 8, ≤350ns, TTL voltages?
>>
File: solderinmg.jpg (2.1 MB, 2340x4160)
2.1 MB
2.1 MB JPG
soldering is hard(using just a laser)
>>
>>2073472
>a laser
You're soldering with a laser for heat? I've heard of it before, but haven't seen anyone do so here. What kind of speed do you act with, and is there temperature feedback somehow?
Here am I soldering onto the micro battery connector of my dead phone's ribbon cable for data recovery lmao.
>>
Happy April fools ohmmies
>>
If I don't respond within 15 minutes assume my house burnt down
>>
>>2073499
battery isn't toasty, output is 3V, no suspicious sizzling noise, we're good so far
>>
>>2073504
Got up to 4V, tried turning it on, but the screen displayed a "low battery" indicator and shut off again. Better leave it charging for longer I guess. I'd feel more comfortable leaving it on a fixed voltage regulator after getting up to 4V, otherwise I think the TP4056 may just cut off once it hits 4.2V and leave be in the dark again.
>>
File: it works.jpg (1.78 MB, 4032x3024)
1.78 MB
1.78 MB JPG
let there be life!
>>
>>2073535
I’m proud of you anon
>>
>>2073535
Perfect! It stayed on 13% without issue, and I was able to get off my 2-factor auth shit without a problem. Well it was a bit shitty using the touch screen and front button with it just floating on top of my wiring, and it was pretty warm the whole time since there was probably half an amp constantly being wasted through the battery's parallel resistance, but it held up fine enough.

FYI, I couldn't do it without the battery because the shitty proprietary phone requires the battery's third gas sensor pin connected for it to boot. So I just put the TP4056 in parallel with the whole lot, and it worked.

>>2073542
Thanks, ignore the colour coded folders and masses of mail alerts now please.
>>
File: 1616083633433.jpg (20 KB, 480x480)
20 KB
20 KB JPG
>>2073448
dad....

>>2073458
a lot of foundries do custom requests. it couldn't hurt to call up microchip and ask. expect MOQ in the 4 or 5 figures, it's not worth bothering for much less than a whole wafer. expect them to sell you 28C256 type parts if your application will allow, or maybe offer you custom packaging services if you can't just leave pins dangling off the end

>>2073535
rescue duty do be like that sometimes
>>
>>2073458
Why do you need ROM? Is there some advantage over various PROM technologies, aside from not having to program them? Because plugging into the address pin(s) and burning some fuses or setting some bits with a (HV) microcontroller board is going to be a lot cheaper than getting some chip manufacturer to do it for you.

On that note, what are the best PROM technologies for hobbyists? I'm guessing it's EEPROM, but it's not like there aren't OTP PROMs still being made. Also for some reason these OTP PROMs are called EPROMs, even though they're not erasable:
https://www.ti.com/lit/ds/symlink/bq2022a.pdf
And Mouser doesn't even have a PROM category, just EPROM and EEPROM. And not a quartz window in sight, from a glance, though that isn't surprising.
>>
>>2073553
Price. For my application, OTP and EEPROM chips are at least $2 each in small quantities.
>>
>>2073555
Shit you're right. Check LCSC, maybe they're better.
What do you need it for? Parallel PROM is a bit of an oddball requirement, especially with that kind of intermediate value. Is it for some sort of discrete logic circuit?
>>
>>2073559
yeah the AT24C08 is 10c each on LCSC:
https://datasheet.lcsc.com/szlcsc/1910291306_HGSEMI-AT24C08N_C434604.pdf
But it's not parallel, it's 2-wire serial. If you're using an MCU anyway there's little reason to use parallel, but if you're using an MCU anyway you'd not need an external ROM IC in the first place.
Might actually be cheaper to get a serial ROM IC and a very cheap MCU (think padauk) and program the MCU to be a parallel interface.

Also:
>$2 each in small quantities
Sure as hell not going to beat that with something made custom for you in a foundry. Unless you get free samples by pretending to be from a company.

