I'm going to add some commentary and thoughts.
well...
i feel that if i'm making references to a 12V system and asking about level shifting that maybe some inferences could be made, but ok.
I think this is precisely one of the issues that comes up very often when trying to diagnose a problem: that inferences have to be made. The goal, I think, for anyone requesting help is to make it so almost
no inferences need to be made. I mean, where that boundary is could be a sliver of a grey area (for example, you don't need to state that connecting wires need to be conductive, but you should state your voltage, and what's connected to what, etc.), but it's very often the assumptions, implicit or explicit, that are the source of difficulties when helping to diagnose.
One of the biggest challenges, I find, when there's any difficulties or frustrations, or even misdirected anger, in most conversations, is identifying these assumptions, especially the implicit ones, or even identifying that there
are any implicit assumptions.
are we talking this over a level shifter or in addition to a level shifter?
it's not clear if i am shorting 12V to the data line, but i have a lot of flickering and artifacts in my LEDs, so i am assuming that i need to have a level shifter, even though i have been trying to avoid the extra parts and real estate use.
Flicker on LED strips, in my experience, is most often power-related, sometimes related to grounding, and sometimes it's just a bad power supply. In fact, just this past weekend, after swapping two "it couldn't possibly be the power supplies" power supply inputs, it was identified that one of the supplies had to be replaced.
Also, the power-grounds must match the data-input-grounds.
this brings me more questions:
i had to look up what a zener diode is because i wasn't familar. since i'd like to block any return voltage, can i use a regular diode?
would i have a simpler circuit if i did that?
i'm more of a relays, buttons, and wiring type (electrical vs. electronics), how do i determine which diodes and resistors to use?
It is my experience that sometimes people can get frustrated when something "they already know" is explained back to them in an effort to help. That's another thing I encounter quite frequently, that someone might even get insulted because they assume that the helper "should know that they don't need to explain something because I already know it!" Looking a little more deeply at that, how is the helper supposed to
assume what the other person knows? Back to assumptions.
So... when I help someone, if there's even a hint that they might know something, my strategy is to append a parenthetical that "you may already know this, but I'm explaining 'just in case' and 'for future readers'," or some such — just in case I'm "supposed to have known" I didn't need to explain the thing to them.
It all goes back to assumptions: Which ones are safe to make (almost none in many cases), which ones might trigger somebody, and which ones don't need to be made (eg. wires are conductive).
further...
i am familiar with loading up a parts cannon and firing for a fix; since i have thermal images and have identified the chip (GD E230F8 JJS515; the ARM microprocessor?) that is heating up, could i replace that chip and fix this pair of boards or is the whole board FUBAR?
and on that note, considering that i am experiencing an apparently terminal problem with the WS2815s (which have been installed for over a year at this point and not successfully powered up, would i be better off just tearing out the WS2815s, which are the problem, and replacing them with WS2811?
since i'm pretty sure that i've identified the microprocessor is the failure point, is it possible i'm forcing the processor into an overclock and that's what's burning it out?
Are you specifying the build to use an overclocked speed?
I don't think replacing the processor will work because the bootloader chip is paired to a specific processor, once that pairing process is done once.
noted. i am currently adapting to space requirements, but the thought hasn't escaped me and i probably have room and a space on the bike to place a DIN rail fuse box in a place i recently discovered.
since the system is frying the Teensy instead of powering the lights, i believe i am experiencing a feedback problem in the data pins and i am prioritizing protecting the board in the future. the bike is an '82... it's kinda on fire all the time.
One thing that could, in theory, happen is a data line that's shorted to the power or something. To me, that would fall into the "I didn't even think of that" category, similar to my "it couldn't possibly be the power supplies" assumption above I didn't realize I was even making.
i appreciate that you're attempting to help.
first thing to address here is that it seems you're talking at me like i might know what i'm doing wrong.
clearly, i don't know quite enough to keep from burning out two Teensy 4.1 boards with the same LEDs. the time frame between posts here is about six months, but the time between dead boards is supply chain, plus life, plus about twelve minutes of diagnostic time; the second board died in the same moment that it was plugged in to those LEDs for the first time after i did the most basic tests that i have tools to do.
i'm frustrated, so i will iterate what i have previously said.
I appreciate that you're identifying your frustration and feelings here, and that you're acknowledging the help. It can be so frustrating when a project just isn't working, and sometimes that can bleed out into interactions that others pick up on. That can even spiral.
@jmarsh is quite technically competent and capable, and probably one of the top contributors and insightful helpers on this forum. When replying to what sounds like a very frustrated person asking for help, it's easy to bristle and respond back in a mirror-like fashion, or it could be perceived as such, even when no intent is there. Crafting a perfect response is difficult, especially when feelings get involved.
The bottom line is that questions you're asked here are very likely relevant to helping solve the problem, especially when someone like
@jmarsh is asking them. This is a very good and very technically capable forum, and people here will get just as frustrated as you (alongside you, that is) when something isn't working right. Heck,
@PaulStoffregen will even buy components and try to reproduce an issue if he thinks there's a true problem, and then fix it. Not many designers or manufacturers will do that. Also keep in mind that everyone here will try their very best to try to help, and especially that everyone who gets involved in a discussion here really does want to solve this with you.
i am powering the Teensy off of the onboard USB port with a USB cable with a USB 5V charge port on the 12V system of the motorcycle, which has been used to charge my phone, power a dashcam system, and a few other things i have used to confirm proper working operation. i don't know how i can explain that in any better method. the USB cable is only powering the board.
the LEDs are directly powered from the 12V system of the bike. i do know enough about things to understand that trying to power a 12V system from a 3.3V power supply will result in a nothing burger.
so let's redirect...
i am trying to figure out why my test bench system (which worked perfectly) and my installed system (which fails by becoming a high output handwarmer) are acting differently; while the bench system works fine, the installed system is torching the Teensy board with prejudice when i try to operate the WS2815 LEDs in the system.
the only connection to the board and the LEDs are through the data pins. they are power supply isolated.
Not saying this is your problem, but check out that Hackaday article and video on how power can be "un-isolated". (Hence the fuses.) It might add some more thoughts you may not have considered.
Thanks for reading all this. I'd love to see pictures and video of your completed project. I'm curious, can you tell me more about your pattern generation software and algorithms? I like to connect on this subject with others who do this sort of thing. LED controls and patterns is one of my main hobbies.
I guess I'm just curious if you're doing all your own patterns, and what kind of mathematics you're using. For example, do you use regular/periodic patterns or add some noise? Do you use 2-D or 3-D mathematics? Etc. Totally respect the trade-secret nature of what you're doing.