Results 1 to 10 of 45

Thread: WS2811 flicker but only at low brightness - Help please?

Threaded View

Previous Post Previous Post   Next Post Next Post
  1. #1
    Join Date
    Jun 2020
    Post Thanks / Like

    Default WS2811 flicker but only at low brightness - Help please?

    Hello all,
    Firstly, thank you all for a fantastic forum. This will be my second year in the UK with pixels and whilst this is my first post, I've used the forum for many resources already. We have expanded for the second year, and this is where the problems started - please help. (And sorry for the long post)

    I've searched for the same problem but can not find the exact problem we have. I have tried many of the fixes for similar problems but nothing has worked and we are at our wits end.

    With some left over 1st year ws2811 5v pixels we made 4 x matrix panels- each 30x33 pixels spaced at 25mm. The string zig-zags up and down panel. 2x matrix used ws2811 from one supplier (on Ebay) and 2 matrix used ws2811 from a different supplier on Ebay. Each panel is driven from a single F16v3 data output. Power is injected every 66 pixels. All 4 of matrix worked fine initially during burn in testing however 2 of the panels started to suffer many failed pixels after a few hours, usually blue or failure to pass data. All failures were on the 2 panels from the same supplier. The other 2 panels were perfect after full burn in.
    We decided to bite the bullet and replace all the pixels on the 2 dodgy panels that were failing and purchased another 2500 5v pixels, this time from Ray Wu as he appears to be respected on here and we did not want another dodgy batch.

    After replacing all 990 5v pixels on the first matrix panels I am getting terrible flicker on it. Test set up is F16v3 1 output direct to matrix pixel 1 via 60cm data cable. 60amp PSU for matrix. 6x 10 amp fuses. Power injected every 66 pixels. F16v3 powered from separate PSU (only data and ground taken from F16v3). 0v common for both PSUs. All data and 0v runs together and are of equal length.

    The flicker ONLY happens at low brightness with a steady single any colour but not the whole string flickers. It starts at around 300 pixels in but it is not consistent point where it starts) into the string and manifest as a bright flash of the same colour for the entire rest of the string. the flash gets brighter the further into the string and the rest of the string flashes together. At anything over 60-70% brightness there is no flicker at all. Full white at 100% is rock solid white, no flicker, no pinking. Sequences will play but high brightness flicker is still noticeable.

    Things I have tried to fix it (in no particular order).
    - Power injecting every 33 pixels -no change. (I've measured the wire cross sectional area of the Ray Wu pixels and they are pretty poor - 5 strands of 0.15mm diameter ~ 30-31 AWG).
    - power injecting every 22 pixels - no change. (All joints soldered and heatshrunk with glue lined heatshrink).
    - Replaced the first 495 pixels (half a panel) - no change.
    - Tried another controller (cheap Ebay ws2811 pre-programmed thing i use for testing) - only drove the first 305 pixels but the flashing started earlier (around 150 pixels in) and was worse (more frequent and brighter flashes, sometimes different colours but always in sync together) over all.
    -Completely rewired all power injection 0v and 5v. - no change.
    - Measured resistance of each 0v and 5v 66 pixel section of the panel (after disconnecting all common power and zero feeds) - good continuity (<2.0 ohm per 66 pixel length). - no 0v breaks which is what I was testing for mainly.
    - Measured voltage drop at power injection points throughout the panel - 5.01v at source with no load. 5.01v at furthest pixel from injection with no load. At 100% brightness, all 995 pixels on white - 4.89v at worst point.
    - Reduced string length to 500 pixels (half panel) - no change.
    Swapping back to one of my original (Ebay supplier) pixel matrix results in perfect operation).

    Out of desperation, set up a complete test rig on lounge floor:
    - Laid out 10 x 50 pixel strings (brand new out of the bag from Ray Wu).
    - Set up power injection to the bare wires at the start of each string (different 5v 60 amp PSU).
    - Set up F16v3 direct to first pixel string.
    - added 50 pixels strings one by one whilst F16v3 set to around 10% brightness, solid red. (worst case for flicker).

    Up to 5 strings (250 pixels) - no flicker at all under any conditions.
    6th string added - some flicker on 6th string only.
    adding more strings results in flicker on additional strings. The flicker/flashing brightness tends to move to be most bright towards the last few strings only.
    With all 10 strings, flashing starts around 300 pixels (same as matrix).
    Never flicker or flashing on first 250 pixels, even when remainder of string is flashing.
    - adding an f-amp in between F16v3 and first pixel - no change.
    - adding an f-amp in between every 50 pixels (one by one) - no change.
    - adding 220 ohm resistor between F16v3 and first pixel data - no change.
    - adding 220 ohm resistor between each 50 strands data - (one by one) - no change.
    - Adding a single split choke around various wires at many points (one by one) - no change. (only had one choke of unknown value).
    - programmed a sequence to have pixel 0 to 250 pixels permanently off and pixels 251 to 295 to ramp up and down solid red slowly - flashing on whole of the second half (251-495) at low brightness - exactly the same as if the first half was on as well.
    - disconnecting the data-in at the first pixel when set on a low brightness, causes all flickering to stop.
    (All 10x 50 pixel strands work on their own or in any combination up to 250 pixels all the time).

    All data is run with equal 0v runs but I am pretty sure it is a data corruption problem and not a power injection problem. There must be something different between the Ray Wu pixels and the Ebay ones. The Ebay pixel chips are marked "WS2811" and the Ray Wu ones are marked "WS2811s" but I cant find any technical specification differences for the "S" variant on the various tech sheets I've looked at.

    We built many different props for the first year using nearly 7000 pixels in total and runs of up to 1000 5v pixels in one string on a single output of the F16v3 with no problems whatsoever, but this latest batch of Ray Wu pixels is causing me endless weeks of headaches - Please help if you can.

    Many thanks in advance.
    Last edited by pixel-newbie; 06-13-2020 at 05:48 PM.


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts