Random flickering/color change in WS2814 segments

Any ideas on what’s causing these random color changes/flickers in this WS2814 strip? Being power by Digi-Octa Power-5 and Dig-Octa Brainboard. It’s only this one strip. Power fed from one end and it was cut to length and soldered at the peak in this roof line (indoor deck). My thought is it’s just a bad strip (or section of strip). I’ve set the resistor switcher from default 249R to 3RR since the runs from the controller to the strips are long. I’ve had another bad one where 1/2 the strip was doing the same thing - replaced it and it went away.

Quite possibly a bad strip. I’ll say, though, of the…lots and lots of strips I’ve bought only one was bad and that was from damage in shipping and that was only one ic. More often it’s wiring issue or some variety of user error. While some say it’s only a band-aid on a larger issue, you could try adding another capacitor at the beginning of that strip. That’s been a life saver for me. I know the board has one, but it’s a cheap and easy thing to try adding another since you said there is a good distance between the board and the strip (I’m assuming power is running the same length as the data). Can’t hurt and might save you from more laborious troubleshooting (spring for the Panasonic brand capacitors as they advertise a longer usage life)

What size capacitor would you suggest?

I bought and have used the 1000uF 35v rated Panasonic variety with good success. Also just a note that a cap that large does kinda delay the lights powering up. Only a second or so and only if you’re using a relay to totally cut the power to the strip.

Thanks - and this is across the power line, correct?

Yep positive leg to positive and negative to negative easy peasy. My strip isn’t directly visible so I put it right across the pigtails going into the strip (probably 6 inches out). People seem to say the closer to the start of the strip the better. I’m sure there’s some wiggle room though.

Until you actually solve the power supply issue with your setup, you’re only masking a fundamental problem.

The only way to have a reliable LED setup is to provide it with reliable power.

Not always true though. 24v ws2805 x 4m on an esp32 board with a level shifter. Power coming from a lrs-350 and about 5 ft of 16 gauge copper wire to the strip. Strip pigtails are 20 guage copper wire. Separate 5v power supply to the esp32. Grounds are bonded. Works fine when there’s no relay attached. Hook up a relay to wled OR use an automation in home assistant to toggle a smart plug based on weld on/off state - get a single 6 led segment that misbehaves about 75% of the time when power is applied (relay/outlet toggled on). Replaced the 1m containing that segment and no change. Resoldered.all connections no change. Different board and level shifter no change. Different power supply no change. Add capacitor works fine now. Used the same general setup with sk6812 prior to converting over to ws2805 and had no issue. While I do recognize this is a power issue what aside from the board, level shifter, LEDs, power supply,l and wiring would you suggest changing or upgrading to remedy this problem?

Also, why are there caps on quinled boards if they are such a no no? I think that guy has some generally sound thinking when it comes.to led controllers.

This is one of two similar issues that a capacitor has solved over the past couple years. Both were 24v LEDs. I think there is a time and a place for them if they are solving a seemingly unsolvable problem.

The first thing I would look at closely in the setup you describe is the power relay.
At minimum you need a bypass diode across the coil.

2nd test would be to run the strip at a greatly reduced brightness to try and see if there’s an issue with the power supply not being up to snuff for your maximum load.

3rd test is run the whole setup at 100% White and measure voltage drop at power supply to see if it actually can supply worst case power to the setup. Also measure drop at start, middle and end of strip to see if injection may be required.

The use of a strip capacitor is not typically going to hurt your LED setup. The problem with suggesting that you need one is that it just covers up some more basic issues with the setup that can just get worse over time. Very often that’s a power issue and poor power just about guarantees poor LED results.

I understand why Quindor would include those on his board - it adds a small layer of help for people who may not know any better. In the end he wants people to be happy with his boards (which are generally quite good IMHO). I’ve built my share of electronics that “just work” when I kludge it together. But in the long run you’re always better off solving the basic problem.

Thanks for all the help and comments so far. A few comments based on suggestions so far:

Power supply is a Mean Well 24V 500W supply. I have 3 5M 24V WS2814 strips (from BTF Lighting) and if you do the calculation, there should be no issue with current draw. The strips are fed from one end (no middle or end power injection). We’ve done tests on the strip to measure the voltage under full brightness, and don’t see any issues (drop of about 1V).

Sounds like a reasonable power setup overall.
I suggested the live measurement test (especially at the end and middle of the strip) as mfg. are notorious for giving optimistic specs for their strips power needs and current carrying capacity.

Are you using the relay to switch 24V power or mains to the PS?
Mains switching is much easier on the relay contacts, but typically requires the ESP to send “off” data to the strip before you’re sure the PS is fully up and running (could be up to 1s+).

Roger that. I will admit I did not do any metered testing as I assumed 4m of 24v LEDs could be powered by a single injection point. It seemed pretty clear the relay was/is the issue. I bought a ssr but got the capacitors in the mail first so just stuck with that.

I DO appreciate knowing that everything is running correctly… But sometimes at 2am a solution is a solution lol.