Thread let the smoke out:>>2068247>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.comhttp://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 ElectronicsGeier, How to Diagnose & Fix Everything ElectronicKybett & Boysen, All New Electronics Self-Teaching GuideScherz & Monk, Practical Electronics for InventorsHorowitz and Hill, The Art of Electronics>Design/verification tools:LTSpiceMicroCapfalstad.com/circuit/circuitjs.htmlNI MultisimCircuitLabiCircuit for MacsKiCAD (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 distributorsladyada.net/library/procure/hobbyist.html>Related YouTube channels:mjlortonw2aewjkgamm041eevblogEcProjectsgreatscottlabPhotonvidssdgelectronicsBigCliveBen Eater>I have junk, what do?Shitcan it
Threadly reminder to serve the ohmnissiah
Thread brought to you by the MAX5079 ORing MOSFET controller with 200ns turnoff. An ORing MOSFET here means a MOSFET being used with feedback as a diode for high-current systems. This IC provides the feedback, has an included charge-pump for powering itself, and has safety systems to prevent shorts.Also yes the MOSFET should be pointing in that direction.
>>2075079Mercury arc rectifiers are really cool. Especially the 3-phase ones.
>>2075100>Mercury arc rectifiers
>>2075103Why is it that having the audacity to like anything makes you a soγboγ these days?
first for the poynting vector.
>>2075122well you're not first, but you have a poynt
What do you think of this design?https://github.com/wagiminator/ATmega-Soldering-Station/blob/master/hardware/SolderingStation2_schematic_v2.7.pdfWant to replace my station controller with a stm32 based one but no one made a controller with one yet. Ironically there already are custom firmwares for those chinese stm32 T12 controllers.
>>2075146>but no one made a controller with one yetso make one
>>2075151ehh I'd have to make the software too and I kinda don't want to
>>2075122Shouldn't E and B be 90 degrees out of phase?
>>2075146>Want to replace my station controller with a stm32 based oneI assume you mean a controller specific to your particular soldering station? Any reason a generic T12 controller won't work? I assume all stations use the same thermocouples, and adjusting for different pinouts and different cold-side thermistors would be relatively trivial. Change the MOSFET if you're on a different voltage. Also I'd probably use a dedicated gate driver.Wonder if I can get rockbox on my station?>>2075158They're 90 degrees out of phase in space orthogonal to their direction of travel, not in time or in their direction of travel. Derive it from Maxwell's equations yourself if you're curious (you only need one vector product identity, use the differential form), you'll find that the e^i(wt - kx) the same in both B and E fields in free space.
>>2075079is that pic based on the mercury method for rectification of high voltage the brits were doing?thats what it reminds me of
>>2075167Well for one a generic controller probably won't fit.Also there are too many T12 controllers out there and the popular ksger ones have only had their corners cut over many revisions.>rockboxDo you even want to work on rockbox? That thing has some arcane design built 20 years ago.There's like one dude working on native ports and I don't think he even got very far supporting the JZ chips that players used for 5-6 years before they moved on to rockchip and Android.
>>2075170It's a lot more common than that, it was (and possibly still is) a common method for high-voltage and high-current rectification for transmission stations. Like the conversion between AC to DC for the english channel, or those two halves of japan, or wherever.>>2075172>a generic controller probably won't fitReasonable concern.>had their corners cut over many revisionsStill no reason to reinvent the wheel. Have a look at all their schematics and use them to design your own, into a form-factor that will suit you. Then for the firmware end, use the custom firmware made by white people and it should be pretty reliable and with plenty of support. Do it right, and you won't need to adjust the firmware at all.The pmp guys on /g/ still use rockbox for their modern chifi players, I think it's still in reasonably good use. Not sure if any portable media players use an STM32 on them at all though.
>>2075179eh guess I'll clone their design thenrockbox doesn't really support any "modern" players, and not any of the current ones. The latest one to get support was the T6, it ran linux and rockbox was ported as an app port, not a native port.The current ones just run android so there's no need for rockbox.An stm32 DAP with rockbox might be nice, but whoever making one would have an easier time building the software from scratch.
>>2075179It's a lot more common than that, it was (and possibly still is) a common method for high-voltage and high-current rectification for transmission stations. Like the conversion between AC to DC for the english channel, or those two halves of japan, orTell me more
>>2075222>that dude holding his crutch like a rifleGiggle
>>2075223He was the real villain. Eric & Dylan were patsies. lmao
>>2075223Is it bad that I do this while walking to and from the break room with my drink bottle at work?
>>2075103Like they wouldn't throw a fit about it using mercury. Fuck off.
Is there an easy way to strip these joined wires without pulling them apart?
>>2075257The wide blade multi-lever strippers do it ok IIRC, so long as the blade is wide enough.Are you inserting the wires into some sort of connector or crimp that's the same pitch as the wires are wide? All too often I see ones where the ribbon has to be separated at the end anyhow because the connector has a wider pitch.
>>2075259Not inserting it anywhere, soldering it to these addressable leds, I'll make sure I buy wire with the same pitch as the pads on the leds.
What is the typical lowest voltage of phone batteries, anyone know from looking at:cat /sys/class/power_supply/battery/voltage_nowOr:dumpsys batteryOr something like that? I'm using 3.2V for my custom battery script
>>2075259do you have a pic of the style you're talking about?
>>2075262Maybe 2.7V or so
>>2075261>I'll make sure I buy wire with the same pitch as the pads on the ledsIf that's 0.1", the ribbon wire will be pretty wide, which I don't see often.>>2075276these
>>2075276>>2075278also probably these, though i've never used them
>>2075278>thesecool, I already have a pair
>>2075097Why is Vbus connected directly to ground?
I don't know if this is even possible but I have a question.So I have a 3D printer. The hotend heatsink is cooled by a small 24V fan. Typically, you run it where the fan blows through the heatsink and you don't have to worry about the airflow. But, there are some filaments that you don't want any air to hit and cool it like ABS for example.So, I am wonder if there is a way where you could have a circuit or something where you flip a switch and reverse the flow of electricity. I may be wrong, but I'm pretty sure with these DC fans, positive to positive and negative to negative makes the fan spin one way, but if you connect them to the opposite like positive to negative and negative to positive, it spins in the opposite direction.Is this possible?>pic related
>>2075313Wait, you don't want air to hit the extruded filament, but you still want air passing through the extruder's heat fins? I'd just print a duct to keep the air out of the way.>flip a switch and reverse the flow of electricityThose DC fans are not brushed DC motors, they have integrated brushless motor drivers and will die if connected backwards. There is no way of reversing the direction of one of these, besides modifying the internal driver circuit by flipping the hall-sensor polarity or something.
>>2075279>though i've never used themI used them extensively at my old job and they're great: if you have them dialed in correctly and they're not cheap chinsy shit you can strip tons of wires in no time, however you can expect to make a mess of the first few wires, especially if they're not solid core.
>>2075315Well for something like PLA you want to cool it as it prints (except the first layer). Ive heard that with PETG you don't need to cool it unless you are bridging gaps. ABS warps to fuck if it gets cooled while it prints, so most people print that in an enclosure. I'm not sure about the other filament types. But you generally have two types of fans, the hotend heatsink fan and the parts cooling fans.But yeah you need to cool the heatsink by and sometimes keep the air from flowing downwards. My other idea was just manually flip the fan around and have a duct blownit upwards instead of wherever it wants. Some people use water cooled hotends too so I guess there is that option but I could only imagine how much your day would be ruined if one of those pipes came loose. Easy enough to do in a PC but with something that moves around so quickly, I don't trust it.
I desolder the XL4015 to put it on a heatsink, because unnecessary heat on the pcb + better cooling.Does a toroid ,pic related,need cooling?On 3A it's fucking hot ( can't touch ) and for the future i want 5A, i thought to stick 2 heatsinks on the sides.
r8 my station
>>2075424A Quicko/Ksger T12. Great for the money, I went for the one with an integrated power supply.
If an "on" transistor (i.e. no depletion zone, free flow across) can be considered an open circuit, can I pass an analogue signal across it (emitter -> collector) without noise/signal degradation?
I thought that was a purple hair anime girl with a very fat butt for a second.I need to neck myself.My libido aside, does having a good understanding of systems and signals, as well as DSP and its applications, help in electronics? I have a spot open this summer to take the courses.
>>2075515electronics is a very broad fieldif you want to work on things with signals then obviously having a good understanding of them should help
>>2075525Good point, is there a particular subset of electronics you focus on?
>>2075313no such thing these days. instead, print ducts that deliver hotend cooling air to/from the air from above the hotend>>2075328PETG is nice material as long as you keep it dry>>2075359>better coolingby how much? modest airflow pulls a lot of heat away from an assembly, maybe 2x compared to still air>it's fucking hot40°C temperature rise is the usual maximum allowed>want 5Ayou'll probably need a different output inductor, then. happy shopping>>2075508a closed circuit, you meanand that arrangement is called a transmission gate, but distortion is possible near the rails due to transistor nonlinearity. nonetheless many musical instruments and signal processors used them for signal path control, mostly JFETs, almost never bipolarsiirc there used to be MOSFETs with the device body brought out as a fourth lead instead of connected to the source, which could be used as transmission gates without severe nonlinearity issues, rare even in their heyday>>2075515get enough of that mercury into your system and she'll be anything you want, sweetie>does having a good understanding of systems and signals help in electronicsabsolutely, they will probably be your leading subjects of interest throughout your entire career or hobby>as well as DSP and its applicationseh, only if your specific interests include it
Hello everyone.A couple months ago I made a post asking for help as to why I couldn't connect to (or discover) any of my Siglent equipments trough lxi-tools, and I was told it was because I needed a network switch between the ethernet port of my computer and the one of the device.So I purchased one for cheap, plugged everything in it but I still fail to make it all work. The network switch works fine since I'm able to use my internet at full speed and the siglent device seems to succed at connecting to the network switch because the LAN indicator turns blue (and yes, the LAN is enabled on the equipment). Running lxi with sudo won't make a difference, I tried to manually connect to the equipment with the IP adress it advertised in its menu using telnet and the port 5024, it won't connect. I also tried to disable the linux firewall. Feeling pretty desesperate right now, does anyone have any ideas ? And does DHCP need to be enabled ?
>>2075508Yes. That's essentially what an analog switch does. True analog switches use a parallel N and P channel MOSFET though.
>>2075508Yes but remember MOSFETs have a body diode in them. JFETs are better for this.>>2075547>there used to be MOSFETs with the device body brought out as a fourth leadWould be damn handy to have those, as they also get rid of the body diode to some extent.Not that I have much reason not to just use CD4051/52/53s.
>>2075627it's more that they allow you to put the body diode where you want it, rather than having it tied to one end of the channel which causes load-dependent distortion in low-Z transmission gate applications>not that I have much reasonsame, or the various improved analog switches suitable for modular synth applications and power rails, like DGxxx series
>>2075651dropping loads all over your junctions
>>2075651I think I understand that circuit. Interesting. VL and GND make a logic-level power input, say 0-5V around the IN_X input, while V+ and V- can be ±15V or whatever.I guess those twin long-tailed-pair looking things bring that 0V to -15V and 5V to +15V.Looks pretty handy for my circuit with ±2.5V for logic and some higher rail for audio op-amps and such, though now I'm wondering if the higher voltage is really necessary.
So I have assembled this and it fucking works. The problem is, I don't understand WHY. This is a variation of Turn-ON Turn-OFF device where instead of one of mosfets there's an LDO http://www.mosaic-industries.com/embedded-systems/microcontroller-projects/electronic-circuits/push-button-switch-turn-on/microcontroller-latching-on-offMy DSO138 only had one probe to I can't simply get how the current flows here and there and what makes it to work. Anyone cares to spoonfeed me? I understand that I ask too much, but PLEASE, HELP
>>2075738Just to elabore on how it works, you push the button it turns on, you then hold it for like 5 seconds it turns off!
>>2075682yepthe first four-transistor bits look like CMOS comparators to me>>2075738>how it works>implyingyou are grounding out your MCU output through D1 and S1 when holding to power off. is that what you want? hope nottry falstad if you want infinite probing. make the EN input a logic gate or comparator with an LED of your choice
>>2075777U6 is not an MCU but an LDO, the POWER_LATCH line is supposed to go to a MCU's (now shown) pin that would detect power off and prematurely (read before hardware cutoff) turn off the device when everything has been saved, by driving the same pin low, the button is there to physically turn off the device in an event the MCU crashes.
>>2075781>POWER_LATCH line is supposed to go to a MCU's (now shown) pin that would detect power off and prematurely (read before hardware cutoff) turn off the deviceI think you're saying the MCU never strongly sources current on the POWER_LATCH pin so the diode is no danger, which is fairwhen presesd, C9 ac-couples the low pulse into the EN, which turns on Q2 briefly and charges C8 quickly, holding EN on via R11/R12/C8 until POWER_LATCH or SW1 pulls downwhen held, C9 develops a voltage through R10 on the one side and SW1 on the other. C9 will again be enabled briefly. when C9 is charged, turning off Q2's strong high, the EN line's state is now governed by C8. once C8 is sufficiently drained through R12/D1/SW1 or R12/MCU, the regulator shuts off and C8 is drained fully through R12/R11/loads
>>2075788Wow, that was on point. Thanks.
>>2075547>absolutely, they will probably be your leading subjects of interest throughout your entire career or hobbyOh, what the hell, I didn't know that signals and systems would be one of the leading subjects. I can't slack on the concepts for it then. Any recommended text for that? Oppenheim's book seems to be the go-to but damn is it math-heavy. Though that's to be expected heh.>get enough of that mercury into your system and she'll be anything you want, sweetieWe'll have one less poster then...
What is actually meant by “signals and systems”? Sounds like two very nonspecific words next to one another.I’m guessing it encompasses things like control theory and signal conditioning and maybe transmission lines or modulation methods?
>>2076010My balls hurt. Is there a circuit I could build to help alleviate the pain in my balls?
>>2076028this one works
>>2076056>building your own TENS unitfucking hell they're already so cheap as it is
>>2076056The power level is too damn high. I don't want to make my balls time travel.
>>2075961in most electronics, you're building systems to process signals, sooo>We'll have one less poster then...but the departure by yiffing 60 kilovolts would make for a welcome change from the usual 240Vac. boooriiing>>2076028yes, but it's a bit advanced, see link below>>2076056>akim tipi sechow forward>>2076074there's a difference between nerve stimulation to deaden and nerve stimulation to excite. this here is a clone of a $700 unithttp://clx.freeshell.org/mk312bt.html
>>2076082>there's a difference between nerve stimulation to deaden and nerve stimulation to excite. this here is a clone of a $700 unitkek that's such bullshit. there is no proven difference in effect between ordinary TENS and any other form of muscle stimulation. the bigger, PT practice units might be more robust, i'll grant that, but apart from that there's literally no study proving any clinically significant difference in pain relief, short term or long term outcomes between the different muscular stimulation technologies that are getting hawked out there.yes this is even true for modalities as radically different as TENS and galvanic stimulation.
>>2076082>>2076093oh wait this is some stupid sex thing?jesus fucking christ >>>/d/
>>2076095an application engineer is damned if they do, damned if they don't
>>2076095>oh wait this is some stupid sex thingyeah they have electrode butt plugsmaybe even sounding rodsbigclive took a look at a few of them, if that means anything
>>2076115lord have mercy
Rolling to select a project from that link in the OP. Also, does anyone have experience with making weather instrumentation? I'd like to try to make a relative humidity sensor. Been a bit inspired lately from working with all these Campbell Scientific weather stations at my job wanted to see if I could make something similar.
>>2076132Rerolling, done the organ before... If it's another one I've done or am not interested in I'm just going to double down on my idea of making rugged datalogger + weather instrumentation
>>2076132the BME280 is an exquisitely sensitive environmental sensor. they are available on chinkboards for $2.50 each, an amazing deal compared to Digi-Key retail
>>2076141We have BME280s on our sonic anemometers at our kilometer testing range and the fuckers die rapidly. I've kinda been turned off by them, but to be fair it could be the case we've got them in though. I'll give them a second chance.The EE181s (temp + relative humidity) from campbell come with this solar plating shield. Maybe I can 3D print something similar or even press some sheet metal on a cone and stack them in that way.
>>2076132>make a relative humidity sensorWhat sort? Looks like the common ones are where a dielectric or resistive material has its key property altered by absorbing moisture, though ones that measure the thermal conductivity of air may also be common.
>>2076154>What sort? Looks like the common ones are where a dielectric or resistive material has its key property altered by absorbing moistureI'll probably experiment a bit. Using a dielectric sounds like a good starting spot for some research.
spoonfeed me again.I found the BJT is likely the culprit, failed closed circuit with 90Ohm G-E and gets extremely hot when device is powered on.Transistors shouldnt die without external cause.I ordered replacements but im afraid the new ones will die the very same death.Doodled out the schematic. Is there something obvious wrong? checked diodes with the pcb layout and they match.Diodes are all functional, cant test D4 im too dumb
>>2076241oh and1 ohm in the transformer seam like its fried, right?or is the primary on those really just a handful of winds
>>2076245That board layout is so confusing I can't tell which winding is the primary and which is the secondary. If it's 1Ω primary 60Ω secondary then that's feasible enough if the transistor is oscillating fast enough. Wait while I reverse-engineer the diagram by hand.
>>207624190Ω doesn't mean anything to something that's non-ohmic, to say nothing of something that's in-circuit with resistors around it. And judging by the heat-sink, they expect it to dump a couple watts at the least, so heat is to be expected.For proper troubleshooting you'll need either an oscilloscope (or a $5 logic analyser and some passives), or to simulate/analyse the circuit and figure out what kinds of average voltages it will have to measure each by each. If your DMM can measure high frequencies then you may be able to use it to confirm the AC voltages too. It's likely running at 10k-300kHz.Oh and you drew your diode backwards.>>2076245>>2076255Yep it's definitely a self-resonant converter, and the 1Ω is the primary. Assuming the transistor is dead, I'm guessing that it was used for too long and the transistor overheated. Or perhaps the small blue capacitor across the two windings failed, causing the resonant frequency to be so high that the transistor's losses were too great.Also that blue cap might be something like a MOV instead, so check the markings with it. A MOV could be a method of voltage feedback, or at the very least overvoltage protection.Diagram here, though I've assumed the transistor orientation.This should be enough to simulate the first half of the circuit and have it work, though naturally you'll need to experiment a bit with inductances and relative inductor polarities.
for a 24v plc monkey this shit is magic>>2076262>90Ω doesn't mean anything to something that's non-ohmici desoldered it for measurement and just realised i assumed the wrong C-E-B layout. But the bjt is definitely fucked, C-E is shortedyep, the zener is backwardsBlue appears to be a 221pF 1kv cap.The transistors Base is connected to the secondary winding, the collector to the primary.and emitter to gnd.i got a scope. So im understanding i should probe the primary side and see if it is oscillating, since it is the primary getting switched by the bjt
Any handy guides for PCB Layouting? I don't have any idea for connecting the bulbs in series.
>>2076028Furfags are there to help.
>>2076309>Internet connected shock collarThis can only end well
>>2076323You should read the reviews and marketing wank. What a time to be alive.
Redpill me on wire-wrapping. Is there any significant advantage using the $14+ tool (WSU 30) vs DIY drill hole in small tube?
Doing an FM hi-fi and I need a big antenna. Been thinking of just getting the largest TV antenna I can get and using that, will it give me better results than an actual FM radio antenna?As far as I know bigger antenna = more reception. I want to be able to reliably pickup signals from 200+ miles away.
>>2076481alternativelyI'm expecting to buy something in the 4-6' range
>>2076241>Transistors shouldnt die without external cause.lel>>2076245>primary on those really just a handful of winds0.5mm solid wire is only about 85mΩ/m, you can reverse engineer the math>>2076302tip 1: turn on snap to grid>>2076411yes, look closer at the wrapping end, there's more going on than just drilling a hole. by itself, the unwrapping end could be worth $14 sometimes>>2076481the real antenna anons are on the ham radio general, check there
>>2076302its like a jigsaw puzzle you try something that looks right and see if it fits, then if it doesn't you try something else.mostly its just practice practice practice to get a feel for it. yes its boring, yes its tedious, yes you will feel great when you find something that works.try to keep your lines straight, break 90 degree angles into 45s (don't > me), other than high frequency lines that require specific impedances, try to keep your traces neat and tidy, line them up at set distances and run them together, leave little space as possible because you are going to flood fill with a (usually ground) plane and you want a nice big pool not lots of little tiny ones dotted about.just little things like the trace through Q1.1 to C2 could be a single straight line through, you don't need that junction going up and around. rotate the diodes so they straddle tr traces. you could rotate your SW components to line up the pins with the relays? or whatever they are, see if that makes the rest of it easier. just try things out and eventually you develop a sense for it.
>>2076411I don't know about you but I personally strip all my wire and then re-wrap it to help the economy.
How would I prevent the output of a CMOS LDO from rising with the input until the reference voltage is reached?I want the value of Vreg to remain as low as possible until VDD is around 1.2 V.
Is there a website or resource that gives pre-made LabVIEW programs for programmable equipment like power supply, DMM, and oscilloscope? It'd be nice if I had more control over what I can do with them on LabVIEW.
Is there any reason vacuum tubes have to use a glass housing? I was thinking of a metal instead. Maybe even use it as an electrode. How hard would it be to make your own tubes?
>>2076302Is that a single-sided board? Remember to flip it for etching.Also all those right-angles aren't the best for etching, you should have 45° angles (or meme rounded corners) where possible.Oh and KiCAD is better.>>2076481The antenna should to be sized to the wavelength of the radio. FM radio has a wavelength of ~3m, that piddly little yagi won't work for it. But you could calculate and design a relatively simple yagi for FM, and build it out of some random metal pipe or stock you've got lying about. But yagi antennae are directional, so if you plan on listening to stations that aren't in the same direction, you're gonna have a bad time. You also have to pay attention to the polarisation of the radio, which I think is vertical for FM broadcasting, so sideways dipoles may not be ideal anyhow.I'd personally stick to a standard vertical quarter-wave monopole and put my effort into low-noise high-gain receiver electronics instead.>>2076698You need something that can crimp the wires going into and out of the tube without making electrical contact with it, without leaking the vacuum. Glass happens to be relatively easy to do this with. Making just the wire glands out of glass would mean more manufacturing steps than just making the whole envelope out of glass, not to mention having to deal with differential thermal expansion of the case potentially cracking the glass.
>>2076635use this high-side switch but add an 11V zener in series with R1. cathode to the left.
>>2076582>you want a nice big pool not lots of little tiny ones dotted abouty tho, vias are free for prototypes>>2076635put a voltage divider on the enable input>inb4 no enable inputmaybe get an LDO that has one>>2076698>I was thinking of a metal instead.no problem>>2076704>Making just the wire glands out of glassthere are ceramics and other materials with more compatible thermal coefficients of expansion>>2076706not them but 1.2V, not 12V
>>2076715>there are ceramics and other materials with more compatible thermal coefficients of expansionDefinitely. But as someone asking "How hard would it be to make your own tubes" I don't think he's going to like crimping ceramics around metal.Also how do you suck out the air of a metal can vacuum tube and seal it?
