|
Charles Ford posted:I suspect they weren't "recycled" (like they weren't pulled off a board or anything), but programmed at the Microsemi factory (as a service they offer), placed on the reel, then at some point they fell into the hands of PCBWay's suppliers. If you want to count just selling pre-owned, unused otherwise original parts as counterfeit then sure, of course. I feel like I'm in bizarro land, how are you not upset about being sold defective parts? Parts that are, at best some form of supply chain injection or (IMO more likely to be) some sort of ewaste counterfeit?
|
#
?
Apr 26, 2024 21:04
|
|
|
|
| # ? May 8, 2024 00:35 |
|
|
Are you surprised? Did they have a contract that specifies preferred suppliers? Did they ask where their stuff comes from? You can be upset, but like, garbage in, garbage out. Pcbway is really good for what it is, but they're only going to do as well as you force them to. They don't know whether your primary motivator is "cheap" or "legit". Most of their customers only care about the former. Especially at the tail end of a supply chain crunch
|
|
#
?
Apr 26, 2024 21:21
|
|
|
Exactly - I'm not really bothered as it's a bit of a gamble. This isn't the first time they've done something wrong (though last time it was just using the wrong IC entirely - same 74 series chip but different parameters so it'd burn out being fed the wrong voltage in a few minutes, which they also blamed on the mystery "suppliers") and they'll give me a refund on the parts + an amount to cover extra effort in fixing it. Though last time they did also offer to fix it, so I'll be interested to see what they actually say this time. It's such a saving over the US companies I've used before that even with the need to swap out those chips (which I already have a stock of) it's still going to have been cheaper. The real main issue is my SMD soldering is shoddy, but it's also a skill I want to improve, so I can live with that aspect. I've also actually sent them parts before too, which obviously makes it much more likely it'll be fine, so possibly for more exotic parts I'll just do that again next time. The first time I did it I was a bit worried they might just steal my parts, though, as you need to send extras in case there's trouble. e: I will say I am actually mildly interested in who PCBWay's suppliers are. There's hints on their site they'll use Digikey/Mouser but during the semiconductor shortage they were also able to source seemingly legit CPUs/etc. that weren't in stock anywhere I'd buy anything from. They wouldn't tell me when I asked. Charles Ford fucked around with this message at 22:15 on Apr 26, 2024 |
|
#
?
Apr 26, 2024 21:27
|
|
|
They definitely buy most of my parts from digikey - that's what takes three weeks to clear their customs
|
|
#
?
Apr 26, 2024 22:24
|
|
|
Speaking of aftermarket chips, does anyone know anything about *selling* into that market? I have maybe $5k worth of STM32F4xx chips at work that we bought when they first started appearing on the market after covid. They weren't the exact model I wanted but they could be made to work in a pinch and after two years of absolutely NOTHING being available we just said 'gently caress it, buy them all' and ended up only using a handful before the processor I really wanted started being commercially available again. So I have ~350 factory new processors that I'll almost certainly never use, it would make sense for our lab to sell them off as surplus and get a bit of cash back if we could, but I have no idea about where to even start looking for reputable buyers. Also I'm sure the university I work for is going to have some major red tape involved in trying to surplus a bunch of non-trackable items.
|
|
#
?
Apr 26, 2024 23:02
|
|
|
Cyril Sneer posted:Quick question. I lost the IR remote for this RGB LED string: You didn't actually ask a question  Anyway that says it requires an RF remote, not an IR remote, so that's probably why the others didn't work 
|
|
#
?
Apr 26, 2024 23:05
|
|
|
PDP-1 posted:Speaking of aftermarket chips, does anyone know anything about *selling* into that market? eBay?
|
|
#
?
Apr 26, 2024 23:12
|
|
|
Shame Boy posted:You didn't actually ask a question Hah, good catch. Yeah, I think you're right. I never realized they even used different technologies for this.
|
|
#
?
Apr 26, 2024 23:33
|
|
|
Cyril Sneer posted:Hah, good catch. Crude RF transmitter/receiver hardware has gotten cheap enough that a lot of things that previously were IR now just throw that in there instead. Confusingly they sometimes just re-use the plastic mold for the remote, and to fill the hole at the top where the IR LED would have gone they put a normal LED that just blinks when you press a button.
|
|
#
?
