Prev thread at bump limit. I don't often hangout here, forgive me if this isn't perfect.Prev thread: >>2816756>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.Read the datasheet.>OP source:https://github.com/74HC14/ohmOP>Comprehensive list of electronics resources:https://github.com/kitspace/awesome-electronics>Project ideas:https://hackaday.iohttps://instructables.com/tag/type-id/category-technology/https://adafruit.comhttps://makezine.com/category/electronics/>Books:https://libgen.rs/>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 Inventors (arguably has minor issues with mains grounding)Horowitz and Hill, The Art of Electronics>Recommended software tools:KiCAD 6+CircuitmakerLogisim Evolution>Recommended Components/equipment:OctopartLCSCeBay/AliExpress sellers, for component assortments/sample kits (caveat emptor)Local independent electronics distributorsladyada.net/library/procure/hobbyist.html>Most relevant YouTube channels:EEVblogW2AEWMoritz Klein>microcontroller specific problems?>>>/diy/mcg>I have junk, what do?Shitcan it>consumer product support or PC building?>>>/g/>household/premises wiring?More rules-driven than engineering, try /qtddtot/ or sparky general first>antigravity and/or overunity?Go away
>>2822343toujours
>>2822343Sexo with leg
>>2822343What the heck is in this can?
>>2822398After 2 min of searching, I think it's this.FWIW, Mechanic is a well-regarded Chinese brand when it comes to soldering supplies. I've used their replacement C245 tips and their soldering flux. Both were good. 1/10th of the OEM prices.
>>2822402>PCB board
>>2822402Thank you, Anon. Is it isopropyl alcohol?
>>2822407>Thank you, Anon. Is it isopropyl alcohol?I don't know. After much searching, I did manage to find a pic of the backside of the can.In comments, people are saying it works well and some use it in ultrasonic cleaners.I've never tried this. I didn't even know PCB-specific cleaners existed before today. I always just used some IPA to clean with. But IPA fumes are definitely not good for you....If anyone has used this, pls let us know how well it works.
>>2822412Translation. Mechanic seems to be from Hong Kong... had no idea.
>>2822343Hi DIY. I'm building a syringe pump for my lab, i need microfluidic mixing of liquids for mRNA lipid nanoparticle formulation. The commercial machines cost 30000$ and i'm just not willing to pay that for a fucking glorified syringe pump.I'm completely new to building machines, i'm a molecular biologist, EE is about as far a discipline from my field as it gets. Most i've ever done is mod my microscope, telescope and build a few rockets. Anyway, i put the concept together on paper, read up on what works with what, then i built it in fusion360. Pic related.I'm here to ask if it seems good to you, if there are any glaring problems, if there are any useful suggestions.The sceleton is made of 20x40mm extruded aluminium profile 220mmx180mmx400mm. The stepper motors are NEMA17s, controled by a DRV8825 each, all controlled by an Arduino Uno REV3, and powered by a Mean Well LRS-75-12. The stepper motors are attatched to a threaded rod with a flex coupler, the spinning of the rod moves the central 3d printed part by acting on a nut. There is also an axl to prevent lateral movement, the moving 3d printed part has a linear ballbearnig in it around the axl.Thanks in advance for any help.
>>2822414>99.5%IsopropylAlso, Mechanic is a state-owned (commie) company.
>>2822415I think you should use the extrusion as a heatsink for the stepper drivers.
>>2822417>Isopropylit doesn't match any of the IPA characteristics. Melting point is way off and even the pH is off.
>>2822421Hmm. So cut a hole in the top insulation, turn them up side down, and let the heatsink touch the aluminium frame? Seems like an interesting idea. I'm just kind of afraid of shorting things up there, i really don't have much experience with elecctronics. SHould i be putting insulation between the power supply and the aluminium frame?
>>2822424>SHould i be putting insulation between the power supply and the aluminium frame?Yes, you could use Kapton tape or a piece of plastic. Then use thermally conductive adhesive to bond it to the aluminum.
>>2822424Not that guy and, honestly, my advice is to avoid premature optimization. The step stick derivatives you appear to be using (or at least have modeled) don't generally have problems with overheating. Not with adequate ventilation, anyway, which open air qualifies as.>SHould i be putting insulation between the power supply and the aluminium frame?No. Actually, you should ensure that those things are in good electrical contact. The case of the power supply is grounded. For the sake of electrical safety and electrical noise reasons, the frame should be, too. The easiest way to do that is to just screw the power supply directly to it.
>>2822429>insulation between the power supply and the aluminium frameI assumed Anon was talking about the motor driver PCBs shorting against the frame.
>>2822426>>2822430Thanks, i was talking about the actual power supply. I may just leave the drivers as is for now, i'll keep your solution in mind (which i like, it's creative), and if i ever get overheating problems, i'll use it. >>2822429Thanks anon, that makes sense. The lab is well ventillated and airconditioned, so overheating really shouldn't be a problem.I have one more question. If i misswire something, how much can i fuck up? Will plugging the wrong wire from the arduino into the driver or from the driver to the motor really fuck either component up?
What's the advantage of a "low Vce(sat)" BJT vs. FET? Which one would be better for controlling a DC fan?
>>2822437>I have one more question. If i misswire something, how much can i fuck up? Will plugging the wrong wire from the arduino into the driver or from the driver to the motor really fuck either component up?It depends very heavily on exactly what gets connected. Accidentally connecting a low-voltage power rail to a high-impedance input will do nothing but confuse you on why things aren't working. Connecting the same power rail to a low-impedance output logic pin will probably destroy the chip, or at least that output pin. Hooking up the wrong pair of I/O pins usually just results in undefined operation, but no damage. Shorting out main power rails could do anything from blowing up thin traces on a PCB to simply causing some sparks before the power supply detects the fault and shuts itself off. Hooking up power backwards will destroy anything that doesn't have some kind of reverse-polarity protection.In other words, yes, you could destroy hardware by miswiring things. However, the Tarduino and related accessories are designed such that it's sort of obvious where things should go. It's far from idiot-proof, especially when involving hardware that isn't designed specifically for that ecosystem. Paying careful attention to the relevant datasheets and manuals and double-checking your wiring before turning the power on is all you can really do in any electronics project. You're not exactly working with high-dollar parts, though. Even if you do ruin something, most of the consequence you'll suffer is having to think about what you've done while you wait on shipping.
>>2822415Why would you want to make mRNA? It has no targeting ability, it will transvect any cell it comes into contact with, even cells that have limited healing capacity like heart cells.
>>2822439In your specific case, none that I can really think of. Saturation voltage in a BJT is essentially the amount of voltage drop you get when the transistor is fully on. That is to say, once you increase base current to the point that an additional increase in current has minimal effect on collector current, Vce(sat) is what you would measure across the collector/emitter of the BJT.In theory, you could bias the BJT such that it just barely hit saturation while in the ON state, which *might* make it turn off faster than a FET in specific conditions. However, this would only matter if you're trying to PWM your fan in the hundreds of kHz or MHz range. There's no good reason to do that (and a couple good reasons to NOT do that), so there's not going to be any appreciable performative difference here. You're likely going to get less heat dissipation from a similarly-specced FET and, as a voltage-controlled device, a FET is more easily driven from modern microcontrollers or PMICs (though, again, not a huge difference in this case).Tl;dr: There won't be any real difference no matter what you choose, but some theoretical factors make the FET a "more correct" option. For reference, almost all commercial low-to-mid-power motor drivers use FETs.
>>2822439> boutique one-off, expensive, hard-to-find part that was probably binned into a different part number or an IRF44Z like everyone else in the world.Hmm. I’d go with the boutique part so I can tell the fascinating story to chicks at parties.
>>2822415Get or make solid couplers, no need for flexible in this app and will add error.McMaster sells accurate and inexpensive acme screws which is what I would use rather than threaded rod. Other than that your plan looks solid.
anyone in here fuck with /diy/ eurorack kits or making your own designs?
>>2822343s e x x o o o s
I wanted to share my excitement with you guys because no one else I know gives a rat's ass about electronics (or any hobby for that matter). I recently designed and submitted my first custom PCB. They arrived today from OSH Park, I put all the shit together, and it fucking worked on my first try. Granted, it was just a little board so that I could have either 5v or 3.3v logic on an Arduino Pro Micro without a fuckload of wires, but I did it, goddammit. Now I can work on making shit that looks and works well. Fuckin too hype.
>>2822546I definitely have an interest in doing so, but I've got too many other projects such that I gave up on doing audio stuff for the foreseeable future. I still really like Moritz Klein's videos though.How do you actually get into doing modular synth stuff for eurorack? Just buy an enclosure and start making boards? What's the hardware behind them like, for connecting power up to each module. I assumed there was a card edge connector or other socket that plugs directly in when you slot the module into place.>>2822594>it fucking worked on my first tryGood on ya lad. First one I ordered had the MOSFETs backwards.
>>2822598majority of eurorack runs on +/- 12V via 16 pin IDC headers and ribbon cables. plenty of options to make your own boards, DIY kits or pre-built stuff. Doepfer is a good one to look at, given he invented the format: https://www.doepfer.de/home_e.htm
Here's the T12 station schematic as reverse engineered by a guy on EEVblog forum.Does anyone know how I'd modify the hardware and firmware so it goes to sleep when I put the iron in a stand? Like one of those stands that have a wire going back to the board somewhere, I've seen them in a few places, and some other T12 stations have them (the ones that go on tool batteries).
what dress and wig are you guys wearing when doing electronics?
>>2822626Don't need a wig. I get lipstick though.
>>2822626my problem is that the programming socks will melt if solder splashes onto them
>>2822626I wear a lead-free Mechanic brand wig and a tutu with built-in buttplug made from recycled toxic waste I got from Temu for 49 cents.>>2822629Are you a K-pop sensation?
>>2822630labcoat and don't put the feet on the desk?
>>2822626old pants>>2822620https://github.com/deividAlfa/stm32_soldering_iron_controller
>>2822632Programming socks and labcoat is an extreme combination, I'd never considered such a thing. Mad scientist basement larping fused with compsci femboy core. They are opposite ends of my wardrobe. I feel much power within joining both ends of this ouroboros...
>>2822633That was the firmware I was planning on using. It doesn't seem to have an option for both sleep AND stand, they seem to use the same input. Even though one of the boards I have has seperate inputs for that. Well I guess I'm fine with still having a timeout timer, and just tapping the iron's tip on the stand to wake it again. I think I could put a parallel R+C in series with the tilt switch, so the tilt switch closing would only give a short pull-down spike, but it wouldn't do the same on opening. If it could use this input as an analogue input and use a voltage divider to tell the difference between stand and tilt switch that would be easy, but modifying the firmware sounds tough.I bought three different T12 boards for this project, only two have arrived as of yet. One has a small linear regulator and small main PFET, but a dedicated pin and removable jumper for controlling T245 tips. The other has a large PFET and regulator, but I can't see an easy way to convert it to T245 tips. The T245 option means decoupling the heater positive connection from the thermocouple sensing stage (see R11 >>2822620), but while I could remove the 4.7k series resistor from this station there's no nice place to solder an extra wire onto besides an 0603 pad, which is a bad idea. The same problem exists when trying to solder an external high-power PFET or regulator to the boards, no solid places to solder wires onto. Both of these support this custom firmware.The third board, that has yet to arrive, has a nice big output transistor and has a T245 option, but has an off-brand STC32 MCU, so isn't supported by custom firmware.What should I do?
>>2822415Looks a lot more like a /3dpg/ project than a strict /ohm/ one. To give my 2cents:>FrameMake it taller and turn your electronics look. Gives you room to add a table for automation or an enclosure later on. If neither of these happens you still got space to print a shelf or additional tool holder. Add angles to your motor brackets to reduce errors.Maybe add angle plates to your base's sides as well for ensured rigidity.>Smooth rods Don't do less than 10mm. If you're into it, you can also use Igus clone bearings here.>MotorsSkip the couplers altogether and use stepper motors with integrated trapezoidal spindles. For your application, no additional cooling should be required, unless you're pumping nonstop literal syrup. Get some cheap aliexpresss stepper coolers and put them on top, if you're getting paranoid about it.>ElectronicsDepending on how clueless you are you might wanna swap the Arduino for a Pi Pico. MicroPython is very beginners friendly and while technically less efficient, more than enough for this project. Swap your DRV8825 for TMC2209, your ears will thank you.If you're using this whole thing for a tray like i am imagining, you might also want to consider a pedal to keep your hands free during operation.
>>2822473Thanks anon. Yes, the wiring does seem pretty obvious, everything is labeled. Is there any reason anywhere i look or ask i'm told to use digital pin 2-4 and 4-7? Why not start with digital pin 1?>>2822477I work with nematodes, they don't exactly have hearts. The point of mRNA delivery is that it will make large quantities of the protein i want expressed, when i want it expressed and for as long as i want it expressed. It allows easy cotreatment with various protein medicine based approaches. My area of research is the molecular mechanism of ageing, and i'm trying to figure out which ageing hallmarks are overlapping, additive or multiplicative, and by what factor.The mRNA LNP approach will alow me to express various ROS effectors like SOD, epigenetic remodellers like ALKBH family proteins, etc...I know mRNA tech is not popular on 4chin, and the vaccines were infact questionable at best, but it is a really usefull research tool. >>2822545Thanks for the advice. Yeah, i have no idea why others who have built similar machines online used flex couplers, i originally wanted to just buy a bit of aluminium pipe, drill a hole in it an use a pair of screws to tighten. These couplers cost basically nothing, so i'll go out and buy solid ones. I have never even heard of ACME screws, but they do look better. But also more expensive. What would i use instead of a nut? Also i'm not in americas, so i'm not sure mcmaster would bring me anything.>>2822666Thanks for the advice. The angled suports i am considering, i think i'll add them later if i feel it necessary. I don't really want to make it any taller, but i do want to put some covering on the top to protect the electronics from any spils. I don't know how anyone would spill anything up there, but undergrads are completely retarded.I never even considered this thing could be loud, i'll be sure to try out different motor driver if it is annoying. Why should the rod be 10mm?
One more question regarding this project. What is the best method for attatching the electronics to the frame? I'm thinking of using electrical tape on the top for insulation and then supergluing plastic spacers on top, to which i will glue the electronics. But his sounds kind of.. permanent and too simple.
>>2822546Made a VCO a while back, but never completed the output circuitryThis shit's surprisingly hard. Guitar effects are more chill
>>2822712>Why should the rod be 10mm?It's the smallest thickness where bending issues by shipping/mishandling disappear. >>2822713Make your electronics work, throw them in a box, box can be mounted via t-slot nuts. The fancy way is basically the same but with a proper pcb/hat.
Sorry for the noob question, but I am thinking on start learning about FPGA and RISCV, does anyone has experience on either that would like to share, such as tools, books and etc?
>>2822742>FPGA>RISCVHow is the former related to the latter? Or do you mean FPGA softcores that implement RISCV-V?
>>2822751Two different interests I will get into at the same time. Sorry if I was not clear.
>>2822742We could put 10 million Z80 CPU cores on a single die now, it’s kind of hard to see the point of FPGAs nowadays except for mid-speed hand-placed asynchronous/unclocked “jungle” logic which most people poo-poo because their design wizard doesn’t support it. Trend/meme/cool factor? Sure.Make sure nobody can copy your design? Sure.RISC-V is a quarter the speed of ARM, and RISC itself, is pretty much dead—it’s life extended by “thumb” instructions stolen from Super-H. They even made dumb decisions with decades of retrospection in `V “because it was easier” but that’s about the worst reason from a designer/implementer/programming perspective.China will fix these problems though, since they use RISC-V as a means to avoid technology embargoes and whatnot. They will eventually run it. Of course the usual clowns are all members of the RISC-V foundation to proprietarize, it if not outright kill, or poison it anyway.
