2811 data interference from dumb string

AlexB

New elf
Joined
Nov 2, 2012
Messages
45
Location
South of Brisbane
Ok - I know, its totally my fault. As a sound guy from way back, I know power cables are supposed to cross at right angles to data \ sound lines to avoid interference... I'm pretty sure this is my problem, and i'm wondering if anyone has come across it \ has an alternative solution to relocating the whole shebang.


We have WS2811 strip run on conduit under our eaves. This year, the wife wanted icicles, and thanks to somebody on here designing a set of dumb RGB icicle strings, she got her wish. With time and budget limited, we decided the easiest way to mount them was to cable tie them to the existing conduits.


This puts 12V dumb RGB string parallel to the strip.


What i'm seeing now is random coloured strobe effects pulsing from the strip at random points. All very low intensity (i'd say 10%) - but very noticeable in a 'blackout' scene.


The controller is brand new, as are the icicles. the effect only happens when the controller (Ray27) is outputting, but it happens on sections of the strip that aren't plugged in.


I've done a little diagram attached to this post so you can see what i mean.

icicle problem.png


As a key - (ignore the red line and orange star):
  • green line is output 1 of 2811 (2 strings)
  • blue line is output 2 of 2811 (3 strings)
  • i've numbered each segment 1-5
  • the icicle segments are indicated by the white lines, and they follow the same numbering.
It's important to note that the strobe issue is only happening on output one (green line) it doesn't happen on output 2. (Segments 3,4,5 - blue line)


To trouble shoot, i've done the following:
  • tried the Ray27 board in test mode, and driving from DA E1.31 and LSP
  • tried an older Ray27 board from another element in test mode
  • tried unplugging icicle segments 1, 1+2, 1+2+3, 1+2+3+4
  • even with only segment 5 running, and all others unplugged it still causes interference on segment 1+2 of the strip
  • all of the extension leads from the controller to the icicles run on the ground. all the data extensions to the strip are in the air
  • There is a setup difference between output 1 and output 2 - output one has a longer cable run, so the speed for output one is 2600, whereas output 2 runs at 3000. I haven't tried changing this yet.
In summary - has anyone else experienced this, and how did you fix it? any gurus have any ideas that don't involve relocating the icicles from the conduit (this is my last hope before i do that, and I really don't want to do this as the icicles are all cut to length based on their current location)...


thanks in advance for any assistance.
 

AAH

I love blinky lights :)
Community project designer
Joined
Dec 27, 2010
Messages
4,188
Location
Eaglehawk
As leds are quite efficient devices and potentially only a few uA us required to get them noticeable at dark it may be worth trying to put a dummy load across them. A 5W resistor that is selected so that it's only dissipating about 1W when your lights are fully on could be worth a go. As a short term measure to see if it sorts the problem you could try hanging an incandescent light across the line.
 

AAH

I love blinky lights :)
Community project designer
Joined
Dec 27, 2010
Messages
4,188
Location
Eaglehawk
If I read right that it's the dumb rgb lighting up then I'd put a low wattage globe between the +VE and green output connection. Assuming that the leds are 5V or 12V then something like a 3W globe as used for an interior light in a car would do for testing.
 

AlexB

New elf
Joined
Nov 2, 2012
Messages
45
Location
South of Brisbane
Sorry no - it's the 2811 strip that is lighting up randomly across the strip, i'm fairly certain the interference is from the voltage EMF on the dumb strings corrupting the data...


i've tried unsuccessfully fiddling with the data speed on the strip - so i'm now in the process of stripping out the extension leads for the strip and relocating the P2 for the strip to see if i get less interference with a stronger signal )(as output 2 is having no problems - and it was only 2-3m from the controller).
 

dale82

Senior elf
Joined
Nov 19, 2010
Messages
581
Location
Kingaroy
to me it sounds like you are having a data problem, I would try adding a null pixel half along the red line (From the controller to the start of your WS2811 strips ) so the data regenerates as it will be data loss.
 

AlexB

New elf
Joined
Nov 2, 2012
Messages
45
Location
South of Brisbane
UPDATE:
So I've resorted to taking the icicles down off the conduits on strip 1 and 2. the strip no longer randomly strobes at idle, but it continues to pulse \ flicker when colour mixing at less than 100% (eg: during fade up in any colour other than RGB - so yellow, white, purple, cyan etc... but ONLY when the icicles are outputting.


ONLY when the icicles are outputting.


Currently - icicles 1 and 2 are on the GROUND. Strips 1 and 2 are completely isolated from the icicles by at least 2 metres. 3,4&5 are still attached to each other as it's high side, but the problem occurs primarily on 1&2. this makes no sense.
If I turn the icicles off, or just leave them at idle - no problem - it is only when they are outputting at the same time.


I've relocated my controller. shortened the extension lead (and singled it up), ive tried everything I can think of - and most things you guys have thought of - and no go.


There is only ONE PSU for the entire strip. The icicles run off a separate PSU, which is 2 metres from the nearest strip. All icicle leads go straight to ground level before going towards the controller - all strip leads run in line with the strip.


I'm out of ideas.
 

