0
\$\begingroup\$

I need advise on the following problem: I am connecting a 5 metres WS2812B based RGB-LED strip to an extension cord. This is a 12V version, where two LEDs are coupled in terms of their addressing and IC. There are 100 addresses and 200 LEDs. I am generating the control signal from a Raspberry Pi.

  • I connect the ground of the Pi's GPIO and the ground of the external 12V power supply (a strong 150W laptop power adaptor)
  • I connect GPIO 18 out (3.3V) to the DI of the LED strip
  • I connect the 12V to the LED strip
  • This works

Now I want to insert an extension between power supply and Pi on one end, and the LED strip on the other end. I bought a 3x 0.75mm 20 metres cable for that. (Here is the product description in German if you are curious).

  • I inserted two wires between the 12V power supply and the GND/12V pins of the LED strip
  • I inserted the third wire between GPIO out and the strip's Digital In.
  • This doesn't work

Simply powering the strip seems fine; all LEDs go on. If I run the modulating program on the Pi, the first three or four LEDs are erratically flickering, that's it. Now I tried all sorts of combinations to figure out what is going on. For example,

  • I inserted a LLC to lift the digital output signal from 3.3V to 5V. Makes no difference. Does not work.
  • I extended only the power but connected the digital signal directly without extension. Makes no difference. Does not work.
  • I extended only the digital connection, leaving the power supply directly connected to the strip. This works.

This result was very surprising to me. I kind of feel relieved that I won't have trouble extending the digital signal, but now I'm baffled why the power extension apparently doesn't work. Using a volt metre, I saw that over the 20 m, the voltage drops from 12V to 9V. However, compensating by setting a higher voltage at the adapter, e.g. 15 to 18V to measure around 12V on the other end, did not change the situation.

My question is: What is the reason for this problem, and what should I do to properly extend the power connection by 20 metres? Is this simply a fault of the diametre, or am I looking at the wrong type of cable? Because even a standard 240V euro cable only has 0.75mm diametre. (In German this type of cable is called H03VV-F). The cable needs to work in outside space.

\$\endgroup\$
  • \$\begingroup\$ I suspect signal margin is reduced due to ground shift from shared ground current. They need to be separated signal returns. (0V) \$\endgroup\$ – Sunnyskyguy EE75 Aug 28 '18 at 20:25
  • \$\begingroup\$ @TonyEErocketscientist Could you explain this in more detail to me? I'm not an expert electronics person, so I don't really understand what "need to be separated signal returns" means, or how I would accomplish this. Thanks \$\endgroup\$ – 0__ Aug 28 '18 at 20:28
1
\$\begingroup\$

The voltage drop over 20 meters of cable will vary based on the load, which you will no doubt be varying based on the number and color of leds you turn on. The change in voltage will cause issues with the led controllers. Use a local regulator, or capacitors, or both. Or use standard AC extensions and keep the power supply local to the strip.

The diameter of the cable will affect it, because a smaller diameter wire has a higher resistance and a higher resistance means a bigger voltage drop as the current increases.

\$\endgroup\$
  • \$\begingroup\$ I cannot keep the power supply local to the strip because it must be inside a building, whereas the strip is mounted outside and at considerable distance (20m is like my worst case scenario). So given this constraint, do you think simply using a larger diametre may solve it? Would 1.5mm work, or do I need to jump already to 2.5mm? \$\endgroup\$ – 0__ Aug 28 '18 at 21:02
  • \$\begingroup\$ Use an outside rated power supply, or an outside rated enclosure. You could increase the diameter of the wire but copper is expensive. Use an online voltage drop calculator to figure out what gauge wire you would need. \$\endgroup\$ – Passerby Aug 28 '18 at 21:06
  • \$\begingroup\$ Thank you. Why do you think the problem persisted as I changed the source voltage to 17.5V and measured 12V at the strip? \$\endgroup\$ – 0__ Aug 28 '18 at 21:18
  • \$\begingroup\$ (I could verify the observed drop of around 3V using an online calculator. I didn't know that this is such a problem for DC, because if I put AC 230V into the calculator, the drop is negligable at 0.75mm diam.) \$\endgroup\$ – 0__ Aug 28 '18 at 21:24
  • \$\begingroup\$ Because higher voltage and lower current. 0.5A through a 3 ohm wire is only 1.5V drop while 9.5A through a 3 ohm wire is a 28.5 volt drop. Both are 115 Watts (230V * 0.5A vs 12V * 9.5A). \$\endgroup\$ – Passerby Aug 28 '18 at 22:14
1
\$\begingroup\$

Just as a follow-up: Before ordering new cable, I tried a hackish solution with putting together 2 and 2 wires of a 1.5mm speaker cable at 27m. This just works, you can see occasional flicker in a few LEDs, but all addressing is correct. The voltage at the beginning of the strip is like 11.3V. So that confirms that the diametre of the cable was probably the problem, and I will now go for 2.5mm H07RN-F cable connected to both ends of the strip, and hopefully that preserves the current "just working" situation.

\$\endgroup\$

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.