>>2822816Build the bow into or over top of the bridge and engineer a clutch to engage/disengage the strings that you control with your picking hand like a whammy bar.
I'm going to rework an SNES controller clone, using the Dpad and 4-buttons, along with a Pico or Micro to do a USB interface. I designed a basic controller shell to use the left (Dpad) and right (4-button) parts and membranes, then planned to cut the board apart for the circuit pads, basically keeping the left and right parts of board. Looking at this more closely... I don't get the D-pad. The 4 buttons on right make sense, all grounding to an outer ring. I can't make sense of the Dpad function... the pads seem interlinked in a way that don't make sense to me. Other Q is whether I can tie into the shiny pads, perhaps drilling them for through-hole wire. They aren't used afaik. Board is single layer; they aren't vias.
>>2822905Back of board. I remember what confused me; the Up and Down on Dpad seem to be in series... but maybe the other end is tied to 5V or something and goes high, rather than low, on actuation... Assume this is meant so Up and Down can't be pressed at same time...
>>2822713You can use a fine thread rod but regular rolled threads can be crusty. Last thing you want is a rough spot causing issues. Someone in your country will have acme/leadscrews and associated nuts. They shouldnt cost much more and are well worth it.3D printers were made for printing electronic enclosures and 80/20 was made for mounting shit. Just put the 2 together.
>>2822905A little confused. Are you just trying to interface with the controller? Or are you doing some kind of mod to it? Because communicating with the connector is piss easy. It uses like five wires.
>>2822951Mod. I plan to cut board in 2 places and remove current mcu. I'll tie back to board fragments with wire to micro.
>>2822712>Is there any reason anywhere i look or ask i'm told to use digital pin 2-4 and 4-7? Why not start with digital pin 1?It's about the hardware connected to the digital pins of the arduino. D0 and D1 are the RX/TX lines of the serial comms that are being used to program an arduino. D2 and D3 are interrupt pins, D10-D13 are used for SPI comms, while A4 and A5 are used for I2C comms. There's also specific pins that need to be used for PWM or other counter/timer purposes. You can use all of these pins for other stuff, but if you might want to use that specific stuff you should keep this in mind.>it is a really usefull research toolNeat, I guess it can be pretty handy.>>2822742>>>/diy/mcg/>>2822793FPGAs are used for precisely emulating CPU cores. Either to use as cores that you can't easily buy, or just for testing. The retro gaming crowd have taken to FPGA products for precise emulation.RISC-V is cheap, and works acceptably. That is all.>>2822905it's hard to tell from the connections due to the silkscreen and hotsnot covering areas. I'd beep it out to draw a proper diagram. I assume the node connected to the bottom of SELECT and START is ground, since it is tied to the two caps of the crystal and to the black wire. R1 is a 1kΩ pullup for D-pad tied to the positive rail (the red wire), but while R2 might also be connected to the D-pad, it has a different value and isn't tied to the positive rail. I'd bring it into gimp and draw colour-coded lines atop the traces as you beep it to build up a full picture. Some connections might be made beneath the blob. Then you can draw a schematic by rearranging the features. Powering it on and measuring the voltages may help too.
>>2822912>3D printers were made for printing electronic enclosuresthat sounds really inefficient. why not just buy plastic boxes?
>>2822970Bc why buy boxes that are never the right size, and are expensive, when you can design one that's 100pct right in about 30 min and print it? And most common stuff have pre-designed box designs free.
>>2822970>pcb standoffs in the right location>cutouts for a display and/or connectors in the right location>not having to carve square holes in plastic boxes>able to make sizes and form-factors for unusual board shapes>able to integrate custom mounting hardware, like flanges, handlebar or pole brackets, captive or heat-set nuts, etc.>not having to wait ages for boxes to arrive or have a large stockpile of existing boxes>makes design revisions quicki will note that the advantages of a 3d printed case begin to dwindle when the scale exceeds about 150cm or a litre or so of volume. the square cube law starts to make injection moulding stronger for the price, and the time to print gets longer. waterproofing 3d printed enclosures is doable, but not trivial.
>>2822984What's the qrd on waterproofing (to a reasonable depth) 3D prints?
>>2822912> 3D printers were made for printing electronic enclosuresLoooloolEven the smallest box cost more in PLA than buying one. Not including printer, time, re-dos, “designing” etc.Don’t leave a 3D printed box in your car. It melts. PLA is one half-step up from hot melt glue.All 3D printed things look like shit. Covered in the hot-melt glue nozzle lines.You can fix it, with body filler, sanding, paint, etc,… but at that point it’s more work 20x time, money and effort instead of only 10x. We call “turd polishing”Any electronic project—especially so-called power electronics—, audio projects, RF projects, computer projects, etc; should be in a METAL case anyway. For durability, heat and fire resistance (e.g. Li-Ion batteries), shielding.If you’re making a quick-and-dirty temporary active filter from 2 9 V batteries and an op-amp… sure go ahead and use a plastic case for that.Otherwise, don’t.
>>2822970>>Buy a boxTell me youve never bought an electronics enclosure without telling me youve never bought an electronics enclosure. Also>>diy
>>2822991You'd have to ask /3dpg/ for a proper response, but in my understanding you want a lot of walls, print in such a way that gives you good layer adhesion (wide walls, PETG, vapour-smoothed ABS/ASA) and have a way of making a sealing gasket. For a gasket to work, the surrounding plastic can't warp under use or clamping force to open up a gap, so spreading out fasteners is sensible. I don't know if you can use printed TPU (or just TPU filament directly) as a gasket, you might be better off trying to cast a gasket from silicone or just smear it against one side of the print using a pane of glass and clingfilm to get it nice and flat. If pressure isn't an issue, you may be fine with just using some thick grease, in which case you should be careful that it doesn't interact with the plastic.>>2823013>Even the smallest box cost more in PLA than buying oneDepends on where you buy them from. China on the slow boat? Probably, but it depends on the size. I've made enclosure boxes for 5cm protoboard that probably cost less than 50c. But from your local suppliers because you need something within the next day or two? God no. A solid 10-15cm wide box costs in the ballpark of $20 where I live, plus petrol or delivery fees, and that's if they have the right size in stock. $20 is a kilogram of PLA, or even more ABS, that's at least 5 equivalent boxes worth. Because you can construct thick walls with an infill pattern, 3D printed structures can have higher stiffness per unit weight compared to solid injection moulded structures, and so use less plastic. They're basically a composite structure, like aluminium and balsa.>PLAABS is cheaper than PLA, and it doesn't have the thermal issues either. You can also vapour smooth it really easily using acetone. If you're serious about rapid prototyping, you'll be printing ABS.>metalSure, if you want it to cost 100 times as much. And turn your lithium ion fire into a lithium ion pipe bomb.
trying to fix the atrocious key fob distance in my old car by adding an external high gain antenna. does this trace look like it would be for a 315 mhz receiver? i cant find any other obvious other antennas (coil or trace) or anything else suspicious, even under the daughterboard, and the backside is mostly bare save for some passives and a small ic (pic to follow).note that the rightmost socket that the trace leads to does not get plugged in my vehicle.
notice how all of the other pins are connected to what i am assuming is the ground plane.oh, and i forgot to mention, i cant find anything else suspicious, EXCEPT where that hole is on the daughterboard there is a square component with a steel cap, and through the hole you can see some brass. it says '0032 320j' (no hits on google), no mfg or anything else and obviously cant see the top of it and its super hard to get a good pic.
>>2823066The daughter doesn't just pop off?
>>2822626crotchless snakebite chaps only, in case a snake bites my leg while soldering.I keep a quantity of soldering flux held by surface tension in the deep creases between my abdominal muscles, and I clean the soldering iron tip using the rough, keratinized skin of my penile glans
>>2823065did you let the magic smoke out of that microcontroller? should we be able to see through it like that?
>>2823068standoff things are soldered on both ends unfortunately >>2823071everything still has it's factory charge of smoke, just a poor application of conformal coating from 20 years ago
>>2823047>A solid 10-15cm wide box costs in the ballpark of $20 where I livewtflolaluminum boxes are like $4-6 and plastic should be half thatand don't forget tupperwares and old candy boxes for temp enclosures
>>2822970>that sounds really inefficient. why not just buy plastic boxes?because they're poors trying to justify owning their shitty $200 3d printers.
>>2823065>trying to fix the atrocious key fob distance in my old caryou arent>does this trace look like it would be for a 315 mhz receiveryou're retarded>i cant find any other obvious other antennas (coil or trace)there are other types, and as the other anon says its probably on that daughter board so it can be reconfigured easily for different regions>notice how all of the other pins are connected to what i am assuming is the ground plane.those arent pins you retard, they're vias, and if its connected directly to ground then it isnt an antenna>>2823072>standoff things are soldered on both endsno shit, desolder them. if you're too incapable of doing this then you sure as shit shouldnt be messing with antennas.
Is it worth making an opamp out of commonly available BJTs instead of using an off the shelf opamps?All the online shops in my country only have absolutely shit tier opamps like 324 and TL082. I just want to measure current through a shunt but these pieces of shit have like ±5mV of offset voltage
>>2822965Thank you, that makes perfect sense. >>2822912I have a resin printer, which is quite expensive when it comes to larger objects. So i'm probubly just going to buy a tupperware box and put it on top the electronics. And cut some ventillation holes on the side or something, maybe attatch a fan to it. I'll se if any of this is really needed, i just don't want some retarded undergrad squirting ethidium bromide all over my electronics. Fucking undergrads, i wish i didn't have to let them in my cozy little lab. >>2822715Yep, the boxc thing is what i will do, As for the rod not being straight, i bought the rods form a 3d printer manufacturer, they are designed to be axl for 3d printers, so they should be good. Still, if i encounter any problems with them i'll examine the matter further.All the parts have arrived, all i need to do now is 3d print my shit, which i can do over the weekend. I do have a new question now that i have the parts.The DRV8825 has pins. The motor has pins (6 of them for some reason). The motor comes with an attachment for the 6 pins which has 4 cables. These cables end kind of blank. How do i get these on the pins? Can i just stip the wires back a bit, strip some male-female wire back too, twist the two together and reseal it with some electrical tape? Or do i need to get some fancy crimping tool or whatever that round shit is called... Also, the DRV8825 comes with a small heatsik, but it's not attatchd, it's like a sticker. Where exactly do i attatch it? Pic related.Sorry for all the completely retarded questions, this electronics shit is completely new to me, i'm used to working with molecules and mice, not wires.
>>2823202>you arentk>you're retardedk>there are other types, and as the other anon says its probably on that daughter board so it can be reconfigured easily for different regionsno one here said that, it possibly being the case notwithstanding>those arent pins you retard, they're vias, and if its connected directly to ground then it isnt an antennaI was clearly talking about the pins on the 3rd unused connector on the backside because if you took 3 seconds to check the pictures you would see the trace goes from a pin to a capacitor and not directly to ground, retard>no shit, desolder them. if you're too incapable of doing this then you sure as shit shouldnt be messing with antennas.the other guy asked if the board popped off, no it doesn't because it's not socketed they are solid standoffs, which while I didn't explicitly say should have been obvious that's what I meant, retard
>>2823202>>2823233
Is there any good free software where i can test my circuit with the specific parts i have? Like the NEMA17 or the DRV8825?
>>2823233>I was clearly talking about the pins on the 3rd unused connector on the backsidenigger you have a big red arrow pointing to a ground trace full of vias. you arent talking about some fucking connector.>no it doesn't because it's not socketed they are solid standoffsagain, get your soldering iron off and remove that board>retardthat projection coming from someone who has no idea what they're doing and came here to ask for help, who desoldering is beyond their skill set, and who things a ground trace is an antenna.>>2823235you're not economically viable
>>2823244I built missiles for Uncle Sam. AI took muh jerb.
>>2823242>good free softwarenot free, but good, and soft enough to rub on your privates
>>2823227>have like ±5mV of offset voltageso add a trimmer to the reference voltage. why are you depending on offset in the first place?
>>2823252I can't trust my 2$ multimeter to measure 5mV correctly
>>2823260and you can't figure out any other way to measure offset with an op-amp?
>>2823198All of my projects have a display of some sort. I have a bridgeport but even still cutting accurate holes for displays is a fuckin bitch. Seriously try diying an enclosure once and you'll be singing a different tune.
>>2823275I'm making 10s of these devices, I can't manually trim the offset on every single one even with a jig to help me out. I also want to use 4 in 1 package opamps to do a couple of other things and they don't have an offset pinTruth be told, even 5mV is good enough but the newest opamps have crazy good specs and I'm just salty I have to shell out money to digikey for that.
>>2823275I think he means make a op amp amplifier with 100 gain, and use your cheap multimeter on that.
>>2823013>smallest box cost more in PLA than buying one.lol. Why do people talk about things they know nothing about? >>2823089>aluminum boxes are like $4-6 and plastic should be half thatsource for landed $5 Aluminum boxes, and $2.50 plastic ones. >and don't forget tupperwares and old candy boxes for temp enclosuresOr, you know, design a correct one and not screw around with trash.
>>2823230For all your plastic cases, go on Thingiverse, and download your favorite Arduino Uno case, then print that. Personally, I'd mount those stepper drivers onto veroboards and mount the veroboard to a case that you design, and add heat sinks to them. That said, I had to design something really similar to what you're doing; basically a Arduino Uno driving a stepper motor given input from some sensors. I don't know if those motors will be strong enough or not; if not, and you need to run larger motors, you're going to find those DRV8825s get hot really fast. Set them aside, and buy one of these industrial stepper drivers for each motor. They are good for 3.5A, stay cool, and aren't really that expensive given you're avoiding buying a $30K machine.As for mounting all this crap... lots of ways to do it. T-nut, U-bolt, zip ties, etc. Just make sure it's all solid. You could glue as well but I'd make a solid mechanical connection b/c I expect this thing is meant to run without supervision.
My magnification arrived. In one of the previous threads some anon recommended this style when working with SMD. <$10. Has two magnification levels with a flip of a lever. Extra 3rd for right eye only. Works well. Recommended.
>>2823299Thingiverse Uno cases... in any shape and configuration you could want. The machine I designed used an LCD Keyboard shield (see bottom left), and the printed case contains the 6 buttons that come with that shield. If you've not thought about how your device is going to be controlled, you might want to go that route, so the Uno can show user what it's doing, and you have some buttons to change settings on how it works. G/l. My machine replaced a lmao $6000 desktop bag feed machine that had a 12 week leadtime to get. It took me about 40 hours to design, build, and program the thing. It's been running for 4 years now, used daily.
>>2823304Keypad shield. They are a little fucky to set up, and use some of your I/Os, but expect you'll still have enough after its plugged in. There are other ways to do same thing, but these just work.
>5 minute project>welp let's add this feature...>and this feature...>and this feature...gonna need a PCB for this>and this feature...>and an option to add this feature in case I want to...>and an option in case I want the kitchen sink on it...>"Argh it's been 2 months already!?"It's a weird feeling when your passion works against you
>>2823315Isn't that how all hobbies work? At least you have the time to do what you want. Some people can't even get past the planning stages of a project.
>>2823315the trick is to find a hobby you're less good at
>>2822343heyyy, im this guy from the last thread >>2819982so i got some OPA683s but they just arent fast enough for this. second mouser order time.i want to purchase some super high-speed op-amps meant for differential amps and see if plugging one of those in magically fixes it (analog devices has a 10GHz GBW one and im just gonna hope its good enough since i cant be arsed to figure out the fourier transform function on my oscilloscope).i might buy some high-speed flip-flops like the other anon recommended, but i see a lot of issues with that approach. for starters, unless they can trigger on 10mV signals then i would need some (high-speed) amps on the inputs, adding more propagation delay (although maybe theres some high-speed transistors that can turn on, or turn on "enough", with 10mV). also, the propagation delay between the two flip-flops has to be long enough for the transient to die down, but if the pulse happens too quickly then we get an additional (variable!) prop delay on the signal out, which doubly sucks.i wanted to ask if putting together a high-speed diff amp on a breadboard is possible, or would the stray capacitances simply be too much for it? would a perfboard be good enough? making PCBs with proper ground planes while i figure this out would be too onerous; is there another option?