On a barely related note, I think it would be cool to have some sort of meme circuit where I have a drawer full of BQ2022s with different memories that I just swap for different purposes. Like different DSP effects or something.
>>
>>2073559
I'm looking into cloning an old design for a retrocomputing peripheral that's moderately in demand but is massively price-gouged pretty much everywhere. I was just trying to price the thing out with new parts before I even start seriously working on it, or on a redesign. Anyway, the old design has a 2048x8 bit parallel ROM on board. I don't know if it's for logic or code.
>>
>>2073562
>Might actually be cheaper to get a serial ROM IC and a very cheap MCU (think padauk) and program the MCU to be a parallel interface.
Yeah that's honestly my next step if sourcing parallel ROMs is prohibitive.
>>
>>2073563
>>2073564
I'd just use a cheap 14-pin MCU instead. Or an ≤8-pin one plus a 74HC595/equiv. Write the ROM into the MCU itself, instead of getting an external IC for it, assuming you don't need to swap the ROMs. If you do, those BQ2022s tickle my fantasies.

At least do a price comparison between small cheap MCU + ROM, vs less small cheap MCU with sufficient internal progmem. And then do the high-pin-count MCU compared with the low-pin-count MCU with a latching SIPO shift register. With a spreadsheet. Ensure you're not picking an MCU and/or PROM so obscure that you can't program it easily.
Looks like an ATtiny13 doesn't have enough progmem, but the ATtiny45/85 have 4/8kB of progmem, so will work. For comparison, that is. AVRs aren't necessarily the cheapest way to go, but they're one of the easiest to program.

Also check that all the parts you're evaluating are 5V compatible.
>>
>>2073564
skip the external ROM IC, just use half of the program flash as data flash and sit in a tight loop reading address pins and writing data pins. a Cortex-M0 would be enough
>>
>>2073573
>a Cortex-M0
Are those cheaper than $1 each? Can get a 595+tiny45 for under $1, on LCSC.
>>
What the hell are these? These triangles pointing into one another, and separated semicircles (which seem to correspond to the triangles pointing away from each other)?

If it matters, this is from a serial interface card.
>>
>>2072909
Anyone?
>>
>>2068872
>>2073307
I used to work in an cable manufacturing factory as a technician for crimping machines, do you really need crimping tools when pliers with teeth work just fine?
The machines basically just hammered a tiny point into the middle part of the crimp.
>>
>>2073637
solder jumpers? haven't seen that style footprint before but the rounded symbol looks like it.
>>
>>2073688
Ah that gave me the keywords I needed. The triangles are normally-closed and the semicircles are normally-open. You can use a knife to break the trace on the "bow tie" and solder to bridge the "circle".