So big dummy question here, why is the resistor and capacitor needed? It kinda makes sense for the original idea since they used a 9v battery and would want to limit the current draw correct? But why is it needed for wall connected power supplies?
>>2076411I've made these by cutting a little bitty hole in the ink straw from a pen. It sort of worked, but just buy the real thing. They work real nice and would be very difficult to machine a properly yourself, especially the unwrap end.
>>2076718same way as one does any other ampoule: pump down via capillary tube, heat neck, twist, crimp>>2076866no, nothing like that. the point of the power supply is to provide loop current so that the modems sense a live telephone line when off-hook. the resistor is there to limit that current similarly as it would be limited on a real telephone line
>>2076877So I guess what I'm asking is does it need to be current limited? I guess I'll run it off the bench supply in current limited mode and see how much it tries to draw. I don't currently have any resistors or caps. Need to put in an order. Haven't really been building stuff lately. Shouldn't the modems only draw as much as they need?
>>2076909yes, it does. unlimited current through the transformers will burn them out
>>2076916Well thanks for educating a brainlet like myself anon. I'll have to come back around when I'm working on something actually interesting.
>>2076909Depends on the modem, I’m guessing the old analog telephone systems required it, but for your modem who can say? Without seeing the surrounding circuit, that is. But not having a resistor (or maybe a constant-current diode) would put your circuit out of spec with the standard, meaning it may not necessarily work on other systems, even if it works on your one.Connect up 9V with the series resistor and measure the voltage across the resistor, that’s a good start.
>>2076909>Shouldn't the modems only draw as much as they need?no, that only applies to pure power supply rails. these are signal lines, and the power is incidental, mostly provided for (historical) line supervision purposes at the central office, but also used by modems as a cheap alternative to a dial tone detectorcheers, good luck
>>2075079Just learnt how a type-2 phase-frequency detector works, it’s neat. Better than an XOR for sure.Any other recommendations for simple but effective circuits to learn?
can someone help a noob understand what's happening when I use a 9V battery as my power source for this PSU schematic? I've been too scared to solder up my 12v AC to DC wall wart thingy, so i tried putting a 9v into the bread board and it lit my first led.
what is the female end to something like this called? and what size do i need to be able to fit it into a standard breadboard? there's a name for that size, last time i bought ICs they were way too small for my breadboard for example
Looking for some recommendations. I have a rock collection that I want to display under UV light so they floresce. I bought some strip style UV LEDs off of Amazon but they kind of suck. So I bought some individually wired UV LEDs (with pre-wired resistor) but they also suck. Can you guys recommend any good UV LEDs?I kind of want individual lights so I can put them where I want, aim them at the individual rocks. Otherwise I might just get a UV floodlight or something like that. But it's funner to do it this way I think.
>>2077109>>2077116The current went through D3, but not D4. You have a +12V rail, but your -12V rail is just sitting at 0V.>>2077130The round connector is called a barrel jack in general, female barrel jack is the term you want to search for.The "pitch" of standard breadboard sockets is 0.1 inches, or 2.54mm. Just buy a bunch of 1x40 and 2x40 pin headers. I just snapped off a 2x3 piece, soldered the + and - wires from a decapitated wall wart across three pins each, and epoxied across the top for strain relief. Plug it into my board, easy power. Do the same for a USB lead if you want 5V.Also know that there are dedicated breadboard power supplies that slot into the rails on the top and bottom of the breadboard and have a USB and/or barrel-jack input socket. Had one die on me though, so ymmv.>>2077134For bright individual lights, you'll probably want the standard-ish 1W or 3W LEDs. They're round and have two SMD legs on either side, and are often found pre-soldered to heat-sinking PCBs on alibay. Come in plenty of colours, 405nm almost certainly included. If the beam angle is too wide (find and check the datasheet on lcsc or sparkfun or something) then you may want to put a reflector around it. I can't remember for sure, but I think there are dedicated reflectors for that particular form-factor. Not sure about lenses though. This should go without saying, but I'd want to use a switching current-limiting power supply/converter for these, especially at 3W. Consider putting 3 in series and running them off the same regulator, for money saved. A common LM2596 module (get the one with 2+ trimpots) goes up to 36V, so will run 9 or 10 of them in series.If you hunt about on alibay you may find little spotlights in their own heat-sinking enclosures with lenses and/or reflectors, maybe even current-source too, though whether they're cost-effective for you is another matter.
>>2077134Oh also ensure that the rocks will fluoresce under 405nm light, should be in their wikipedia page, if not some rabbit hole of a geology or chemistry wiki/engineering data site. If not, you'll likely want to use mercury lamps instead, UVB or UVC LEDs are likely still too expensive.
>>2077109that circuit is expecting AC. it uses diodes to steer positive voltage to positive regulator, and negative voltage to negative regulator. so, if you plug in the battery one way, you'll get around +7V out, and when you flip it you get -7V out.if that works out, then go ahead and plug it into an AC transformer. if you're too chickenshit to do that, you've picked the wrong hobby.
>>2077146>chickenshitlol, it worked as in it lit an led with the two outs at the end (what was supposed to be +/-12v). I'd check it with my multi but it's broken, so I'll get a new one tomorrow before slapping together the real psu with the transformer.>>2077141thanks for the help
>>2077141>You have a +12V rail, but your -12V rail is just sitting at 0V.Wait no you have neither, because 9V is insufficient to pass through a 7812. They have ~2V dropout or something, so expect to have 9-7V on the output at a very high impedance. The ~0V at the -12V rail is also at a high impedance. If it wasn't obvious, the circuit requires AC as the other anon said.But the power LED will turn on in any case, so long as the input voltage is positive and above 2-4V, depending on colour.Also it's kinda weird that they just let reverse current flow through the LED in what I assume is breakdown condition, though not intrinsically bad.It's even stranger to me that they drew the circuit diagram with the GND rail below the -12V rail, barbarians. Always do it this way, rectifiers notwithstanding.
>>2076704>Making just the wire glands out of glass would mean more manufacturing stepsI'm kind of interested in making my own tubes for a ~kW RF amp. I'm just trying to avoid buying glass working equipment. Worth it for me since I'm not mass producing anything.>deal with differential thermal expansionLol I had this idea of using plastic here. I somehow totally overlooked the fact the terminals are going to conduct heat.>>2076715>there are ceramics and other materials How hard of they to work with? Next idea after plastic is an epoxy like JB weld. Like make the sides and tops metal, and have an epoxy base. I'm kind of curious about the RF properties of JB weld now. It's microwave safe, so there's hope.How precise does the geometry need to be in a high power VHF tube? Like if my grid is 2mm farther away from the cathode, is it going to ruin performance? I don't mind having to custom tune my circuit to make up for lack of precision. Gain and distortion are all that matter to me.
>>2076718>Also how do you suck out the air of a metal can vacuum tube and seal it?I was thinking Schrader valve. I'm crossing my fingers those can hold the vacuum I would need.
>>2075079>everyone in family is an electrical engineer (father, grandfather, uncle, cousins etc etc)>study EE as well>all through childhood and youth pops brought home different circuit boards that his company made for products>just studied them when I was young>literally have a garage full of PCBs that do different stuff>graduate.exe>employment.exe>buy random hot air station off Amazon>desolder all the interesting components off year old PCBs for using in my personal hobby projectsIt's like Christmas every day, I don't know what I'll desolder and make with the part
>>2077533>be homeless dude squatting in a flood damaged house>people are throwing out electronics like crazy after flood>salvage what I like>studying switch mode power supplies>can't find parts I need to make my own>fuck it I have lead acid battieries, switches, transformers, diodes and caps>rig up circuit to run battery current through transformer coil and charge a cap when the switch is cut off>become human pwm machine that monitors voltmeter and tries to maintain cap at 100V>junkies amazed I can zap them with a 12V UPS battery and a mess of wiresNice gondola :DDD
>want to get an ee job but don't have a relevant qualification>play shenzhen io>really like the idea of working by designing circuits>start to play a lot of the solitaire game instead>finding the quickest way to disassemble a disorderly stack of cards and put them in the right order is kinda relaxing>get job at supermarket because covid>end up having to move trolleys>now finding the quickest way to disassemble a disorderly stack of trolleys and put them in the right orderpottery
>>2076582>>2076547>>2076704Sorry for my late response but thank you for the advice you gave to me, I will gonna do what is the best in return.
It's me once again. After 16 days of waiting for Mouser to put 5 sensors in a fucking box and mail it to me I finally got the shipment only to find an open static bag with no parts in it. WTF has Mouser become?! QA has gone to shit. Handling has gone to shit. Processing times have gone to shit. Customer service is going to send me replacements 'soon'. Fuck this. I regret trying to save 5 bucks not going to Digikey.
>>2077661>an open static bag with no parts in itSounds like you got robbed by the postal spics, good luck next time.
>>2077678Sadly no. Not opened as in cut or ripped. Opened as in never sealed.
>>2075079Should i be worried about 16MHz quartz arduino crystal cross talk with this reset line?
>>2077810XTAL2 is the clock oscillator input line, right? MCU_RST isn't much of an aggressor and only changes when you least care about clock irregularities, so I wouldn't worry about itMCU_RST also isn't much of a victim with the internal 50k pullup or w/ejust on principle I'd move the 5V up to make room for the MCU_RST, and not sure what all those GND pins are all about
Why is the typical leakage current so different from the maximum leakage current, for a CD4053? It’s 0.1nA to 50nA, what the heck?
>>2077820with modern material purifications, required for the ultra-fine leading-edge processes, regular parts get really good, much better than the industry spec from the 1970s
Can someone help me troubleshoot?I was making a full wave bridge rectifier but my waveforms shows that it is a half wave which is a bit weird. I looked through my diode configuration and the positive and negative cycle paths are right, but I've also fucked up on seeing obvious shit in the past...
Here's how my thing looked like before I added in a ground note to the picture (Green drawing attachment) which then changed it into a half-wave rectifier, super fucking weird because if I didn't, the bottom part of the waveform is super nasty. I think I'm fucking myself somewhere
>>2077855>>2077840FUCK, I forgot to reply to myself. I'm gonna kill myself at this rate with how dumb I am.
>>2077854seems the problem is the scope. you cant display the 2 waveforms together like that coz they dont have a common ground. you should disconnect the AC from the scope, and just look at the DC - if the peaks are at 120Hz (100Hz in Yurop) then it's working fine.
>>2077860>you should disconnect the AC from the scope, and just look at the DC - if the peaks are at 120Hz (100Hz in Yurop) then it's working fine.How do I check the Hz on the peaks for DC, sorry anon I'm about to go bonkers
>>2077877dunno what software you're using. it's not something i've ever seen. somewhere in the options should be one to label the horizontal axis in seconds, milliseconds, etc.
Hey /ohm/,newfag here... can someone recommend me a good all-around soldering tip? I'm buying a soldering iron and need to choose a tip and I know fuck-all about what's actually useful. I'll be working with mostly SMD and some throughole stuff too.Thanks!
>>2077896best all-around shape is the small screwdriver tip, like D12 or D16 in yr pic. for long-life composition, you want copper core with iron coating, and maybe a chrome layer over the iron layer.
>>2077908Thanks anon! Any idea why there's so many different tips?
>>2077912the more tips you own, the more productive you are, and the more profitable the makers are.
>>2077815the GND is for a current limiting resistor for an LED, and also one of the xtal caps. I was going to say i can't move the 5v line up since that's the edge of the board but i moved some stuff around and switched some layers and got it to work. thanks for the suggestion anon.