>>2823202The trace with no ground plane looks much more like an antenna than anything on the daughter-board. It even has an impedance-matching network at the top. It could be a feed-line, but I'd expect ground plane on both sides of the trace in that case. But it's either the trace or there's an antenna plugged into that 10-pin connector.But you're right that trying to modify this would be basically impossible. The best thing someone could do is removing metal shielding from the enclosure around the circuit board.>>2823227A discrete op-amp will be even harder to adjust the offset voltage of than a monolithic one, not to mention less consistent and less thermally coupled. The only thing I'd consider is making a discrete chopper amp circuit using conventional op-amps as part of it. Pretty sure that's doable, but not trivial.If you can't trust your meter's offset voltage, just flip the probes around and average the magnitudes.>>2823230Image search "6 wire stepper pinout" and "drv8825 wiring diagram". Use a multimeter to figure out which stepper coil is which. If it spins the wrong way, flip two wires or just change it in software.IIRC the heat-sink is usually stuck to the PCB on the opposite side of the driver. Either crimp dupont terminals on the wires, or just solder.
how can i find pin-compatible new chips for long-obsolete devices? pic rel.
>>2823340>i wanted to ask if putting together a high-speed diff amp on a breadboard is possibleno>or would the stray capacitances simply be too much for it?yes, plus inductance>would a perfboard be good enough?not without proper ground planes>making PCBs with proper ground planes while i figure this out would be too onerous; is there another option?yes. get some copper tape and stick it to the perfboard. cut away the spots where you don't need ground planes. this is a classic RF prototyping trick.that, or just deadbug it.
>>2823363>yes. get some copper tape and stick it to the perfboard. cut away the spots where you don't need ground planes.interesting, thanks.
>>2823065>>2823066>>2823068>>2823071here is the other side of the daughter board. square component looks like weird a pot or trim cap or something, but idk for sure.
>>2823384the only markings at all on the square. probably a lot/bin number, not even a part number.
>>2823384>>2823386variable/adjustable inductor
>>2823390I can believe that. here's a dumb question, is it possible that it itself is the antenna? they did bother to route out a hole from the PCB under it to expose its underside after all.
>>2823390also, it is 2 pin only (not counting the cover mounts), idk if variable inductors can or can't only have 2 pins (meaning, if they can't then it isn't)
>>2823395I imagine it's part of the tuning circuit for the PCB antenna. There are such things called loading coils which shorten the length of an antenna.https://en.wikipedia.org/wiki/Loading_coil>>2823398>it is 2 pin onlyYeah, that's normal. Two pins are the metal case and two pins are the coil.
>>2823353Daughter board with castellated vias, and a sufficiently compact chip soldered atop it (e.g. QFN, DFN, LGA) assuming the most recent datasheet doesn't refer to any pin compatible replacements directly. You can always email the manufacturer, but a look through their own parametric search (or on octopart) filtering by specs and footprint should give you a short list of potential candidates that you can browse through the pinouts of manually. I've never had any luck for equivalent footprints for anything that isn't standardised like an op-amp or has as few pins as a transistor.>>2823395The hole is so you can adjust the coil without having to desolder the board. They're often designed with a screwdriver profile on the bottom of their plunger for this reason.
>>2823340Back in the 70s we’d use LM311s to clean up incoming signals for digital. Not sure how fast you need it to be.
>>2823408200ns aint great, but at any rate a comparator wouldnt really work since both inputs are held around ground for most of the time, the noise would cause it to slam against the rails unless it was really good at rejecting anything less than maybe 2mV (and was also really good at accepting anything above 5mV).
>>2823399i wonder if it's a loading coil or just a tuner, I'll have to fiddle with it (not the adjustment, that would be dumb) later>>2823406I'll take a better picture later, but there is no drive on the underside of the inductor, it's just a flat brass face
>>2823424>it's just a flat brass facemaybe they cheaped out on parts or changed suppliers?
what's going on?
Have you taken the SMTpill? If not, what's stopping you?
>>2823455yes get on my level
>>2822608Are the odds of being able to sell stuff good? Or is the market saturated?
>>2823450Nobody wants electric vehicles.
>>2823450Why do they still want $200 for a ryobi battery pack with roughly $10 worth of cells in them?Probably one of things where they don’t sell so they start to double the price to make their money. Then “g wagon” type customers will finally start buying them.
>>2823450deep sea mining for lithium is going pretty well right now.
>>2823441I was mistaken, it's not brass it's just the other end of whatever is in the middle up top, but as you can see yeah no adjustment is possible from this end
>>2823483>no adjustment is possible from this endactually it is, if you really wanted itunscrew ferrite screw from the top sidescrew it in from the bottom sidethis assumes you can re-tune itwhich isnt hard, just turn it until you get best distancetho you gotta use a nylon/plastic screwdriver
>>2823482>deep sea mining for lithium is going pretty well right now.a drop in the bucket.>>2823450>>2823464>Nobody wants electric vehicles.correct. nobody Chinese are not buying EV cars as much anymore.
>>2823455everything's so fucking small.few weeks ago I was trying to fix my brother's old PS3 and I didn't expect every fucking cap & resistor to be of the smallest possible size!It's like 15 years old. I didn't think we had such advanced shit that long ago.
>>2823450>LFP is cheaper than NMCWait why would anyone use NMC for anything other than a phone or tablet? I cannot imagine a car or home storage with only 1000 full charge discharge cycles.>>2823522It's basically a computer, computer motherboards have used tiny components like that (far denser too) since 2000 at least.
>>2823450They're afraid the west will actually deliver cheap sodium batteries and are slowly price dumping to avoid any foot holding.
>>2823557>west>manufacturing anything physicallyYeah nah mate
>>2823558Are you ok, anon?
>>2823559I'm just upset 80 year old manufacturing and industrial design engineers with a lifetime of experience (or at least until the outsourcing started) are sitting in retirement homes and we have a generation of 20 year olds who have no interest or experience in manufacturing
>>2823455My shitty eyesight.>>2823522What is there you can even service on that thing? >>2823558> develop battery tech =/= manufacturingWest still tops out R&D. It then sends all the designs to China for them to build, and then steal. Don't get me started on that whole dynamic. >>2823557> sodium batteriesI thought graphene batteries were going to be the Next Big Thing? I guess these are easier to make, use existing processes, and sodium is cheap af to source.
>>2823582>I guess these are easier to make, use existing processes, and sodium is cheap af to source.That's only part of the supply chain though, the nickel and cobalt for NMC cells are still major contributors to cost. I'm not sure if there are sodium iron phosphate cells in development, but I haven't seen any.Sodium ion stores less energy per unit mass or per unit volume compared to lithium ion, but has the promise to be cheaper and last longer, probably safer too. So they make more sense for home storage than for EV use. Not that I'd recommend sodium ion for home storage, I doubt they'd ever get cheap and reliable enough. That's going to either liquid metal batteries, or to flow batteries, in both cases you have liquid electrodes so they can't grow dendrites.
Are there products on the market that sense current on a 240 plug that then turn on a slave-paired 120 plug? Gotta wire up a table saw that turns on a separate dust collection system, but can ONLY find 120-master 220-slave systems.If I can't find that, isn't this just an ESP32-plus-a-relay situation?
This is probably the best place on the internet to ask this question, as we get shitloads of "got junk? bin it" questions.I want to desolder lots of power caps off of ATX power supplies and TV power boards I've accumulated so I can actually bin all of these fuckers.My question is what method you think is the best. I have an atten 862D hot air station I've been using, and people who say "oh yeah just heat it up and they'll fall out" don't know that so many of these caps have bent pins that retain them.I also have a t8280 board preheater, but haven't used it for this purpose, yet.Current method doesn't involve my engineer solder sucker, but I have one of those and a pinecil/ts100 that I could try using if you guys think it's genuinely better than hot air.
>>2822415>NEMA17If I had to guess the reason for this device to cost $30000 is because the precision of the motor and drive mechanism is significantly more precise than the lowest price home 3d printer
>>2822415this is gonna cam over like nobody's businesslead screws are not meant to support anything, only move it on one axis. And a single linear bearing is not going to keep it from shifting side to side or twisting
>>2823609Not that onion but dual rods and a stiffer connection should it it right?
>>2823610Four rods plus the lead screw.
>>2823610two rods would help a lot yeah. I dunno if there's some requirement that it sticks out that far but the closer the lead screw is to where the force is applied to the syringe, the less torque there will be on the platform
hey I made this post on /g/ and they sent me here. :I wanted to know if there are good resources to learn what the little components on pcb's are, I'm not tech illiterate but i am pretty retarded when it comes to this stuff still. i just want a really rough general understanding cause these things are still alien to me
>>2823614do you know what capacitors, resistors and inductors look like? there are a gorillion youtube videos on the subjectmost other things are "ICs" which come in standard packages so they could be anything, you have to look up the part number to tell
>>2823584>saferI feel like people are seriously underestimating this point. A comparatively small bang and that's the worst to happen? It's OSHA's wet dream and will be pushed by insurances everywhere. They won't even need much more than some few videos of burning teslas.
>>2823609I think it will be fine for 2 reasons: 1) Forces involved are small. It's plunging a syringe and expect F won't be that high. F at syringe would prevent camming. 2) OP is likely copying a similar designI'd be more concerned about twist (orange arrows) but that will probably be fine too. OP will know 100X more than we do once he actually builds and tests the thing. Critical parts are 3D print and he can always revise it. I always assume (and tell anyone else doing same) that you can expect to build 3x of any new / novel device before you get it right, so just plan on that. 3D print massively speeds up that process.
>>2823614https://github.com/tioluigidev/Tio-Luigis-Components-Codification-Guidesupplemental:https://pinouts.org/
>>2823590>the reason for this device to cost $30000 is becauseI'd have to see the machine, but expect it has to so with the certification involved in it (which begs Q whether anon needs same cert for what he's doing.) They certainly sell these cheaper; pic related. Any new device really would need to be tested and statistically verified to be accurate... but then anon is a scientist and would know how to do that... right?
>>2823651everything associated with biotech has a 10x markup, on top of the obligatory 5x markup that comes with manufacturing. Never buy anything new, unless you're being backed by a 7 figure VC investment
>>2823487WHERE DO I BUY THESE IN BULK FOR LESS THAN $0.50 APEICE??
>>2823455>what's stopping you?being cuckchained to KiCAD and PCB fabs and having to replace all my jellybean partspoint-to-point ftw
>>2823656Eh, anything associated with industrial production is like that. At some point, you're pay for "it just werks" bc the cost of the machine is trivial (even at 10X) to the cost of taking down the plant due to breakdown.
>>2823588I usually walk them out with a soldering iron.You want to avoid getting them too hot.Your dad’s soldering gun came with a tool with a little slot in one end to help straighten the pins.
>>2823563>a generation of 20 year oldsthe tiniest generation ever. or maybe the next one. nobody is breeding anymore. they just watch tiktok streamers. and migrants from africa have no interest or experience in anything. maybe AI will fix this.
>>2823673AI is only here to guide humans to extinction and document all of it.
>>2822415>>2823230Should've really gone to /3dpg/ first, anon. They already manufactured your electronics for 20bucks on ali..
>>2822415Not sure why you're trying to reinvent the wheel, that's not a good design in any case. I assume you want it cantilevered like that because you want to put petri dishes underneath or something, if not then don't make it cantilevered.What I would do is get two of pic related and mount them to an aluminum plate. They're columns for table top milling machines and are made to be precise and handle a ton of force.https://www.sherline.com/product/mill-column-cnc-ready/#instructions
TIL tantalum capacitors are polar
>>2823694nta, but i don't really get your post. You could've just said you'd use linear rails instead.
>>2822343We're powered exclusively on solar and the initial setup was done professionally and cost quite a bitI've since done a smaller/budget setup (for a shed - pic related) and need to do another one for another shed.Does anyone have any suggestions on things that I could've done better here (while operating on a small budget)?The one thing that pains me the most about this setup is that the battery cables are hard-wired to the switches (all connects are using lugs).But, I cannot find any kind of plug that would be able to handle the amperage needed.In case relevant:>Victron is a 100/50 with 8 panels in total (250W/24V, wired in parallel of sets of two)>Inverter/Charger is 2400W max output (cheap Chinese shit, so we don't use the charger component)>Batteries are 4 x 12V (~1000W/h each, wired in parallel for 24V)I wouldn't mind doing a quick tutorial at some point for other people trying to do similar (it's actually painful to find good info for a small budget setup like this). But wouldn't mind an audit for shitness before doing that.
>>2823812Use XT150 or AMP connectors. You could also probably get away with Romex instead of the heavy gauge stranded wire.
I need to tie into this circuit board to get access to the two switches. I've figured out the traces, you can see the access points opened up with sandpaper. I plan to run through wires from back of board to front, in order to keep the wires out of the switchgear. The rest of circuit will be abandoned; I may cut it off if it starts interfering with operation but for now plan to leave it. What's the best way to do this? I was planning to drill 1/32" holes (smallest bit I have) near the traces, then bend wire to trace and solder it. Is there a better way to do this? I could cut and snap board and tie to the ends of the traces but I think that would just make the trace lift over time. And should I coat the exposed traces when done, and with what? Or does it even matter?
>>2823824>XT150 or AMP connectorsThanks, will do some research and see if I can source these where I live.
>>2823812Anderson has 120A and 175A for common sizes, and you can get panel-mounts and waterproof boots for them. 50A Anderson is even more common. They're all big and sturdy connectors, they can be a pain to unplug sometimes but you can get handles that bolt onto the connector to make it easier. Only like colours can plug into one another, so if you want to have 12V stuff and 24V stuff, or some other separation of systems, you can do that with colour coding.>250W/24VThat's each panel? So it's 2000W total? At 24V that's nigh 100A, but if I'm reading correctly the charge controller (and your MC4 wiring) is only 50A rated.If anything stands out, it's the lack of fuses. A main fuse/breaker (which you might have out of shot, idk) on the battery is basically a requirement, but seperate fuses for your inverter and charge controller are generally recommended also. Same for a fuse going to any other DC loads (lighting, fans, etc.) and coming from any other DC sources (AC charger).Also the two red wires coming from the victron charge controller and going to the battery is not how I'd have done it, if it was done using extra wire I'd have used some black heatshrink (or paint or even a sharpie) to colour it, at least in a few black stripes.>>2823825Drilling holes for strain relief is a good idea if you have the room for it. Tying a knot in the wire on the copper side of the board to stop it getting yanked out will help too. But those shitty FR2 boards are still really prone to delamination, so I'd definitely want to gunk over the solder joint with something to stop that. Like neutral cure silicone or hot glue.The over the top method is to cut away all but 3mm from around the edge of the blob, sand off the mask, and solder its traces face-down in a cutout in a well designed custom PCB.
>>2823833>At 24V that's nigh 100A, but if I'm reading correctly the charge controller (and your MC4 wiring) is only 50A rated.The solar is wire in 4 sets of 2, so it's ~48V, so I think that becomes effectively 50A?>A main fuse/breaker (which you might have out of shot, idk) on the battery is basically a requirementI don't have this. I haven't been able to source DC breakers locally that can handle the amperage, so I think I might need to go online to get those.>(or paint or even a sharpie) to colour it, at least in a few black stripes.I might go do this now actually. I wasn't happy with that either, but it was really hard to find 16mm cable (which the Victron expects). And I could only get it in red.
>>2823862>The solar is wire in 4 sets of 2, so it's ~48V, so I think that becomes effectively 50A?Oh you've got 48V of PV going into a 24V battery system? It works with an MPPT controller that can handle it, but it may be marginally less efficient. Make sure all your panels are at the same angle with respect to the sun.>I haven't been able to source DC breakers locally that can handle the amperageA fuse is fine, the big ANL wafer fuses, bolt-down fuses, or even MIDI fuses should do the trick. Or mega-blade fuses. The sort of thing that's used for high-power car audio systems. If you have a spare or two, you really shouldn't have to worry about it popping, it's basically just in event of an appliance dying short-circuit, or something shorting out your wires. Might be a nice idea to cover the battery terminals with something in case someone drops a spanner.>I think I might need to go online to get thoseWatch for non-brand-name sources, even amazon is likely to have shitty fakes that can't handle as much current as they say on the box.>16mm cable6AWG should be fine, especially for a short run of cable like that. You're using ferrules in the screw terminals, right?
>>2823769>TIL tantalum capacitors are polarWhen you got them in a pack, didn't you notice one leg was longer? Isn't that obvious?
Jesus fuck are these things small! How do you guys work with shit that's 0402 or 0603 or even 3014??? When building boards, is it even possible to solder these fuckers by hand? How is it done???
Hey guys I bought one of these to reuse some atx power supplies I got for free https://www.aliexpress.us/item/3256805806875839.html and I removed the stock on/off switch and routed it the switch so its more convenient to turn the psu on and off. I use this https://www.aliexpress.us/item/3256805747053818.html to power LED strip on my bench, and it seems to have an on/off on the potentiometer, and I wanted to see if its possible to remove on/off switch and just use the pwm module to turn the psu on/off.
>>2823867>Make sure all your panels are at the same angle with respect to the sun.These are good, they're all lined up right next to each other.>A fuse is fineI'll probably go this approach actually. Simple/cheap.>Might be a nice idea to cover the battery terminalsYeah, these we'll be moving (and possibly putting in a cabinet). We were in a bit of haste in putting this system together.>Watch for non-brand-name sources, even amazon is likely to have shitty fakes that can't handle as much current as they say on the box.This was actually part of my nervousness with them. We've bought chinkshit in the past that we run BELOW spec that has blown (Amazon's not a big thing here, but most things are Chinese imports).>You're using ferrules in the screw terminals, right?We're not actually. We looked for them, but one of those situations where we could source appropriate ones locally again. We did make sure the cables are held in there tight (and the MC4 will be better secured to avoid any pull/tension once we've actually mounted that board). We do want to use ferrules on the new setup though (and might fix this one to use them too).Might go look at ordering some Anderson and Ferrule crimpers (unless I can get away with using my lug crimp set?).
>>2823871i thought they were ceramic until it fucking exploded. i almost threw out the bag thinking i had a bad batch of exploding ceramics.>>2823872tweezers, hold it down while you solder one side.
>>2823873>possible to remove on/off switch and just use the pwm module only if on/off contacts are NOT used on the PWM itselfsee if there's traces going someplaceanother likely problem is that the PWM switch is not momentarywhat happens if you permanently hold down the power switch in ATX?my guess is that it's gonna gonna flip on/off every few seconds
>>2823773What I posted is a complete z-axis unit that directly takes a stepper motor. Normally there would be a milling head attached to the black block using a grub screw that digs into the round post that's sticking out. A little expensive compared to rolling your own, but there wouldn't be any fucking around with trial and error, just bolt it onto a surface from underneath and away you go.
>>28238720805 is definitely possible with bare eyesight. 0603 you need good vision or a magnifier, 0402 even more so. I've soldered an 0402 by hand before, surprisingly not that difficult if you have your SMT passive technique down.>>2823874I'd be a bit skeptical of nobrand ferrules, they're meant to be tinned copper, but you can never quite trust chinky brands. Ferrules definitely need a dedicated ferrule crimper, possibly a hex crimper of exactly the right size if the ferrule happens to be a sufficiently snug fit. Andersons should be crimpable with the same tool as lugs.>>2823883The power switch on an ATX itself is usually supposed to be a latching one. There's a pin that gets tied to +5V or GND, doing so turns the supply on so long as the pin is held to that voltage. Computer motherboards may use a different method for power, using a momentary switch instead.
>>2823889I've done 0603 by hand before. Made oversized pads and pre tinned them, then put the LED on top and reheated each pad. The surface tension actually pulls the LED and lifts it up on the opposite side so it looks like the titanic, had to hold each LED down with a screwdriver.Problem was how uneven the hand placed grid looked, never ended up using it.
>>2823872Wait until you drop a tray of them.On a shag rug carpet.Costs about $70 bucks for a small shopvac, sucked up the area for an hour, and then washed them, then had my kids sort them.I still hear them come up sometimes to this day. <tink>…. <tink>
>>2823890Looks good. You’re a natural. You should make a couple of gigabytes of core memory next.
>>2823558The west is doomed.
>>2823890why would you do this
>>2823833>(and your MC4 wiring) is only 50A ratedHow are you meant to conceptualize wiring from an Amperage-Rating point of view?Wouldn't it depend upon the length of the cable?I would've thought the most accurate metric for cable suitability would be something like Ohms per Meter or something?
I just finished designing a 20W synchronous boost converter circuit board with an MCU to handle all the switching and stuff. Wish me luck bros, hope i can code it properly and get the thing working. I'm going to use it to drive a 20W cob LED, basically making a flashlight with a decent light output and real battery instead of cheap Chinese torches with 1W LEDs on them. Quiescent current is going to be ~3uA so I'm pretty excited
>>2823897There's two current limitations for wire. For long runs of wire, the cross-sectional area needs to be high as to not incur a high voltage drop. This is the case in the DC power wiring for security cameras for example. The calculation is pretty simple, bulk resistivity (ohm meters) is a characteristic of metal, if you divide that by the surface area of the cable you get the resistance-per-unit-length (ohms per meter). Voltage drop V=I*R = resistance-per-length * current * length. You can look at it either from the perspective of how much voltage drop will cause loads to work less well (e.g. a CCTV camera can work down to ~10VDC), or you can look at it from the perspective of efficiency (e.g. 10% voltage drop between panel and battery means your battery gets 10% less energy over a day as it could).On the other hand, for short lengths (less than 2-10m or so) the topic is governed by thermal issues. Roughly speaking, heat dissipation scales with surface area, which for a given size of wire scales linearly with length, and it also scales with temperature. But heat production P = R*I^2 scales with resistance, which for the wire also scales linearly with length. Maximum heat dissipation is governed by the maximum temperature the wire can get up to before it begins to melt, so the thermal equilibrium between resistive heating and heat dissipation at this maximum temperature directly tells you the maximum current that can go through this wire. This isn't a function of length, because the length cancels out, assuming the wire is approximately in a long length. If the wire is bundled against itself, then this assumption changes. If the wire is enclosed in conduit or cooled by airflow or liquid, the proportionality factor changes.Practically, people have already written up charts for voltage length per metre or amps for certain gauges of wire, I find the american wire gauge chart on wikipedia very useful.
>>2823897>>2823900I gauged that wire as being 50A rated because where I work we sell some solar cable as "50A wire" that looks like the same thickness.FYI, ampacity is the maximum current a gauge of wire can handle, there are a few ways of measuring it based on the wire getting up to a certain temperature (I'm not sure if this includes insulation) or the wire melting. As you suggest, ohms per metre is the important metric for calculating voltage drop.>>2823899It's for an LED? The design has current feedback, right? I don't trust chinese COBs to have any form of current limiting, even though I only ever see people wiring them directly to voltage sources. If you already have the COB, try and plot yourself an I/V curve for it with a CC/CV PSU, or some resistors or whatever.
>>2823892>You should make a couple of gigabytes of core memory nextI would have not been upset at all if we hit a technological brick wall around that time and were still hand weaving core memory.>>2823895>why would you do thisIt was going to be a mini audio spectrum visualizer display for an analog equalizer I wanted to build. It was a time when making LED matrices by hand and running them with arduinos was a big thing.
>>2823911>hand weaving core memoryEven if that was the case, we'd be using automated looms for it at the very least.>a mini audio spectrum visualizer displayWhy SMD LEDs though? You could have done THT LEDs easily using perfboard. The distance between those LEDs doesn't look smaller than 2.54mm, but I could be wrong, it's hard to gauge scale.Also what's the black gunk?
Here's a fun read:https://hackaday.io/project/183302-a-usb-c-connector-on-a-flex-pcbUSB C connectors have a 0.7mm connector board on the inside, so I guess if you can get a PCB the right thickness or stack some layers you can make something like this.
>>2823875>tweezers, hold it down while you solder one side.Thats fucking impossible. Your component will get stuck in solder and the tip of the iron is several times bigger than these components.And good luck even seeing this shit with your naked eyes.>>2823891I was once repairing my wristwatch and one of the little springs went flying. Took me literally days to find it. I ended up buying a large magnet and finally I found it.
>>2823917neat! reminded me of this.also, have you seen NSA's ANT catalog? lots of good ideas in there that could be applied to USB-C.
>>2823918>Thats fucking impossible. Your component will get stuck in solder and the tip of the iron is several times bigger than these components.i dont know what to tell you bro. i hand soldered some 1206 LEDs a couple weeks ago with a chisel tip and some tweezers. maybe its just something that takes practice.
>>2823917this one is cool too: https://hackaday.io/project/28516-business-card-gamepad
>>2823914Because visualizers are bar graphs and those SMD LEDs are sort of already in the shape of a bar. Back then (13 years ago) I couldn't just 3D print a bar graph mask or light pipes out of resin.It's funny how technology gave us the ability to easily make shit, but then also made the act of doing so completely pointless; I could probably find a complete visualizer unit with an OLED display from china if I looked, but who the hell needs a graphic equalizer anymore? These days I just play music from youtube videos of varying quality and mastering.
>>2823902>It's for an LED? The design has current feedback, right? I don't trust chinese COBs to have any form of current limiting, even though I only ever see people wiring them directly to voltage sources. If you already have the COB, try and plot yourself an I/V curve for it with a CC/CV PSU, or some resistors or whatever.Unfortunately no anon, I did think of having a current sensing circuit but I want this to be really simple so I'm going for a simple voltage feedback. I tested the LEDs and have characterised their VI curves which I am going to hardcode in the firmware. The torch only needs to work at certain fixed duty cycles and I don't forsee any surprise transient loads so it ought to be fine
>>2823914>You could have done THT LEDs easily using perfboard.maybe he didnt want his board to look like an assembly kit for 6-year-olds.
>>2823923Pic related.>>2823924>I tested the LEDs and have characterised their VI curves which I am going to hardcode in the firmwareThat drifts as a function of temperature. So good luck. Maybe you can add a sense resistor to the input or output and feed it into a spare I/O, or try to gauge based on the voltage drop seen as a result of the battery's ESR. If you're not measuring the battery voltage, then you can estimate it based off the duty-cycle.>>2823931>maybe he didnt want his board to look like an assembly kit for 6-year-oldsHe's way ahead of you, it's covered it in black goo, he's already got that sorted.
I want to capture a very short signal and process it on a microcontroller. The signal should only be about 100uS long but I want to capture it with precision so I need to sample at high speed - ideally something like 300mhz or at minimum 50mhz for a few thousand samples. Ideally analog though I may be able to do it with just 2 or 3 digital bits from comparators. Any of you have ideas for this? My ideas:1. abuse a QSPI controller or something on the MCU (but 300mhz is pretty fast for SPI)2. plug a parallel ADC into one of those FIFO memory chips and drive it with some kind of oscillator (but FIFO memories seem to top out at 160MHz or so and are expensive)3. FPGA or CPLD - I've done only a little bit with FPGAs. Would a MAX II EPM240 be enough for this? That's what's cheap on ebay and I don't want to spend muchIf you're curious I'm trying to make a very basic time domain reflectometer that tells you where a 1-10 mile wire is shorted
>>2823957define "process it". 100us is not that small, is it a digital pulse? or some form of fluctuating signal? if the former just run it though a interrupt based timer. if the latter, should run it through the onboarf adc first to see if it will handle it, otherwise id say an external adc.
>>2823957Assuming it's just a digital pulse and the voltage doesn't drop below logic-level, there are a variety of different things that might work for this:>high-speed STM32 or overclocked RP2040>high speed digital logic like shift registers>a time-to-digital converter ICIf the assumption is wrong, idk bro just get a high speed comparator.
>>2823961>define "process it"I think all I need is timestamps (and maybe magnitude) for each peak and valley. I don't know exactly what the signal will look like but yeah it will be some big pulse and then one or more smaller positive or negative pulses some time later. Here's a pic from wikipedia.>>2823962>>a time-to-digital converter ICoh wow those do exactly what I want, I didn't know there were general purpose timer chips like that. Now I just need to figure out the analog side. Thanks!
/o/tist here, I know fuck all about 3 phase power, but are all the 3 phases supposed to show continuity between each other in a frequency coverter?L1, L2 and L3 are separate before the fuse block, but after it they all beep together, in the frequency coverter the L1 L2 and L3 pins beep, showing continuity between each other.
>>2823970Fucking cunt, forgot my pic even though it doesn't really tell much
>>2823970>>2823971The phases overlap. You should measure 220-240V between any two of the hot leads.
>>2823970you're measuring resistance while it's off? yeah it wouldn't surprise me. connecting all the phases together is a "brake" and may be what it goes to when off.
>>2823991>>2824011Thanks, I'd rather not fuck around with this shit honestly as I know jack shit what I'm doing with AC power, I'll leave it to the "professionals" for monday.I was just trying to find if there was a short circuit or something easy that causes the fuses to go off, it's from an electric turntable and they accidentally lowered it on top of its own cable, cutting the cable inside.Replaced the cable, but for some reason it keeps popping the fuses inside, so I gave up.
does anyone know what material they used for the contact springs in old e27 lamp sockets?Is this ordinary brass or did they electroplate spring steel?because i need to manufacture a hand full
>>2822343I'm trying to desolder chips from old routers as practice, why the fuck will it not melt?
>>2823825>>2823833>Drilling holesSo, what I ended up doing on this was using existing hole (and their pads) and running wire wrap from trace to those pads. Board is full of abandoned pads and avoids me having to drill the board at all. I don't thing I'll even need to cut it down. How can you tell FR2 from FR4, anyway? The color? This is a cheap toy so not surprized its a cheap board.
>>2824071lol what kind of chips and how are you trying to desolder? Through pin, smd, iron, heat plate, hot air, etc.
I have an LED fixture that evidently has a non dimmable driver. Can I just swap out that driver for a dimmable one? Or can I put a DC dimmer on the output from the driver? Best way to make non-dimmable led less bright
>>2824071Heavy ground plane. Turn your temp and air speed to maximum, and keep the board off of any surface that sinks heat i.e. tile or metal. A PCB preheater is perfect for this application.
>>2824041I think it's brass. why not put a magnet to it?
>>2824074FR2 is brittle trash that stinks of toasted plastic and FR4 leaves itchy splinters in your hands (and lungs)
>>2824080I dimmed an LED shop light by replacing the current sense resistor with a big one. That light uses a linear current regulator.I think it depends on how the particular light is made, you can't just assume all the drivers work the same. I found the datasheet for the chips in mine.
Never realized power JFETs are a thing. I always thought they are used with very high impedance signals like op-amp inputs circuits
>>2823957Again. Pulse stretcher.This was a common DIY circuit back in the day, when you’d make your own “logic probe.” I bought mine though.You need a pulse stretched out to see it on a LED. The little lights on your ethernet router have something like that too, as does the “collision” light on your hub.