Apparently if you toggle them, you can replace a ROM chip with a RAM chip. I don't know what the advantage of that would be given the serial card would presumably no longer work correctly.
>>
File: download.png (52 KB, 851x352)
52 KB
52 KB PNG
I tried asking on /sci/ but they're useless and I didn't think this was worth it's own thread, but I was wondering if there was another way to do the demonstration where a chocolate bar is put into a microwave without a rotating tray, and the radio wavelength the microwave uses shows-up as the melted part of the chocolate bar. Is there another way to do this, with something like a magnetic field viewing film instead of chocolate, and preferably without needing it to be inside a microwave or similar cavity/chamber?
>>
>>2073706
This question was more for /qtddtot/. Anyway the wavelength shows up because in a common cheap microwave oven standing waves form so energy concentrates on the same spots and other just never receive it, hence the spinning plate; more sophisticated ovens instead of spinning a plate actively reflect the waves around so the hotspots won't stay stationary in space. It can be done outside an enclosed cavity, I guess, although you'll likely won't form standing waves, HOWEVER I can't stress enough how dangerous that would be, and I suspect that if you're asking such a question you lack both the knowledge and the understanding to safely handle a live microwave without enclosure. If you are interested in the ondulatory nature of the phenomenon you can use a rope, tie one end somewhere that won't move, and wave the other end with the right pacing until something like pic related appears: where the rope doesn't move is where the chocolate wouldn't melt if the rope was a EM microwave, where it moves the most is where it would melt.
>>
File: download.jpg (39 KB, 764x764)
39 KB
39 KB JPG
>>2073735
Yes. I understood all of that without you needing to post that wall of text. I basically just wanted to know if there was any other way to visualize actual radio waves without needing it to be dangerously powerful or using that dumb rope analogy.
>>
>>2072136
Shoot them.
>>
>>2073592
>≤350ns
even with hardware SPI at (nonexistent) fCK/1 prescaler at 20MHz you're not gonna make it
>>
>>2073742
You could maybe do it with a sound wave in a tube, I think the nodes are a slightly different temperature to the anti-nodes. See Nighthawkinlight's latest video series for more on thermoacoustics.
>>
Is it worth salvaging DC-DC converters from laptops? They have wide input range, and good efficiency.
>>
>>2073742
Then why didn't you ask this directly instead of using that dumb microwave and chocolate analogy? Anyway the magnetic field of an EM wave is eight orders of magnitude smaller than the the associated electric field, so if anything you're going to see the electric field; a wave with an intensity on the order of 1 mW/m^2 (that's the intensity that you have at a distance of 1m from an isotropic radiator of 10mW of power) produces a maximum electric field on the order of 1 N/C and a maximum magnetic field of 10^-8 T (a fridge magnet is in the order of 1 mT and the Earth's magnetic field is 10^-5 T). These are incredibly small numbers, to be fair I don't see how you can practically visualize them.
>>
>>2073817
>Is it worth salvaging DC-DC converters from laptops?
not really, buck regulator IC's are a few bucks in singles from American catalog suppliers. if you go china LCSC it's even cheaper.
>They have wide input range
so do other buck regulators you can buy new and know what you're getting.
>and good efficiency.
same as every other buck regulator out there, around 90-92% max
>>
>>2073819
The inverse square law doesn't exactly apply in a resonant cavity, and even conservation of power doesn't work up-front if it's got a high-Q factor; you need to integrate as the energy in the cavity will accumulate. I think there's some hope of it working.

If nothing else, bolting a magnetron to a metal pipe is hardly too dangerous.
Though I'd lean towards 100MHz or so; it's slow enough that the electronics don't have to be particularly specialised, and small enough that you can have a 1/4 wavelength resonator inside. But you'd need to do testing with an SDR or whatever to ensure that little of it leaks such that there's no possibility of an RF burn. Or of the FCC bashing your door down.
>>
File: solder2.jpg (1.91 MB, 2340x4160)
1.91 MB
1.91 MB JPG
>>2073482
he connector in my pic is the micro battery connector dor the phone im dicking around with.

, no feedback exactly, this is more of a laserwelder that im using to solder. , so i can control the aperture of the laser, time, frequency of shots and volts. everything has to be lowest setting power wise. if you miss, it blows a hold in the fr4, or it makes the ceramic smt components explode, even with nitropgen my joints oxidize p fast. i think i would have better results with solderpast and not using regular fluxcore solder because its kinda big
>>
>>2073862
are you a self replicating robut?
>>2073861
if you're looking for danger, try bolting a magnetron to your ass.
>>
>>2073862
Hope you're not blasting holes through middle layers there
>>
>>2069497
So it's a lifepo4? Use li-ion, you'll get most of it's capacity between 4.2-3.2v.
>>
>>2073817
no
if your time has any money value at all, you can buy dcdc converter ICs and associated discretes for less, and know the good condition of what you're getting
>>
>>2073861
The initial post specifically asked to not use a resonant cavity though. Beside, I was giving a few figures to give a rough an idea of how small the fields generated by EM radiations are. Energy integrates, yes, that's why things melt, but I was thinking more about directly measuring the EM wave. In any case microwave ovens work on the resonant frequency of water molecules, which is how EM energy is converted into heat, so the frequency is slaved to the resonant frequency of the molecule of whatever you want to use to integrate the energy.
>>
What's the best way to characterize /diy/ capacitors and inductors?
>>
File: superior-z-cross.png (52 KB, 480x360)
52 KB
52 KB PNG
>>2073889

fancy name
gold lettering
exorbitant price points
>>
>>2073900
what?
>>
>>2073900
lel, this is a correct answer

>>2073889
>the best way
test equipment
>best
who knows, your requirements and capabilities are invisible to us
the AADE L/C meter is not too ridiculously expensive but good enough for many designers to have stolen for their own LC meter designs, both commercial and community
>>
>>2073879
i was, but not through the copper, once the laser hits the copper, (on the lowest setting) the heat dissipates completely. thinking now, it could be useful to create access to middle layers without blasting the copper. you could probably even do it with a co2 laser
>>2073866
i had just woken up i see what you mean .
>>
>>2073025
4.2x3 is 12.6v and this is what you should supply. What are you using to charge it?
>>
>>2073900
>straight solenoid inductor
I know it’s air-core because that lets you reach higher Q factors and lower hysteresis and such, but isn’t it feasible to make an air-core toroidal inductor? Be handy for RF.

I’d also tend to think that, for audio at least, there are high-speed enough ferrite cores that would be perfectly good without producing distortion. But maybe that’s just audiotards I'm arguing with.
>>
>>2073742
So, not quite a demonstration of a standing radio wave, but I have always found this demonstration kinda interesting. It is a visualization of a standing wave in a wire: https://www.youtube.com/watch?v=M1PgCOTDjvI

the actual wave vizualization starts around 8:38, but the rest of the video is also quite interesting.
>>
help me make electronic whoopie cushion pls sir
>>
>>2074115
Either make a linear bellows-style device, or rotary air-raid siren device. Bellows will have quieter motors but will be led continuous and harder mechanically. Best method might be a relatively silent radial blower fan feeding the end of a balloon or whatever. Some piece of rubber to constrict the airflow and make some noise.
Could also use a sucking fan I guess, perhaps just half a balloon (the half with a hole) held across a vacuum cleaner nozzle will work, but you’ll find it rather noisy.

I prefer the good old days of wiring up someone’s automatic doors to a 1/2-second clock, that you activate remotely.
>>
File: wtf.png (32 KB, 819x243)
32 KB
32 KB PNG
How does backlight work on laptops?
I remember hearing they boost voltage, like 50 volts or so... But this laptop doesn't do this.
Just straight 19V which are 7,4V when on battery...
>>
>>2074158

old laptops used CCFLs which are, essentially, tiny fluorescent lights, and these use high voltage to ionize some gas. newer laptops get light from white LEDs, so no high-voltage needed.
>>
>>2074158
Ok, NVM, modern panels are OK with 7-ish volts
>>
>>2074168
Yes, they have CCFL invernet in bezel, but LEDs...I mean, you can wire them in series, and get 50 odd volts
>>
>>2074170
>invernet
inverter. daaamn.
>>
>>2074168
>old laptops used CCFLs
Pretty sure no CCFLs will run at 50V; television CCFLs run at hundreds if not thousands of volts. I think the 50V backlights we often hear about are just series LEDs to lower the current going through the ribbon cable.
I think I recall Louis Rossman complaining about a 50V line on a macbook being right next to some low voltage line without any protection, and it was a frequent short in the event of water ingress. And I'm pretty sure macbooks have been using LED backlights for a while.

>>2074171
¡ʇǝuɹǝʌuı ǝɥʇ oʇ ǝɯoɔןǝʍ
>>
>>2073771
Then he's gonna have to use a high-speed MCU or an OTP/EEPROM and deal with the price.

On the MCU path, assuming the parallel output bus updates once per 350ns, that's ~2.9MHz. Multiply that by 8 for the simplest serial input and you've got ~23MHz, which is indeed infeasible with an old ATtiny. But if he's shunting the values directly out of PROGMEM and onto 8 pins at once (say, an ATtiny44), I'm pretty sure that wouldn't take many clock cycles. I'd write it in assembly just to be sure. At the max 20MHz, that's exactly 7 clock cycles per read (assuming you can get the clocks synchronised), which should be feasible writing in assembly so long as you can write a whole byte to a whole byte's worth of pins simultaneously. I think you can, but it's been a while since I looked into doing such a thing. That would be a ~$1 solution.

The M0 route may be even cheaper, I think I saw one ~16 pin SWM050 something for 50c or so, but I'm unfamiliar with ARM. Definitely both worth looking at.