>>2077896D12, d16, d24, d52. If you are desoldering with wick something in the c series. I use the k tips for tinning wire just because they are wide and you can trap the wire easy.
>>2077810Lmao nice ground plane>>2077855You can’t common the grounds from the AC side to the DC side, you’ll short half the wave. Leave the AC side ungrounded and measure it’s voltage with reference to the DC V- and DC V+, that should help you understand.>>2077908Seconding this, at least for small components and SMTs. Though if you’re doing large components, ground planes, etc, then you may want a D24 or K or whatever. K is fine for basically all THTs, D16 might be more comfortable to you though depending on how you like to solder. Conicals suck ass.t. owns K and D12
>>2077896D-1.6 is a good all-around tip but I've come to like C-2 and C-3 for QFP/QFN, and they're also good for any kind of flat surface or lug. the hoof tips (CF-x) are supposedly nice for larger fine-pitch QFPs but I never got the hang of the things. I like K for cleaning up bridges>>2077912as a rule of thumb, the tip should be about the same size as the work
>>2077855you're diode bridge is fucked up.re-draw it correct, note the direction of the diodes and what nets connect at which points.reverse D2 and connect the anode of D4 to D2 and the cathode of D2 to the anode of D3 which is also the original ground.remove the 2 spaghetti nets under D2.
Ohmbros, why is it that in the Laplace domain, when we do mesh or nodal analysis, we can treat inductors and capacitors as resistors when doing the math? I don't get the concept or the theory that leads to us being able to do this. It's super neat tho.
>>2078017take the equations for inductor and capacitor in the time domain and write them out.now perform a node or mesh analysis on it, you now have a differential equation.now look up the definition of a Laplace transform and what it does to differential equations.
>>2078012>you're diode bridge is fucked up.nope, the bridge is actually correct. it's just drawn so stupidly, that it looks like it should be wrong.
>>2078028>actual maththe /ohm/let's greatest enemy
>>2078032Differential equations are my bane. Complex numbers are my nightmare.And I love(hate) how electronics has both, so now I'm studying ODEs and Fourier Analysis...
Say I have a standard 12V computer fan that I need to extend the wiring on. What actually happens if I use wires that are much thicker? Like 12AWG instead of 20AWG? Sorry for the stupid question.
>>2078070nothing much, really. less voltage drop along the larger wire, but at a fraction of an ampere of current, E=I*R may not extend noticeably far
>>207807012awg is pretty thick, but nothing bad will happen.
>>2078030Apparently it's not because you arent getting the result you expect. I dont even want to look at your spaghetti mess because it's making me more angry the longer I look at itm 9/10 troll I want to strangle you.
>>2078105>it's not because you arent getting the result you expectI suspect the bridge is correct, but as seen in the green squiggle on the picture here >>2077855 he's commoning the grounds of the AC and DC side as I pointed out here >>2077953.More specifically, it means that when the top of the output transformer (node 5) is negative and the bottom (node 3) is positive, D4 and D3 will be conducting, meaning D1 and D2 will be reverse biased at ~12VDC each. But by connecting both ground connections on either side of D2 (nodes 0 and 3), he's shorting the 12VDC across D2 through the "ground clips".This is why I use single diode rectifiers.
how the fuck do these even work, I thought LEDs didn't have "resistance" because they're semiconductors and semiconductors don't obey Ohm's law. So how are these not effectively equivalent to a dead short?
>>2078151>So how are these not effectively equivalent to a dead short?
Anonymous 04/07/21(Wed)03:05:26 No.2078152▶>>2078133>This is why I use single diode rectifiers.which means you're only using half your transformer, and getting half the rated current, you half-tit nitwit.>>2078151you are seriously confused, and need to hit up wikipedia or youtube.
>>2078153Do you think I'd have to ask if I weren't such a retard?
>>2078154I did, but I wasn't able to find a useful answer anywhere. My last resort therefore is a norwegian ice fishing forum.
>>2078156well the VI curve above kinda explains everything, like only VI curves can do. if you feed a blue LED less than 2V, nothing much happens, but as you feed it a lil bit more, the current starts to shoot up, faster and faster, until it burns itself out. so it does have resistance, of a sort. a resistor would have a straight line climbing up; LEDs have a hockey stick shape. when you put the two in series, you get a compromise which works well most times.
>>2078154>Anonymous 04/07/21(Wed)03:05:26 No.2078152▶>>>2078133(You)>>This is why I use single diode rectifiers.>>which means you're only using half your transformer, and getting half the rated current, you half-tit nitwit.To clarify, I use single diode rectifiers more often because I find myself in one of two situations. Firstly, I don't have a transformer at all, hence I care about having one rail remain neutral, for safety and noise reasons. Secondly, I've got an overpowered transformer rated at hundreds of mA, when I'm just making a dinky little solid state audio circuit.>>2078156What you want is the schockley diode equation. Long story short, for any given current there will be a particular voltage across a particular LED. But unlike a resistor, this isn't governed by a simple equation like V = I*R, rather it's an exponential. Taking the slope of both V/I graphs, you'll find the slope of a resistor is just the resistance, but since the slope of the diode's graph is not constant, the effective resistance changes as a function of its input.For a more practical look at diode driving, you mainly need to know three facts:A: a small change in voltage will produce a large change in current because of this exponential curveB: increases in temperature cause an LED's forward voltage (at a particular current) to decreaseC: LEDs, alongside most electronic components, have a maximum power ratingAlso important is the following thought experiment. Say you have an LED that wants 1W of power, with a 2.5V forward voltage at room temperature. This will take 0.4A of current to do. If you put a 2.5V battery across it, 0.4A will flow through it as per the exponential graph. But as it runs, it heats up. As it heats up, its forward voltage will drop. As the forward voltage drops, the current will increase. As the current increases, it heats up more. This is a self-reinforcing cycle, known as thermal runaway.Using current limiting is hence a requirement.
>>2078162>a dinky little solid state audio circuitA dinky little solid state audio circuit that likes to have two voltage rails, let me state. Having two rectifiers of opposite polarity give me a split-rail power supply with just a single secondary winding.If I'm not making audio circuits, I'll probably be using some shitty SMPS instead.
>>2078161>>2078162I had somewhat of a grasp on this concept, but that's exactly why it's so confusing that you can just replace a car's incandescent bulbs with LEDs unharmed when their current draw *should* be absolutely astronomical at 12v.
>>2078184>their current draw *should* be absolutely astronomical at 12v.The exponential curve is steep, but not vertical. Pick a certain voltage, get a certain current. Maybe a couple of amps in normal operations, and at the extreme forward voltages without current regulation you might find 100A, but there's still ohmic parts of any LED that will limit the current at those extremes. Assuming you're using a sufficiently low duty-cycle that they don't burn out, such things can be measured.I'm 90% sure that common LED bulbs like that have some form of internal current limiting circuit inside them, be it a little CC switcher, a linear CC regulator, or maybe just a resistor for the lower power lights.
>>2078191>I'm 90% sure that common LED bulbs like that have some form of internal current limiting circuit inside them, be it a little CC switcher, a linear CC regulator, or maybe just a resistor for the lower power lights.This. Any drop-in LED replacement for a bulb is not just pic related in a case. Look at Bigclive's teardown of the Dubai bulb, for instance.
If you had a conventional diode that wasn't in an opaque epoxy case, would it emit any light when energized? (I don't mean same as an LED: I assume that if conventional diodes would emit light, they would do it less intensely or in less useful spectra.)
>>2078195>Look at Bigclive's teardown of the Dubai bulb, for instanceI would not conflate mains AC bulbs with automotive bulbs. AC bulbs require some form of power supply or limiter much more so than DC bulbs; I've heard of 12V COBs being connected directly to car batteries and not immediately dying.
>>2078196Well there are diodes in glass packages (DO-35) that may have a slight optical interaction. Silicon has a bandgap of 0.7V, meaning any standard silicon diode will interact with (absorb or emit) light at 0.7eV (one photon per electron), which is 1-2µm, near infra-red. But considering DO-35 diodes are small-signal diodes used in things like ideal rectifiers and sample+hold circuits, and aren't frequently covered in heat-shrink, I doubt there's much interaction with light at all. I imagine the geometry means any light emitted will just run straight into the other side of the junction or the substrate or whatever. LEDs are specifically designed to have their connections out of the way of the light path, after all.But photodiodes (dedicated light receiving diodes) will emit IR light when used as an LED, so will solar panels. LEDs will also act like photodiodes if wired such. But naturally, LEDs will work
>>2078203 - will work better as LEDs than photodiodes will, and vice versa.
>>2078196I believe most "non-LE" diodes emit very faint UV.
>>2078191I'm limited to 3 amps total. I'm gonna be powering a moped's entire lighting system with a chink buck boost regulator. Would a single inline resistor on the negative side work just in case?
Say I had 4 separate 12V devices each on their own 12V line coming from the same power supply. Is there a way or something I could use to switch everything off at the device end of the lines instead of the PSU end? I'm imagining a board that the 4 pairs of wires plug in to with a switch and then the 4 devices plug in to that board. I think it exists but I don't know what it's called.
>>2078206If you bought a bulb rated for 6V or 12V or whatever, and is designed to just screw into a standard light fitting, then it's going to run at that without any extra gubbins. Because such a light will have an internal current regulator.Look for a teardown if you're curious.
>>2078206if you're using the kind of cob shown in >>2078151 you dont need to add anything as they're made to be direct replacements for incandescent bulbs. that is, they will typically consist of (several sets of) 3 series LEDs with a small resistor in series to limit current. oh, and there's also a diode rectifier so they work from either polarity, just like an incandescent.
edison screw vs bayonet, which is better?
>>2078215soldered into place
>>2078215Edisons don't get stuck in place as easily but they're not as easy to service depending on the use case>>2078212Thanks, I'll take your word for it
Does /ohm/ know what "molded use" means?
>>2077855Anon, you've got both ends of D2 shorted to ground. You should probably oughta fix that...
>>2077661I just order by the 100s from 3 different AliExpress/eBay vendors.
>>2078251>I google big boob 12 trillion results
>>2077912Different jobs call for different tools. Sure you can have a jack of all trades tool, or do it with one that is of sub-optimal design, or meant for a different job.Just avoid conical tips, those are all-around terrible.
>>2078070It's better than using wire that's the same or god forbid, thinner.Think of wires like pipes. A larger pipe will carry the same amount of water, a smaller pipe will restrict the flow.
>>2078232Single = by itself (open air)Molded use = potted (covered in plastic resin or other conformal insulator)
>>2078259Thanks. My plan was to submerge everything in oil. Would that count as potted?
what's the simplest way to limit high voltage DC current that isn't a resistor? I want it to be completely passive if possible. I will take AC current limiting otherwise if DC is too annoying. it doesn't have to be super fancy dancy, the voltage can be unregulated I just need current control to keep it within half an amp or so.Im working on a project to make a bright ass light using 3 parallel rows of 6 100w 36v led cobs in series thermally glued to a big ass heatsink out of an old tv. I'm using recitified and filtered ac power just because it's extremely simple and cheap and compact vs 18 100w drivers (they're 20-30 bucks a pop, no thanks). I've tried it and it works, each chip gets 29-30v and the whole deal pulls 6-7 amps which is perfect, however of course being LEDs as they heat up they pull more current and will eventually sudoku.
>>2078264Eh. Have fun googling the properties of the oil you want to use. Might need additives. Probably yes.
Is a battery bank just a massive capacitor?
