|
Is there a Spice program that isn't a complete turd designed in like 1994? Has nobody modernized this software?
|
#
¿
Nov 23, 2020 18:38
|
|
|
|
| # ¿ Apr 19, 2024 19:38 |
|
|
Shame Boy posted:There's falstad/iCircuit but it's not SPICE it's its own weird thing that's less useful for actual work but good for like, dicking around and demonstrating how things work. Other than that, no. Well, is there like.. a commercial something available that does circuit simulation? Or are big engineering companies just using LTSpice or whatever?
|
|
#
¿
Nov 24, 2020 08:17
|
|
|
I'm looking for a type of 'night vision' camera to observe light in the ~940nm range. Preferably something portable, like a handheld device, though a camera you hook up to a computer is begrudgingly acceptable. It doesn't need to create its own light, if that matters. I'm working with components in the infrared spectrum and want a way to 'see' what they're illuminating, and also to check visibility of anything which might interfere with testing (light leaks from other sources, for instance). I'll be Doing A Thing with some other test equipment and generally just trying to look through the camera to check the test conditions. Budget is fairly loose, maybe sub-$800 if possible.
|
|
#
¿
Dec 4, 2020 06:25
|
|
|
Anyone have a take on a desktop pick-and-place? Looking to spend less than $5-6k. Was looking at the CHMT36VA.
|
|
#
¿
Dec 19, 2020 06:24
|
|
|
Cory Parsnipson posted:Little late to the last page discussion, but what does an SMD workflow look like? And what kind of decision making or tradeoffs should I consider before using or switching to SMD components? "SMD" is just 'a device that uses surface-mounted components, instead of thru hole or similar'. You can solder SMD components by hand if you want (I wouldn't), but you generally use a specific workflow to populate those boards. A basic SMT workflow is: 1) Stencil printing -- apply solder paste to a substrate, either with a stencil or a solder paste head on a pick and place 2) Populate board -- place components on the solder paste, either by hand or with a pick and place 3) Reflow -- bake the board to melt all the solder at once, either with a hand heater or a reflow oven If you want to 'practice soldering' you want through-hole kits, not SMT. The only hand soldering you should do during SMT is for rework or fixing gently caress-ups. Or crusty old nerds who try to flex on each other by hand soldering 0402's with a microscope e: there's a lot more to 'real' SMT lines, like optical imaging, x-ray inspection, component test and traceability, visual inspection, bake and outgas, etc, but that's the core of it PRADA SLUT fucked around with this message at 23:29 on Dec 21, 2020 |
|
#
¿
Dec 21, 2020 23:24
|
|
|
Does anyone know where to get those linear guide rails / servos that you use in a Flying Probe / AOI machine? I can't figure out how to get something high speed, there's plenty of them for like.. standing desks. e: I know the standard ‘square x’ rails and the belt-driven carts, but is there a high-precision option? PRADA SLUT fucked around with this message at 09:00 on Jan 7, 2021 |
|
#
¿
Jan 7, 2021 07:30
|
|
|
I’m looking for a small solar panel, like 2x4in, to fit inside a case. The dimensions are pretty tight. Is there a “custom fit” panel that would work, or like a ‘cut to fit’ thin-film panel?
|
|
#
¿
Apr 2, 2021 07:07
|
|
|
I'm looking to decode a DisplayPort signal while it's in between a host computer and a monitor. Would it be possible to buy/build a device that splices the wires such that I can connect a pin header between them (and then wire it up to a logic analyzer), or is there some reason that wouldn't work? I know I can physically build things, I was wondering if the signal wouldn't be readable for some reason. Assume I have measurement hardware with the appropriate bandwidth / Nyquist frequency cutoff / etc e: Just considered if I splice the cable but put a pin header on it.. wouldn't I get some reflections since it's a high speed line? PRADA SLUT fucked around with this message at 05:17 on May 18, 2021 |
|
#
¿
May 18, 2021 05:06
|
|
|
I'm trying to read non-"picture" data signals the monitor receives through DisplayPort, like the power on/off/wake/etc signal. I was reading the AUX speeds are slower, so in theory I wouldn't need a but this is all from reading manuals and specs and the like, I've never actually tried to probe it. I was thinking of the FPGA stand-in but I'm not familiar enough with them use to make it my first option. e: something like this PRADA SLUT fucked around with this message at 07:43 on May 18, 2021 |
|
#
¿
May 18, 2021 07:06
|
|
|
Does this layout look weird to anyone? I'm trying to think of the best way to represent this: I was going back and forth on splitting the grounds e: visually, not electrically PRADA SLUT fucked around with this message at 07:59 on May 24, 2021 |
|
#
¿
May 24, 2021 07:55
|
|
|
What are people using to store microcontrollers, FPGAs, etc? I have about a dozen different dev boards that don't fit in a normal parts organizer. Hoping for some type of box for them all (ie, not like a built-in).