Either way spending $2 or even $5 per MCU/OTP/EEPROM is hardly an issue if you're competing with something heavily price-gouged.
>>
>>2074115
that Dave Jones guy is very good at making farting novelty gadgets

>>2074305
SWM050 is supported by libopencm3, but I don't know about programming parts, usually ARM stuff requires little helpers served from the host or dongle to work the flash on the chip. PSoC is weird, it's in an onboard utility ROM
>so long as you can write a whole byte to a whole byte's worth of pins simultaneously
yes, you can
in the ARM case I think it could be a word port read, a logical AND, an indexed byte load, a byte write, and a branch. maybe like 8 cycles, so a 48MHz core clock to meet the worst possible case
>>
Good read for any fellow PCB fabbers or chemists:
http://turtlesarehere.com/html/through_hole_plating.html

>>2074309
>maybe like 8 cycles, so a 48MHz core clock to meet the worst possible case
Never done assembly before, I guess it depends a lot on how the ROM is being addressed. If it's just cycling through the whole thing sequentially I think it would take 6 cycles or less per read-write, but I've no idea the kind of interface he's requiring.
If it's parallel address, then that's 11 address pins, so like 2 cycles for the address reading I guess if it's an 8-bit MCU, possibly just one if it's 16 or 32 bit, but maybe if there's only 8 pins per port you'd need two seperate read operations anyhow. Need even more pins too, but I still suspect fully parallel MCU is the best option besides a dedicated memory IC. If it's got serial addressing then we're right back where we started.

Also my M27128A EPROMs actually suit the application, assuming he wants a meme CDIP with quartz window for his project. They're on eBay for under $1 each (plus shipping). I think they look cool, and it would be neat to include obsolete parts like them in a practical project, assuming the size isn't an issue. Could customise them a little with a nice branded sticker over the window too, if he's selling them.
M2716-1 will also work, so will M2732A, most of them are under 350ns.
>>
>>2074320
Looks like the M27C256 and M27C512 are the cheapest on eBay. Plus they've got 4/5 extra addresses, to tie some DIP/binary rotary switches to in case you want up to 16/32 different ROMs to swap between. And they're 45ns response time, which is pretty fast. 0.82USD each including shipping.
>>
>>2073819
>Then why didn't you ask this directly instead of using that dumb microwave and chocolate analogy?
Waving a piece of rope and actually visualizing something directly caused by the radio waves are two completely different things.

Really, I'd love to cast a thin block of chocolate with an antenna stuck half-way though it, then amp up the output so high that rings appear in the chocolate around the antenna, but I figured there's got to be a cheaper, easier way and one that'll keep the FCC off my ass.
>>
File: Left or right.png (17 KB, 1136x549)
17 KB
17 KB PNG
Is the left circuit ok or do I need to do the right one or something different entirely?
The goal here is to discharge the capacitor that can be anywhere between +200V and -200V.
>>
>>2071235
The noob with the 12v fan here.
I bought the XL6009 but it seems doesn't work with the fan.
I hook the XL6009 with a 10K resistor and the measure seems ok: 4.85V in, 12V out.
But when I hook it with the fan, both input and output voltage drop to ~3.5V and the module becomes quite hot so I unplug it.
I'm using a 5V 1A phone charging adapter as the power supply.
>>
>want a smooth and reliable pot for a front panel
>can't tell if the $5 bourns pots are actually any better than the chink pots or just marked up for the brand
is there anywhere that, like, reviews potentiometers/switches/encoders?
>>
>>2069275
Skin resistance falls dramatically when wet. I've heard that even 12VDC could be dangerous when salt water is involved.
Current from a lightning strike creates a voltage gradient in the water. Not sure, but I'd guess that the voltage difference in the water from one side of your body to the other could be lethal.
>>
>>2073645
The pressure from a proper crimping tool results in a 'cold weld'. Much more secure than with pliers.
>>
>>2074305
>Either way spending $2 or even $5 per MCU/OTP/EEPROM is hardly an issue if you're competing with something heavily price-gouged.
Yeah that may be the way forward. I'm probably not going to be making 10,000 of these so shelling out for mask ROMs is not likely to be feasible.

