Hi! I am having a little problem, unrelated from FastLed,

Have u tried just pressing on the strip between the working section and the next failing one?

yes. nothing positive

so, update from the day after:

I pass over all the “junctures” from a system comprised by 80 pixels / 240 leds, with a reinforcement of tin.

The problem persists. even though is not as bad as the white leds, there are still some coloured pixels lost here and there. Mostly, when I first turn on the system.

my conclusion so far is that it is a crappy quality strip.

Just a little thing here, I would not solder the strip while it have power to it. When I need to turn on, I apply the power to the strip first and then Arduino. When I turn thing off, remove the Arduino power first and then the strip

ok, and why is that?

In my system, my power supply feeds both the arduino and the strip, so when I turn it on, both receive power at the same time

When power on the strip sucking a lot of power,that bring voltage drop on Arduino, in your case a drop 12v to Arduino is OK, but for 5v strip it does make a difference .

Sorry to hear that… I would really like to help you with this as I know that would drive me crazy !!!

To the best of my understanding, you have a bad contact problem somewhere and you have to find it or junk them strips. The fact that it eventually works by itself for some time would point that way.

Assuming you did a good joint soldering job, where else could you have a bad contact ?

Maybe check the Surface mount LEDs themselves… if a pin (or pins) is somewhat ‘floating’ on its pad that would kind of explain the weirdness.
Can you take a real close-up picture of a section of your strip and post-it?

thanks JP!

It is really crazy-driving behaviour…

Unfortunately It is too late for me to junk them. I am preparing an installation for next week and I only have this strips. No time to get new strips.

I am quite certain to have done a good soldering since it is not the first time I do exactly this job. The only thing that changed from previous experiences, is the supplier to whom I bought the strips…

I also thought and did check pins and components. the problem is that I have 8mts of led strip to check. And dont forget that the problem is not a specific part of the strip, but the whole strip!!! A strong indication of this is the fact that, as explained before in the case of the first pixel working and the rest of the strip on white, is that if I discard this first pixel, the problem is now the new first pixel. and so on…

I think I will have to back up with the idea of some lost pixels here and there, and change my supplier for the future. And the lesson is to order the material with time, and try it out with time to check for possible failures and malfunctions

I may try an other power supply and shorter strip section.

I agree with all trouble shooting ideas @JP_Roy ​​​ has suggested. Here’s a few more things to try of you haven’t yet.
(Sorry pressed send by accident to soon… phone typing…)
Place a 300 to 500 ohm resistor on the data line (at the LED strip end of the data line).
Put a 1000 uF (16V) capacitor across the + - at the begining of the strip.
Also, just want to double check that you have the ground of the LED strip and arduino connected together.

thanks marc!

GND is properly connected

1000 uF (16V) capacitor I do not have to try…

the resistor I should place it between the end of the data line and what?

thanks again

Place it at the end of the data line soldered to the Data In connection on the strip.

ah ok. so between the arduino and the strip. This is something I was doing with the ws2812b, that I also lernt from (http://mrossphoto.com/wordpress32/neopixel-ws2812-dlw/)

nope, it doesnt sort out the problem

Examples (green wire is data):
https://drive.google.com/file/d/0B1DBHDKW1M8uZ0c3OHlDRUhVRnM/edit?usp=docslist_api
and
https://drive.google.com/file/d/0B1DBHDKW1M8uNk9zSGZ1R2xMcGc/edit?usp=docslist_api

and what would the capacitor do?

I have 10uF 50V

The capacitor helps smooth out the power supply to the LEDs as they use more or less current while turning on and off (sometimes extremely rapidly based on pixel animations).
Skip using the 10uF and find a 1000uF when you can.

thank you very much again marc!

By the way!!! (sometimes the greatest ideas come very late): I have found another ws2811 like the ones I am using and fail, but from another provider. They work properly…

for me it was quite clear alreay for some time that the problem was the strip. now remains the question; what is it that provoques this failure?

@JP_Roy
@Luminous_Elements
@marmil
@Nail_ENVY

a new day´s update:

I was thinking a lot on this problem, and I arrived to the conclusion that the problem exists in the whole led strip, because as previously exposed, it always happens that the first pixel works correctly, and all the following pixels fail, and if you cut this first pixel and connect the arduino to the second pixel, which will be now the new first pixel, the same happens; this new first pixel works correctly, and the rest of the led strip works.

As I found another ws2811 led strip that works ok, what I did was to use one pixel from the led strip that works correctly as the first pixel, and the rest of the les strip from the led strip that doesnt work correctly. The result: PROBLEM SOLVED.

My conclusion is then that there is a function that the first pixel of a led strips perform, and the rest do not. My question is then: What is this function?

Wow that’s good detective work !!!

yes. I am actually quite proud of myself :smiley:

It took me several days though.

I still want to understand the whole thing!!!