|
|
#
¿
Jun 4, 2021 16:46
|
|
|
Improvements: LFO with a fart
|
|
#
¿
Jun 7, 2021 07:03
|
|
|
Is there a dev board out there (eg, ESP32 family, not like an FPGA) with >26 analog inputs (and read at least 3.2 volts, resolution doesn't matter)? I know you can use a MUX on it or build out a cape or something, but I was wondering if a board exists that has that many analog inputs already?
|
|
#
¿
Jun 28, 2021 23:16
|
|
|
Is there a name / place to buy those cables that are like DuPont cables, except instead of the square plastic female ends, they have the thin, rounded ends, like the Saleae cables?
|
|
#
¿
Jul 10, 2021 00:51
|
|
|
KnifeWrench posted:I believe those are just 0.1" terminals with heat shrink over them. Or at least they were back when I bought mine. It's not for hobbyist use, so I'm looking for a place to buy them pre-made if they exist anywhere.
|
|
#
¿
Jul 10, 2021 05:38
|
|
|
ANIME AKBAR posted:Your requirements are a bit vague. What do you want at the other end of the cable? How many positions? Doesn’t matter what the other end looks like. Could be a naked wire, could be another female header, could be a jack, whatever. Could also be a single cable or a bundle. I just need a way to connect to some small pins in tight spaces that a square DuPont can’t fit.
|
|
#
¿
Jul 11, 2021 23:21
|
|
|
Anyone have experience with FCC regulation/approval? I'm building a battery-powered product for commercial use that will have wireless functionality . If I use an already-certified dev board (eg, like an Arduino whatever), to what extent can I modify it or attach parts to it? For instance, if I manufacture a shield that just has some minor parts on it (no radios, nothing clocked >9Khz). and attach that board over the dev board, can I sell that "system" without requiring certification / EMI testing (because the dev board has it), or does the addition invalidate the original certification? Is there a specific set of "allowable" attachments to a dev board, or does any modification change it? I tried looking this up but couldn't find it anywhere. e: I'm aware I could design a custom PCB and get it tested/certified, I'm just wondering if repurposing an existing dev board would be an option
|
|
#
¿
Jul 12, 2021 03:54
|
|
|
Foxfire_ posted:You have to test the actual product (everything put together in the actual configuration an end-user would get & use). Compositions of certified subsystems still require test (they can and do fail). You can skip testing for some things if can write a reasonable engineering justification (i.e. "Rev A passed, then we changed suppliers on this resistor for Rev B" or "we changed the shape of a plastic case"), but an Arudino+shield combo that's never been tested seems more than that bar to me. For example, I could see that radiating more because it's creating bigger antennas for noise already on the Arduino signals. And on EMC side, I doubt there's any existing claim about what the Arduino does because it's up to you to define what 'functions correctly' means for your product. Do you know what testing / certification this system would be? FCC? Some other?
|
|
#
¿
Jul 12, 2021 22:37
|
|
|
Is there a heatsink tape that's removable? I'm looking to attach some minor heatsinks to power MOSFETs for prototyping but would like to be able to remove the heatsink when I'm done.