I've got a UPS second hand. It's working enough that it's able to give an error code signalling battery or charging failure.I pull out the battery and it's a lead acid unit, supposed to be at 12V but it reads less than half a volt. All I know about it's previous life is that it sat for who knows how many years. Could the battery just have died it own peaceful dead in that time or is it still likely that the charging circuitry failed and the battery died as a result?
>>2823883Thanks for the reply, I did a continuity on various points, the power+, output+, closest pin you have circled in red have continuity to each other, when its "off" when I turned the knob and hear the click (on I guess) the the two circled pins, power+, and output+ have continuity to each other. the Vcc pin on NE555 doesn't show direct continuity but there is numbers that fluctuate (idk if this signifies something) sorry I am new with this so I am not sure if this proves anything or not, Anything else I should check?
>>2824074FR4 is fibreglass epoxy, it’s usually clearish with a green tint, but I’ve seen opaque black and grey before. It usually has a fibreglass weave texture poking through the unused side a little, making the glossy epoxy ripple. FR2 is wood fibre phenolic, it’s brown, and the unused side is usually flat matte.>>2824080Yes, if it has current limiting resistors then a normal variable voltage or fixed voltage PWM will do the trick. If not, then a constant-current supply with PWM or variable current dimming will work. Check out MeanWell’s line of LED power supplies.You can also dim it by interfering with the feedback loop like the other anon suggested, but read the datasheet first, they often have limitations that aren’t too convenient.>>2824166Woah that’s sic!>>2824180Lead acids, especially old ones, can self discharge like that. Buy a new battery, but check with your multimeter that it charges properly.
>>2824203Is it possible to easily test the recharging circuit before buying the battery? Just to be sure and not run the risk of wasting money on the new battery only for that to not be the issue?
>>2824205You can try reviving the dead battery on a normal lead acid charger, if it gets high enough for the UPS to treat it as a potentially good battery then you can measure the output current. If the voltage just drops below 11V after you take it off charge, it might even be possible to connect the battery to the UPS while it’s being charged, though you might need some diodes.
>>2824233>normal lead acid chargerI've got my bench top power supply, would that work? The battery is at .4V so I don't put much faith in this but whatever, doesn't hurt to try. What sort of current should I set it to?
Anyone here knows the max frequancy limit for a 2N2222 relaxation oscillator before it becomes unstable?
>>2824245If it has a current limit, then sure go up to 14V for a day or two. The battery should say the charging current on it, my 9Ah bats say 2.7A max, so 2A should be fine.>>2824255
>sent a design out to a fab house for work>accidentally put a a metric 0603 resistor for an imperial 0603 footprint on the BoMthe fab house techs are laughing at meI know itI can feel it in my bones
>>2824166Are they cheaper than normal MOSFETs? Since they're less complex to manufacture I would guess so but then again since their demand is lower, I don't think they will be actually priced lower
>>2824114nice idea anon.It is brass, some screws appear to be stainless (i have 5 of those in different states of decay) and the center tap on all these is supposed to have a fucking coilspring below.I will tin coat all of the ones still decent to prevent further decay, they will never see a 100w bulb anyway
>>2824336They're SiC, my guess is they're high spec, and either used because they can handle more current than SiC MOSFETs, or specifically because they're normally on with zero gate voltage. The lack of the body diode also makes them usable for switching bipolar signals, but the driving circuitry to allow for that isn't simple. I guess that would make them fail-safe in some applications (e.g. in series with an inductor that stores energy).>>2824337Nickel electroplate them instead. It's easier and harder-wearing than electroless tin plating. Though tinning in a solder pot or with an iron is pretty easy I guess.
>>2824346>Nickel electroplate them insteadthats a shitload of prep work, meanwhile the solder flux at high temp would do that in an instant. and i have yet to setup my plating setup
>>2823971>>2823970you’re not going to like what I’m about it to say but I googled it and some identical questions on Reddit came up and the responses said that it’s normal and yes it is supposed to do this.three responses to various questions are copied below>The 3 phase wires should have continuity, yes. The resistance between them should be very low, in the neighborhood of 1-.20ohms. Most multimeters don't show readings like that properly and sometimes it even shows as a short circuit.>You are reading through windings of a transformer or motor or similar. This is normal. >There's a control transformer that operates the control circuit between 2 phases. This is what you're seeing. And while it is drawing some power, due to the way reactance works, the impedance will be much higher than resistance and the amount of power it's using when not operating is negligible.also wash your hand
>>2824337I have an e26 (6 amp 125 volt rated bare-bulb-to-wall-plug adapter) that I just checked and like yours, the spring contact is brass, the light bulb screw threads are copper, and the screws holding everything together are magnetic steel. The wall plug blades are nickel-plated and non-magnetic.
>>2824264Just woke up and put it to charge at 14V 2A, the power supply says it isn't taking any current but I'll leave it there for the rest of the day so I can pretend I'm being productive somehow.
>>2824166Solid state tube. Finally. >>2824080Depends on the circuit. Switch mode driver? Oof. prob easier to change driver with a dimmable one. Linear driver? You can add capacitor (0.1 - 2uF filter cap, class X. Or 630V film) in series. You can also try adjusting current by reverse engineering driver. They usually have a feedback current shunt or whatever.
>>2824380After 3 hours, it's reading 3mA now. I guess that's something?
>>2824435lead dendritesIt's dead, Jim.
>>2824447I'm sad.
>>2824453You can recycle the battery or throw it into the ocean to make yourself feel better. Sorry, Anon.
>>2824180Beside the battery. If I try to plug it into 220V it starts rapidly shutting on and off again very quickly.Would that be a signal of a fucked charging circuit or could the battery also be to blame for that?
>>2824474It might be normal behavior due to a missing battery, or the charger is burned up. If you're lucky you'll be able to see heat related PCB discoloration and/or burnt components.
>>2824182>Anything else I should check?noyou've determined the PWM switch is already in use, and has 12V on it, so you cant use it to switch the power supply(the ATX power supply turns on when you connect pin 16 (green wire) to ground)you'll need to keep using the lil switch on the adapter board or solder a new switch in parallel for easier access
I'd like to set up a simple FFT analyzer on my computer to run audio checks on some audio microphones I'm working on. Idea is to run white noise into them and measure frequency response across audible spectrum (20 - 10K Hz.) Anyone have any recommendations for software? Looking for somethign stand alone vs. dumping it into Matlab (which I no longer have anyway) and ideally would just work with my soundcard. Not concerned with calibrated in/out levels, just the overall frequency doman response. Stand alone instrument would be fine as well, but I'm not seeing anything less than $400.
>>2824474>could the battery also be to blame for that?dont suppose you have some old 12V drill battery in NiCd or NiMh?should work fine in place of lead-acid for testing
>>2824505Cordless stuff isn't too common in this country due to cost, so no.
would it make sense to buy a cheap $30 home security camera and add some DIY features like solar power? in terms of learning / cost saving.
>>2824508>so nookaystill, you can borrow an SLA from grandma's apartment building fire alarmnot like she's indispensable
>>2824522>fire alarmsUnheard of unless you're in an industrial setting or some hospitals.
>>2824504Audacity
>>2822414>>2822423it's probably a blend of isopropanol and some unnamed petroleum distillate.
>>2824541Does audacity do spectral analysis? I'll check it out. That would be ideal.
>>2824500Sorry anon I guess I should have mentioned that I did in fact already removed the stock switch so I can turn the psu on at the front of my "workbench" a while back, but since I do need it, then I will keep it. I appreciate the help!
could an MCU be powered from a solar panel indoors? it would probably only need a few microamps in a low power mode. although not very practical, that could probably work, similar to how solar powered calculators work? is there a way to estimate an indoor current capacity of a solar battery based on its maximum performace outdoors in direct sunlight? say if a short circuit current is about 1A, surely it could supply at least a few milliamps indoors? i couldn't find any videos experimenting with that.
>>2822546>>2822598Idk what a eurorack is but regarding MIDI I just finished this system which automates the pipe organ in my local church, it goes hard. I play a midi file on a pc and it sends it to 3 pcb's which convert the midi signal to 64 contacts per pcb, which operate relays which function as OR gates parallel to the existing contacts inside the organ (so it remains playable as it was)
>>2824626Man I keep telling you to redo your board with an rj21 jack. You can splice 25 pairs of wire into one connector with a 80$ tool and it'll be 1000% tighter
>>2824628>80$ toolpick the right connector type and you just need a viceor strong thighs, which is what i useif my thighs hurt, i'll use the solder-on type
Do I need to desolder a transformer from a circuit in order to diagnose it?I've a got an automotive battery charger that blows the fuse. The primary coil reads around 50 ohms and the secondary reads 2 to 3 ohms, though that is with it connected to the board and maybe that path is through the diodes.The board itself has a couple transistors that seem fine, an op-amp I can't test while in circuit, a relay which is working and a two old electrolytic capacitors, both of which have a bit less capacitance than they should and one seems to be leaky.Seems to me the secondary is shorted out, but can I say that for sure without desoldering it from the board? Cause the way they built this thing makes that a real pain in the arse.
>>2824747>Do I need to desolder a transformer from a circuit in order to diagnose it?Almost always, but you can sometimes get around this. With a low-voltage ring tester you can measure inductance and capacitance without going over the 0.6V threshold for any logic or transistors or diodes to conduct. With some extra hardware, you could measure one winding while directly sending an impulse into another winding, and measure the turns ratio.You only need to desolder all but one wire of each winding to measure it, but if it's a PCB-mount transformer you'll probably want to fully desolder it. That said, transformers don't often die, and if they do it's almost always an open-circuit from a broken wire or dead thermal fuse. Usually capacitors and semiconductors die before transformers do. A secondary could be shorted out, but it's hard to say if it is or if that's just the low ESR of the windings. Generally the resistance of a winding should scale at least linearly with the voltage (or the turns), so if a high voltage winding has a lower resistance than a lower voltage winding that's a good indicator it's got a short somewhere. If you have a good measurement of its resistance (either it's above ~10Ω or you have kelvin sensing ohm meter) you can calculate what the resistance should be for the volume of wire, after measuring its thickness.
>>2824071clean ur tip :3
>>2823351So the NEMA17 datasheet was really usefull, i managed to wire it the motor correctly on the first try. For splicing wires i just ended up stripping back a bit of the insulation, twisting the copper (is it copper? that's just a guess) together with a pair of plers, folding it in half, twisting again and then taped it up.>>2823590You'd be wrong. The reason for the price is that a grant typically is between 0.5 to 2 million dollars. A fraction of that is allocated to machine purchase. So they can come up with ridiculus prices, and you'll still pay. And thes biotech comanies are almost monopolies and the few competitors that exist for cartels. So you can't go to the competitor. And they are often in bed with publishers, so if you didn't use the "proper machine" nautre may not publish. But fuck'em, i'm building my own shit, and they can't stop me. The only thing that can stop me is my own retardation and ineptitude. I wired everything up it's working (the machine is not put together yet, i did the wiring just on my kitchen table), except i burnt out one of the DRV8825s by turning the potentiometer the wrong way. I know, i'm a genius...>>2823609I saw several similar designs, and they seemed to work fine. If the distance between the syringe and the threaded rod turns out to be too big, i'll reprint that part for peanuts. We'll se tomorrow or the day after, when i put it all together.
>>2824895>i burnt out one of the DRV8825s by turning the potentiometer the wrong wayalways adjust those while measuring their reference voltage live with a multimeterthis is as good time as any to swap to a driver with more microsteps like the TMC2209, less noise and more precision
>>2823691That sounds interesting, could you elaborate on this further? As it stands now, the two motor drivers and the arduino cost me 32 dollars, and since i have to replace one of drivers due to a severe case of my own retardation it'll cost me an aditional 2.7$.>>2823694I chose this design because many others have built similar cheap syringe pumps, and did the same, and it worked for them. The only modification i made to the other designs is that i made it vertical instead of horizontal. I need it vertical, because i'll be working with small volumes of liquid, around 1-3ml in each syringe. I need that pushed through a microfluidic mixer slowly, and i need all of it to be pushed through. So the idea is, the liquid at the bottom will be pushed out and then through the system by the air behind it in the syringe, and for that it needs to be vertical.Using a mill columne is not a bad idea, but it seems more expensive then my current setup. This version cost me 16$ for the aluminium frame, 2$ for the threaded rod and nuts, 8$ for the axls 10$ for the ball bearings and couplers and 10$ in resin. Thats 46$, the columnes you suggest seem to cost around 200-300$. Now you may say, why does it matter, you'd still be saving money. The thing is, i'm not even remtely sure i'm going to be reimbursed for this machine. I don't know if you have worked in academia, but these are all stingy rat bastards. The head of the genetics department where i work is especially a piece of shit. He steals ideas, and i'm fairly certain he embezzles money. I love the work i do, but i hate the system and most of the people. Anyway i'm trying to keep costs to a minimum. So far i'm a hair above 100$.Anyway i have one more question for anyone who may be interested. Apparently the DRV8815s are sensitive to fluctuations in the supplied voltage, and are best protectedy a 50-100 uF capacitor. My question: how close should i wire the capacitor to the driver itselfe?
>>2824901>how close should i wire the capacitor to the driver itselfe?cops, like secret service bodyguards, work best when as close as possible
what are the best strippers for the thin wire?
>>2822343Please help me >>2824902
>>2824913If you're doing 26g or thinner all day, then hot wire strippers. You might be able to jury rig something out of an old soldering iron.24-28g you can get some speciality shit. I haven't had a great experience with these but the only set I got is old.There's a different style that wraps arpund your hand but I can't find any
>>2824920>helpmillions of youtube videos on the subjectseems easy enoughtry this one and look for more on the RHS panelhttps://www.youtube.com/watch?v=rvLzWW9Ot6I
>>2824944>millions of youtube videos on the subject>seems easy enoughI know about the subject, I work in security and previously worked fixing laptops and phones... And no, none of those videos will help, unless you or I manage to find the right one, of course.
>>2824913Some really thin and tough (plastic?) wire is just to touch against a butane flame to make the insulation creep back a couple of millimeters.
>>2824901Note sure what's there to explain. On the pro side it's a finished product with known drivers. Aready tried and tested by the annex engineering folks, meaning you may ask for help on their discord. Less chance to fuck up wires and looks better than an open mess. On the down side you already spent time and 30 bucks on your current approach.
What’s the point of making one of these:https://musicfromouterspace.com/analogsynth_new/TRANSISTORMATCHER/TRANSISTORMATCHER.htmlCouldn’t you just get a good 5V supply, a known resistor, and put it in series with a BJT with its collector and base tied together? Why use an op-amp to buffer the output? Why use a dual rail supply?
>>2824974>Why use a dual rail supply?#1 reason: it makes the ardunio kiddies freak out and run away like it's radioactive technology
I am an absolute noob at electronics and i want to make a video sync detector using arduino nanoJudging by various sources, video sync is -300mV and analog pins on the arduino are 0 to 5V only. Can i use a buffer and put a 300mV voltage divider after it to raise the voltage to appropriate levels?
>>2825072Just use an LM1881.
>>2825072>raise the voltage to appropriate levels?no, that's sillyyou use a hi-speed comparator to trip on a given analog voltage, and turn it into a digital signalonce you're in the lands of 1's and 0's then you can manipulate it at will to fulfill all your depraved needs
author of >>2824902 hereI'm now 99% sure about the chip. it's MX77L12850F, spec sheet here:https://www.macronix.com/Lists/Datasheet/Attachments/9023/MX77L12850F,%203V,%20128Mb,%20v1.3.pdfnow I need to learn how to use the CH341A programmer..from what I can see here: https://www.openhacks.com/page/productos/id/2955/title/CH341A-24-25-Series-EEPROM-Flash-BIOS-USB-Programmer-with-Software-%26-Driver24xx = EEPROM and 25xx = SPI. is this correct?
