Just upgraded to 16.0 and noticed a number of WLED drivien lights fixtues have “gone nutz”. Lights start normal and then start acting nuts (recognizable as patterns or actions, not the flickering of bad LEDs).
When trying to see if maybe the settings for some effects trying to figure out what is up, I keep seeing popups about ADALight and receiving sync an option to disable once or for a bit (whic never seems to “stick”). I never enabled anything called ADALIght (never heard of it til this and googling).
I do not have any sort of syncing enabled on these fixtures. Never have. They are both DigiUNO units that have been running for years.
Any idea why this popped up after the 16.0 update and (more importantly), how to disable it?
I used the standard ESP32 build (basic one, nothing special, just **WLED_16.0.0_ESP32.bin)
**
Appreciate any pointers to “getting back to normal” 
Gerry
ADA light is a serial protocol. anything sending serial data to your controller?
I read that (USB/serial stuff) in a few places and the answer is a solid “nope”.
This is a DigiUNO unit (which doesn’t have any explicit serial support) and is attached only to a 2 wire 5VDC power supply (no USB, no data lines). As for hookups/pin-outs, there is only a 2-wire 5VDC +/- going in and 3 wires out (5VDC +/- and Data). Nothing else (and physically, nothing remotely close to it - it’s a sealed up light fixture).
And it has been running earlier versions of WLED for several years without anything like ADALight visible/enabled or affecting stuff.
I think I have to down-rev to 15.4 again. This is causing the light to blink off and seemingly “restart” whatever ADLIGHT is trying to do every 10-30 seconds (I tried to use the “Override Once” button ,but it does not “stick” and after the light blinks out (uptime does NOT reset, so it’s not rebooting), it seems to “come back”.
The “off/restart light cycle” thing related to this is causing a lot of complaints from the household
it may be receiving random noise, AFAIK there is a pull-down enabled on serial pins to mitigate that but hard to say what is causing it, I have never seen it in any of my devices.
@coderabbitai please check if there is any reason for the serial input to receive noise, i.e. check if the pull-downs are actually enabled correctly and if there is any obvious reasons for the described issue
edit: lol, thought I was on github
Well, it’s something in the 16+ code that is a change. I tried the 17-dev nightly and had the same problems.
I down-reved to the 15.4 again and it’s working 100% fine and correct. So something only 16.0+ seems to be responding too.
The unit is physically pretty sealed up - opening it to get in and try to do something physical like add pull-down isn’t something I can do anytime soon (it’s an afternoon to unmount, tear apart, dig in, etc).
It is a bog-standard DigiUNO and I have a number of these running in the area.
Even if there is some physical interference (which I agree, seems likely), was there any changes in 16.x code base that would start “acting” on that noise?
Like some serial port monitoring being enabled that wasn’t in the 15.x series?
I THINK I can work around this by adding in an extra LED output that I don’t use to force pin 3 (which I think is the default serial-in pin) to be a 1-LED output. I did this with the 17.0-dev and so far, it seems to work.
I do think there is likely some interference and I think the version of the DigiUNO I’m using doesn’t have a pull-down on the unused pin, so this forces the pin to output mode and “bypasses” the whole thing.
Might be nice to have a more formal way to do this - either a checkbox to disable serial or a checkbox to disable the adalight stuff for cases where existing hardware is now not entirely compatible, at least “as is”.
can you check if the bin from this PR is any different?
Installed and running. Removed my work-around “fake” LED output (so nothing using pin 3 again).
It has been running for 15 minutes now on this hardware (which was the first I saw the problem on) without a problem. Previously, within 30 seconds, I’d see a “reset” and a refresh of the UI would show the ADALIght popup.
I’ll let it run for the next few hours (it’s right near me, so easy to keep an eye on), but first blush is showing success!
Thanks!!!
Gerry
So about an hour in now and not seeing any of the serial port influence I was seeing before with 16.0. That is, I’m not seeing random effect restarts regularly or popup/warnings about ADALight.
I AM seeing a noticeable “hard reset” on an effect pretty much every 20 minutes. I’m using the “Random Colors” effect with the sliders set to speed 130 and fade 255 and using the “Rainbow” pallete. The smooth color changes made the previous problem very evident.
The effect works exactly as expected, though at about 20 minutes (cannot be sure it’s exact that, but within a minute of that), the color fade abruptly stops and “jumps” to red (which I think is the starting color). Its a hard jump.
That may be what the effect does or something else. I can’t say I’ve used this effect long enough before to note if this is just “normal”, but in case it is anyway related to this this (serial buffer gets junk and finally “gives up” after 20 minutes - entirely unfounded speculation/thought ;-)).
The unit is NOT rebooting - uptime does not reset in the Info.
So just reporting this in case it’s related.
Gerry
Well, unfortunately, while it didn’t happen right away, the ADLight popup and weird effect restarting is back. I suspect there is electrical interference and the change might have made it less susceptible, but it still is there.
Sorry for the bad news.
