Okay so my first test was to unplug the PI from the end of the string since that was the easiest and now all 9 light up!!! Any ideas why the PI at the end could be causing that? I did make the cable myself so was thinking I somehow reversed polarity but I checked that and no issues.
Could something actually be wrong with the last pixel where it won't pass the voltage? For now I just kept the PI at the end disconnected, if I go all white there is some flickering so I would definitely need that PI at some point, but luckily none of my sequences have full white so it shouldn't be a problem.
If I was in your shoes, I would just not use all-white, leave the PI open and move on to something else. I have AC's for all-white; if I want whites, I'll SSR that.
Did you probe the PI leads with/ and without power with a voltmeter? How about swapping the power supply going into the other end? I've sometimes had power supplies fail to power and actually behave more like a load than a power. If your voltage actually dropped when you connect it, then I think you're seeing that. But it's after Thanksgiving, no time, maybe you just need to call it done.
Agreed this sounds like a January problem! haha
This 12v PSU is brand new and dedicated for now to only that series of props but it does seem as if it's adding load and pulling power away from those 9 pixels. It's a Mean well too so wasn't even thinking it could be that. I haven't probed anything either but again for now I'll wait till January. But when I do how do I check them without power? Assuming I could just probe the PI wires lead with the PSU turned on and see if i'm getting ~12v. And then maybe only connect the PI line to the pixels at the end but not connect the data line at the beginning (pixel 1) to see if voltage is passing backwards through all of the pixels? Not sure that actually checks the whole setup with all 392 pixels though.
The PI leads will be one voltage with the PSU on and another with the PSU off. That's not too much work to probe, even in December. Switching PSU's is, though, especially if it's a 2nd-story bedroom window.
There is an uncommon PI problem that Bill Porter only vaguely touched on in his pres (and never specifically said), where if you inject a little too much with different PSU's and their voltages are slightly different, it will actually cause a short. For example if your Meanwell is delivering 13.2V but your controller is providing 11.9V, but when your pixels are mostly off, the voltage never quite dips down to 11.9V by the time your meanwell current reaches back to the controller. That's why the rule of thumb is to isolate the VDD line coming from different PSU's. But that, too, is only a rule-of-thumb that can be broken, if you know your voltage drop best-case (actually worst-case this time) is sufficient.
This may be a dumb question, but wouldn't there be no voltage on the PI leads with the PSU off? Again my controller is on a separate PSU (PSU-1) so only V- and DI are going from my controller to the first pixel. I'm then injecting from my second PSU (PSU-2) at pixel 1 on the string and again at pixel 160. The PI lead with the issue is running to pixel 392 to inject. Would I be probing the exposed leads at pixel 392 or are you saying back at the PSU, or in this case I'm also using a distro block? or do i need to actually use the probe from the multi meter to pierce the cable casing itself? Sorry I've never really tested like this before.
Actually, I just thought that I could possibly put a power injection tee on there and be able to check the voltage there when the PSU is on right?
I did see somewhere the mention of two different PSU's voltage fighting against each other which is why I chose to just completely separate the two and only send V- and DI from the controller.
I also thought about possibly moving my PI point to be ahead of the window in question, which would be at pixel 272. If that worked then it could be something with pixel 392 itself.