>>2824974Yes.https://www.dragonflyalley.com/synth/images/TransistorMatching/ianFritz-transmat0011_144.pdfI've made this with a regular 12V barrel jack.According to an old Roland document, you only need 1mV match for most applications.
i am designing a simple pcb where i will have 5v micro hooked up to a lipo battery 4.2V but i also have a small vibration motor which needs 6V and 200ma to work. What is the easiest way i can turn 3V - 4.2v lipo voltage into 6V on the pcb? some sort of very simple boost converted? is there something like that in one chip where i only add few reservoir caps to it?
I have a circuit board that I want to sleep and wake up from multiple CAN sources/transceivers (the TJA1043T). I attached a screenshot of the block diagram of the component. Would it be acceptable to gang together the INH lines of multiple ICs together and attach them to a wake circuit or would there be a problem back feeding the lines from other sources? From what I gather the internal transistor probably functions as a P channel mosfet. Any ideas?
>>2825148>some sort of very simple boost converted? yes>is there something like that in one chip where i only add few reservoir caps to it?yes. search Digikey for DC-DC converters
I spent more time I want to admit trying to resin this stuff. After a while I gave up and just applied the resin after soldering everything down using a brush. I also opted for making PCB single side, used wires to connect stuff on front side. Otherwise curing the front would be too trickyIt looks so easy on videos but whenever I try it, either some parts were under curing and peeling off. Or mask doesn't work and I get cured resin on pads. In future I will go for smaller PCBs, hopefully they will fare better. I also ordered paper for silk screen, maybe it will do better at absorbing inkNow I need to design a box for this
Good old-fashioned web site about diy electronics, hope you like https://www.djerickson.com/index.html
>>2825170>either some parts were under curing and peeling offdo you scrub the surface with light sandpaper first? that could help with adhesion
>>2825251Yeah I got it shiny with a hand drill and those spongy wire ball attachment.I don't think adhesion was the issue. It was more like some parts were slightly thicker than the rest and bottom of the resin were still liquid. Maybe my board got warped too much after several tries and washes. Even if I managed to get a good adhesion on entire board, then I had the issue of some pads having cured resin on top that does not come off so I gave up on this board
>>2825072>>2825076Comparator is the right answer, but it doesn't need to be particularly high-speed. A cheap generic LM339 or similar is plenty fast enough.Be warned that composite video signals are AC coupled. This means that when the video has a lot of white in it the sync pulse drifts down and when there's a lot of black it drifts up. You solve this with a termination resistor and diode clamper circuit before the comparator.
>>2825170Check it.https://mgchemicals.com/products/circuit-board-design/electroless-plating/421a-liquid-tin/You will love it.
>>2825254might want to level it with a spatula then? idk this is just one of many reasons I use point-to-point as much as possible
>>2825091I’d at least want to make a low noise high-repeatability PSU for it, probably with a REF3325, maybe some capacitive multiplication. But good to know.>>2825148First run it at 3.2-4.2V and see if it’s good enough. If not, yeah get a boost converter, maybe one of those monolithic modules.>>2825157If the output isn’t push-pull, and all the outputs pull/push the same way, then yes it should be fine to connect multiple together. That’s usually the only reason you’d ever use an open drain output, other than level shifting, which definitely isn’t the case for a Pchan. Should only need one pull-down.>>2825170What resin?>>2825295Or you could just dissolve some lead-free solder in HCl, brush some thiourea-containing brass polish on your board, then brush on your tin chloride solution. Sprinkle some baking soda in the tin chloride before you store it, the CO2 will prevent O2 dissolving into the solution and oxidising Sn(2+) ions into Sn(4+) ions.
>>2824927>>2824949TY! Here's what I'm dealing with. It's basically thin braided wire that has PTFE sheathing. It's slick as snot on a doorknob. My automatic wire stripper just slides over it. Burning it with solder just creates mess and doesn't work. Using a lighter just darkens it and creates a charred blob after a while and the remnants of the wire don't take solder at all.I've used my teeth to remove the sheath. I only did it 4 times. I won't do it again. I'm terrified of creating a $4000+ dental bill. But teeth work and the wire solders nicely.I don't know what AWG this is but it slips past generic strippers. My automatic strippers can handle the PVC wire of this size but not this wire (which I assume is PTFE because it's so damn slick and tough). So they don't do shit to it.HALP BROS! What can strip this????PS: forgot pic. But here it is.
>>2825319Thanks dude (tja1043t guy). I'm scratching my head, i already had the design like this. Had two instances where one of the ICs gets zorched during plugging in. Added some esd tsorbs to the lines affected between board revs but no dice. Seems like the last culprit would be some transient on the ground line between the PC with test interfaces and the circuit board. That's the only path unprotected. Thing draws 500ma steady after the failure, pretty wild. It's always associated with one of these ICs and if you remove vss/vbat the rest of the board continues to operate.
>>2825341It doesn’t have a diode on that line like the CAN lines though. Maybe there’s an equivalent input diagram for that pin, with more detail like in an MCU datasheet, but it really should be safe. Add a series diode and it will definitely be safe, at least for that pin. Or try using 1k series resistors instead of diodes, then a 10k+ pull-down. Ideally, this is stuff you can test without ordering another lot of boards.Post schematic.
>>2825340Hot tweezers or a soldering iron won't touch teflon easy. They'll also make death fumes. You're probably best off with some sort of tensioned-jaw stripper, one with sharp enough jaws to grip the insulation, but with light enough tension that it doesn't just cut through the copper. Though if the insulation is thin enough, the small amount of area on two sides might not be enough, and so you may need to go for a twin-V wire stripper, not that I've ever seen a tensioned-jaw one.Conventional automatic wire strippers should be able to go down to a low enough pressure (assuming they're the adjustable ones) to not damage that sort of wire, but I'm not certain. Is it the cutting or the holding jaws that are slipping? Try using a file to sharpen the slippy jaws of your current pair to see if that works. If not, maybe you can 3D print a jig to hold razor blades in a fixture or something.
Forgive my ignorance but I dont know where else to ask. I just picked up an old coin acceptor which came with a usb interface. It is a simple board, sends 12v to power the coin mech, via 4 pin loom (top), usb out (bottom). Its recognised as a generic usb input device under win10. I've used a usb sniffer software and this thing is 100% generating some kind of event via usb when I drop a coin in the mech, but I cannot for the life of me figure out what data (if any) its sending. I just want this fucker to trigger a keystroke in mame. I know I could snip the wires and use an actual keyboard encoder, but since the coin mech and this interface (whatever the fuck it is) is working, now I'm trying to puzzle out exactly what it is and what its doing. Strangely the chip has had the logo roughly sandpapered off so I cant identify that either. Any suggestions?
>>2825340>HALP BROS! What can strip this????i have 2 sets of the most basic strippers you can buy: the one with a sliding screw and the one with a cami modified them both to go as low as i wante.g. see how i filed down one flat on the cam so i can go down to AWG30 or less on the original AWG24 settingand that's the only setting i use as i strip every gauge based on smell and feel, like an expert lockpicker feeling for the pins
>>2825380>Any suggestions?i'd use an integrator (cap + resistor + CMOS gate) to convert the USB data pulses into 1 long pulseuse the pulse to drive a tiny reed relay, or a CMOS switch like the 4066wire the relay or switch into a PCB pulled from some old keyboarda reasonable computer should react like it has 2 keyboards, and accept input from both at same timeuse this article for inspiration: https://garrysblog.com/2024/01/09/making-a-game-controller-from-an-old-usb-keyboard/
>>2825380Where'd that board come from? If there's an ebay listing (or other source) with a picture that doesn't have the IC name sanded off you might be able to find a datasheet or user manual, though it's probably just a microcontroller. Or maybe a manual for the "USBJK.PCB BS0622" somehow. Worst case it needs proprietary drivers. In that case you might still be able to use some sort of (hardware or software) USB packet sniffer to figure out what's being sent, and bodge together a driver yourself. Best case it's some sort of USB-serial signal.>>2825386>convert the USB data pulses into 1 long pulseOr he could just take the pulse from before the usb chip? Worst case it won't even output a pulse if it isn't connected to a usb master, though I guess if there's a spare USB port that probably isn't a problem.
>>2825386Sound like an interesting project but its a bit beyond me, I'm an absolute novice with electronics, computers good, pcbs not so good.>>2825387Thanks but nevermind. Evidently they still sell these things on alibaba etc. I just happened across a free one from an old junk box. Its apparently a 3rd world internet kiosk controller (comes with cd rom for win xp/2k lol) so chinese proprietary it is.http://www.happmart.com/Pro_Details.aspx?id=394&px=31&sid=44I could easily give it the snip and just run the coin mech itself into an ipac, but I don't like letting shit like this beat me (especially since everything obviously works) so I'll probably go full autism and spend some time really tearing that usb interface apart with a sniffer.
>>2825401You could use the optical pickup from an old mouse to simulate a coin mech.
mBed is closibg in 2026. What now? There is no RTOS alternative that is as convenient to use.
>>2825421Why don't you just learn to code and fork it?
>>2825340candle and a window, just burn it off
>>2824628I don't remember posting this before as I finished it very recently but then again I am retarded so it could be that I did. I chose to do everything with rj45/utp for ease, availability and cost effectiveness. Also mamy music related things (as well as the 16-relay boards) are multiples of 8.
>>2825474Look mum no computer has done several identical projects and he continues to struggle splicing the massive number of wires involved in his projects.I guess ecelebs jave their own discords now and don't need to ego post on 4chan anymore.
>>2822594fuck yeah, i love making PCBs it's so satisfying when it all comes together. well done anon.
ayo, i just inspected my headphones, and noticed the cord has exposed wires. i've had these headphones for about 2 years.can i just put electrical tape around this and ignore the problem, or do i need to buy new headphones?
>>2825491first, you need to kill yourself
>>2825493do you love china
>>2822343listen OP I'm going to need an explanation of that image.because that is taken from a man's POV, and she has her log on his legs, which means they do the sex and she solders.where do i find one.
>>2825494do you love being retarded, because it seems like you do
>>2825496so yes i'll use electrical tape
>>2825497no
>>2825491Desolder the wires and install a 3.5mm jack in your headphones. Use a locking 3.5mm stereo patch cable to connect the headphones to a source.
>>2825501thanks bro
>>2825502no problem, anon
>>2825478Yeah I saw that project, it served as an important inspiration for this project. Tl;dr my local church has an enormous pipe organ which isn't used as nobody can play it. It's in need of repair and we got the money to do it but the parish council said we aren't allowed to as it'd be a waste of money to restore an organ which isn't played. I meant this project as a cheeky checkmate to that argument. Shit is cash though, pumping some Vivaldi through this thing sounds fucking awesome.
>>2825495>she has her log on his legsw-what?
i have this capacitor that instead of having one side with a negative sign painted, has a positive on top then a > in the middle and then a negative at the bottomis this just another way of indicating the negative?t, noob
I took apart an old electric motor from maybe 1930s and I was removing a field coil with a chisel until the end of the coil moved and I was able to see the crosscut of the field coil and its insulation. The insulation was made out of cardboard that was light blue inside. I am not sure yet but I think that in worst scenario that is blue asbestos. I will test a sample of the cardboard to be sure. The motor didn't have a type plate but it was manufactured by Siemens-Schuckert and looked like one in the pic. Be carefull when fixing old machines. The insulations were mostly made out of wood until you cut the coils.
>>2825563Yeah.That must be very old, though. Way back when they would mark the positive instead of the negative. If you can do test it to make sure it still works.
How can I test a circuit that's supposed to run on low voltage AC while also having OCP? My power supply can only put out DC.
>>2825604Over-current protection? I'd use a conventional step-down AC transformer to get the desired voltage, then put a fuse in series. Maybe a breaker, but those can be kinda slow to trip. Another option is to have a bridge rectifier in series with the power, with a MOSFET and a current sense resistor inside it. You'd have an amplifier on the sense resistor, and if the instantaneous current ever goes above a threshold, it would trigger a latch that turns off the transistor. You'd have a button to reset the latch. An NE555 is a good choice for this, it has comparators and an SR latch inside it, can run at 12V, and can source/sink a decent amount of current in and out of the MOSFET gate.If you want constant-current regulation though, that's gonna be harder.
>>2825601thanks anon, picrel is the cap.
What’s the first step on how to use a scope? Got this cheap ass little one to figure out if I’ll use it enough to justify buying a real one. I never thought I needed a multimeter until I got one and learned how to use it.Can I see like fuel injector pulses on a car with the thing?
>>2825792I got the same one. The nice thing about battery powered units like that is that you can ground it wherever without causing shorts.First you calibrate the probe, don't bother with the cheapo wires they give you, you probably bough the pack with the precision probe so get that one out, hook it to the calibration signal on the top of the scope and calibrate it. Remember it has to be on 10X, the manual will cover that if you need more info.>Can I see like fuel injector pulses on a car with the thing?Sure. Ground it somewhere on the chassis and hook the probe on the injector wire. Most all scopes these days have that auto button to try and fit the signal on the screen, on this one it is quite slow and I often do it myself manually but it might help you get started.
>>2825795I have no idea what any of that means. I’ll learn. I just hooked it up to one of them smart battery chargers because the box says “High Frequency” and you can see a funny wave every second or two where it pulses. No clue what any of the settings are really, mostly the X1-X10 etc. but saw how the 5V ans the 500mS are divisions on X and Y I think
>>2825801A typical probe will look like this. When I bought mine it had the option to pay a little extra and come with one of those or just get the basic black and red wire you have on the left on your pic.X1 and X10 refer to the attenuation of the probe. That basic two wire one won't have that, obviously. But most proper probes do, a few very specialised ones even have X100. That indicates the amount of attenuation of the signal, that is how much of the signal is reduced before being sent to the scope. The more attenuation the smaller the signal will seem but the higher the impedance of the probe will be. In order to see that signal on your fancy screen, you must steal a little bit of it, higher attenuation means less of the signal energy is stolen by the probe and thus lessen the influence of your measurement on the signal itself. Rarely that is actually a problem and you'd want to use 1X for those rare occasions, but most of the time you'd work on 10X. They all use BNC connectors and your probe should have come with an adapter for those.The tiny wave you're seeing on the screen seems like noise, not any real signal.The X axis is time, the middle setting you have selected on your pic changes how much time each square on the grid represents. Y is voltage of each square too. You want Y to be as low as possible so you get better magnification of the signal while also keeping the whole wave on your screen. X should be as low as it has to be to get 2 or 3 cycles on your screen. At the top of the scope, between the jack for the probe and your selector button is a little metal tab with a hole in the centre. That outputs a square wave and is used to calibrate the probe. You don't need to ground it for that because it's already grounded inside the scope. That's a good signal to experiment the settings on. Calibration is done at the base of the probe with a small screwdriver, the reading should be as perfect a square wave as possible, without rounded edges.
>>2825810I got the probe.And that was noise on the screen. I went to a longer time increment when it was hooked up to the battery charger and could watch a funny pulse roll by. I’m trying to figure out what I fux with next
>>2825801>>2825810The menu on the left has some additional functions that took me a while to discover. The run button at the bottom freezes the scan, pausing the signal you have on the screen and keeping it there. It also unpauses it. Holding that button removes the text from the screen so you can better see the signal.Mode cycles between the trigger modes. Normal runs the scope only when the trigger is met, Auto is the same but it if no trigger is present then it forces the scope to trigger anyway, showing that baseline trace you see on yours. Single mode means the scope with run once the trigger is met, but only for one cycle.beside the mode on the top left side of the screen, you have an arrow falling down. That indicates your trigger mode, it means the scope will see a falling edge on the signal as a trigger, starting at its topmost point and going from there. You can change that by holding the mode button (took me so long to find that...).That top scroll button is used to select the functions on the bottom and you already found X and Y scale. In this scope as well on most I've seen the scope doesn't figure out on its own whether the probe is on 1X or 10X mode, so there's an option to set that and tell the scope what you're doing. This will adjust all the relevant parameters the scope calculates for you and displays on that text in the screen.You'll also notice two arrows on either side of the screen they're also adjusted with the top scroll and the arrows on the side. The first one on the left is your zero, that determines where in the screen is the zero voltage level. The one on the other side is to make the signal stationary on the screen, if it is on the screens but appears to be rolling, adjust that one till it stays still.
