0
\$\begingroup\$

I am facing a strange issue while building a simple LED matrix. I am using the SK6812 LEDs and an ESP8266 controller. The LEDS will be controlled over Wifi. My matrix contains about 180 LEDs on a PCB of 25x25cm.

After about 120 flawlessly working LEDS there are strange lightning artefacts, like flickering and wrong LED color. All LEDs and the ESP are connected to the same ground. Also the logic shifter (3.3V to 5V) is working as it should. It converts the 3.3V data form the ESP to a 5V signal for the LEDs.

3.3V to 5V level shifter

Additionally to the level shifting I added one 470 Ohms resistor (R2007) to the data line as suggested on this page.

Long story, short: If the 470 Ohms resistor is removed and the data line is bridged with a 0R Ohms resistor all LEDs are working flawlessly.

  • I am not much of an electronics experts, and asking myself what is the purpose of this resistor? Is it safe to remove it?
  • And, can this be the issue for my problem, or is it only a coincidence that removing it solved my problem?

EDIT:

The LED_DATA line is pretty long, as it travels over the whole board, front and back. My roughly length estimate would be around 13 meter. The track 0.25 mm thick and copper has a thickness of 1 Oz. Based on my calculations the resistance on the data line should be 27 Ohms.

\$\endgroup\$
  • \$\begingroup\$ What does your PCB layout look like? It sounds like its a noise issue with return currents, assuming that the Din and Douts are all very short. \$\endgroup\$ – Aaron Apr 15 at 19:55
  • \$\begingroup\$ The whole data line is about 13 meters, but between two LEDs there is a maximum fo 20 cm data line. The track is 0.25mm thick. It has a clearance of 0.2mm and 0.4mm to the ground copper fills. \$\endgroup\$ – Markus Apr 16 at 7:49
  • \$\begingroup\$ I have experienced the exactly same problem (with WS2818B LED though). The resistor is the problem, and it's not needed. Just remove it. \$\endgroup\$ – Codo Apr 17 at 8:45
0
\$\begingroup\$

As your LED_DATA trace gets longer and more LEDs get added on, the capacitance of LED_DATA will increase. It's likely that in combination with the 470ohm resistor, you're creating a lowpass filter and rounding your LED_DATA signal. As the capacitance increases and the signal gets rounder and rounder, eventually it will stop working. If that's the case, replacing it with a jumper would fix it. Try looking at the LED_DATA signal (with and without the resistor) with an oscilloscope.

\$\endgroup\$
  • 1
    \$\begingroup\$ My understanding of the data sheet is that each led is retransmitting the signal. If it's not then you are right. \$\endgroup\$ – Aaron Apr 16 at 14:06
  • \$\begingroup\$ "plastic forward strengthening technology, the transmission distance between two points over 10M". Yeah, I'd say you're right, this is probably not standard digital output stuff. \$\endgroup\$ – Aaron Apr 16 at 16:00
  • \$\begingroup\$ Unluckily my oscilloscope is not able to capture such fast protocols. If I can find someone with a good oscilloscope I will check the signal data. \$\endgroup\$ – Markus Apr 16 at 22:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.