>>2078336caps block DC
>>2078163you mean a Cockcroft-Walton voltage doubler?>>2078197well, yes, they're at a favorable point on the V-I curve and localized heating is minimized by the substrate so thermal runaway would be less likely and take longer to occur>>2078307>what's the simplest way to limit high voltage DC current that isn't a resistorwith any simple way you will be turning excess voltage into heat and/or bouncing it back into the mains. you need to think primarily in terms of managing that heat to avoid the new thermal runaway hazard created by the regulator. decide if you really want to be doing that, or whether an xboxhueg buck or boost converter sounds easier to handle>>2078347>*blocks your path and converts your current into voltage*nothing personnelt.battery
>>2078336batteries are usually chem-based and capacitors are plate-based, but super/ultracapacitors seem to blur the line a bit
>>2078209So nothing like this exists? I guess a switch could only work if they were the same, probably a bad idea to connect 12V rails together, huh? I guess I could just use 4 switches, I just thought one would be easy.
>>2078307I assume you're on 240VAC, rectified to 340VDC? A capacitive dropper is definitely an option. At 8.3A total, anything linear is going to be experiencing 1kW of beans, not recommended. Any switcher, even three seperate 2.8A ones, is probably going to be somewhat expensive because of the high voltage, but still worth looking into. A standard line switching controller wired up with a beefy MOSFET and current feedback is what I'd look into first off.>>2078389>a Cockcroft-Walton voltage doublerno, this:>>2078562Draw a diagram, your explanation isn't very intuitive. You could use one switch that controls 4 relays or MOSFETs or whatever. Connecting rails together is not considered a good idea, at least not without linking the feedback networks somehow so they can share the load. But it's also considered bad to turn off different power supplies of to one circuit at different times. So you need to ensure your timing is sufficient. Relays might be fine, FETs definitely will be.
>>2078209why not just put switches on the + lead at the device end. most 12V type devices won't care. some devices get pissy when you plug signals into them while unpowered, but usually they will tell you so in the manual or you already know because you designed them
>>2078336>>2078347>>2078413In AC analysis you always short any DC power sources, do you only do this because batteries act like a giant capacitor and present a low impedance to any AC signals? For non-battery based switching and linear supplies they tend to have large output capacitors which short any AC. If you remove the capacitors will DC power supplies present a significant AC impedance that would need to be accounted for in AC analysis? More over does the frequency response roll off as you decrease frequency and should you account for this effect? Especially with linear regulators they tend to have small output capacitors which makes me wonder if you can really treat them as short with low-ish frequency AC signals (DC-200Hz or thereabouts).
>>2077953>you’ll short half the wave.This is exactly what happened wtf, is there a reason why I can't common the grounds from AC to DC? I'm pretty new to electronics sorry, but I'm trying my best! >Leave the AC side ungrounded and measure it’s voltage with reference to the DC V- and DC V+, that should help you understand.Huh, I'll leave it ungrounded like you said but I really don't get what you and the other anon meant by DC V- and V+ desu.>>2078012>>2078030>>2078105>it's just drawn so stupidly, that it looks like it should be wrong.PLEASE STOP BULLYING ME AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA, MULTISIM LIVE DOESNT LET YOU DO 45 DEGREE ANGLES OKAY, IT'S ONLY BACK AND FORTH, I WISH I COULD MAKE IT LOOK NEATER I REALLY DO>I dont even want to look at your spaghetti mess because it's making me more angry the longer I look at itm 9/10 troll I want to strangle you.I-I'm sorry... I wish I could do it neater too... and it's not a troll I swear! I wouldn't even have asked if my friends didn't all come across the exact same issue too.>>2078133>More specifically, it means that when the top of the output transformer (node 5) is negative and the bottom (node 3) is positive, D4 and D3 will be conducting, meaning D1 and D2 will be reverse biased at ~12VDC each. But by connecting both ground connections on either side of D2 (nodes 0 and 3), he's shorting the 12VDC across D2 through the "ground clips".Ohhh, that does help explain why my thing keeps fucking up, thank you anon.>>2078242L-listen bro, I'm not good at using grounds...
Ok what voltage rails do I pick for my modular synth? I know I want +/-2.5V for any 7400 logic or other 5V parts I want to include, but I’ll need higher rails to give me room to use JFETs and stuff, and better avoid the rails of my cheap op-amps. I’m thinking +/-5V or +/-6V would be good for compromise, but I’d like some other opinions. Many analog switches seem to not go much higher than 12 or 15V total, which is a pain. Meanwhile, JFET op-amps seem to want like 24V total for optimum use.
>>2078749>do you only do this because batteries act like a giant capacitor and present a low impedance to any AC signals?ideal voltage sources have zero output impedance. since you are not interested in dc potentials, you set them all to zero. a zero-volt, zero-ohm voltage source is equivalent to an ideal wire>>2078751there's a neater way to draw it>>2078757±15V is a common standard. +5 is usually relative to GND, which is fine as it is very easy to use a single opamp and a few resistors as a level shifter and gain block on the way out of whatever logic-level analog you are doing. >many analog switchescheck out the DG2/3/4/5xx series of analog switches which is an old part but well-sourced, they have independent supplies for the logic and signal path and have 36V or higher absolute maximum Vsupply range. DG21x and DG41x are popular. don't discount point-of-load regulation down to ±12V at the card or even lower, just be careful with the heatonly half jpoking, consider a small, quiet fan and baffles inside the cabinet to stir the air and promote thermal equilibrium and/or serious thermal management with water-cut front panel heat sinking plates and sheeit if you are serious about stability
>>2078778>+5 is usually relative to GNDI’ve decided not to do this, because it’s somewhat important that my CMOS logic outputs have 0V net DC, and saves level shifting.> an old part but well-sourcedJust what I’m after!> don't discount point-of-load regulation down to ±12V at the cardWhat do you mean? I plan on having 1-3 cascade-able motherboards, each with 8 slots for sequential effect daughter boards. Having one master power supply board before the first mobo is my intention.>thermal stabilityLmao the only transistors I’ll be using will be in saturation, I don’t think I’ll have to worry about temperature stability.
>>2078850>it’s somewhat important that my CMOS logic outputs have 0V net DC>implyingare you sure? complementary ≠ symmetric>I plan on having 1-3 cascade-able motherboards,ah, I was thinking that, in those places where you do need something lower than ±15 in a rare spot and you don't want to provide a connector pin, just stick an appropriately rated LDO right there from the appropriate 15 rail>Lmao the only transistors I’ll be using will be in saturationyou're probably ok then
>>2078874> complementary ≠ symmetricActually I had some issues with that in sims, when feeding a digital output into an integrator for a triangle wave. But high-pass filters might not be the answer either, I’ll have to think about what situations I’ll want to be using uneven duty-cycles. Guess I’ll do that just on the inputs of the digital-to-analog circuits, on a case-by-case basis.I still would find +/-2.5V to be a nice lower voltage rail to clip against or whatever other low voltage analog shit I gotta do, not sure whether it would make a difference to my circuits’ noise.>you don't want to provide a connector pinI’ve got nothing but connector pins. My rack system will use no patch cables whatsoever, just audio, pulse, and control signals already on the pinout for each board. Got like 20 pins on my connector. Was going to go for common 0.1” headers, but now that I think about it, I could use PCIE-style slots. What do generic edge-slot board connectors cost I wonder?
>>2078751Don't feel discouraged anon, this is a very easy mistake to make, and I'm surprised I don't hear of it more often. Any label you add to your circuit, be it GND or Vcc or a custom label, will signal to the software that the two points are connected. This is a convention from drawing circuit diagrams by hand, and carries through to digital layout diagrams and sims alike. This is actually pretty handy to remove clutter, such as in pic related.A rule of thumb I stick by is "only one ground reference/connection per circuit". If you have an isolating transformer in your circuit then you can have one reference on either side of it. Otherwise, if you want to measure between two points, measure the voltage across each with respect to ground, and subtract/take the difference of the two. All sims should have difference functionality. I've heard that Multisim is a good simulator with a lot of good parts from one anon who shills it here, so stick to it and learn its ins and outs.
>>2078565>Draw a diagram, your explanation isn't very intuitiveHere is a shitty drawing... Hope it makes sense.
>>2078934It's not stupid, and while an existing solution might not exist off the shelf, it would be reasonably trivial to make one.Are there 4 different 12V output power supplies, or just one? Does the main board need to be running even while all 4 peripherals are off? Also it's very likely that you can just have one 0V/black wire, and just switch the 12V wires.
>>2078937Yeah the main board needs to stay on at all times and theres only one power supply. I hadn't considered a single black wire but I don't know much about electronics beyond soldering. I think I could probably decipher a schematic and solder the circuit but it would look like hell. I've done that a couple of times but that was years ago. Most of my current electronics hobbywork is just changing plugs or extending cables.
>>2078966I think you'll need to elaborate what's in the main board. Because from what I understand you could just have one red wire and one black wire coming out of it, going to a single switch, and branching into 4 afterwards.If you don't want to do soldering for that, consider using terminal blocks or wago blocks or whtaever.
I am looking for some good general purpose micro compatible with arduino language to stock up for my projects. So far i have been using the atmega328p which is amazing and robust, but sort of stone age nowadays, and i want to do some shit like encryption in my new projects and 328p is hitting its limits on that one, so what would be a good replacement?I really like the esp32 but for most projects i don't need wifi or bluetooth so it seems a waste to use those, i really like the stm32f103 but for some reason they are not in stock much these days, and when they are they cost like five bucks, compared to the blue pill dev board which costs like $3 but i ain't buying and desoldering those, thats retarded.So which to pick?
Does anyone have a great idea what the most effective DIY heater on 12v would be? When I google I only get pajeet shit.I'm trying to keep my greenhouse above freezing temperature with a car battery during the night.
>>2079008They make ready-to-go 12V heater modules like:https://www.amazon.com/Fdit-Constant-Temperature-Heating-Element/dp/B07B26KKH7/...but you're not going to be able to do what you want to do with a car battery.For one thing, car batteries aren't made to be cycled. They're constructed such that they can provide very high currents for brief periods, but the tradeoff is that their electrodes turn to mush when deeply discharged. For another, even large ones only hold a few kWh. Heating takes a shitton of energy to begin with. Adding in the fact that this is going to be a very poorly-insulated greenhouse only compounds that problem. I doubt a single car battery will, at MOST, last more than a couple hours at the required power output, even if you abuse the thing by draining it beyond 80% DoD (which we've already established will quickly ruin a car battery, anyway).
>>2079008>DIY just run current through a wire with correct amount of resistance
>>2079007STM32s are a good choice for power, but they're expensive because of the ARM licensing from Acorn computing. Yeah, the guys who made the shitty computers you had in primary school with the circus game. The bluepills are cheap because they usually use knockoffs that may not necessarily work perfectly (https://github.com/keirf/Greaseweazle/wiki/STM32-Fakes). But if you know what you're doing, you should be able to buy a bunch of knockoffs that you know to be fully functional. IIRC there was even one knockoff F103 that corrected a bug in the original. One I'm hopeful for is the GD32V, a RISC-V implementation of the STM32, thus avoiding all the ARM licensing, though I don't think it's easy to buy yet.If stock is an issue, there's also the STM32F4/G4/F7 series, which are better at DSP among other things (https://stm32-base.org/guides/getting-started).Look on LCSC for more legit parts than you might get from alibay if you aren't already, possibly better stock too.
>>2079008>above freezing temperature with a car battery during the night.forget it, not enough energy density.you need propane.and fuck global warming.your tomatoes are more important.