>>2825810>>2825815Dope. I’ll keep playin with it.What is the difference between the AC and DC setting? Seems like it just moves the bottom of the wave to zero or puts the X axis through the middle of the wave.
>>2825818AC and DC are couplings modes. DC is as normal as you'd expect. AC puts a capacitor between the probe leads and cuts out on DC. If you have an AC signal riding on a DC one, that mode will filter out the DC and you'll see only the AC wave.The " How to Diagnose and Fix Everything Electronic" book has a good section on scopes and their modes.>>2825815Good. Use it.Forget about that basic one, this is the one you want.
>>2825819So can I stick the probes in a 120V outlet and watch the sine wave? Maybe I’ll try it on my $500 non-inverter generator.
>>2825825>>2825819… to see how dirty a cheap generator AC wave is compared to FPL’s grid
>>2825825You can do that on both modes and the result will be the same.But let's say you have AC sine wave that's riding on a 10V DC one, meaning the "zero" of the AC wave is actually 10VDC. On DC coupling mode, you'd have to shift the zero on your scope all the way down to fit the whole thing on your screen and maybe mess with the scale too, all to see a sine wave in a situation where that 10VDC offset might not matter to you other than knowing it is there. If only you could remove that offset and have the sine wave oscillate around the zero on your scope... That's what the AC mode is for. The capacitor will block DC and you'll only see the AC component of the signal.For measuring outlets you can use either. Since this is a battery powered scope you can put ground on one lead and probe the other if you have an ungrounded plug. You can't go grounding your scope willy nilly like that on a real scope, else you risk shorting it and causing quite a bit of damage.
>>2825833Ok this would’ve helped looking at that battery charger, because I had to set it to 5V increments so the pulses were a small squiggle. If I set it to AC, then the 0 on the X axis would probably be like 12.5V (battery voltage) and I could more clearly see the pulses up to 13V?
>>2825543leg* on his legsi was very very tired, i am still very tired.
>>2825783the printing is slightly worn in a really inconvenient way, it's supposed to be a negative. you can see more wear lower down.perhaps it's even a printing error.
I have a box full of old vacuum tubes, most seem to be from old television sets so they're useless for things like audio amplifiers, but could there be fun things to do with them?
>>2825902>most seem to be from old television sets so they're useless for things like audio amplifiershow so?
>>2823340update: got it mostly working.
>>2825815>>2825818Smokings bad, mkay
>>2825841No. Your pulses are still DC even if they oscillate (AC goes negative in its oscillation, that's what makes it AC). AC mode just shouldn't do anything in your case.Battery chargers are usually between 13-14VDC. You have 8 squares on your grid so you could put your voltage scale on 2V (so 16V from bottom to top) and put your zero arrow low at bottom and that way you should see a straight line up top representing your DC battery charger output. Use the lowest possible voltage scale that still allows you to fit the entire signal on your screen, 5V is way too much for most stuff.If you suspect that DC output is noisy and want to look at that noise in particular then you could use AC mode and ignore the 13-14VDC, freeing up screen space so you could reduce the voltage scale even further to zoom in on that AC noise.
>>2825958>If you suspect that DC output is noisy and want to look at that noise in particular then you could use AC mode and ignore the 13-14VDC, freeing up screen space so you could reduce the voltage scale even further to zoom in on that AC noise.I think that's exactly what he's saying by:>If I set it to AC, then the 0 on the X axis would probably be like 12.5V (battery voltage) and I could more clearly see the pulses up to 13V?
>>2825959Yeah, that does make more sense.At first I even thought his charger was outputting AC. I just got home and I'm too tired.
Prime day ends in 6 hours. What kind of controller should I buy if I want to make a battlebot in the near future or RC car/toy/tank. Should I just buy something off of ali express weeks ahead of my project?
Gentlesirs, I got a 12V LED "lamp" and a photoresistor that I want to supply with no more than 0.7V.My objective is to use the photoresistor to sense light so as to make the LED "lamp" dim down in proportion to the light in a room. I'm thinking of using an op-amp to compare the signal passing through photoresistor against a fixed voltage signal and then having that signal pass through the base of a PNP transistor, with the LED "lamp" connected to its collector terminal.I'm thinking of using a 15V power supply for the entire circuit and then "splitting" the voltage by passing it through a pair of resistors in series in the form of a classic voltage divider, but I want to know if this won't end up being extremely energy-inefficient, and whether there might be a more efficient way to do this.t. retard
>>2826086>photoresistor that I want to supply with no more than 0.7Vso why does this requirement exist?cant think of a single reason why this would be neededwhich means you've created a problem from a non-problem, requiring work to find a solutionjust use any voltage you like, and use a large resistor in series to save currenti always use 1 million volts or more for my LDRs so i can limit the current to pico-amps while getting good resolution between dark to bright
>>2826086>>2826124Totally unrelated but what software do you use to make those diagrams?
>>2826127>what softwaregoogle imageslet some other crumb do the work for you
i bought two of these lightbulbs from chinks https://www.aliexpress.com/item/4000548976800.htmlbut neither of them works when i put it in the socket here at 240V, any idea why? they don't seem broken
What would be the best (and preferably, easiest) way to discharge a capacitor that has been charged up to around ~500 V DC? I would need to produce individual short (maybe ~10-50 microsec) pulses into a device under test. An SCR seems suitable for this because it only needs a single trigger pulse to start conducting and with just that it discharges the storage capacitor until it's empty. But the complication is that the SCR is at the high side with respect to the DUT. Is the some ways to drive an SCR from a logic level control signal in a way that's isolated from the high voltage parts? I tried to sketch something with Falstad circut sim. The "Discharge" switch attempts to represent the SCR gate trigger signal. The "charge" and "discharge" switches should never be closed at the same time. There are scopes for cap voltage, DUT current and gate current.
>>2825170After you etch the board and before populated it with your components use the liquid tin chemical to tin the exposed copper preventing the copper layer from oxidation over time. The resin also called conformal coating is used for enviroments where dust (contamination) may short out the traces on your board.
>>2826138>buy chinkshit>get chinkshit
>>2826209>his lightbulbs are 100% local and free range doubt
LM358 is not very good for audio
>>2826257What kind of gain is the op-amp set to? I can’t imagine anything having that much THD as a buffer at least.
>>2822343I need help. I am a retard with electronics, but a project of mine has led me down this path.I need be able to open a solenoid valve for a short amount of time (1/100-1/6 second) with the push of a butto. The challenging part for me is figuring out a circuit that can do this, as an Arduino is not suitable. the specs of the valve are as follows:>12V operating voltage>6 Ohm resistance>19.22 mH inductance Thanks in advance!
>>2826298Use the arduino to activate an relay.
>>2826301Arduino not possible:)
>>2826306Amplify the signal from the arduino with a transistor.
>>2826298Are you trying to build a supersonic Nerf blaster, Anon?
>>2826311funny idea, but no. I'm building an on-demand direct seeder for corn.
>>2826298>figuring out a circuit that can do thisyou can use a monostable like a 555 to create a short pulse every time you press the switchbut you need to add a power transistor and diode to handle the 2A currenti'd choose an NPN or PNP rated for 5Aand dont forget the diode across the solenoidit's super important
>>2826314thanks for the info! I now have something to go off.
>>2822546Came here to ask about this actually Getting a semi-modular soon and looking to make some of my own modulesThink I'll either do a convolution reverb or a probabilistic drum sequencer. Something digital since I've got easy access to a lot of MCU and SoC devices, still very bad at anything analog.Any experience/tips/advice?
>>2826257wtf is your input voltage?
>>2822714what ICs did you employ? I've tried CEM3340 clones and they're not too hard, although getting them to output a solid sine wave is a bit tricky>>2823462if you're just making basic stuff, through hole soldering onto some pcb then you'll be able to sell to people who are just getting into it and need some basic modules for cheap. a majority of the market is people who've already dumped heaps and heaps of money into it, so you'd have to have a pretty standout design to sell heaps.>>2824626nice, you'll enjoy this playlist: https://www.youtube.com/watch?v=8PwwRR8deHk&list=PLluPQLh1xzlI7EMB5qIxDd_1OLE-Z_kyC>>2826329nice, which semi-modular? i've had a couple of cheapy behringers and they're not bad. been meaning to buy a moog one, but they're a bit expensive. if you're going to get into building your own shit, get a simple output module DIY kit. you'll get a bit more experience with the basic analogue circuitry that way. also, put together link rel: https://pichenettes.github.io/mutable-instruments-diy-archive/module_tester/or something similar; makes testing your builds a fucktonne easier.
>>2826362>standout design to sell heapsThat's too bad. Seemed like a good idea. Any other markets for electronics that aren't saturated?
>>2826366Take an existing product, and make it worse by adding wifi and Bluetooth and calling it smart and making it phone home to a Chinese server. People pay out the ass for this kinda stuff.Being more serious, I think there is room for innovation in a lot of areas without becoming hugely more complicated. But any good electronics design should go hand-in-hand with good mechanical design too. And good mechanicals can be expensive. Same for good power supplies. Synth modules and guitar pedals are rare cases where mechanicals don’t matter much and you don’t really need an internal power supply. I think there’s still a lot of room for novel and interesting synth modules and guitar pedals. I also want to design a piano keyboard where the keys can be pushed in different directions (forwards or down, front or back) and with differing amounts of speed and force, so you can get a whole bunch of nuanced information about play style to feed into a polyphonic synth of some sort.
>email a bunch of US fabs to get a quote for a fpc run>every single response is from an indian
>>2826298>>2826314This, but I'd probably use a MOSFET instead of a BJT.FYI, using an NPN/Nch transistor means it's on for the short period of time and off by default, while using a PNP/Pch transistor means it's on by default and off for the short period of time.
>>2826400>every single response is from an indianhow do you know they're indian?do they email using smoke signals?
>>2826441They use "sirs" a lot in their emails
>>2826428 And if you use a PNP/Pch transistor, you will have to place it on the high side of the load (between Vcc and the load) instead of on the low side of the load (between the load and GND). Usually it's better to design some control circuitry so that you can use a NPN BJT or N channel MOSFET, since they have better characteristics than their complementary counterpart.
>>2826362>what ICs did you employ?A TL072 for the input buffer and exponential converter. The rest was transistor-based.
>>2826400hope you didn't redeem.
Hi /ohm/, /3dpg/ newbie here. I have a 12V PWM fan that I want to use on my printer, the issue is that the fan header is a 2 pin 24V one; in practical terms voltage ranges from 12,6V at 1% to 24V at 100%. For the time being I'm using a el cheapo buck converter and running the fan at 100%. I'd like to create a circuit that regulates the voltage down to a fixed 12V and use the difference with the source voltage to command a VCO for driving the PWM of the fan. The issue is I don't know how make the regulated 12V, please send help.
>>2826817Use a buck converter or 7812 regulator wired directly to your power supply to run the VCO & fan.
>>2826820and don't forget the flyback diode across the fan terminals.
>>2822343>tear down wifi "smart outlet">offline buck (non-isolated) topology>has a push button for human interface>"UL certified"i always read that non-isolated offline topologies were inherently unsafe and i assumed they were non UL certifiable... but then i look at every "UL certified" "smart outlet" on the market and they're of similar size/form so i assume all of them are likely non-isolated topologies.are these actually UL certified?
Trying to wrap my head around this circuit from an old lm317 datasheet. I'd like to develop something like this into my own circuit which is to consist of a switching pre-regulator + LM317.When the output starts drawing current, R3 develops a voltage across its terminals. This voltage is amplified with an LM301. Will the LM301's output become more and more negative with increased output current? When LM301's output reaches a certain negative voltage (maybe about -3 V), it starts pulling current away from LM317's adjustment pin. Is this how it's supposed to work? What's the purpose of D1?
>>2826970You already have a pass transistor and an op-amp, why bother with an LM317? You’re just getting a shittier feedback loop with thermal protection that won’t protect the pass transistor. The only thing the LM317 is uniquely providing is a voltage reference, which isn’t very good even compared to a TL431.
Anyone else feel like they're in a weird spot in their abilities? I'm decent at soldering. I can read schematics and I can make PCBs. When it comes to microcontrollers, I can read a half-dozen tutorials and Frankenstein the code together into something functional. But I can't create from scratch. I have no fucking clue how people reverse engineer stuff. I can break an idea down into pieces that I can figure out, but fuck me if I could make something without help. I'm kind of just playing LEGOs. I don't have any engineering background, so I don't feel comfortable going to makerspaces and shit. Anyone else feel similarly? Like an imposter? (Not looking for validation, just some thoughts and opinions.)
>>2827047do you know how different components work? or what they do?
>>2827085I couldn't tell you the scientific process behind how they do what they do, like how a capacitor physically stores a charge or how a transistor switches, but I understand most of their functions and how they work together.
>>2827047Starting from scratch can be tricky because shit might refuse to work if a detail is wrong, and knowing what is wrong requires somewhat deeper knowledge. I'd start by learning some ASM. Use whatever resources there are out to learn the ISA (probably ARM) and read the datasheet of your MCU to understand how GPIOs work. Get some LEDs to blink with a minimal firmware, where you know what every byte does. Then continue with C.
>>2826820>>2826821Thanks anon. After posting I realized I was being retarded: my DMM wasn't picking up the PWM that was coming from the 24V fan header and was just giving me the RMS value. In essence my step down converter is shitting itself because of the input being far from DC. I hooked up my scope to the fan header, discovering that the negative pin sits at 20V in respect to ground, and I got confirmation.The good thing is that now I just need to pick up the PWM signal and transfer it to the PWM pin at the proper voltage, the bad thing is that simply hooking a large cap between positive and negative triggers the short circuit protection of the motherboard, so I need something a bit more fancy than that.
>>2827136You could use a 12V Zener diode after the buck converter to clamp the voltage. Maybe you could PWM the ground side of the fan with a MOSFET.
>>2827160Forget the diode after the buck. I'm retarded. Maybe a resistor divider on the PWM line would work.
>>2827047no, that's normal. modifying known projects is the non-retard way to learn.but of course tryhard 4channers always try (and fail) to do the opposite
>>2827162I checked and the PWM frequency coming from the motherboard, about 300kHz, is out of spec for a PWM pin. My solution is the following: LC filter to turn the PWM into DC, the regulator to give me my 12V, a 5V 25kHz sawtooth generator (thanks w2aew), a comparator shitting out a 5V 25kHz standard-compliant PWM signal. Pic related is my idea in DaveCAD.
I was reading the recommended literature from the OP's pasta. Read Getting Started in Electronics and finished How to Diagnose and Fix Everything Electronic yesterday.I feel like I skipped a step in between one and the other. The first one was a good introduction to the components I wasn't familiar with and had some cool circuit schematics to analyse and think about how they operate and maybe test a couple. The second didn't have as many schematics and specially the later half where he explains common failure and signal tracing, I still don't understand how the signal gets produced and shaped and what not from one state to the other and some uses for certain components are still obscure. I know about amplification and whatnot but I can't just form looking at the circuit understand what the signal is supposed to look like to then make sense of the signals I'm probing.Should I go back and read something else is this normal and I can continue to the next book in the lest, the All New Electronics Self-Teaching Guide?
>>2827186https://electronics.stackexchange.com/questions/674983/right-way-to-step-down-a-pwm-signal
>>2827196I can't just step down the voltage because frequency is off by an order of magnitude. I thus need to rebuild completely the PWM signal, and the easiest way I came up with is using the rectified DC voltage off the original PWM which is a function of the PWM duty cycle, and compare it with a sawtooth with the correct frequency and thus generate the proper signal for the fan.
