Pixels Only working correctly when I touch the data line! Help!

canigian

New elf
Joined
Aug 10, 2010
Messages
2
I built custom pixel icicles 2 years ago and have always had what appeared to be "data" issues with them. This will flicker at weird times, work normally at others. The prop is 804 WS2811 pixels, broken up into 134 icicles that are 6 pixels in length. I built this prop in a modular manner in "bars" that can be taken down and easily stored. Each bar can be tested solo, which I am doing now, and I am finding some VERY strange things in some of the bars, not all.. I am running these on an e6804 sandevices board. there are test patterns that you can run when you plug pixels into it. 1 is all red, 2 is all green, 3 is all blue. When I wire up a bar to the board, I am running 1, which is all red. and find that the first pixel is blue, and all others are magenta. Now, if I touch the data connection on the pixels, suddenly, all pixels turn red, as they should be. if I take my finger away, they go back to 1 blue, and all magenta. I've built ALOT of pixel elements, but most are between 20 - 150 pixels. Not sure if this a symptom of a "big" build, or what. Has anyone ever experienced this??? I have no idea what is going on, or what I did wrong. Please help.
 

AAH

I love blinky lights :)
Community project designer
Joined
Dec 27, 2010
Messages
3,517
Location
Eaglehawk
How long is the run from the E6804 to the first pixel?
Are you touching the copper of the wire or just the insulation?
 

scamper

Dedicated elf
Joined
Jan 5, 2014
Messages
1,092
Location
collie
As Alan said, more info is key.
But I would be looking at earthing problems. maybe either poor connections, or even a floating earth. Which basically is if your controller and lights don't have a common earth.
Otherwise, if the run is too long etc.
 

canigian

New elf
Joined
Aug 10, 2010
Messages
2
Touching the bare wire. The run is about a foot. Can you recommend some reading up on the earthing issues? Thanks for the replies!
 

scamper

Dedicated elf
Joined
Jan 5, 2014
Messages
1,092
Location
collie
Touching the bare wire. The run is about a foot. Can you recommend some reading up on the earthing issues? Thanks for the replies!
Sorry, I can't help with reading lists, I was a technician years ago and learnt a lot of this stuff at uni. The ailing brain and the fact I have been out of the game for 20 some years leaves a few blanks.
But earthing is probably the most misunderstood part in all electronics, as everyone thinks all they need is power and data.
but, if you don't have a good earth, then power is really nothing, That is why birds can sit on a live power line, they have no reference to earth.
Also, the same with data, there must be a reference to something.
And that is why the earths must be connected if using more than 1 supply over props, and the earth from data and power must be joined.
If they are not, you can get what is referred to as a floating ground, which just means it is not 0v it could be just above for example you get a data earth that is floating at .5v and the data is 1v, then instead of having a nice strong 1v peak data value, you only have a .5v peak data value, which the electronics cannot decipher. (please note these are examples, not actual figures)
When you tie the earths together, then they all have the same potential so you eliminate many problems.
A floating earth can be induced by a "noisy" power supply or similar, so as you touch the wire you could be acting as a filter.
Again, this is just fuzzy theory. But none the less..... Earth is key!
 

TerryK

Retired Elf
Joined
Feb 9, 2020
Messages
52
Location
West Central Ohio
A timing issue perhaps? I'm not familiar with the E6804 although I did download its User Manual and glance through it. If I understand correctly the E6804's refresh rate is dependent upon the type and number of pixels. I'm wondering if the controller is trying to update the string too quickly or at least faster than what the WS2811s are able to do.

Do you have any idea or know what the maximum refresh rate of the string is and what the controller is attempting? There is also a SanDevices note dated Jan.2014 about a jumper on the controller pertaining to a 3.3 versus 4.0 volt data signal.
 
Last edited:
Top