re: eBay chips, yeah that's a possibility. I'm a bit concerned about yield dealing with those suppliers (I'd kinda hoped to use new manufacture everything) but that's probably not that realistic the more I think about it.
>>
File: sallen_key_lpf.png (27 KB, 1385x953)
27 KB
27 KB PNG
Why doesn't my transistor based active filter work?
>>
File: bode.png (46 KB, 2548x1192)
46 KB
46 KB PNG
>>2074595
here's the bode plot, it's clearly all wrong
>>
Ohmmies, I know there is a good chance that the question don't belong here but I'd like to try anyway.
Is there any good textbook/resource on sensor design/remote sensing technology?
>>
>>2068247
This isn't as sperg of a question as this thread is probably used to, but I just came into possession of a GE 7-4640A radio clock and I have absolutely no idea how to operate it. I checked every component and it seems in order, no broken buttons or dials, display functions fine, speakers works alright. Despite this, none of the buttons really seem to work too well. Both the Alarm and Sleep buttons do nothing, the Snooze button seems to work I think. I have no idea how to set the alarm.

I admit this is one of my first experiences with an older GE model so I'm out of my depth here. If anyone has experience with this model or can point me in the direction, I'd be much obliged.

tl;dr bought an old GE alarm clock, i'm a retard and dunno if it's broken or if i just have no idea how to use it
>>
>>2074633
>sensor design
you mean the entire science of physics? yeah, go to school
>remote sensing
the remoteness concern is orthogonal to the sensing concern. usually the two concerns are decomposed along a defined electrical interface (e.g. 1mA per 1m/sec speed, 40.8µV/°C...) and treated individually with regard to the defined interface
the CIA did some experiments with remote sensing in the 1960s, might look them up

>>2074651
/ohm/ is not a consumer electronics helpdesk. try /g/
>>
File: s-l1600.jpg (168 KB, 1134x1134)
168 KB
168 KB JPG
Hey, guys, quick question from a noob.

The connection pads for these little guys are a little confusing, am I correct in thinking the idea is to put a screw through the hole and into a plastic standoff or similar, with a wire between the screw head and the pad (possibly with a washer)? I may be blind but I can't see any way to mount them otherwise.
>>
>>2074652
thanks, sorry for posting in the wrong place
>>
>>2074652
>you mean the entire science of physics? yeah, go to school
Oh, my. Took a while to find a textbook and wow, you're completely right. I guess I can't ever avoid physics (shhh, I know, dumb statement for someone who's going to pursue electronics and possibly signal processing.)
>along a defined electrical interface (e.g. 1mA per 1m/sec speed, 40.8µV/°C...) and treated individually with regard to the defined interface
:^)
Anon is very smart! I need to read more!
>>
>>2074667
my second line was legit design/engineering advice. type K thermocouples have a 40.8µV/°C slope. the other two lines were snark, it is true, but there was some advice in the first one, namely that "sensors", the science of devices that convert all kinds of physical phenomena into particularly electrical phenomena, or, restated, the science of observing physical phenomena through electrical phenomena, is a significant subset of the whole of physical science, or at least its specialty of material science
idk try Wilson's sensing technology and devices handbook if you're up for some reading
>>
File: 1598373090677.png (209 KB, 715x881)
209 KB
209 KB PNG
>>2074678
>my second line was legit design/engineering advice.
Oh, I know anon! I hope you didn't misunderstand me with that silly emote, you know the face you make when you have no idea what's going on but you smile? I've been here for a while and I know that sometimes we're a bit sassy but mostly well-intentioned! I just don't know what you meant haha with the slope because I don't understand much about it which is why I said that I have to read more!
>type K thermocouples have a 40.8µV/°C slope
Like this! It does sound really interesting though, kinda amazing that I can learn new stuff everyday here, I'm only on MOSFETs right now but I'm going to read this spec sheet since it seems like the topic at hand.
>the science of devices that convert all kinds of physical phenomena into particularly electrical phenomena, or, restated, the science of observing physical phenomena through electrical phenomena, is a significant subset of the whole of physical science, or at least its specialty of material science
Ahhh, that makes a lot more sense now, thank you anon! It kind of sounds like signal processing too, but I haven't taken any courses in that yet (soon). Again, thanks ohmmie.
>>
>>2074690
do you like wonton soup
>>
File: 1599394600108.png (1.6 MB, 2048x2048)
1.6 MB
1.6 MB PNG
>>2074712
Wonton soup is ehh, I don't like Asian soups.
To be more accurate, I dislike soup in general. I'm more of a broth guy, even then, I dislike it. Do you like them, anon?
Maybe it's because I grew up on Coke and Pepsi (the latter being the superior choice), if the opportunity presents itself I'd just go for carbonated drinks like the dumb mutt I am.
>>
File: 1547731384491.jpg (101 KB, 880x870)
101 KB
101 KB JPG
>>2074721
wontons are gross but i like ravioli lmao
>>
>>2073841
>>2073882
What about just dremelling out part of the board with DC-DC?
Not all cheapo DC-DCs are good enough, they have down sides like using BJTs and diode instead of 2 mosfets.
I mean you can find more modern DC-DCs on ali that have all-in-one IC that uses mosfets internally, but...
>>
>>2074735
>What about just dremelling out part of the board with DC-DC?
lame, unreliable, you probably shorted or opened traces on the internal layers, and what makes you think you got everything?
>Not all cheapo DC-DCs are good enough, they have down sides like using BJTs and diode instead of 2 mosfets.
if you're not using a battery power source, those details become mostly irrelevant
if you have standards, why are you salvaging someone else's assembly without knowing their specifications and requirements, instead of engineering and selecting for yourself? how would you even know the current ratings of their inductors, for example?
this is why
>I have junk, what do?
is in the OP
>>
I was wondering is there any other relevant norms then IPCs one when it comes to PCBs, solderings, etc? Like some european one?