|
|
#
¿
Jul 14, 2021 04:57
|
|
|
Weird issue. I have an Adafruit Huzzah32 (Espressif ESP32) dev board. When I used it last, I loaded up some minor I2C communication program to it and it worked fine. Fast forward a month, and I plug it in to upload a new program, and I get this issue: Hooked up to my computer over USB, the board gets extremely hot after about 10 seconds, and then starts to turn off and back on repeatedly (ie, connects and disconnects from the computer). I assume it's overheating and hitting some thermal shutdown. I set it over a fan and it stops shutting off and on, but I can't upload any programs to it (it's not a board that needs a boot hold or anything, it works over just the Arduino IDE). The Arduino IDE and VSCode both recognize the port, as well as Windows Device manager, and the cable / etc works with other boards (though I don't have an extra of same board to test, it works with like an Arduino MKR1500) Trying to debug, I hook a scope up to the pins, and I get a qualitatively similar output on nearly all the I/Os, and a few pins like the ground pin:  Tying Enable to ground correctly turns off the MCU and the chip cools down. Also, I hooked up a USB meter to the cable and it reads 5V 1A when it's hooked to the computer (and enabled--0A when not enabled), but when I hook it to a power-only USB outlet the amperage fluctuates between ~0.3A to 0.8A. Any ideas? I'm just trying to erase it and put a new program on it. This is NOT one of the boards with a boot button, so the "hold boot" or "add a cap from En to Gnd" tricks don't work.
|
|
#
¿
Jul 15, 2021 04:28
|
|
|
Sagebrush posted:"Gets extremely hot on power-up" sounds to me like the MCU itself is fried. No good way to fix that other than replacing it. I considered that, but why would the pins output a signal then? I assume I’d get either garbage or nothing.
|
|
#
¿
Jul 15, 2021 05:40
|
|
|
ante posted:I don't know who needs to hear this, but thermal cameras are also getting pretty cheap now. My $300 handheld one is among my most-used piece of test gear, and I point it at every board I bring up for the first time. I actually am in the market for one of these. Options?
|
|
#
¿
Jul 15, 2021 07:03
|
|
|
babyeatingpsychopath posted:This has been my experience with 3V3 boards. 5V gets in the wrong spot and the internal protection diode shorts closed. There's not enough trace width to let the thing draw enough current to pop the die off the package, but the chip gets very hot. I have a couple of ATMEGA chips and at least one Raspberry Pi that do this. For the ATMEGAs, I was able to voltage-probe around with my scope and find the pin that was sinking current and lift it from its pad to solve the issue. That whole peripheral ended up being burned out, so the other 7 pins on the port became output-only. Let's assume I can hit any trace on the board. How would I apply a correct voltage to the MCU, while still allowing for uploading a program over usb from the computer, without drawing USB power? I can see how I do the former, but not both at the same time. Would I need to shunt USB +V but keep the data pins? e: I don't have a JTAG connector for USB > Serial off-hand to try and hit the serial traces directly, bypassing the USB stage. PRADA SLUT fucked around with this message at 16:46 on Jul 15, 2021 |
|
#
¿
Jul 15, 2021 16:36
|
|
|
Is there a place to get extra-wide breadboards? I have a dev board that's too fat for a regular board (I can only hit one set of pins when I place it over the divider). Or, is there a way to hook this thing up that isn't pulling the power rails off two boards and sticking them together (or grabbing the pins directly)
|
|
#
¿
Jul 19, 2021 02:16
|
|
|
What’s the best way to program multiple microcontrollers at once? I’m programming an ESP32, right now over USB with the Arduino ecosystem (but I have a USB to serial adapter), and I want the same program on many identical devices. Fan out a USB to serial connection? Some method of “forking” a USB signal directly? Do I need to replace the boot loader? Another option? Assume I’m programming so many that it’s not feasible to do them individually. PRADA SLUT fucked around with this message at 07:20 on Jul 22, 2021 |
|
#
¿
Jul 22, 2021 07:11
|
|
|
Quantity would be tens of thousands, so no feasible way to program them individually. I know a manufacturer can pre-program them, but I'm looking for a solution that I can do "in-house", whether that be a custom script + building some fixture, or some piece of equipment that's already set up for board programming. I'm leaning for a fixture because I want to put in a diagnostic test at the same time, so the workflow looks like "put board(s) on fixture, write code to board, run test on board", as one step. My other question is that if programming a board that has a built-in USB vs just a serial connector makes a difference for multiple boards at a time. I assume I would want the latter and just have the USB->serial connection handled by the fixture or computer or whatever, not the boards themselves. PRADA SLUT fucked around with this message at 17:27 on Jul 22, 2021 |
|
#
¿
Jul 22, 2021 17:23
|
|
|
Shame Boy posted:Presumably a bunch will be on a single circuit board panel that you snap them out of right? One thing I've seen is to have test traces running to each board on the panel, and then just slot the whole panel into a test jig and program/test all 16 or whatever of them at once before you snap them out. You have to get a little creative about running the traces in the little left-over areas between the routed-out bits but it cuts your workload by quite a bit if you can do more than one at once. They're not on panels, they're discrete boards. Imagine going to Adafruit or wherever and just ordering thousands of dev boards, with the only options being "has or doesn't have a USB connector on it". It's an unusual situation which is why I'm looking to know the best way to program them as-is.
