WS2811 Not Forwarding Data

Other Hardware LEDs and Multiplexing
1

I will go into further details on code if needed. But I have verified all timings on a scope both at extreme limits of spec and at mid point of spec from the datasheet and experienced this problem on ESP32 using the neopixel library as well as on a full custom PIC project.

I have designed a board using the WS2811N chip available for assembly from JLC. I have followed the reference design and best practice for layout.

I send a string of bits for 2+ devices (that work on a WS2812 strip). But when I send them to the WS2811N, the 3 LED elements light up as expected but no further data is forwarded on.

Has anyone have experience on this specific chip and either had it work, or experienced a similar issue?

Cheers,
Rhodeworx

2

First of all have you increased the logic level of the data line from the 3.3v (ESP32) to 5v (ws2811) ?

Did you make a prototype before you ordered the manufacture ?
Can you post the circuit design and the gerber files.

I have used the standard ws2811 successfully, the 'N' is only different because of the it's package i believe.

A simple example code should work normally speaking. Can you get the first IC to update successfully after the initial data transfer ?

3

Admittedly, from the ESP, I didn't boost it, though I did run it off 3.3v under the presumption that it would behave like a WS2812. I will look at using a WS2812 as a buffer.

The N seems to differ also, in that there is a 9th pad (Presumed ground in my layout) and that the speed pin present on DIP and SOIC is N/C on this package.

I can write multiple different colours to the chip successively though haven't tried analogue values, only 0x00 or 0xFF and combinations there of.

No prototypes, for the price of the tiny board on an existing order, I just tagged 5 off on.

Drive code for Arduino:

Adafruit_NeoPixel strip = Adafruit_NeoPixel(12, LED, NEO_RGB + NEO_KHZ800);
 strip.begin();
strip.fill(strip.Color(0,255,0));
strip.show();

Code for PIC18F27K40 @ 64MHz:

#define send1 DATA=1; NOP(); NOP(); NOP(); NOP(); NOP(); NOP(); NOP(); NOP(); NOP(); NOP(); NOP(); DATA=0;
#define send0 DATA=1; NOP(); NOP(); NOP(); NOP(); DATA=0;

typebool PixelArrayUpdate()
{
    for(uint8_t i=0;i<PIXEL_ARRAY_SIZE;i++)
    {        
        if (PixelArray[i] & 0b10000000) { send1; } else { send0; }
        if (PixelArray[i] & 0b01000000) { send1; } else { send0; }
        if (PixelArray[i] & 0b00100000) { send1; } else { send0; }
        if (PixelArray[i] & 0b00010000) { send1; } else { send0; }
        if (PixelArray[i] & 0b00001000) { send1; } else { send0; }
        if (PixelArray[i] & 0b00000100) { send1; } else { send0; }
        if (PixelArray[i] & 0b00000010) { send1; } else { send0; }
        if (PixelArray[i] & 0b00000001) { send1; } else { send0; }
    }
    return True;
}

image
image543×600 26.3 KB

image
image517×518 4.84 KB

4

The 100 Ohm resistor doesn't belong in the Vdd line. A resistor in series with the data is usually recommended.

You shouldn't need a level shifter if the 1st LED/chip works because the data is buffered when it's passed-through.

It might be worthwhile to try a regular LED strip to confirm if the problem is with your custom board, software, or something else.

5

The best solution is to power the ws2811 with 5v (with the 100R resistor you are using that is. 12v is also an option but then that value needs to increase, and you can power 3 leds in series) and shift the 3.3v logic level from the ESP32 to 5v, using a 74HCTxx powered with 5v. I use a 74HCT04 and run the signal through 2 of the 6 gates, so i invert it twice. That way you make sure that the logic level is correct. A ws2812 can be run with 3.3v, (It is outside of the datasheet spec though) but Blue will not light up as brightly as Red & Green. Using it is a buffer is also pointless, since if powered with 5v, the 3.3v logic input won't suffice, and if powered with 3.3v, the output won't be the required 5v.
A ws2811 differs from a ws2812 significantly with it's external current limiting, and i suspect that the part of the circuit that passes the data signal on is affected by the lack of current provided. Hey that may be another Hack, less resistance on R1, but it is not the way to go.

Yes it does ! read the datasheet. For 12v use the resistor has a bigger value.

All in all add a TTL 5v logic chip and power the thing with 5v (both the PCB & the TTL) and get back to us

I have never even used the speed pin, ever !

Just out of curiosity, what is the point of this board ? It is not the complete board that you've shown us, is it.

6

Have you managed to get this working?

7

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.