Also big question why it is so damn expensive to buy a book about it?
>>
>>2074753
>lame, unreliable, you probably shorted or opened traces on the internal layers, and what makes you think you got everything?
I'm pretty sure all DC-DC traces are on the outer layers.
>if you're not using a battery power source, those details become mostly irrelevant
Source is battery.
>if you have standards, why are you salvaging someone else's assembly without knowing their specifications and requirements, instead of engineering and selecting for yourself?
Because it takes about 1-3 months for parts to arrive. Only parts that are available locally, are found in the garbage.
And why pay some IC manufacturer jew, when you can cut out shit from laptop that accidentally killed CPU with 19V after coffee bath.
As for requirements, they are pretty close to laptop. Fuckton of power, high efficiency. Also I'm pretty sure engineers at whatever taiwanese firm know more about dc-dcs than me and will design laptop with proper distances and filtering.
> how would you even know the current ratings of their inductors, for example?
By checking laptop schematics, they usually tell you current ratings there for each bus. Or by checking what is on that line.
>>
File: 1612985336099.png (211 KB, 315x499)
211 KB
211 KB PNG
>>2074730
>wontons are gross but i like ravioli lmao
Oh you were talking about the food wonton and not the soup lmao.
It's aight, I wouldn't ever buy wonton myself.
Ravioli sounds good, there's a small Italian area near me so I'll go there and try em out just for you. I never ate real ravioli, you probably ate the shitty school food we have here in the states. Can't believe I hated pizza because public school pizza was so trash that I started to hated the real ones because I thought they'd be the same.

Time to read that Wilson book you recommended me (unless you're other anon), will probably pair it up with Fraden's book it too. Always nice to have multiple references.
>>
>>2074421
Rings won't appear without a standing wave. For an antenna not in a cavity, you'd just see the inverse square law; a slowly growing sphere of molten choco.

>>2074422
If the cap is at a negative voltage, then the left circuit's body diode will conduct regardless of the gate voltage. The circuit on the right will work, so long as you have a split-supply (say, ±10V) to power both gates with.
I'm assuming the arrow at the top is going to the power supply that charges the 200nF cap? Shorting 200V through a 470Ω resistor will result in it dissipating 85W, and those 100W resistors are expensive, why do the values have to be so low? If you want to limit current into the cap, I'd use a constant-current feedback power supply. That can do ±200V output. Hmm, that would be quite a difficult project.

>>2074435
Just a guess, but I'm betting the thing won't run very well with an input voltage of 3.5V. Either your power supply is insufficient, or you left a resistor or something in series with it. Don't do that.

>>2074503
If you can get a datasheet (LCSC for the chinkies) and examine the cycles, that should be a good enough metric. Just buy a dozen and you'll be able to avoid any QC rejects.

