Christmas light led strips

Hello all. I have been trying to get my christmas lights running for 2 years now and cant figure it out. Anyone out there can help would be awsome.

I don’t know if its to do with the length of the wires being too long, All the connectors are good. Voltage is around 4.80v at the end of the strips. The first strip runs perfect. Soon as it gets to the second strip it lights up around 10 white LED’s. I changed the wiring to see if it was the strip that was bad. But runs fine.

I have attached a picture of the wires


19ft of wire, that is 6m, at what capacitance p/m ? for the datawire, the thinner the wire the better and preferably solid (not multi-strand) the problem is capacitance, metal on the surface where it meets the air starts to act as a capacitor (equally as a function to the surface of the metal) and this 'saws' the 'square' digital signal out. do you have any 7400 series TTL chips lying around ? The WS2811/2 chips depend on a good square digital signal with a logic 'high' close to 5v any 7400 TTL chip provides that and only requires a 2v logic high. place the chip near the beginning of the second strip, should work just fine. Don't invert the signal or invert it twice. Also 'inject' 5v power into it with a thicker cable near the start of the second strip this may help a bit, 12v ledstrip suffers less from voltage drop since the voltage for the WS2812 chips themselves never drops below 5v, (discoloration occurs below 11v) the ultimate way to cover distance with ws2812 signal is to use max485 protocol, this way you can run separate power supplies and theoretically your cable could be 1000ft long.

O-o i looked at you picture again, you should 'bridge' those power supplies, and put them next to each other (which will make them appear to be only one) and connect power & ground between the last led of the first strip and first of the second strip. Or you should separate the system fully with 2 max485 chips and twisted pair cables. this is the 12v schematic, but the 5v is the same just without the 7805 regulators WS2812via twisted pair.JPG
Capacitors should actually be 1uF just to be on the safe side

WS2812via twisted pair.JPG

I don't know if its to do with the length of the wires being too long

Well... That's easy enough to troubleshoot...

Also, try swapping the strips around. If the last ship in the 1st string isn't putting-out a good signal, everything after it will be bad.

I think what @Deva_Rishi is saying is, if your drawing is accurate, then you must common the ground of both the supplies.

Op’s drawing.
Needs GND to GND on first strip and second strip.


It needs that at the very least...

It needs that at the very least…

It certainly does not need the bridge between the 5V outputs that I thing you suggested in your last posting. At least not as long as we don’t know what power supplies are used. Usually it is a very, very bad idea.

Using 2 power supplies on a single ws2812 signal is usually causing problems, the best is to put the twisted pair signal cable with 2 max485 chips, this always works in multi power supply situations. if you do not want to make one of those, then to have 2 supplies act as one is your best bet ! I know it's not always a good idea, even if it is just for unplugging 1 and leaving the other one on.. or for unwanted switching (i recon they are switching PS) of them, but powering the first strip with one and the second strip with the other is not a reliable solution. Maybe the plan is connect the GND the strips, if that works great !, if not add a TTL chip, if that doesn't work, put the max485 thing. Please understand i work with LED-strip in an environment where something that is not working on the day it should is just as bad as breaking something. It doesn't work, therefore it's broken ! The party is on today, who cares if you an fix it by tomorrow !

Connecting two switching power supplies in parallel will most probably fuck up their switching, possibly first overload one and then, when the other one suddenly is the only one, the other one too.
On the other hand, I absolutely do not see any reason why powering two strips from too power supplies should be an issue. Best I can think of is that DO of the last LED on one power supply has a lower voltage, because VDD is lower there, but according to the datasheet, V_IH = 0.7*VDD, so that should be fine if the supply voltage does not drop too much.

I think before we speculate more @shawno needs to provide at least the following:

  1. is the provided wiring diagram accurate with regard to ground and power distribution? Note the 3rd strip has no power at all.

  2. What is the goal of the 470 Ohm resistor in the control line?


470 Ω to protect the input of the string.

470 Ω to protect the input of the string.

the tutorial i got was that it was to protect the output of the arduino from 'bounce' (or any accidental short of course. I am actually most puzzled by what the 15A fuse is supposed to do ? What the problem tends to be with 2 powersupplies is in my opinion, that there tends to be. And using 1 big one fixes it !

The tutorial I got was that it was to protect the output of the Arduino from ‘bounce’ (or any accidental short of course).

No, the Arduino is robust. It is to protect the WS2812s.

There are many problems with the original diagram, but the single most glaring is the suggestion that a signal wire could travel anywhere separate from its corresponding ground wire in a single cable.

For LED strips of this length and current consumption, there needs to be a heavy power cable accompanying each strip in order to feed power to the “distant” end and possibly a few tapping points along the strip. Obviously unless you can get most of the LEDs to light, it will not be possible to assess the voltage drop.

The best reason not to use an obsolete 7400 series gate as a buffer, is that its VOH is characterised at only 2.4 V minimum, 3.4 V typical which is clearly quite unsuitable for reliably driving CMOS. Yes, it may work if you are lucky. :roll_eyes:

While paralleling two regulated power supplies is not entirely advisable, if they are a stable design there is no reason they would “fight” and if connected to opposite ends of the strings, the intervening resistance will ballast them nicely.