>>2827207>frequency is off by an order of magnitudeMy fault, Anon. Your solution should work. My first thought was to use a ATtiny and thermocouple or thermistor to PWM the 12V fan independently and cut power at low temp.
>>2827195the second book sounds pretty high level, not in a mathematically-hard way, but more of a "you need irl experience" way. i would start breadboarding some simple blinker circuits, so you can see for yourself what poor grounding, connections, and wiring is like in practice
>>2827232>you need irl experienceYeah, it felt like that for the later half. >breadboarding circuitsI did so with the first book in the list because it had those schematics for me to build.
Most brutal thing I saw today... retard takes a fucking dremel with a carbide bit and uses it to "clean" the damage from a battery leak on his retro Amiga computer. The results are as you might expect... carnage.
>>2827255https://www.youtube.com/watch?v=53rbInd47ws
>>2827259kek'dit's hard to top 8-bit guy with his paperclips and retrobrighting misadventures... but this is just as retarded.I'm actually amazed these pre-boomers are posting this shit on YT. Anything for clicks & $$$ I guess...
>>2827263noodle arm needed pliers for bending a paperclipthats next level onions
>>2827186>LC filter to turn the PWM into DCit turns it into a kind of DC, but it's one that's jumping up and down quite a bitbetter add a diode in series so it never discharges into the sourceyou'll get just the peak voltage, 24V
>>2827270300kHz is more than enough for that to matter. with a 1ms time constant, a 4-pole Sallen Key should easily get it down to the noise floor
>$15 food scale from walmart has a 400 mAh internal battery & usb-c chargerwho the fuck decided that this was a good idea? why would you put a ticking time bomb inside a device that hasnt needed one for like 20+ years.there's even a slot for 2 AAA batteries so the internal Li-ion serves absolutely no purpose other than to increase cost and make the thing break in a couple years
>>2827280because people will actually pay for that. though i think they should have used a 14500 cell in a battery snap.
>>2827286digital scales need to be less footprint than spring powered ones, else people might wake up to the scam
>>2827270>>2827272Well, I sourced an inductor from my trash pile, should be 10uH, and a new 300uF cap from my stock, and they seemed to do the trick. I wasn't picking up any significant ripple with my scope either. Now that I think about it I could turn to a Sallen Key: I was planning to use a comparator to generate the PWM, but if I use a regular opamp I could use the second one in the PDIP to run the Sallen Key, and I should have an overall smaller circuit compared to the beefy LC. Thanks for the tip anon.
>>2827333The only advantage of a high-pole filter over a low-frequency RC filter is you can have faster response time. Even an LC filter is overkill, and probably necessitates a bit more care not to cause a spike somewhere.
>>2827349Wait, I just realized I can't use that topology: the opamp would need to rectify a 24V square wave, producing 12 to 24V DC, while hooked up to a 12V supply. That's just not possible. I guess I'm back to the LC low pass.
>>2827359Also because it’s a low side transistor, the average voltage will be 24V at 0% duty, and 0V at 100%. If you can adjust the PWM frequency in the firmware (pretty sure it’s possible) you’d still need an inverter along with a logic level shifter. A logic FET does both nicely, don’t forget the pull-up resistor. You can also avoid both issues by taking the signal from the fan FET’s gate instead of its drain.If you can’t change the frequency will you use to generate the new PWM signal at a different frequency? A 555 circuit isn’t very linear when used as a voltage controlled PWM generator (and it can’t go to 0% or 100% easily), I’d sooner recommend a dual comparator, or just an MCU.
>>2827359Actually if you use an MCU and a level shifter, you could use one timer/counter to figure out the duty-cycle of the input signal, and use it to set the other timer to PWM the output. That will minimise your response delay better than any filter. Also no need to bother with ADC reading.
NEW THREAD >>2822343NEW THREAD >>2822343NEW THREAD >>2822343
>>2827656
What are some ways to cover up the underside of a veroboard in order to avoid something touching it and shorting it all out?I've finished building my circuit but it's still a prototype so I haven't built a proper case for it. The wires coming in and out are insulated but the side of the board with all the soldered connections isn't. Last time I covered everything with electrical tape but after a while a component burned out and I had to remove it and deal with all the glue residue. What are some less messy options?
>>2827720Use a less sticky tape? Or maybe cut a strong plastic bag into shape and tape that around the edges instead
>>2827720Use standoffs or rubber feet. If the board is small you can heatshrink it.
>>2827656kek... we have to wait for page 10 unfortunately
>>2827720Neutral-cure silicone
>>2827765Oi, that one sounds nice. Is it easy to remove later if needed?
Remember that new baker who burned his house down killing his whole family because he sat by the screen updating some ancient pantyhose knitting forum instead of watching his bread?
>>2827773If the board is clean beforehand it's easy to peel off. Test it on a scrap board first and try to remove it.
>>2827774diamond hands
>>2827773>Is it easy to remove later if needed?ignore the crazy suggestionsjust cut to size a piece of acrylic sheet, or cardboard from a cereal boxpoke 4 holes and use wires to attach in twist-tie style
>>2825340https://www.aliexpress.com/w/wholesale-fiber-optic-stripper.html
>>2827929Why is it crazy? Maybe you're just a lazy vagina.
>>2827930lmao
>>2822343I want to build one of those floppy drive to SD cards. what should I use for the pins to hit the floppy drive? it's driving me crazy trying to figure out how to wire it.
>>2827955Saw a guy on youtube do something similar. A card-edge connector or pin connector like thing is going to be high friction and not that good for cycle life. Pogo pins work, but you'll probably want some sort of moving mechanism so pushing the card in causes the pins to pop up and press into the pads. Similar to how a locking IDC connector's side levers swing into place when you push the connector down. If you're going to have a cool rotating mechanism anyhow, you could maybe ditch the pogo pins for springy bent metal pins like what's inside a connector, but pogos are going to be easier to mount.
>>2827956i dont really know what any of that means because i'm retarded. is there an off the shelf product that i can get where the pins are able to be used? i dont really want to regularly have to disassemble the thing but with pogo pins im afraid of the low ammount of contact and also i find it kind of strange that there is no more spring like shape as you mentioned so that they can be depressed better. its just a huge problem. I dont know what part to pick.I also am trying to use UHS II/Express pins too so it's not as easy as a lot of these videos show. the only thing that seems like it would be the right shape is a spring finger but I cant find where to buy one or in the right size.
>>2827955>floppy drive to SD cardsWhy in the fuck would you do that?
>>2827958Don't copy that floppy. You wouldn't download a pair of Air Jordans.https://www.youtube.com/watch?v=wy7ZBX9BTUA
>>2827960sleeper build ricing. also some media is better off in external. and in any case i find that the floppy disk form factor isn't just retro cute, but is superior for display purposes in the sense that you can print off labels for each one. on a regular SD card theres no real reason to do that, theyre very tiny, and best kept in a closed space so they don't get lost or broken. I also have a case designed for optical media too, but I decided to start with the floppydisk pizza box case I have.
>>2827964So you want to put an SD card in a floppy disk and read/write to it with a floppy drive?
>>2827972no, im using a regular SD card controller, but i need to route the pins to the surface of the floppy mounted SD card. while floppy disks that accept sd cards exist, floppy controllers are much too slow to be worth the effort even before you consider the outragious cost of the floppy disks due to the rarity.
>>2828001Why not use NFC tags in the floppies that mount different disk images on the external SD card with a microcontroller?
>>2827972these are so amusing
>>2828002?????i have no idea what you are talking about at this point. no internal logic of the floppy drive is used.
>>2828005I have no idea what you're talking about either, Anon.
>>2828006https://www.youtube.com/watch?v=gNQYbSWBhAs
>>2828001>>2828007>i need to route the pins to the surface of the floppy mounted SD cardTake a length of stranded copper wire and separate the strands to connect the pins to the card. Use something to insulate the wires like lacquer or solder mask. Magnet wire works too, but you have to scrape or burn the enamel off so solder will stick to it, and some people find it to be a pain in the arse.
>>2828019strands of copper wire aren't going to stand in place very well, thats the entire point of the discussion.
How do you fix these switches? I've two sets of these that are getting no connection. I tore both down, cleaned pads with 0000 steel wool and alcohol, then washed caps in soap+water and dried out. One of the failed switches on one bank works; the other bank has 4 that didn't work before cleaning, and still don't work now.
>>2828021>the entire point of the discussionif you can find a thin enough SD socket, you'll get easy and reliable contact with any cardif cant find thin enough, you could cut away one or both sides of the floppyuse matching-color sticker on opposite side to cover ugly cutout hole
>>2828071Reflow the solder joints and check the cable or connector for continuity.
>>2828071>cleaned pads with 0000 steel woolnever use steel wool, no matter how softan eraser is good enough to get rid of all oxidationcould be the carbon pads on the rubber part is worn outyou can check with an ohmmetereverything does look in good order, so maybe wiring or connectors are the problem
whats wrong with this? its doing weird shit like continuing to operate with the 24v rail disconnected.
>>2828099ahh i had the gate and source shorted, fuckin shitty ass datasheet.
Almost got the RJ45->DB9 harness working bros. Kinda worked at first, altho some of the text was garbled... but then it went to complete garbage. Is it because I used jumpers? What sort of wires should I get? 22AWG solid core wires?How should I go about turning this into a "permanent" adapter once I get it working?
>>2828083>>2828077I’ll look harder at the wires. Testing was done at solder joints so that’s a valid possibility.
>>2828120Twist the pairs together and keep the wires as short as possible. Solid core 22-26AWG is acceptable.>turning this into a "permanent" adapterMake a PCB with a RJ-45 and DB-9 connector.
>>2828145>keep the wires as short as possibleand make them all the same length.
Is a pull-up resistor still needed when an open collector output is connected into a PNP transistor like in pic related? Will the PNP base end up in some undefined state without it?
>>2828153>Will the PNP base end up in some undefined state without it?pretty sure an open base is not defined in the datasheetbut it's prob fine 99% of the timeadd the resistor so it'll work 101% of the time
>>2828153I was always under the impression that the PNP's gate has to be pulled high to turn it off.
>>2828073these aren't 17 pin uhs II and I havent been able to find one, thats a micro sd
>>2828145>>2828146Is this better? Sorry I'm retarded. Not sure what you meant by twisting them so I just twisted the grounds together>Make a PCB with a RJ-45 and DB-9 connector.Do i have to learn PCB design or is there a simpler way?
>>2828099Is that a JFET?>>2828153No base current = no collector current, it should be fine. Though if there's any leakage current to ground in parallel with the IC's output that could cause a phantom current. Breadboard it yourself to see.>>2828163What protocol is it? If it's ethernet, lmao, good luck. If it's RS232, then do testing at a lower baud rate until it works flawlessly, then increase it to see if there's any bandwidth bottlenecks (there shouldn't be). Also if it is RS232, consider a twisted-pair line transceiver IC to turn the single-ended signals into differential signals, it's much better for long distance transmissions over unshielded cable.>PCB designMight be a good idea, especially if you need an IC, but if it's just a wire-to-wire adapter, you can get away by soldering wires from connector to connector instead of using those shitty screw terminals. Not that I've seen a non-board-mount 8P8C socket.You can also just buy these, pic related.
>>2828163+- +- +- +-
>>2828167It's RS232. Shitty network switch -> RJ45 -> DB9 -> USB -> PC>testing at a lower baud rateWouldn't the other end(network switch) have to support that? The documentation says to use 115200>twisted-pair line transceiver IC to turn the single-ended signals into differential signalsYou lost me, sorry>it's much better for long distance transmissionsIs this considered long-distance transmission?>You can also just buy theseHoly shit, that looks perfect. What do I do with the wires I don't need?>>2828172So I should run 2 twisted wires between each terminal? My hands are too big for this shit
>>2828176Wait, it's an RS232 network switch with RJ45 plugs? Kinda strange, but ok. Post the pinout.>What do I do with the wires I don't need?Leave them disconnected, I'd probably cut the connectors off and maybe tape the ends up so they can't short against anything. That one, PA0906 is sold by Jaycar in Australia and New Zealand, there's probably similar things sold elsewhere in the globe.
>>2828176>So I should run 2 twisted wires between each terminal?Yes. +- +-
>>2828183>>2828183>>2828183NEW THREAD
>>2828159>>2828160Thanks anons, I'm going to add a 100k pull-up just in case
>>2828178Pic related is the console output and pinout. Goes from somewhat legible to complete garbage.>Wait, it's an RS232 network switch with RJ45 plugs? It's a network switch with an RS232 over RJ45 management port>Leave them disconnected, I'd probably cut the connectors off and maybe tape the ends up so they can't short against anything.Yeah that's what I figured, thanks>That one, PA0906 is sold by Jaycar in Australia and New Zealand, there's probably similar things sold elsewhere in the globe.Yep. Got one for ~$4, it's gonne be here in 2 days.>>2828180Why do you keep typing +-?
>>2827930thank you!
>>2828190>Why do you keep typing +-?So you don't twist gnd-gnd because that would do nothing.
>>2828194i.e. Rx-gnd Tx-gnd
>>2828194>>2828195Just to make sure I understand, I should add 2 wires, one from RJ45-5 to DB9-3 and one from RJ45-5 to DB9-2? And twist each around the other wire going to the respective DB9 terminal?Also, according to the documentation RJ45-4 is actually not connected... not sure why they have it showing in the pinout.
>>2828201Or are you saying it should be RJ45-3 and RJ45-6 to DB9-5?
>>2828167>Is that a JFET?NMOS, the first thing i grabbed out of the bin. is there a better device for making voltage sources like this?
>>2828201>>2828207On the RJ-45 side, twist 3 & 4 in one pair, and 5 & 6 in the other pair. DB-9 side is 2 & 5, and 3 & 5.
>>2828218Ohh i thought you meant to literally connect them. Fuck that was a pain in the ass.
>>2828232If it's still not working, lower the baud rate, and make sure Tx and Rx cross over between connectors.
>>2828244No output at all now, not sure what I screwed up. I'll just wait till I get the 22AWG solid core wires to try again, thanks for the help
>>2828252Swap 2 & 3 (yellow wires).
So i built the syringe pump i was talking about earlyer. The 3D printed desings didn't really work, the resin was too brittle and the metal parts were somehat bent (about 2mm) which isn't a problem except the 3D printed parts needed to be sanded to shape and it was too hard to do. Anyway, i had some hobby glass (which isn't glass its plastic, Poly(methyl 2-methylpropenoate)). So i just drilled some holes, and sanded it to shape where needed. It looks like ass, but it works fine.For the structural elements i used copper brackets, nuts and bolts...The electronics i put together in the way i'll show in pic related. Fuck fritzing btw, i made it in MSpaint. For now i just chucked it all in a plastic box i bought at SPAR.I have a couple of questions. Can i just glue the electronics to rubber spacers? If i put the arduino and the two DRV8825s in a box do i need to cut holes in the box to avoid overheating? Any other advice?
>>2828374Cool stuff. Yeah resin printed parts are not really strong enough for mechanical use. For this use-case, you can probably turn the current way down on the stepper drivers to minimise heat production. Be that via UART or via the trimpot. If the syringe won't droop under the gantry's own weight, then you can even disable the stepper drivers when they're not being used.Looks like a pain to remove the syringes to wash or replace them.Also we have a new thread: >>2828183
>>2828083>>2828077>>2828142Follow up on this. After testing, etc, pulled the cable to closely inspect the board for trace damage. Realized the cable had been soldered in backward at the factory, so the Red marked line wasn't Common. Soldered it back in the right way and all's good.
>>2828583Nice. Good job, Anon.