AAH

I love blinky lights :)
Community project designer
Joined
Dec 27, 2010
Messages
4,188
Location
Eaglehawk
Can you pop up a mudmap of how your controllers and power supply/s hook together Alex?
Did the icicles generate problems with the pixels when the icicles were in test mode and with the dmx lead disconnected from the 27 channel board/s?
How many watts/pixels etc have you got hanging off the power supply and what wattage is it?
 

plasmadrive

Full time elf
Joined
Jan 12, 2012
Messages
248
Location
Elk Grove, CA
Being the originator of the Plasma Icicles I can tell you that I ran into all kinds of this issue. Everything worked great until I put them up and bundled the cables together.. then noise was my enemy. I tried all kinds of things from grounding the conduit to null pixels and nothing worked reliably.. until.....

....... I put Pixel Extenders in the mix. Now for my situation I am using e6804s to drive my setup. The e6804 can be use like a differential output for WS2811s by using the CLOCK output as the inverting output. I put JiSys Pixel extenders right at the first pixel of each universe of Plasma Icicles and the entire problem went bye bye immediately!!! They are not expensive, they are small and work great. I put them in ziplock bags under the eves and with a wire tie holding them closed.. have had no noise issues since (3 years running).

Most likely your RGB strips are being fed with PWM power and that will cause quite a bit of radiated noise. If you don't want to do the Pixel Extenders, which I think is the most reliable way, try using shielded cables for your RGB strips and grounding the shield. Not sure that will work but might be worth a try.

Now I just had one issue that I never expected and that was an EMP from lightning close by. I had my power supply still grounded to the conduit from original testing while trying to get rid of the noise. I never earth grounded the conduit this year and that may have been my down fall during the storm, but I can tell you now that my conduit is no longer tied to my power supplies or to each other. (Just FYI)
 

pmiller

New elf
Joined
Oct 27, 2015
Messages
9
plasmadrive said:
Being the originator of the Plasma Icicles I can tell you that I ran into all kinds of this issue. Everything worked great until I put them up and bundled the cables together.. then noise was my enemy. I tried all kinds of things from grounding the conduit to null pixels and nothing worked reliably.. until.....

....... I put Pixel Extenders in the mix. Now for my situation I am using e6804s to drive my setup. The e6804 can be use like a differential output for WS2811s by using the CLOCK output as the inverting output. I put JiSys Pixel extenders right at the first pixel of each universe of Plasma Icicles and the entire problem went bye bye immediately!!! They are not expensive, they are small and work great. I put them in ziplock bags under the eves and with a wire tie holding them closed.. have had no noise issues since (3 years running).

Most likely your RGB strips are being fed with PWM power and that will cause quite a bit of radiated noise. If you don't want to do the Pixel Extenders, which I think is the most reliable way, try using shielded cables for your RGB strips and grounding the shield. Not sure that will work but might be worth a try.

Now I just had one issue that I never expected and that was an EMP from lightning close by. I had my power supply still grounded to the conduit from original testing while trying to get rid of the noise. I never earth grounded the conduit this year and that may have been my down fall during the storm, but I can tell you now that my conduit is no longer tied to my power supplies or to each other. (Just FYI)
Doesn't the extender need to be used with the E1.31 bridge?
 

plasmadrive

Full time elf
Joined
Jan 12, 2012
Messages
248
Location
Elk Grove, CA
pmiller said:
plasmadrive said:
Being the originator of the Plasma Icicles I can tell you that I ran into all kinds of this issue. Everything worked great until I put them up and bundled the cables together.. then noise was my enemy. I tried all kinds of things from grounding the conduit to null pixels and nothing worked reliably.. until.....

....... I put Pixel Extenders in the mix. Now for my situation I am using e6804s to drive my setup. The e6804 can be use like a differential output for WS2811s by using the CLOCK output as the inverting output. I put JiSys Pixel extenders right at the first pixel of each universe of Plasma Icicles and the entire problem went bye bye immediately!!! They are not expensive, they are small and work great. I put them in ziplock bags under the eves and with a wire tie holding them closed.. have had no noise issues since (3 years running).

Most likely your RGB strips are being fed with PWM power and that will cause quite a bit of radiated noise. If you don't want to do the Pixel Extenders, which I think is the most reliable way, try using shielded cables for your RGB strips and grounding the shield. Not sure that will work but might be worth a try.

Now I just had one issue that I never expected and that was an EMP from lightning close by. I had my power supply still grounded to the conduit from original testing while trying to get rid of the noise. I never earth grounded the conduit this year and that may have been my down fall during the storm, but I can tell you now that my conduit is no longer tied to my power supplies or to each other. (Just FYI)
Doesn't the extender need to be used with the E1.31 bridge?
Nope. I used the J1Sys Pixel Extenders driven by e6804 boards. D+ for Data+ and CLK for D-.
As a matter of fact I have used the same extenders with the Pixlite Long Range when all I needed was a single output. (the Long Range remotes have 4 outputs in Condensed).
 

pmiller

New elf
Joined
Oct 27, 2015
Messages
9
Nice. I'm new to this. Can you show me how you connected the J1sys? I'm using 3 wire pixels. I know there are 4 wire pixels that require clock along with data but 3 wire only needs data and of course (+)(-) power.
 
Top