Currently have 5 m WS2815 strips installed running from 18 AWG wiring connected to 12 VDC Meanwell power supplies and our controllers (SP108,SP102,SP608), power injection on all lines. When we tested with shorter cable, everything worked fines with the LED color and lighting programs, however when I moved the controllers to the main boxes (this is for a pool so 100-200 feet on 18 AWG lines) and the LED strips now only partially light. They will turn on but will not be the correct colors or programs or be controllable, and once we added the amplifiers we bought:

enter image description here

All the lights turn on but flash quickly, will not select the right color, or the correct program will run. I have also tried the basic, 33 or 470 ohm resistor on the data lines but that did not help (per some on line recommendations.

I have also tried another amplifier made for long line transmission with no luck :

enter image description here

Any help with my data line issues to solve the LED control would be great. Power is not the problem, I can move the controllers and run power only, however wanting to keep the controllers out of the wet boxes that are in the pool as not sure once they are all closed up I will get a decent wifi/Bluetooth signal and we didn’t design the system to integrate that way initially.

So an update - the data lines are losing one volt over the distance based on the resistance - since the data lines start off at 1.5 volts from the led controller need some way to increase the voltage for the data

lines. enter image description here

With the resistors and the amp before the led Wes re still getting rapid colors flashing and unable to control from the 608 or 602 controllers ?

UPDATE : we have been focusing on using btf lighting drivers (sp602e, 102e, 110e, 107e. Etc) for the LEDs so we have the ability to control them wirelessly but that combined with the long line issue, are there or is there better LED drivers to use for something like this setup ?

  • \$\begingroup\$ start by moving the initial successful setup to the pool area ... that will eliminate the location as being the cause \$\endgroup\$
    – jsotola
    Commented Aug 29, 2021 at 17:39
  • \$\begingroup\$ We did that - all worked fine - now we want to move the controllers close to the lead of the long line runs (start of the 100-150 foot wires). Once we do that the LED data signal and resultant control goes Haywire and the LED strips flicker, blink and won’t run the correct colors again. Unfortunately I no longer have an oscilloscope to troubleshoot the data lines. I can’t keep the controllers in the pool as they are not waterproof – \$\endgroup\$
    – LGTZ73
    Commented Aug 29, 2021 at 18:04
  • \$\begingroup\$ Output of controller is not good for long distance. It is just TTL signal. For long runs you need something like current loop. Transmitter and receiver should be used. May be better solution make the water proof enclosure. \$\endgroup\$
    – user263983
    Commented Aug 30, 2021 at 0:50
  • \$\begingroup\$ What would be the best controller the. To use or would it benefit to set up a current loop ? I have 7 different lines to drive with different controllers and boxes - which we ran the lines to a centralized location to ‘make this easier’. Ive had no issues with the 24v RGB LEDs, only the WS2815 strips. \$\endgroup\$
    – LGTZ73
    Commented Aug 30, 2021 at 1:43

2 Answers 2


Add 25 to 100 Ohms in series to each data + CLK signals on board. (cut and jumper with R)

I've only proven this works by simulation using LED strip impedance and CMOS driver RdsOn.

The ringing is causing glitches when combined with CM (common mode) line noise when stretched out like an antenna. A suitable CM clamp would help as well as earth grounding suppress. Since the signals are not differential, they can pick up a bit of this stray line noise. Putting your hand around a length of strip and better with bare feet to couple yourself to RF ground with some nF value. If this improves the LED communication significantly , a series a cap to coax shield to the soil may work.

The 1V drop is not too critical as the logic works at TTL levels and the current is regulated for RGB. ( I forget the chip specs). But you may need to feed DC power to the middle of the string or both ends.

  • \$\begingroup\$ Would swapping to shielded or cat cable on the data lines only be better for the long runs then ? Assuming the the resistor series would be at the LED front side or would you recommend putting them right after the controller ? The CM clamp is a interesting thought - I could ground to the other ground lines like the main LED lights I installed. Not sure why there isn’t a true ground on these lines with the length of runs. What about still having the amp installed or swapping to a logic level shifter in series with the resistor ? \$\endgroup\$
    – LGTZ73
    Commented Aug 29, 2021 at 21:02
  • \$\begingroup\$ Resistor must be next to driver to dampen CAT3 cable higher impedance or CAT5 higher capacitance and impedance. If only slightly better, also adding a load R equal to 2x source for 5V to will work with 1.5V threshold. (If I recall correctly) then you can go even further. But verify supply ripple is low otherwise crosstalk so add 0.1uF //10uF (low ESR ) at driver . Keep power cable away from data cable in case of crosstalk from ripple by a few inches and CAT5 for Clk and data. \$\endgroup\$ Commented Sep 1, 2021 at 19:51
  • \$\begingroup\$ So adding the resistor to the driver line or across the cat twisted pairs (line a and b) at the source ? Same with the capacitor - is that the supply for the led or the ? \$\endgroup\$
    – LGTZ73
    Commented Sep 2, 2021 at 0:47
  • \$\begingroup\$ Resistor NOT across, BUT in series with Driver , cut and add SMD R or it none short resistor, Cap is for DC but may not help much \$\endgroup\$ Commented Sep 2, 2021 at 1:25
  • \$\begingroup\$ I did this based on my calculated load for just the simple test circuit I have rigged up with the one string of WS2815 lights, which the way we have the actual ones - the load won’t exceed about .9 amps on one circuit at any given time. When I added the resistor to the driver line, the LEDs have a flickering on some lights in both an off and on state. Same when adding the capacitor to the cap to the DC line. When removed, the flickering goes away. What am I doing wrong ? \$\endgroup\$
    – LGTZ73
    Commented Sep 2, 2021 at 18:05

These chips and associated controllers are commonly built for the three wire "neopixel" which have strict 800kHz control requirements. Your LEDs have a separate clock line (so-called "DotStar" libraries) so it is possible to use a much slower clock speed. A slower clock will allow a longer range without reflections / ringing and associated attenuation of signal or dirty signals. If you are trying to do this with purchased controllers because you don't want to get into the programming, then you'll have to look for one with the slower clock than your existing controller.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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