LED Strip - Raspberry pi - Odd Glitching

Forums › Gadgets, robotics, home automation, electronics (including wearables) › LED Strip - Raspberry pi - Odd Glitching - Help?

pomtom44

128 posts

Master Geek

#250715 22-May-2019 15:36

Hi all
Sorry if this isnt the correct section to post, admin please move if needed

I have two strips of ws2811 leds (at least thats what I believe they are, got off ali express so not 100% sure haha)
And both are slightly different (although i think its just different coloring of the ribon part)

Anyway
Im trying to power them from a raspberry pi
During testing of a short 5 segment stretch, i had zero problems, so I went to install the full length where I had planned (Outside on the stairs)
its about 2x 5m strips, broken into sections for the stair poles
(if that all makes sense)

I installed it and then was having some odd issues, so I went back to testing and found that after about 18 segments I started to have odd glitches.
Doing some googling, i found that the led strip was expecting a 5v digital source, and the raspberry pi was outputting 3.3
So I got a level shifter and that seemed to fix the issues
However when I plugged in the next segment I found more glitching.
It appears to be when I plug in a second segment that I have issues, and im not sure why

Im going to do some more testing and will post more findings on here as they come out,
Just looking for some more help and input from others.

hopefully to find something that I hadn;t thought about, or didnt know about.

Thanks in advance.

pomtom44

128 posts

Master Geek

#2243185 22-May-2019 15:49

Just realised I used the term segment for two different meanings

My led strip is 3 led's per "segment" (RGB values applied to 3 led's at each time) Ill call it just a single LED for now, as it pretty much is one

My stairs are broken into 2 hand rails
Lower and Upper
Im only working on lower at the moment for testing
That hand rail is broken into 3 sections for the downward poles
Each section has a row of about 80 led's with a combining plug going through the poll to the next section of leds.

When I mention testing 18 segments, im talking about individual leds in a single section
and when I talk about plugging in the next segment, i mean plugging in the joiners between two strings of 80 ish leds.

geocom

594 posts

Ultimate Geek

Subscriber

#2243187 22-May-2019 15:51

Do you have the 2 5M segments joined together or are they on different pins on the micro controller.

Realistically 5M is the maximum that you can really use in a continuous strip the voltage drop is too high after that.

Also you will need quite a beefy power supply and the path should be split between the 2 strips not one through the other

Geoff E

frankv

5678 posts

Uber Geek

Lifetime subscriber

#2243190 22-May-2019 15:54

+1 for figuring out the level-shifting issue. But that's not the problem this time.

RPi is not hard real time... timing will vary depending on interrupts and other processes. WS2811 *is* dependent on accurate timing.

Your short strip meant that (probably) every time the processing to generate the pulses out to it all fit within one time slice. The length of the pulse stream is directly proportional to the number of LEDs on the string. So the long pulse stream gets a *long* all-on or all-off section added in the middle, and all the subsequent bits would all be out of sync.

You could use an Arduino or ESP32 or similar to do the actual pulse generation, or get a Beaglebone Black which incorporates (more or less) an Arduino onboard, or break up your strings into sections of 5 LEDs, or Google "Raspberry Pi CNC" to find hardware solutions, or LinuxCNC for a software solution, or do some hard real-time interrupt service writing on the RPi.

skewt

750 posts

Ultimate Geek

#2243191 22-May-2019 15:58

You probably can’t draw enough amps for the strip from a raspberry pi directly

A 1m coloured LED strip I worked with using an ardiuno recently needed 5v 2amps

pomtom44

128 posts

Master Geek

#2243209 22-May-2019 16:18

for the voltage im putting 12v from a 5amp power supply, so I dont think thats the issue
The length im only running the first strip which is about 1.5M ish and when i plug in the second strip is when I have the problems 1.5M ish again

for the strip length, i have seen longer strips then mine running online. so i doubt length is the issue as im only running about 2-3 meters for this testing

and for the timing issues, I found things online talking about that, but I found something online which says if I use pin 18 I can use the adafriut neopixel library which deals with the timing issues and allows for the led strip to be run from the pi in decent lengths
Unless what I found was wrong?

the other thing I have thought about.
How long between the Pi and the first led is acceptable, as my cable may be too long for the signal to be sent and may have degraded before it reaches the first pixel

geocom

594 posts

Ultimate Geek

Subscriber

#2243220 22-May-2019 16:45

pomtom44:

How long between the Pi and the first led is acceptable, as my cable may be too long for the signal to be sent and may have degraded before it reaches the first pixel

How long is the cable? realistically you would want it to be as short as possible.

Geoff E

pomtom44

128 posts

Master Geek

#2245192 25-May-2019 10:07

ok so tested on my workbench.

works fine
So my only guess is my run from the pi to the first led is too long
unforntatly, its a run that I can't shorten as im using exsiting wiring and my pi is indoors, the lights are outdoors.
im trying to guess the run length, and i think its about 8-12 M from pi to first led.

Is there a signal booster I can put either before the pi or before the first led to boost the signal?

Tradepub

Free professional, reference and technical white papers (affiliate link).

pomtom44

128 posts

Master Geek

#2245193 25-May-2019 10:11

I should also add, I can't easily get to the half way point of the wire, as its buried under the garden.
so id prefer something at the start or end of the run