|
|
#
¿
Jul 22, 2021 17:50
|
|
|
I'm aware that there are methods to program multiple boards during manufacturing, or one can make design decisions to make boards broadly programmable (eg, add contacts to a panel). What I'm asking is that is there a 'best' way to program a large number of discrete mass-manufactured boards, either over USB or with serial. ie, the only adjustable design parameter is "has a USB port" instead of serial pins. Eg, if you had a pile of tens of thousands of Arduino boards from Sparkfun or Adafruit or whoever, and they need programming, and it's not just a "get interns, who cares about how they do it" situation. Solutions I was looking for would be along the lines of "make a fixture with pogo pins that has a computer runs the programming upload when you press a board against it then have the board run test code and output the result over serial to a single pin which the computer reads for a pass/fail confirmation" A Proper Uppercut posted:Batteries in parallel act as a single cell and don't need to be balanced, unless even my limited battery knowledge is wrong there. They still need balancing because cells degrade over time and not at the same rates. Multiple clusters in parallel might have specific cells which degrade more than others over time. Cheap BMS's will just take the worst of them and that becomes the default for the entire system. More specifically, "clusters" of individual cells act have sense wires which the BMS reads, and it will infer the entire groups health from the reading on the sense wire. A 12V battery might just be four 3V clusters of like six cells each. PRADA SLUT fucked around with this message at 03:02 on Jul 25, 2021 |
|
#
¿
Jul 25, 2021 02:42
|
|
|
csammis posted:If you already knew the solution why are you asking and then getting pissed when people suggest legitimate solutions that don’t happen to work for you? I don't know if the fixture is a "good" solution, just that idea was just about the only thing I could think of in hardware. Are there like pre-made serial programmers out there that do this for multiple discrete boards faster? Can you somehow fork USB and forgo the need for building a fixture? Can you fan out a serial connection without any issue, or would you hit some limit from signal integrity? Is there a typical way of programming many discrete boards post-manufacturing, or is this an area where there's no real standard method (because, yes, I understand that people typically don't try to program them after they're out of the panels and in a pile). The majority of the solutions were ignoring the initial axiom of programming discrete boards, and instead going "well just design something different". I'm not wondering "what's the best way to design a PCBA so it can come pre-programmed from the factory", I'm wondering what's the best way to program thousands of "off-the-shelf" discrete boards, that isn't "just hire 100 interns and give them Arduino IDE's and a USB cable" Dominoes posted:You are not doing a good job of explaining what you're trying to do and IMO something sounds off. I'll put the problem this way: You have two boxes in front of you. One contains 10,000 Arduino Unos with a USB connector. The other contains 10,000 Arduino Unos with a serial connector. You shall program Blink.ino to each of the 10,000 boards in one of the boxes. You should use a system to program multiple boards at the same time. You may buy or build equipment for this purpose, to a reasonable price (determined through judgment). You shall not program them all individually. You have access to typical equipment found in an engineering lab. "You" denotes any natural person associated with the project. Cojawfee posted:Depending on what kind arduino it is, it probably has a 6 pin ICSP header that connects directly to the microcontroller to program it. You could make a jig that connects directly to those pins to program the boards.You could even program one of them to be the programmer. Connect to the ICSP header, push a button, and have it light up a green LED if good, red LED if bad. Does this just transfer a binary as-is? Like is there a standard library or something that allows for programming between devices like this, or does it need to recompile each time?
|
|
#
¿
Jul 25, 2021 07:40
|
|
Just make ten of those fixtures and crank a thousand through each.
|
BattleMaster posted:Yeah I hope you didn't already purchase the boards because buying hobby grade stuff from a normal hobby retailer when you need thousands of a thing is going to be expensive and labor intensive versus spending some time doing some design for manufacturability and having them delivered to you in a complete state. It's part of a larger project, basically reading some sensor values and syncing them over Bluetooth. The gist is that a supplier has boards "at the ready" that they'll move for basically cost, or I can spin up a custom board but that takes more time and money. So, the tradeoff of getting a bunch of boards that are ready to go and I can have in two weeks, versus a multi-month project of design, test, supply chain logistics, manufacturing, shipping, etc. The "hobby grade" isn't relevant here because it's basically an SoC with some supporting passives.
|
|
#
¿
Jul 26, 2021 04:59
|
|
|
Just get a female/female DuPont and cut one end off. I’ve got a bunch of them from Amazon with no issue.
|
|
#
¿
Jul 28, 2021 06:51
|
|
|
Is there a Windows equivalent to Serial Bluetooth Terminal?
|
|
#
¿
Jul 29, 2021 04:31
|
|
|
Forseti posted:It's been a while for me since I've used Windows, but don't serial bluetooth devices appear to the system as a COM port? In that case I think you could use PuTTY? I’ve used PuTTY for direct serial but Couldn’t figure out the configuration for Bluetooth. I’ll look at it again.
|
|
#
¿
Jul 30, 2021 01:57
|
|
|
Can someone help me understand newlines in PuTTY? I'm trying to get it to send a \n when I hit enter after a command. I'm connecting to Serial (bluetooth). Both implicit CR/LF doesn't help. Using Local Line Editing doesn't help. I'm connecting to a MCU and it has a function that Does A Thing when '\n' is the incoming character, and it's not picking it up (though it picks up everything else fine). It works fine using the Arduino built-in Serial Monitor and the Android bluetooth serial app, but not PuTTY. e: if PuTTY is dumb and I can't do this, what else should I use (for Windows)?
|
|
#
¿
Jul 30, 2021 21:51
|
|
|
BattleMaster posted:I have never had problems with a putty correctly sending \n when I hit enter with the default settings. Hook up a logic analyzer to your serial lines and see what it is sending. I'm connecting over Bluetooth, and I've got a USB cable into the computer. It'd take me a few days to get a "serial" board I could hook to a Saleae.
|
|
#
¿
Jul 30, 2021 22:11
|
|
|
babyeatingpsychopath posted:Press CTRL-J to send just a newline. This doesn't work right when I terminate a line with "CTRL+J" My text line looks like this code:Note that this does work when I use PuTTY in the "immediately send characters one at a time as I type them out" mode, but not in the "line at a time" mode. BattleMaster posted:Aren't there serial lines between the bluetooth module and the microcontroller? No, it's a SoC.
|
|
#
¿
Jul 30, 2021 22:12
|
|
|
AoE is fantastic, though might be hard to understand in parts if you don’t understand basic calculus.
|
|
#
¿
Aug 19, 2021 17:24
|
|
|
Anyone recommend one of those cameras that goes into the top of a microscope, but it also has infrared?
|
|
#
¿
Aug 21, 2021 02:03
|
|
|
So is the take to forget about trying to do IR on a microscope camera?
|
|
#
¿
Aug 21, 2021 04:37
|
|
|
|
| # ¿ Apr 19, 2024 19:38 |
|
|
My concern is just heat actually. Mainly looking at a PCBA and determining if heat sinks are good, if design needs better cooling, etc. But on a microscope scale (leads, solder joints).
|
|
#
¿
Aug 21, 2021 05:00
|
|