WLED Version:
Tried on Versions 0.14.2-b1, 0.14.1 and 0.14.0
WLED Configuration:
SK6812/WS2814 RGBW
Color Order: BRG
Swap: W&G
Length 16
Auto calculate white channel from RGB: Dual
When the color is set to White it Flickers different random colors (Green color mostly) usually towards one end of the LED Strip.
When the color is set to Orange (Default boot up color) then the Flickering is less.
Tried the same setup with D1 Mini and it works like a charm.
Looks like some issue with respect to refresh OR white channel on S2 Mini.
Please let me know if any further details are required to fix this issue.
use a levelshifter, to solve this problem. I know exactly the same with a WEMOS D1 mini.
Or second way for longer distances RS485 busdriver. A 3485 (3,3V type) for controller and a 5V type near stripe (will need 5V) i use for this 7805 compartible step down switcher from 24V rail https://wled.discourse.group/t/once-more-esp8266-based-wled-controller/11130
RS485 is a good choice for long distances.
For the ESP (Tx) side, a 3.3V version is unnecessary. The standard MAX485 5V driver chip has 2.0V VinHi data inputs that are compatible with ESP GPIOs (same as 74AHCTxxx devices).
If I understand it right, it would not be a level shifter issue because the same strip is working fine with D1 Mini which is even lesser (performance wise) a hardware than S2 Mini.
Also the length of the strip is 1.5m (21W/m) which is connected to a 24v 4A power supply.
No, this a a gamble of fortune. Maybe on ESP do have some mV more Voltage, maybe the resistance of used output is lower… maybe die constellation from stars is better
To be in a safe working area with 100% function you will need a levelshifter - a really fast levelshifter, not lame crap with transistors, like for I2C or serial outputs with 9600 baud
Had the same problem.
In my case level shifter does not help and i would even say that it is not necesssry.
What caused flickering? If you connect 24v ledstrip with any esp/arduino you should connect grounds… But if you do it all noises generated by (usually poor and the cheapest) power supply go to esp. There is no filtering. The noise affects data.
I tested several variations and conclusuon is simple: YOU NEED GOOD POWER SUPPLY. NOT THE CHEAPEST FROM ALIEXPRESS.
What i tested with 24v fcob 2814:
2 different Cheap power supplies with esp/arduino - - - flickering
2 different Cheap power supplies with athom wled controller - - - works, i guess that the controller has some filtering built in
Meanwell xlg-150-24-a power supply with esp/arduino/athom — works correctly
My conclusuons are:
Cheap power supply generates too much noise, transmitted via cable or even air
bare esp/arduino do not have anything to filtr out the noises
athom is p&p controller which works on 5/12/24v has implemented some solution for this problem
I am not expert and did everything experimentaly, but i am prettt sure about results. Maybe there is any other solution to work with cheap PS. Meanwell is 3or 4 times more expensive?
For people from Gernany, Pollin do have a offer for MEAN WELL, LRS-150-24-DDP, 24V-/6,25A, 150W for 6,95€ up to 4 - 5 to 9 5,95 and more than 10 4,95 https://www.pollin.de/p/schaltnetzteil-mean-well-lrs-150-24-ddp-24v-6-25a-150w-353178
You can also try to get a cheap, used PSU like use for industrial PLC
I have 24V 10A used PSU at work, with more than a dozend years “in first life”
When I found out that power supply is the reason, I wanted to buy something I am sure about, so picked more expensive option buy with active PFC filtr( do not know if it is necessary). I saw PS mentioned by you but did not want to risk.
However great to know that the cheaper mean well works for you!
But I agree with you, that most cheap testboards are more sensible for instabilitys. This boads must be small and cheap.
For first tests it’s ok and you will have fast a succsessfull result. But for a stable solution a exta PCB the better way - if it’s made right