>>2074547
>I'm a bit concerned about yield dealing with those suppliers
I would be too, but it's still really surprising how much cheaper than anything that LCSC has they are. I guess nobody makes parallel ROMs. I'd make a preliminary design around them and see how many you can order at once. 100-1000 pieces should be sufficient for an initial product, since I wouldn't want to hand-solder any more than that.
On that note, the W27C512 EEPROMs on eBay are even cheaper at <50c each. But they don't have the cool quartz window and ceramic package.

>>2074656
Solder. Glue. Luck. They're designed to be compact and solderable, not to be mechanically stable. If you're not in a position to make a case that it just slots into, just glue it to something. Or buy the IC and make your own PCB.
>>
>>2074503
>can't tell if the $5 bourns pots are actually any better

used them for years, and they're excellent pots.

>>2074656
>with a wire between the screw head and the pad

no. the holes are for meant for soldering only. to mount it, use glue or double-sided tape.
>>
File: file.png (71 KB, 658x669)
71 KB
71 KB PNG
Retard here. Could pipes be used to negate false echo in ultrasonic distance sensors that have independent receiver and transmitter? I need to avoid false echo from the water tank lid I'm mounting my sensors in but the current sensor has such a wide sensing angle that the lid itself causes wrong readings. It's also a single transducer that's set to receive or transmit based on the state of a trigger pin so I can't separate receiver from transmitter. My idea is to get a sensor that uses separate transducers and mount each individual R and T at the end of parallel small diameter pipes. That way the tank lid can't just bounce ultrasounds directly into the receiver regardless of sensing angle and the only sound that should make it into the receiver is what bounced off something perpendicular to the direction the pipe is aimed towards, which would hopefully be the water surface rather than the edge of the lid. Pic related for a poor attempt at a drawn explanation of what I'm trying to do, right side being the current setup with echo from lid causing problem. I have no idea if this is even possible. From what I've read this could reduce the max sensing distance due to reduced power of received sound waves but I'd be good even if the max range is cut in half.

Another workaround I thought about would be to force the sensor to completely ignores all received echo for a certain amount of time creating a true blind zone, but I'm not sure the cheap shit sensors I'm using can do that. I've seen a bunch of videos for high cost industrial ultrasonic sensors that can be configured like that, some apparently could even focus or widen the sensing angle based on distance, but they're mostly made for low range production line sensing and I really don't want to spent 500 bucks on a single sensor.
>>
>>2074836
Pipes wouldn't be very space efficient, you'd likely want to use parabolas instead. Also do some testing to ensure whatever material you use reflects sound well and doesn't allow much transmission at all. Maybe there's some funky total internal reflection or other refractive index shit to be done here.
With the inverse square law being as it is, you may find that an attenuated transmission through the walls of your pipes across ~1cm to the receiver leads to a stronger signal than the transmission reflected off a wall 2m away.

It's also an option to use a phased array to determine what direction the signal is coming from. I think a cheap modern MCU is
more than fast enough to pick up phase information from 40kHz transducers. Personally I'd have a + shape of 5 transceivers, send a chirp to them all (possibly with the help of something like an 8-output Si5351), and wait and listen to those same transceivers for a reply. Not only can you tell directionality from the receptions' phases, but you can also phase the transmissions as to aim at or around particular objects.
>>
>>2074871
Thanks for the ideas brother. Space isn't much of a problem and the drawing has poor scaling too. Those pipes wouldn't be more than 3 inches in length, just enough to get below the lid I'm picking up as false echo with my current sensors. As far as receiving sound through the pipes, I'm guessing a sensor that works using this method of individual T and R like the cheap HC-SR04s have some kind of delay to prevent picking up the initial sound rather than an echo. I'm assuming this is why the 2 cm blind zone is needed. Phased array sounds cool but maybe overkill for what I'm trying to do here but I'm a very big brainlet so I'll need to look into it a bit. Best case scenario I'd just be able to find a good sensor with a narrow beam angle that runs for a reasonable price.
>>
>>2075079
>>2075079
next thread people
>>
Threadly reminder I love my ohmmies
>>
>>2068247
I *need* the pillow from pic rel but I'm afraid it's a shoop.



Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.