Erratic Pixel behav...
 
Notifications
Clear all

Erratic Pixel behaviour

7 Posts
6 Users
0 Reactions
2,372 Views
Matt_Edwards
Posts: 2664
Topic starter
(@matt-edwards)
Famed Member
Joined: 18 years ago
[#30901]

I consider myself a recalcitrant Pixel user, hey the ones I have had for years are only just going out this year.
As mentioned in this thread I am getting a DMX dual bridge going.
The 6803 work fine.
I member sent me some WS2801 strings to test.
The Code works well but one pixel string is erratic.
Solid On and Off are fine. But when the channels ramp up or up, some of the pixels flash all sorts of colours. it is not all pixels on the string, but the same ones and some of the dodgy ones seem to flicker when ANY of the pixels are ramping, not just them selves.
If I swap the strings on the controller ( it is a dual string controller) the fault moves with the string.

I have a short video of the problem

Has any one noticed this sort of issue, and if so is there a fix?


6 Replies
kingofkya
Posts: 2165
(@kingofkya)
Noble Member
Joined: 15 years ago

Check your power suppply it is very likely it is too low power or your string is too long.


Reply
Posts: 6728
(@p-short)
Illustrious Member
Joined: 20 years ago

I had some issues with the WS2801 strings as well when I was debugging the PX1. Some individual pixels in certain strings had problems, and it seemed to follow the string. Splicing the 'bad' pixels out seemed to help, temporarily. At the time I was trying to use the same code to drive both WS2801 pixels and WS2811 pixels by employing the trick that involved sending a WS2811 signal to the WS2801 pixels while simultaneously supplying a clock with a low-high transition in the middle of the cell. In the end I had to split the code up into two separate sections, one for each type of pixel. For the WS2801 pixels I had to remove all extraneous transitions from the data signal. I never figured out exactly why the WS2801 pixels were acting strange, other than to surmise that the chip designers did something in the design that was subtly different from the published spec.


Reply
tucsonlights
Posts: 667
(@tucsonlights)
Honorable Member
Joined: 15 years ago

Try using a lower resistance on the resistor networks on the outputs. I think Jim uses 220 ohms...Try 120 or lower.


Reply
Matt_Edwards
Posts: 2664
Topic starter
(@matt-edwards)
Famed Member
Joined: 18 years ago

Ok so this is what I will try:
1) measure supply by piercing the cable between pixels
2) adjust series resistor on controller
3) scope the data and clk lines between the pixels
4) throw them out and buy a new string 😉

Sent from my SM-N9005 using Tapatalk


Reply
Posts: 6
(@camper)
Active Member
Joined: 13 years ago

Yeah probly power. High Vf on some of the leds at the end of the string, low V+ due to end of string, V changing due to updates, dropouts occur, V oscillates, flicker.
You could probably get away dropping the total brightness a few bits in software, looks like you are cutting it close.


Reply
Posts: 210
(@sandy)
Estimable Member
Joined: 19 years ago

I've just got the WS2811 5v IP68 led pixel module 50 Count string and the 5v 60W Waterproof Power Supply,from DIYLED, and am having framing errors on the PC1 and erratic LED operation.
so I went back to the 5V 1A power cube that I was originally using, the the problem went away.
The cube is a linear supply, and the 60W is a switcher.
Seems to be the noise from the switcher.
Has anybody seen and cured this before?
I've tried a ferrite on the output cable and a .1 disc and 100 uF across the power connecter to the PC1.
so far no luck.
I should add, I'm only testing with the first 15 bulbs.
When it falses, other bulbs in the string down line can come on.

Seems to be the lack of a common ground between the RS-485 output adapter and the PX1.
By connecting the ground of adapter to the pin 1 of the RJ-45 via the CAT5 cable, everything settled down.


Reply
Share: