Anybody else seeing this? I have a project that is running perfectly happily most of the time. Using Teensy 3.2 and about 200 APA102 leds. The leds are in two different form factors; standard strips and what Ray Wu calls beads. The interface between the two requires a custom connector, but not that big a deal, just soldered wire splices…they all work, most of the time.
Every once in a while some of the bead leds glitch…wrong colors, persistent colors when they should be off…and then they recover and go back to nominal behavior. My first suspicion is that the data stream is getting interrupted, but how? I’m not running any interrupts, or even reading inputs.
What are you using for a power supply and how are things wired up?
Are you using a level shifter (74HCT245)?
Have you tried re-soldering your custom connector wires?
@marmil Power is standard 5v 10 A supply. Yes, level shifters, common grounds, common clock line, separate data lines. The soldering seems secure, no effect when I wiggle the wires. 100 ohm resistor on each data line.
I had seen some of the posts on data rate issues, but I think I have that solved…as noted, most of the time it runs just fine. The most promising idea, I think, is that I’m seeing RF signal picked up in the data wires. It sounded like Stefan Petrick was seeing some similar behavior a couple of weeks back. Since the build is sitting next to my WiFi router, it seems like something to investigate.
The leds in the bead form factor do not like high data rates at all. They are in a different package than the standard 5050, with a smaller IC. They will run at 2Mhz, but not reliably at 4 Mhz. I think Daniel has noted in several posts that not all APA102s are created equal.
I’m about to install the larger version, which will get tested outside, away from the router…we’ll see what happens.
@Paul_Guthrie Ive had similar experiences with bead pixels. they have a lower drive frequency because they have longer cables between each other. more noise etc…
)