>>2079008>>2079041A refurbished volt/leaf/aqua battery block might be a good option, plus a whole bunch of solar or a large wind turbine or whatever. A heat-pump like an old fridge compressor (even peltiers) will be more efficient than resistive heating, if he wants to go that way. Heating is pretty energy intensive in general.Easy to do the calculation if you log the temperature inside and outside the greenhouse over time. Calculate the total thermal resistance of the greenhouse from both: the rate of change of temperature divided by the temperature difference (should be approximately constant), and an estimation for the total thermal capacity of the greenhouse (volume times air's constant). Then integrate over the difference between the outside temperature and your desired temperature to figure out the total energy that will be required to keep the greenhouse warm enough each night. Your battery bank or propane tank will need to support that energy, multiplied by 1.5 or 2 or some other safety factor.
>>2079008Put huge barrels full of water inside your greenhouse.They store up thermal energy during the day and can keep it from freezing during the night.
>>207905190% of water's buffer energy is in the latent heat of fusion, so you'd want that to be significantly above the temperature you want to retain the greenhouse at. So maybe there's something cheap you can add to water to increase it's freezing point to 5 or 10C. Or not even use water. Whatever it is, it needs to reliably melt completely the next day, otherwise it's useless.
>>2079031do you have experience with stm32? one thing i don't like is that all minal stm32 circuits i can find all feature like 10 fucking external caps. are stms capable of running standalone without any externa components like for exaple atmega328p does? (apart from adding external oscilator)
Should I go with 0.020" or 0.015" diameter solder of pic related? I've only gone through a couple of tiny 18g rolls of 0.031" solder and noticed it was a bit thicker than I'd like. When I wanted to lightly coat the soldering tip with solder, I would struggle to get anything other than a giant glob onto it. I'd be using it for through-hole rework, mostly, if it matters.
Update on the Timing light.The BJT was indeed the culprit, after swap i immediate noticed the circuit only pulling 4 Watts instead of 30, then measured the HV rail which got charged to around -400V.
>>2077912>Any idea why there's so many different tips?I've been soldering for 20 years off and on. I was taught by someone who supervised the construction of the radios on the Apollo missions. I still have no idea what most tips are for. Most of the time I find myself preferring whatever one is already on my iron. A tip has to be extremely unsuited for the job, before I will decide to not use the one that's already hot. The guy who taught me had an old school 60W Weller kit with about 5 tips and a heated solder sucker attachment. If you have no idea what that tip is for, you're in the best of company. The only wrong tip is the one you can't make work.
>>2079129they're about the same. the smaller one is usu preferable except when you need a lot of solder, like to tin a fat wire, in which case it's annoying to have to feed so much line to it.>>2079132>The BJT was indeed the culprit,so i was right 2 threads back."come on dude, power semiconductors are almost always the first to die, not caps, and never ever transformers. coz semis are faggets."
>>2079141>so i was right 2 threads back.and i waited multiple days for the spare part. Well at least it was no cascading fault.Still baffles me how that thing is toast on a light out of the box (and going by amazon reviews that being a cripdeath is not uncommon)
Tips for learning how to solder? Everytime I have to wire header pins or similar microcontroller stuff, I end up with a niggerlicious mess and giant blobs of solder.Recommendations for cheap soldering equipment would also be nice, I only have a crappy old 5W iron and cheap solder.
>>2079158all the skill in the world wont help if you dont have the right equipment: clean quality tip, hot enough iron (15W+), brand name rosin-core leaded solder. as for how-to's, youtube has 'em by the millions.
>>2079158-Keep tip clean and shiny with a fresh coat of solder. Wipe off oxidation frequently>header pins-The pin is most of the mass, so focus the heat on it-Don't melt the solder with the iron, Melt it with the part you're heating-Pre tin parts and traces >5W iron 5? doesn't seem like that'd be enough heat. I still use an old weller 40W? I think-A higher temp Iron can put less damaging heat into parts If you're quickfind an old circuit board and remove componants to get the feel of heat transfer
>>2079158step 1: (the most important one)buy TS100
Most everything about frequency domain transforms make sense to me. All but phase.For example, with these diagrams from a DSP textbook, why is it that the phase is variable despite the sample being a single impulse? I know it's periodic because the discrete, sampled nature of the signal causes the frequency domain to wrap around itself from circular convolution. But even if you unwrap the phase so that it's a single line... why is it not constant?I guess what I'm asking is, where do they get the phase from? The same thing happens with filter charts. I understand the frequency cutoff, I understand that filters offset the phase. But how can the phase change with frequency and not just fuck up the entire waveform?
>>2078311>>2078565how would a capacitive dropper work against an ever increasing current? I know they get used in cheap and dangerous Alibaba phone chargers and shit but in those cases it's a fixed current draw that essentially never changes.the reason I chose a the six chip series configuration is so that voltage control is not necessary since my power supply (the wall) is that a fixed voltage that cannot increase, current is the only concern.no 120vac->170vdc which is how each chip gets about 30 volts.>>2078389I could just get 18 LED drivers and be done with it, but that would run like $600 in which case I could just buy 2 or 3 brighter than the sun studio lights and not have the fiddle fuck around.yeah it seems unfortunately passive regulation without generating a million watts of heat isn't likely. I saw on scamazon some cheap switch mode constant current regulators using an LM something or other, pretty cheap can get 10 of them for like 15 bucks the only problem being that they have a 3 amp output limit and while that's up 30 volts which would underdrive the LEDs and that's fine cuz it's already what I want them at I need to double check the current draw at that voltage because I know running them at full powah will make them pop in short order. I don't mind having a regulator for each chip in fact it would probably be safer since if one failed open it wouldn't let the others take in the excess voltage and explode, it's just like I said I don't want to spend a million dollars on it and I don't want to be changing blown ones every 2 days.
>>2079007the GD32F103 is almost 100% compatible with the STM32F103 and could be easier to find. also STM32L0 and F0 are pretty good chips (about 50% slower maybe) which might have power enough for the stuff you want to do>>2079031>expensive becausethat, and the world domination pissing match>>2079072not that anon but I do>10 external capsdecoupling caps are normal for actually powerful chips. just learn to solder 0402 caps right onto the leads if you're so tetchy about having a board>>2079129I do a lot of 0402 SMT and 0.5mm is what I use>>2079158>I only have a crappy old 5W iron and cheap solder.>I end up with a niggerlicious mess and giant blobs of solder.cause and effect. for cheap irons that are capable of producing professional results, look up a 936D clone on alibay>>2079195I am sympathetic to your aims but you have no idea of the magnitude of what you are trying to do. you are creating a lot of heat and you don't seem to have any realistic plan for dealing with it. a free-air heatsink is not going to keep LED dice within operating spec while dissipating 600W of thermal power. you need water blocks or custom metal for thatit's not so simple, but Microchip has some interesting LED driver controllers they acquired from Supertex a long time ago, in the HV99xx range. HV9910C is classic, unfussy, 42 cents at Digi-Key and 15 cents on ali, requires an external MOSFET which you would have to select. but you could build six separate 300W buck-style drivers with accurate (and dimmable) current regulation for about $10 each depending on board house and parts source. or you could throw caution to the winds and build one hueg 9A constant current driver using the same principles and beefier parts (if you can find an accurate current sensing shunt resistor)
>>2079234but can it work without the caps or no? because there is no way in hell i'm soldering million caps every time i want to use one of those fuckers
I have a project where I need to use a tubular hall effect sensor/magnetic pickup like pic rel. it's a diy efi system on an old points small engine. I am forced to use this type of sensor because the engine controller requires it, unfortunately you can't just use the points opening and closing as a trigger.so my idea is this: the intended sensor is normally open and the magnet passing by closes it, so I'm thinking of using a normally closed sensor and wrapping it with some wire to turn it into an electromagnet. I will run this wire to the output of the primary ignition coil which is always energized by a ring of magnets in the flywheel and so the sensor will remain magnetized, then when the points gets opened the field collapses and the sensor closes and it triggers the ignition, then the points close again and re-energize the electromagnet and open the switch.does this sound like it would work? and no unfortunately I can't use the Ring of magnets as a trigger, they're on the inside of the flywheel and there is no space in there for the sensor and you can't get a good signal from the outside. Plus I doubt the software would be able to be configured for something like that.
>>2079237>can it workstick to dev boards until you're willing to do electronics>>2079256can you just logically invert the Hall sensor output
>>2079259so you don't know?i'll guess i'll give a try and see if it works or not, it would be great if it did since stm chips have tons of features
>>2079260you'll be soldering 48 pins anyway, what's 10 morenot following manufacturer recommendations is a stupid game with stupid prizes. it's not that we don't know, it's that capacitors are cheaper and faster than the alcohol needed to troubleshoot intermittent power-rail-related problems
probably not the best place to ask but i don't have many options so here i goI recently picked up a nakamichi receiver and it came with a remote, which unfortunately doesn't work. i opened up the remote earlier and found some small corrosion and a part of the contacts clearly had scorch marks on it. i was very hopeful thinking cleaning the corrosion would be all it needs to work again but unfortunately that is not the case. the pcb appears to have no damage to it so i'm not sure where to go from here and i'm not too knowledgeable about this. any suggestions on what to try next?
>>2079259the point isn't normally closed versus normally open, it's that the engine does not have a magnet to trigger the sensor unless I glue one on which I don't want to do.
>>2079271Replace the resonator (blue component).
>>2079274ahhhh>output of the primary ignition coilwhat about a small permanent magnet on or near the sensor somewhere, maybe even behind it? so that you have some consistent level of flux variation that could be sensed that doesn't vary with RPM or crank position, or run the risk of EMPing your sensor due to misfire>>2079277that's probably a tag tant, look at the board
>>2079277>>2079288here's a closer look at the board if that helps
>>2079288I'm sure if I put a magnet on the sensor in any way to be significant it would just stick it open (closed) or mess with the waveform in some way. plus it's meant to rapidly go in and out of a magnetic field anyway so I don't think there would be much point in trying to clean or stabilize it like that.if i were to end up with enough of a power variation to affect operation I would just switch to battery power across the points, I'll need a battery anyway to power the electronics. points can certainly take the current seeing as it's meant to be closed for 80-85% of each crank rotation.the primary is the low voltage side so I'm not worried about any emp effect, plus I plan on using normal jacketed wire for ruggedness and wear protection since it will be on an engine/vehicle after all.I guess calling it the primary is a bit of a misnomer since it uses a remote ignition coil whose primary is fed by this coil under the flywheel, it's more or less a lighting coil like pic related that just supplies constant power to the primary of the ignition coil through the points. essentially a mechanical electromagnetic battery.
Are there trim caps in the nanofarad range? I can't find any.
>>2079256do i understand correctly you need a high signal when the spark fires?Because that would be pointless as fuel has to be injected in the suction airstream, plus if we are talking 4 stroke simple engine the spark plug will fire 2 times per cycleyou can steal a signal from the spark wire as done on all revolution meters ever build for small engines.You could even attach a wire directly to the sparkplug and run it through a series of resistors into a transistor, without inductive pic-up
>>2079168>>2079175>>2079186Thanks, I'll look into buying a proper iron. How terrible are those no-name chink stations (Yihua, Yaxun, whatever)? I'm a poorfag student, so can't afford a Weller/Hakko. I wanted a rework station to do small repairs on phones and stuff, but if the quality is too shoddy, I'd rather get a better quality iron.