Apr 26, 2024 23:40
|
|
|
Anyone have a versatile charger they're happy with? I need to charge a variety of 18650 lithium and the like, sometimes SLA ups batteries below 10Ah, that kind of stuff. * I have a lovely one meant for RC/drone lipo packs, which works ok for those specifically, but can't tweak voltage limits on different chemistries, the fan is dying, and the UI sucks. * I have a really good victron charger for car batteries or larger sealed lead acid, but it only has 3A and 10A settings, and SLA batteries below 12Ah or so shouldn't be charged above 2A. * My benchtop power supply will oscillate weirdly around the CV/CC transition point as a battery charges, and I'm worried about it being damaged. I suppose as it nears the CV setpoint, it backs off the current from the CC limit, which causes the voltage across the battery to drop slightly, which lets the current rise again, and so on in oscillation. Also a regular benchtop supply doesn't do absorption phase time limiting and backoff to float, etc..etc.. The main things would be having actual charge profiles with bulk/absorb/float and so forth, as well as actually customizable current limits and voltage setpoints as weird chemistries come and go.
|
|
#
?
Apr 27, 2024 21:32
|
|
|
I would like to confirm/deny my hypothesis about why this lamp stopped working. I have this USB work light https://assets.cdn.jula.com/preset:...=19790101000000 https://www.jula.se/catalog/el-och-belysning/armaturer/arbetsbelysning/arbetslampor/batteridriven-arbetslampa-led-024720/ It runs on a rechargeable battery and charges with USB-C A few days ago my girlfriend plugged a USB-C laptop charger into it. It ran for a minute or so and then went out and now it won't turn on. The specs for the lamp says "Input: 5V 1A" and the laptop charger says "Output: 20V 3.25A / 15V 3A / 9V 2A / 5V 2A" So my guess is that the laptop charger gave it 5V and 2A and fried it due to too many amps? Pictures of the disassembled lamp and of the charger: https://imgur.com/a/Dh4wOmN I can't see anything that looks fried though. Would this be something that could be fixed with some tinkering?
|
|
#
?
Apr 27, 2024 21:47
|
|
|
USB-C stuff negotiates voltage levels and poo poo before turning on the juice, so that should work fine. USB-C laptop charger powering a desk lamp is A-OK.
|
|
#
?
Apr 27, 2024 23:16
|
|
|
A power supply never gives more than the device "asks for". That lithium cell can deliver dozens of amps, but the circuitry ensures that there will never be too high of a current flowing through the LEDs. There are exceptions - extremely dodgy AliExpress/wish/temu devices that don't have an actual charge controller for the lithium cell - but this device looks like it has some generic charge regulator IC (by looks - i haven't verified it with the datasheet) so you can hook it up to a 100a power supply and it will still be fine. The only thing i can think of, is that the power supply does not correctly switch between voltages. Measure.to verify. Also check to see if the cell isn't deep discharged. Sometimes the charger locks out in such a case.
|
|
#
?
Apr 27, 2024 23:21
|
|
|
LimaBiker posted:Also check to see if the cell isn't deep discharged. Sometimes the charger locks out in such a case. I think it could just be this - if it was switched on and charging, the LEDs probably used more current than the battery was receiving from the charger, so despite being "on charge" it went flat. This could be because it didn't correctly ask the laptop power supply for 5V 1A with the proper USB process, so the laptop followed the USB spec and only have it 250mA. This is a common issue for cheap electronics as they don't want to pay for a chip to do USB power negotiations, so both the charger and device just offer 1A without the negotiation. I can't think of any reason why that charger would've damaged the device though. I think it's extremely unlikely the laptop charger put out more than 5V - the USB PD spec is designed specifically to stop that happening, and USB laptop power supplies are either compliant or need to be recalled.
|
|
#
?
Apr 28, 2024 04:56
|
|
|
Is there some way to look up multiple parts at once? I have these piles of chips from grab bags and I want to cut down on all the turnaround on looking them up.
|
|
#
?
May 1, 2024 08:22
|
|
|
Rocko Bonaparte posted:Is there some way to look up multiple parts at once? I have these piles of chips from grab bags and I want to cut down on all the turnaround on looking them up. There's a DigiKey API you can sign up for.
|
|
#
?
May 1, 2024 12:14
|
|
|
eBay finally delivered the weird connector I needed to create a canbus splice for my motorcycle but I've got a very rudimentary understanding of canbus termination requirements. This arguably speaks to "you don't understand canbus so please don't monkey with your vehicle's central nervous system until you do" and I'm very cognizant of that, which is why everything I'm doing right now is designed to be removable when I'm actually riding the bike. Anyway, I effectively bought the connector required to plug into my bike's tire pressure monitoring system canbus signal, so my intent is to create a Y fork and provide canH and canL to a split off connector that I can use to either probe, or to eventually hook up an esp32+can interface to READ data. The new offshoot twisted pair will be fairly short, maybe a few inches at most.  My question is about termination requirements when I'm just using the new offshoot connector for probing with a DSO or LA. I guess I just don't understand the requirement, actually. Ideally this is a cable I can keep in place even while nothing is hooked up, at which point I presume I WOULD need to terminate it? What would be the potential effect of just leaving the connector unterminated while it's not being used? I have a sinking feeling that this is a super basic question with a really obvious answer but somehow I'm just not putting two and two together. I'll take my lumps but I'm thankful for any thoughts here. My end goal is to install an esp32 that can send basic bluetooth HID to my phone based on the canbus messages from the scroll wheel on my motorcycle's handlebar which is effectively useless without BMW's own proprietary GPS unit on my older bike. The messages are all documented by enthusiasts, I'm just trying to figure out the connection now without making a mess of my bike.
|
|
#
?
May 1, 2024 14:24
|
|
|
You don't need to terminate the stub you added. It's fine. You can leave it in place. Every car has a stub without a node connected: the diagnostic connector. It is supposed to be there. It doesn't cause problems. If you are asking about a terminator or terminating resistor: No, do not do that. The bus is already terminated, probably in the ECU and TPMS. Some definitions: Terminating resistor - 120 ohm resistor, one at each end of the bus Backbone, to suppress reflections Backbone - the longest length of the bus between two Nodes Node - ECU or device communicating with the bus Stub - length of wire between the Node and Backbone
|
|
#
?
May 1, 2024 15:10
|
|
|
A+, thank you. Yes, I was worried about terminating with resistor. Totally forgot about the unterminated diag port. Thanks!
|
|
#
?
May 1, 2024 16:15
|
|
|
babyeatingpsychopath posted:There's a DigiKey API you can sign up for. Octoparts also has an api that’s free for even significant hobbyist request levels, I used it to find parts during the chip shortage for a low run product with a Python script. Salami Surgeon posted:Every car has a stub without a node connected: the diagnostic connector. It is supposed to be there. It doesn't cause problems. On my car the diagnostic connector is its own bus, straight into another module, to act as a kind of firewall. It’s annoying as I’d like to be able to control the doors with OpenEVMS but that needs access to I think the body can bus (the car has like twelve can buses, according to the maintenance manual). I too was thinking about hooking into one but I haven’t decided a good place (eg near the obd2 port where the module is) to hijack for the signals. I also kind of want to fiddly with the HUD, which only has two can buses going into it (instruments and adas) but there’s nowhere good to get access to those signals either.
|
|
#
?
May 1, 2024 23:08
|
|
|
I think diagnostic connector on its own bus might be driven by a ISO 21434 thing. Fairly recent. So more accurately, almost every car in the past 30 years has a stub without a node connected on the main CAN databus. Cyber security is going to drive a lot more individual databuses and burying safety critical ones behind firewalls and gateways like you are seeing. Note: almost all my experience is on the SAE J1939 side and not directly with ISO 11898 CAN. J1939 is probably a bit more robust than ISO 11898 on the message side, but I think the physical layers are identical. CAN is extremely robust. Don't worry too much about screwing something up. The J1939 spec does not explicitly forbid empty nodes, but it does have one explicit mention of an empty node: the diagnostic connector. In practice, I've seen other empty nodes. A prime example would be a car with manual and automatic transmission options; one main harness part number would include an AT ECU stub that gets left empty on a manual. I've seen lots of crazy stuff too. Stubs should be under 3 meters, seen a lot over 3. One terminating resistor. Three terminating resistors. UNtwisted pair. Too many nodes. Stubs off of stubs. Things Just Work. Until they don't, then everything spits out dozens of fault codes. If you are not getting fault codes, let it run.
|
|
#
?
May 2, 2024 02:38
|
|
|
Thanks for the replies!Splode posted:I think it could just be this - if it was switched on and charging, the LEDs probably used more current than the battery was receiving from the charger, so despite being "on charge" it went flat. But shouldn't I then be able to charge it if I plug it back in? I tried with a phone charger as well but the charging LED on the lamp doesn't light up either.
|
|
#
?
May 2, 2024 11:22
|
|
|
Not necessarily. If you deep discharge a lithium cell below a certain voltage, the charger can lock out because deep discharged lithium cells can become unstable. Some chargers allow charging at a very low speed (phones usually do, they then appear not to charge but in fact do), some don't.
|
|
#
?
May 2, 2024 18:01
|
|
|
As a complet novice I can attest to that… I’ve seen plenty of “dead” batteries for laptops or anything you can imagine that wouldn’t charge. Pull them apart and hookem up to a power supply and they’ll except charge, hold it and discharge just fine. I find that the actual cells (like 18650s) themselves are crazy robust… but I’m not using them very heavily at all so it may just be that they’re out of spec. Vvvvvvvv very true two posts below. always look it up on batteryuniversity.com… but I think I set voltage limit to 4.2 and current limit to 1 amp for 18650s since I’m almost never trying to quick charge or push anything to limits. CC/CV are a definite must namlosh fucked around with this message at 14:19 on May 3, 2024 |
|
#
?
May 2, 2024 21:52
|
|
|
I always liked what farmers did in the old days for dead car batteries: zap them with a welder!
|
|
#
?
May 2, 2024 23:48
|
|
|
namlosh posted:As a complet novice I can attest to that… I’ve seen plenty of “dead” batteries for laptops or anything you can imagine that wouldn’t charge. Pull them apart and hookem up to a power supply and they’ll except charge, hold it and discharge just fine. I find that the actual cells (like 18650s) themselves are crazy robust… but I’m not using them very heavily at all so it may just be that they’re out of spec. And it must be explicitely stated that 'power supply' means a current and voltage limited bench/lab power supply, and you want to charge it at 1/10 or 1/20c until you reach the lower limit of a lithium cell. Also be aware that you can blow up the power supply if it's set to like 1v, and you hook up a battery up to it that delivers 3v or something. The control electronics will attempt to lower the voltage and get very confused.
|
|
#
?
May 3, 2024 08:02
|
|
|
Hi all. I'm back with some extremely basic questions about troubleshooting. For reference, my previous posts:The Wonder Weapon posted:I'm trying to do some basic led strip lights and I've hit two in a row that I'm having trouble with. I can't tell if it's bad product, or if it's something I'm doing wrong. Since I initially posted this, I've ended up with three separate LED strips from three separate manufacturers, and two separate LED controllers from two separate manufacturers. This is the second LED controller I bought, for reference: (https://www.aliexpress.us/item/3256805081032487.html?spm=a2g0o.order_list.order_list_main.112.28671802GtMaMJ&gatewayAdapt=glo2usa) Every combination of LED strips and controllers is giving me the same result: red works, but the other colors don't. I can't imagine that all three LED strips are defective in the exact same way, and it would be surprising if both LED controllers were causing issues in the exact same way. So I presume I'm doing something that either doesn't work, or is just wrong, but I can't tell what. I hooked up my latest batch as follows: LED strip > 4 pin connector > led controller > usb 5v adapter. (Same thing as in the picture in my original post above) Previously someone in the thread mentioned testing with the diode tester on the LED strip. I don't have that setting, so I tried measuring voltage at the copper pads, but to be honest I don't really know what I'm looking for. Here's my voltmeter, and here's the strip. What exact points do I need to be checking, on what setting, to test the other colors?  
|
|
#
?
May 3, 2024 17:26
|
|
|
The Wonder Weapon posted:
You can't do it with your multimeter I'm afraid. You need a DMM with a diode test. A diode test basically puts 3 volts out through the probes and measures how many volts return. 3 volts is enough to make most LEDs glow. Most colors will light fully while white and blue LEDs will glow dimly. They like around 3.5 volts to be at their brightest. Keep in mind that a diode test is a directional test as diodes are polarized devices. Even a good, functioning diode will fail a diode test if hooked up backwards. The same goes for Light Emitting Diodes (LEDs).
|
|
#
?
May 3, 2024 19:49
|
|
|
kid sinister posted:You can't do it with your multimeter I'm afraid. You need a DMM with a diode test. A diode test basically puts 3 volts out through the probes and measures how many volts return. 3 volts is enough to make most LEDs glow. Most colors will light fully while white and blue LEDs will glow dimly. They like around 3.5 volts to be at their brightest. Ah, well. I will have to see about borrowing one. But testing the strips aside, there's no way I got three strips in a row that all had exactly the same malfunction right? What else is going on here?
|
|
#
?
May 3, 2024 19:58
|
|
|
You're trying to run 12V LEDs with a 5V power supply Phone posting but I'm sure others will be able to run with that information
|
|
#
?
May 3, 2024 20:56
|
|
|
Gahhhhhhh I knew it had to be something stupid I was overlooking. That's probably it! I will hunt one down and see how I fare. (also drat these 5050 RGB 120LED strips use a lot of power; looks like 5a for a 2m strip)
|
|
#
?
May 3, 2024 22:00
|
|
|
Anyone want to offer opinions on workspace organization stuff they like/hate? My work is moving offices and I have a chance to change things around. Current plan is: - Index card filing boxes for IC storage in ESD bags - Cover one entire wall with bin rails and plastic bins - Power strip screwed into the top of benches for portable things, cords for everything semi-permanent (soldering irons, scope, bench power supply, ...) down to power strip behind and on the floor/back to leave the top plugs free - A few test lead hooks on the wall by every bench to try to lure people into at least hanging things up somewhere instead of letting them get tangles, and a ~5' stretch with labels by kind for them to get eventually sorted back onto - Wire spools on portable caddys (I don't love these since they get tangled, but also don't want to have to walk to get wire) - Per-person "keep track of your own poo poo" toolboxes for screwdriver bits & hex wrenches instead of trying to have common ones that inevitably get lost
|
|
#
?
May 4, 2024 23:22
|
|
|
Anyone found sets of crimp tool & 0.1" headers that they like and aren't total crap? Stuff for making your own custom female-to-female connectors, using stuff like this: https://www.amazon.com/CHENBO-Connector-Housing-Assortment-Terminal/dp/B077X8XV2J/ Also interested in larger stuff like molex sizes or PC fan headers, that kind of thing. I've been using screw terminals everywhere, and they get fairly expensive and take up a lot of PCB space as opposed to just some 0.1" male pins or horizontal connectors sticking off the board edge. "Proper crimping tools" always cost a fortune but these days I feel like there must be stuff that works OK without costing $100+
|
|
#
?
May 5, 2024 01:39
|
|
|
I just squeeze 'em with needle nose pliers.
|
|
#
?
May 5, 2024 01:45
|
|
|
I got some Engineer PA-09 crimps recently for that kind of thing, they seem to work pretty well but aren't too expensive.
|
|
#
?
May 5, 2024 03:00
|
|
|
Charles Ford posted:I got some Engineer PA-09 crimps recently for that kind of thing, they seem to work pretty well but aren't too expensive. I also have these, I use them on JST PH-2.0 connectors and they work pretty well, way better than the crap pair of cheapos I had where it would miscrimp about 50% of the time despite painstakingly lining it up.
|
|
#
?
May 5, 2024 03:03
|
|
|
Charles Ford posted:I got some Engineer PA-09 crimps recently for that kind of thing, they seem to work pretty well but aren't too expensive. Yeah Engineer crimp tools are the closest you're going to get to good crimping without buying the real deal.
|
|
#
?
May 5, 2024 03:39
|
|
|
|
| # ? May 8, 2024 00:35 |
|
|
Foxfire_ posted:Anyone want to offer opinions on workspace organization stuff they like/hate? My work is moving offices and I have a chance to change things around. Put side cutters and tweezers in the "your own poo poo" toolboxes. People always try and use side cutters to cut through fuckin bike locks or something and ruin the blade edge. Similarly they use tweezers as crowbars and gently caress up the tips. I've just implemented a similar system for smd components (ESD bags in small boxes), so far so good, but a bit too soon to tell. The rest sounds great, take a photo when it's done if you can!
|
|
#
?
May 5, 2024 06:07
|
|