I need some help troubleshooting a megatree. Here’s my setup:
Software: Xlights
Tree: 360-degrees, 32 strands, 95 12v, WS2811 RGB nodes per strand (Ray Wu).
Organization on Tree: Odd number strands start at the bottom of the tree at end at the top of the tree; even number strands start at the top of the tree and end at the bottom. Strands (1,2), (3,4), (5,6), etc. are joined with 3-core pigtails at the top of the tree.
Distance from controller to first pixel at beginning of strings: between 10 and 20 feet.
Controller: Falcon F16v2 in 16 string mode.
Organization on controller: 8 “strings” of 4 strands, 1 string per controller output.
Organization on Xlights: 8 strings of 4 strands, start at lower left.
Controller channels: 380 nodes per controller output; 1140 channels per controller.
Null Pixels: NONE
Zig: No
Power Supply: Meanwell 12v 350w
Power Injection: V+ and ground at beginning of strands 1 and 3; V+ and ground at ends of strands 2 and 4.
All nodes were tested before hanging. Only one bad pixel out of more than 3,000.
All power and data cords tested good before installation. One minor short found on one power cable, which has been discarded.
Overview of Problem
I am connecting the strings to the controller one string (4 physical strands) at a time, then adding each successive string once the prior strings are working. The Falcon is outputting “color ramp” in test mode. I have also used RGBW and White Ramp to max out the current and see what happens.
String 1 (strands 1-4) work properly. No issues whatsoever.
String 2 (strands 5-8): Strands 5-6 work properly. Strands 7 and 8 flicker, and pixels are different colors throughout the strand. On white particularly, some pixels will be white, some will be yellowish, or other colors. The pixels at the end of the strand 7 and 8 are very dim.
Troubleshooting Steps Thus Far
1. Figured it was a power injection issue. Found one injection cable that was warm; figured it was a minor short, and replaced the cable. No change.
2. Still thinking it’s a power cable problem, I swapped “known good” cables from the first strand. Power seems to flow better, but the problem persists.
3. I also tested and swapped the Tees used to inject power and connect strands 6-7. All tees test good.
4. Thinking it may be a connection issue at the top of the tree between strands 7-8, I move the input from the controller from strand 5 to strand 7 to test it. Strands 7-8 light up perfectly.
From the above, I infer that I have an issue with data transmission between strand 6 and 7. They connect at the bottom of the tree, so it’s easy to reach them.
I have read a couple different posts that describe problems similar to mine. Madsci1016 indicated that it may be a “voltage drop to reference data line issue,” and the solution to the problem is either (a) reduce the brightness to keep current down, or (b) install a null pixel.
I’ve reduced the brightness on the controller to 75%, although I haven’t tested it yet (daytime, and I’m at work now).
I have several of Ray Wu’s null pre-made pixels. Here’s what I don’t know: (a) where should I place the null pixels, (b) how to configure the Falcon F16v2 to account for the addition of a null pixel in the string/strands, and (c) what changes to Xlights will need to be made to account for the null pixels.
My thought is that I should place the null pixel between the last “known good” pixel and the first pixel of the strand that displays the problem, i.e. at the end of strand 6 right before strand 7. On the controller, I assume I need to add one more channel to the string, and break it into two virtual strings with the null pixel being the last pixel on the first virtual string.
Is this correct, and if not, what am I missing?
ETA: clarified V+ and ground being injected.
Power injection means V+ and V- at every injection point. Ground carries just as much current as the positive side.
Sent from my Moto G (4) using Tapatalk
Yes, I know. I have a common ground at all injection points with V+ isolated for different power supplies. I've been using "V+" and "V+ iso" as shorthand on my wiring plans for +/ground injection, so sorry it carried over into my post. Have edited the post.
You definitely seem to be describing a power problem, especially with the dimming. Have you checked with a meter? It sounds like one if your injection points isn't working.
Sent from my Moto G (4) using Tapatalk
I haven't put a meter on it yet, but I'll grab the multimeter and check the current at the end of strand 6. It's entirely possible that I botched the soldering job on the pigtail. Which is going to suck, since I'll have to resolder it while its on the tree.
Now that I think about it, I seem to recall that the last couple nodes on strand 6 were very dim. I'll have to run that down, too.
You might want to check the voltage at each end of each strand. Color wash won't be very stable though, you might want to use white ramp to get the worst case. Even that won't be very stable as the intensity changes during the ramp. Your voltage drop will change in proportion to the brightness (current draw). You may know all this already, but maybe someone else will benefit in the future.
Sent from my Moto G (4) using Tapatalk
After testing the entire tree without issues, I think I have a bad pixel at the end of strand 6. The last pixel on that strand glows a dull orange--even with good power injection right into it--and only downstream pixels from that point are affected. The one bad pixel I found during testing glowed dull orange without power being applied, and this one has the same appearance. The downstream pixels will flash, turn on and off, and change color randomly. I may have a short on the data line or may have broken the solder joint on that pixel when fabricating the pigtail (the end pixels took the worst beating). I'll replace it in the morning and see if anything changes.
I need some help troubleshooting a megatree. Here’s my setup:
Software: Xlights
Tree: 360-degrees, 32 strands, 95 12v, WS2811 RGB nodes per strand (Ray Wu).
how may leds per roll? What controller are you using?
If you have the same as me set the node/string to 100
I have the 5m 300 leds per roll, one ws2811 chip controlls 3 led lights, hence in xlights set the node/string to 100 and in the controller set to 100 nodes as well
In my controller the F16V2 I add virtual string and check reverse B/C I also have 2 strips joined at the top of my tree.
if your controller does not do that, you can set node/string to 200 and strand/sting to 2
remember that there are limitations on each output port on the controller. my F16v2 is 680 Pixels so in theory one port I would be able to drive 6 strands. when the 7th strand is added 80 of the 100 pixels will light up and the other 20 pixels will flash all sorts of weird colors.
Also I ordered 35 rolls of lights and one roll was different I didnt even notice it, all the rolls looked the same. But boy it cause me headaches. one roll was ws16703 and it messed with my data causing issues.