>>2079376as was said, get TS100 it costs around fifty bucks even pajeets can afford that and it's one of the best irons on the market
>>2079342no. ok forget everything else, let me just boil it down the the bare essentials. in the included pictogram would the following be true or incorrect:1) would the coil of wire around the metallic tubular hall effect sensor (PNP, NPN, NC, NO, doesn't matter) create an electromagnet when energized2) would the tubular hall effect sensor pick up this magnetic field and go high or otherwise change from its default output state3) if you were to press and hold the momentary switch to open the circuit between the coil of wire and the power source, would the magnetic field collapse and the tubular hall effect sensor go low or otherwise switch back to its default output state4) if you were to rapidly press the momentary switch multiple times a second, would the output of the Hall effect sensor be a waveform similar to the one shown, same as it would be when it is passed in and out of a magnetic field as intended
>>2079397f u c k i n gp i c t u r e
Anyone have any ideas why my scope's bandwidth is so unexpectedly shit compared to my SPICE sims? If I put a 100nF in parallel with C5 the bandwidth gets significantly better, but I'm not sure what's going on exactly, or why I would need that much in the first place. Bottom table is the probe specs, not that I think they are the dominating factor in this case, but I figured someone might be able to correct my probe circuit if I got it wrong.
Got some silly questions about amplifiers if anons don't mind answering. Please excuse my mouse drawing.1) Why is gain in decibels? Is it because it gets hard to understand what a number means without units so we attach dB to make it clearer? dB had always been used as a 'sound' measurement unit for me so it's weird seeing it in electronics.2) Why is this called A_vo sometimes called the open-circuit voltage gain? I get the voltage gain part but why the open-circuit in particular? Is it because there is no load attached to it yet?3) Why is R_out placed in the basic amplifier circuit? I know that R_in is so that we can figure out the voltage difference between the inputs then amplify it but I'm not sure what R_out is for.
>>2079195>120vac->170vdcOh, you're doing it like that? 170/6 = 28V each, so you're not underdriving them too badly. A capacitive dropper will absolutely not help here, because even if you wanted to drop ~10V across one at 5-6A, that's a reactance of like ~2Ω; a massive value capacitor would be needed. Now you may have guessed this, but there's a non-zero chance that you can run your array with no current limiting whatsoever because the LEDs are being underdriven. They'll definitely heat up and consume more current, but this may not be enough to start thermal runaway. I don't suggest you test this though, as there are other fluctuations to care about, like mains voltage changing from loads on your local grid, or temperature changes.What I'd do is make some linear constant-current drivers, using darlington/slizzy transistors clamped to a heat-sink, and TL431s or whatever. Designed such that, if insufficient current is being drawn, their dropout voltage is insignificant, only if thermal runaway starts to happen and they draw more than ~5A or whatever do they start to increase their current draw.>>2079237You can probably get away with just one cap from 0V to Vcc, assuming all your digital rails are the same as your analog rails.>>2079256Sounds like you're designing something like a guitar pickup. Putting a permanent magnet next to an analog hall sensor and measuring the fluctuations in the magnetic field as the thing rotates is what I'd do. Need a high-pass filter and amplifier after the sensor. Or use an actual pickup instead of a hall sensor in the first place. An open-core inductor being fed with a constant-current source would also work, instead of having the coil and magnet seperate. While the field would be weaker, you'd be able to tweak it in case the engine spins too fast or too slow for that magnetic field strength.>>2079392Plus the cost of a 24V power supply. Running it on 19V is only 70% of max power. T12 is cheaper, if not portable is fine.
>>2079446>Why is gain in decibels?Gain accumulates over stages. An amp with a power gain of 10 followed by another with a power gain of 4 will have a total gain of 4*10=40. 10x is also 10dB. 4x is also 6dB. The total gain is 10+6=16dB. You turn a multiplication in to an addition. Makes mental math easier, so it's easier to understand what's going on at a glance in a circuit with 10 stages.
>>2079446>Why is this called A_vo sometimes called the open-circuit voltage gain?The voltage gain goes down when you load it. Yeah the dependent voltage source sees an open circuit.>Why is R_out placed in the basic amplifier circuit?It's the theveinin equivalent output resistance. It models the fact that voltage gain goes down with loading. A short circuited ideal voltage source creates infinite current. A short circuited circuit creates finite current. It's about turning a complex nonlinear circuit, and estimating its behavior with a simpler linear equivalent. No amplifier is linear, but under certain circumstances it's valid to think of it as linear.
>>2079405You're on fucking AC triggering>>2079446>Why is gain in decibelsdB are useful in the context of signal propagation, especially radio. They're a logarithmic scale too, meaning they give a certain order to a bode-plot, where both axis being logarithmic give a linear plot.>open-circuit voltage gain>because there is no load attached to it yetNo load, but also no feedback.>R_outInput impedance tells us how much current will be drawn into the input. Output impedance tells us how much the output will sag, which results in a decrease in gain, and a wasting of power. All circuits have both. For low-frequency circuits, you want to maximise input impedance and minimise output impedance (see op-amps). For high frequency circuits however, you want to match them to prevent signal reflections.Also yeah >>2079468 thevenin.>>2079462>gain of 10 = 10dBDon't you mean 20dB? This is V, not W.
In this schematic the opamp is supplied with the 3.9V power from the buck converter and not the cleaner 3.3V from the LDO. Any idea why they did this?
>>2079500They've got decent PSRR either way, and the A input can be as high as 3.9V when Q1 is turned off.Also what a mess.
>>2079187The phase is measured in degrees/radians, not time. So if you move a 1khz wave 1 second, you moved it 1000 cycles but if you move a 2khz wave 1 second you moved it 2000 cycles.
>>2079129Is that what kester looks like now? I haven't bought a new roll in years.
>>2079338that's kind of big. largest I've seen are in the 100pF range. there are circuits to multiply capacitance which may suit your application>>2079376they're actually not bad, and they're pretty durable. tips are cheap. TS100 is a meme, don't bother if you aren't a field tech or a weeb who's run out of Hello Kittyana to buy>>2079500well-designed buck outputs are quieter than you might think. also minimum PSRR on that device is 80dB (1/10000 of supply ripple bleeds into the output), no need to waste too many fucks on it>>2079514if you're gonna pay $30 for a roll of wire, they better make it pretty
>>2079376I have a Yihua, It's pretty solid, I haven't used the hot air gun much though other than for heat shrink. The iron is nice though.
I'll be honest I still don't understand why N-MOSFET can work as an inverter with:Input low -> output highInput high -> output lowIs it because if VGS > Threshold voltage, it will conduct and then lose a bunch of voltage through the drain resistor? If that's the case I can understand why input high -> output low, but I don't get input low -> output high, that doesn't make sense because there are no amplifiers, but I'm retarded so I know I'm missing something crucial.
>>2079662>Input low -> output highFET is off, treat it as open switch. Your output is pulled up to VDD by Rd.>Input high -> output lowFET is closed switch. FET in saturation is much lower resistance than Rd so the output will get pulled to ground.Don't think about inverters like linear amplifiers. That is completely the wrong way to approach it. In an inverter your FET is a switch, you are always operating in either saturation or cutoff.
>>2079662>doesn't make sense because there are no amplifiersthere's most of one
um, is that fet depletion mode, or just simply drawn?
Haven't dabbled in electronics for a loooooong ass time was getting into opamps and was wondering how youd calculate the effects of bias and offset current for this thing.Ive tried writing KCL equations for the negative opamp input but all of them result in values WAYYY off the money.Can anyone help a spergy retard out?
>>2079766can you be more spec?
>>2079771Pretty much im asking what the bottom 5K resistor does. The grey KCL line.I get the circuit on the left, its easy because it simplifies down due to the ideal case there the negative input voltage (v-) is equal to the positive (v+) which is shorted.However once the bottom 5K resistance is added in v+ is no longer 0V and I can understand how it effects the KCL
>>2079295that scorch mark could be blocking current, so you need a voltmeter to verify 3V is actually getting to the PCB. you can measure across legs of C4, which is in parallel with battery. push hard coz C4 legs are prob dirty/oxidized. next, dip q-tip in 75% alcohol and lightly rub/clean all the black contact points for the keys on the PCB. that's those interlocking L shapes. also, clean the black dots on underside of rubber keypad with warm water and dish soap. just rub with fingers, dont use anything harsh like toothbrush.when all is dry, test it by pointing LED at a working camera or webcam. if working, you'll see the LED flashing on camera when you press a key.
>>2079776I haven’t read the specifics of your post, but since V+ = V- then I+ = I-, hence the voltage drop due to the input resistance and that current, multiplied by the op-amp’s gain, should give you your offset. I’m assuming that’s what you did and found it didn’t correspond with reality? Perhaps it only applies to the first of multiple gain stages in the op-amp.I think, for a qualitative understanding anyhow, that you’d be better off looking at an equivalent transistor diagram of an op-amp, specifically at the long-tailed-pair.For quantitative, I couldn’t tell you, I’ve never even used such an extra resistor.
>>2079809Ok yeah I just calculated that, with 10k input resistor and 20nA input current, a 741 op-amp’s 200k gain would result in 40V extra on the output. Though of course the negative feedback of the op-amp would counteract that. I wonder what effect the extra resistor even has, aside from setting a perfectly precise gain?
>>2079691lel, it's NMOS>>2079817it develops that 200µV which is now part of the input and will be multiplied by your feedback gain. the idea is that the inputs see the same impedance so that the input bias currents (which are equal by design and testing) will develop the same offset on each input, nulling that source of offset
>>2078307>what's the simplest way to limit high voltage DC current that isn't a resistor? I want it to be completely passive if possible. I will take AC current limiting otherwise if DC is too annoying.https://en.wikipedia.org/wiki/Current_limiting_reactor
>>2078307>what's the simplest way to limit high voltage DC current that isn't a resistor? I want it to be completely passive if possible. Difficult requirements.See if you can find a lower steady state voltage that puts you well out of thermal runaway territory but still emits light.I tried to trigger thermal runaway when I was playing with LEDs, I was using a precision voltage supply and I wasn't able to get my LED to commit sudoku even if I was blasting it with a heat gun.
>>2079832>lel, it's NMOSI know, still doesn’t tell me if it’s depletion mode or not.>multiplied by your feedback gainOh that makes more sense. So the result is a permanent DC offset if it’s a constant current leakage, but if it’s more like a resistance to ground then it would be a slight difference in gain, right?>nulling that source of offsetYeah I already understood how adding that extra resistor would make the two voltage drops equal and hence trivial for a differential amplifier.>>2079854There’s a really neat demo, where you get a high-power incandescent lamp, a couple of metal stands, and a rod of soda-lime glass. Make the circuit with mains such that the rod sitting on the metal stands is in series with it. Blast the glass rod with a propane torch until it starts to glow, both the rod and the lamp. As the glass gets hotter and hotter, the lamp will get brighter and brighter, until at some point you’ll be able to take the torch away and the lamp will keep getting brighter, until the glass rod melts in the middle and goes open-circuit.
Does anyone have experience with DAC/ADCs? I've bought a 16-bit ADC for a project but is that a suitable resolution for audio signals? Youtube is full of audiophile videos so I don't think I'd get a good answer there.
>>2079503>>2079593If so then I wonder why they even bothered with an LDO in series.
>>207987016 bits is what audio CDs use. nobody needs more unless they're audiophool types.
>>207987016 is standard for consumer uses.Pro audio is usually 24 or 32.
