LED control circuit WS2811, PC817, P-Mosfet - malfunction at 11.5V

I made a circuit based on WS2811 , PC817 optocoupler and P-Mosfet to control more leds. The schematic is for one 14led light element controlled from an ESP8266 based controller. All elements (10pcs) are connected with a 3-wire 3x1.5mm2 cable every 50cm. Individual elements flash correctly when connected close to the ESP8266 controller. When connected with a 3x15mm2 wire (approx. 5m) to all lighting elements and a power supply of approx. 14V is applied, it lights up relatively well. However, when reducing the voltage to 11.5V, the Christmas tree starts and all the elements light up as they wish. I have done tests on a shielded cable - it is even worse. On the DO wire from the WS2811 there is a 2kOhm resistor, because only at this resistance the other light elements light up. What could be wrong ?
.

Hello,
as I understand it, the ESP8266 controller is about 5m away from the WS2811 and it doesn't work?
Everything is fine, strange as if it works, the basics of transmitting digital signals bow,
with curiosity, if you have an oscilloscope, see what you have at the beginning of the cable and after 5m.
Greetings

You have the Q1 control strangely implemented: 0 logical is ground through a 20om resistor and 1 logical is +12 through 220k - strange that it works anyway. Look for an integrated driver or "on foot"
Or vice versa....

>>20732981
Works at 5m but at 14V, at lower voltages below 12V it goes crazy.

Hi,
Because you are close to the margin of error,
Can you explain this "interesting" way of powering this WS2811 chip? He requires a 3.5-5.5V power supply and you are powering it with current, admittedly such a "fabrication" is in the datasheet?
Greetings

https://www.tme.eu/Document/ed561c5587a172a1d73c47508d3f27db/WS2811.pdf here is the datasheet . I have this chip in the 8Dip version and there the power supply of the chip is via 2k7 from 12V. In practice the circuit supply is in the order of 5.8-5.9V

Overdrive

Directly control N mosfet, +12V > CL2+LED+LED+LED >> D MOSFET, S to ground

To OUT (WS2811) , resistor (not to exceed approx. 16mA) to +5V or +12V >> 10 ohm resistor to G MOSFET + to GND 10k

For 5V, Logic Level, for 12V should be enough.

It is necessary to reverse the logic because there it negates, so software or hardware


How to be with the optocoupler:
+12V (+5V) > resistor (16mA) >. LED (A) of optocoupler > LED (K) of optocoupler to OUT (WS2811).

Transistor of optocoupler; +12V (+5V) >> Collector >> Emitter >> 10 ohm resistor to G MOSFET + to GND 10k

+12V > CL2+LED+LED+LED >> D MOSFET, S to GND


When 5V, Logic Level, when 12V should be enough.

No negation needed

HD-VIDEO Is this how it should look like ?
.

I have modified the diagram slightly. Everything works properly. The individual circuits are connected to each other by a 3-wire wire every metre. This gives the effect of a light hose. However, after some time (at least about a month), a single circuit in the cascade can fail. The WS2811 always fails. I have only one outB output used in the circuit . Can the other outputs " hang" without potential ?
.

Hello,
I remember there was once a discussion somewhere that the "Chinese" were saving money on capacitors and adding them in the path of the WS supply reduced the risk of damage.
Since you have long wires between the circuits, be aware that they can act as a coil, something can be induced in them or they can generate pins that can cause problems.
Regards

Well there is 22uF and 100nF , do you think that is not enough ?

Hello,
I don't remember exactly, but there was something about adding a resistor of about 33 ohms on the data lines and something on the power supply, I don't use these circuits, just read something once. You'll have to do some searching.
Greetings