>>2079870CD quality is a good standard, it's 16 bit at 44.1kS/s. Some people go up to 24 bits, and it's nice to have audio masters stored in a higher quality for mixing and mastering, but you almost certainly won't notice the difference as an end-user. 12 bit sounds acceptable if you've got shit speakers, 8 bit is landline quality or worse.Sample rate matters also, if you want it to be audibly indistinguishable from reality, it needs to be at least twice our maximum range of hearing; 20kHz. This is Nyquist's sampling theorem. Hence these sample rates being a bit over 40kS/s. S/s is just samples per second, btw. Some ADCs are obviously just meant for low sample rate data acquisition, so ensure you avoid those. Note that Nyquist's sampling theorem in this instance kinda implies that the ADC will assume that high frequency waves are sines, and that the DAC will produce high frequency sines. For this purpose, you need analog filters on both devices. The antialiasing filter before the ADC needs to strip away everything above 20kHz to a significantly low amplitude before the f_ny/2 point (22.05kHz in a CD's case), and the DAC's output filter should do something similar too, especially in the event that the DAC outputs square waves. In the DAC's case you can spoof it with some sort of notch or elliptic or whatever filters targeting the harmonics of the sample frequency. There are also some ADCs and DACs that deliberately oversample and have internal digital filters to cut down on aliasing without needing a high order analog filter. You can also manually oversample them and use DSP filtration in your MCU.There's also something to be said about protocol. I2C isn't terribly ideal for ADCs/DACs, especially not continuously sampling ones at a high frequency. You'll want SPI or I2S, or parallel I guess.Then there's nonlinearity, distortion, noise, etc. I'm not too sure how to classify these, so read datasheets and appnotes and such.
>>2079873The 3.3V is fed to the OLED, which will have a potentially noise-sensitive controller in there. It also goes through the control header alongside the USB data lines and the I2C lines, probably to an MCU. They even went through the effort of adding a double-sided LC pi filter, separating the noisy SMPS's ground from the MCU and OLED's ground. Though noteworthy is the fact that the USB ground is the same as the SMPS's ground, I guess the MCU has a seperate ground for USB.
>>2079877According to this it'd have 4 groundsKinda doubt it though, wouldn't be surprised if they're all connected to the same ground
>>2079879Wait how is the element being switched here? The only thing I can see is Q1 being switched by Q2, but Q2's base is AC-coupled, which would make PWMing it rather difficult. Unless it relies on C24 to charge up in a certain time, and by varying the duration between Q2 pulses, you vary Q1's duty-cycle. But feeding that capacitor ramp into a MOSFET gate is asking for switching losses.
>>2079881The iron runs on DC though
>>2079884It still needs to switch the element on and off depending on its temperature. Doing this very fast makes it a lot smoother to use. Why do you think there's an STM32 in there in the first place?
>>2079593>there are circuits to multiply capacitance which may suit your applicationThanks for the heads up.
i have bought induction cooking stove for my apartment but the manual says you have to hook it to a 3 phase power but my flat has only 1 phase, where can i get some sort of phase splitter to turn it into 3 phases?
>>2079881That circuit is pretty nuts. C24 will keep the gate low for a little while, so Q2 only has to come on every now and then. I would guess this is a safety feature so if the MCU stops toggling Po (due to software crash or whatever) it will turn itself off instead of possibly staying on forever.The gate isn't driven very strongly so there is a danger of thermal runaway, but I guess they have run the numbers and figure it will still turn on and off fast enough to avoid blowing up.
>>2079879Similar design but easier to read
>>2079928>the manual says you have to hook it to a 3 phase poweri doubt it says that, unless you're in Finland or some such place, where 3-phase is used in homes. much more likely is that it requires 3 wires coz it needs split phase 240V plus neutral. this is the norm in N. America.
>>2079949i'm not retarded it's definitely 3 phase this is from the manual
>>2079968ok, from the writing, seems you do live in Europe where 3-phase is used at home in some places. as to your question: phase converters exist but given the power requirements of a stove, they'd cost like $800, which is probably more than the stove.
>>2079606>yihua>yi is pronounced yee>hua is pronounced hwa>yee hwa>tfw
>>2079928rotary phase converter or a 1 to 3 phase vfd. but your best bet is just returning it and getting the proper one.>>2079968it looks like it's 3 phase and 2 phase/split phase capable from the picture.
>>2079405My rise time on a 1982 scope. 1ns@50 ohm
What are my choices for a tiny color OLED these days? Still just this thing?
>>2080080I like this 128x128 TFT
Buy BAKU soldering station locally, or buy pic related from china? Difference is just 25 bucksIs pic related any good? Baku seems to have 24V iron, while this seems to have mains iron and mains hot air.
>>2080146picrel looks ok, actually. only concerned about how to set air flow on the air handpiece>mains ironthe air handpiece is almost always mains anyway. 900-M style irons are typically 24Vthe set of accessories is decent. the brass wool is excellent for tip cleaning and the tip assortment is good enough for a broad variety of work
>>2080207>picrel looks ok, actually. only concerned about how to set air flow on the air handpieceButton thingy.And on local chinkstation I get 3 knobs. Which might be better for BGA rework, because I'd be able to follow temperature profile much closer to avoid issues with dead cpus.
>>2080213oh>knobsbuy this one
I've never seen non-electrolytic caps bite the dust. Should I simply replace or is this indicative of more problems? The device 100% works.
>>2079929So I was right? Shit, can't believe they'd torture the FET's gate with that kind of capacitive-ramp waveform.>>2079968Eh, it's probably fine if you just feed it single phase from a high-power outlet. Or split phase, in case there are any euros with that. It will just be rectified in the input anyhow.How many watts or amps is the thing rated for?
I have a project to build an audio-amplifier with tubes, and I can get all parts but for one: A LTEI19/KD-0703 audio transformer which isn't produced anymore. It's 300 ohm, and the only alternative I can find has 600 ohm. Can I use the alternative or does anybody know something better?
>>2080298>I've never seen non-electrolytic caps bite the dustpowerful self-healing properties>>2080367no way to know without seeing the stages in question. better yet, post whole schematic
>>2080441I can't do that due to copyright laws and shit, but it's this one: https://www.elektormagazine.com/magazine/elektor-201401/24350I'm a bit worried though as people say that the way it's built there may be noise on the output from the power supply...
>>2080367Any 300 ohm audio transformer with the same ratio should work, doesn’t have to be a similar model at all. If you can’t find even that, then you could wind your own, or chain a few in series/parallel/whatever.As for the PSU noise, I’d consider swapping out the PSU circuit for a different design, one with more linear regulation and passive filtration.
>>2080298>I've never seen non-electrolytic caps bite the dust.Rifa caps are the bane of every vintage computer restorer. Yes, those fuckers do fail. I don't know how certain they are to fail—basically everyone in the retrocomputing forums will say they're guaranteed to pop as soon as you turn them on—but yes, it's likely just your cap and nothing else bad.
Hey /diy/I wanna know if it's safe to communicate with multiple 74 series logic chips over 20cm wires at 4MHz speeds. Will I lose signal integrity? Has anyone else had a similar issue here?
>>2080589Which 74 series, I’m guessing 74HC? Twisting wires together is advisable, but otherwise you should be fine. Look for some appnotes if you’re curious.
>>2080452>I can't do that due to copyright laws and shit,what a sissy, pantywaist, lily livered, cream puff. the entire global internet was built precisely for the purpose of sharing copyrighted stuff.
>>2080677>ideas are freeanon gets it. I love you
>>2080677No, it was built for cyberpunk particle physics researchers to share data from lab to lab. Could write a cool story about that.
I know PCB fab houses are commonly used by you guys, but has anyone used a CNC fab house? For metalworking mainly. I want half a dozen prototype parts made of both aluminium and steel, and investing in a CNC mill is simply not an option at this scale.
>>2080781metal shops dont deal with shrimp. they will snigger and call you 'Miss''get big and dirty or forget it.
>>2080792I'm talking about the dedicated one-off prototyping shops. They charge a premium, but there's still certainly a market for that.
>>2080781yeah people do this all the time in the keyboard community
>>2080489This anon is helpful. >>2080677This anon is a toddler
>>2080498Thanks! I had no idea what a Rifa cap was, I've just really with polyester audio filtering caps before.
>>2080813I see, might I have luck asking in /g/'s mkg general, or are they not /diy/ enough?>>2080817Note that a different transformer and/or PSU may require redesigning the PCB somewhat. I assume the article itself has the actual circuit diagram in it, which I'd copy across into KiCAD and turn into a PCB layout that you can edit. Would only take a couple hours. I'd consider converting it to a more SMT-styled board too.
>>2080823/diy/ has a machinist general that could help. they would probably know the details of that ordering process, the expected file formats, etc.
>>2075079Im a newfag here, but not really new to electronics. Gonna try a project from the roll list. ROLL
what's /ohm/'s favorite connection for board-to-panel-mount switches/pots? i can't decide between 1/8" QCs or a couple of 3 pin molexes
Threadly reminder I love my ohmmies
Follow up on the AC-DC common grounds diode issue I posted way back (>>2077855)>tell professor I finally got it to work>professor goes "huh, but yours doesn't look like mine and my ground is also connected to AC mains and the circuit side">he has AC-DC common grounds too>he shows me his waveform>looks way too clean>realizes it is way too fucking zoomed out>"please zoom in for me">he had the same fucking issue but didn't see it because he was so far magnified out that it looked like a clean waveformHeh.......Here's your silly greentext and an anime reaction image, I'm never doing it again to preserve the integrity of this place.
redpill me on building floyd sweet's vacuum triode amplifier from readily available parts
>>2081452>vacuum tubes>redpillanon, i...
>>2078778>check out the DG2/3/4/5xx series of analog switches>30 datasheets later I'm still tabulating the relevant specsWHEN WILL IT END
Ok now to do the other common series
>>2081634Oh and damn are those DG417, 18, and 19 good. 20 (35max) Ω and 0.1 (0.25max) nA, should make for a very comfortable sample+hold follower + shorter all in one. In my case, I plan on making a meme synchronous voltage follower, pic related.
you know how newton supposedly saw an apple falling and got his idea to mathematically define gravity?i like thinking that einstein did the same thing when he thought up general relativity, except for him it was seeing a man commit suicide by jumping off a buildingbecause of the "consider a man falling off a building" thought experiment</offtopic>
>one NC stands for "normally closed">the other NC stands for "not connected"what the fuck vishay
>>2081395gtfo from that uni/school
>>2075079https://www.youtube.com/watch?v=q43tZ6DjuIETim hunkin is back baby.
>>2081497>>2081604it's not vacuum tubes, try not being retarded and/or glowing. also name a device with better coefficient of power if it exists
>>2081771>Tim hunkincoffee and donuts
>>2081846The good tubes are CRTs, neon, and gnarly waves.
>>2081870neck yourself nigger
>>2081882Hang ten, brah.
I've got myself a PCB off of ebay based on an old syntom II drum machine. It's simple enough but for it's white noise transistor it calls for a BC108B which is priced much higher than other general NPN BJTs. Having looked at the datasheet, nothing stood out in particular. However, I see a few guitar pedals for sale that make a point of containing one of them. Does anyone have enough experience with BC108Bs to explain why they're the preferred choice?
>>2081937because most musicians that deal with audio electronics are retarded and swear on their mothers they can hear the difference between transistors and anaconda cables vs regular cables.>I bought another transistor and it did not workProlly because it is in a different operating point (different transistor hfe or something) or if it blows up it means whatever you bought was of lower power rating than the one they used.
>>2081846>NOOOO YOU ARE LE GLOWIE>>>/pol/
that's not a knife tip. THIS is a knife tip
>>2081846>self-powered>energy that resembles electricity>negative energy>anti-gravity characteristics>God revealed to Floyd sufficient information>buzzwords like zero point energy and triode being used improperly>oh we did this experiment that could revolutionise transport, and it works i promise, but it made a loud noise so we will never do it againquackery detected, don't come back
>>2082078>>2082078>>2082078>>2082078 New Thread
>>2081937no. but you should build a replica of the noise generator (or add three machined sockets where the xistor should go), order a selection of cheap modern transistors, and pick which one is closest to the original by ear.Be sure to post it on a blog or music forum if you find one. Chances are, somebody has already done just this, so google first.