LATEST RELEASE:  FPP 7.5 - Download from here -



Welcome, Guest.
Please login or register.
Forgot your password?

+-Site Stats

Total Members: 16397
Latest: Markr90
New This Month: 0
New This Week: 2
New Today: 0
Total Posts: 132079
Total Topics: 16371
Most Online Today: 68
Most Online Ever: 7634
(January 21, 2020, 02:14:03 AM)
Users Online
Members: 2
Guests: 44
Total: 46

Help Request: Malfunctioning LED strands/strips. Pointers appreciated!

Started by khanklatt, October 06, 2020, 04:49:32 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.


Update: See below for some improvements that have helped!

I have five strands of WS2815 LED strips (of varying lengths) that are inconsistently lighting when I try to do a color wash test pattern from my Falcon F16v3 controller. I've posted a few videos of four of my strands that seem to have basic continuity of signal and power, and what I'm looking for is advice to minimize time spent either taking the strips down to repair them, or in specific strategies to use to replace elements or entire sections of the strips.

I would be most appreciative if someone with more experience than me can inspect my videos and make some suggestions about how to approach recovering/fixing each strand!

After thinking about my 5th strand (always lit on white) I thought that the problem was likely a short between V+ and a signal line. This being one of the longest strands (longest lead to the start of the strand) I started thinking that this could be a signaling issue. Maybe blue and green are not supposed to be wired to the same terminal for WS2815s?

This mode of thinking also got me to mentally model why there'd be a color flicker like I saw on the other lines? I explained it to my wife where the signals from the controller are entering two wires which have slightly different lengths, just like your index and middle fingers aren't the same length. I said "if you taped your fingers together and tried touching your touchscreen, the touches would come a fraction of a second apart; this is probably why the lights are flickering, due to a slight delay in duplicate signals to the lights".

Sure enough, I don't know why I thought I should wire the WS2815's two signal lines to the same signal port on the controller. Fixing that has substantially improved the situation. In case it helps anyone, all of my youtube videos are here: , otherwise I'm taking down the old videos below.

- Do you see any obvious problems of individual pixels or portions of the strands that need replacement?
- Do you see any opportunities to fix "in situ" (in place) as opposed to opening the channels, removing the strips and then replacing them?
- Do you suggest taking down the aluminum channels and remounting the strips something I am hoping to avoid?

What I've tried:
- It's not a controller problem because the problems persist regardless of which port I plug them into.
- I put the controller into the color wash test pattern since that's how I tested the strips before mounting them.
- Either because of water incursion at the connection joints (which I tried to waterproof) or damage during mounting (some channels bent 30-60 degrees while lifting the channels up), I suspect there are shorts or other faults along the strips.
Update:Picking either the blue or green wire (e.g. signal or redundant signal) but not both has greatly improved many of the strands.

Finally, I should point out that the lower channels (strips 1-3) were operating fine as you can see in the next pic (taken back in November) so the dysfunctions appeared after the lower strips were exposed to the elements for a while. I'm coming back to fixing these for Halloween, but these have been problematic for a while (it's been dry these past few weeks, so these issues have persisted through wet/dry conditions and being unplugged for many months):

In case it is useful, my original plan and post is here:,11708.0.html

Here's the diagram of the plan I posted at the time:

This is fairly close to what I ended up doing, except the yellow top tier I split into two strands (two ports on the controller), I skipped the green area (too complex for too little benefit). I refer to each strand by number, and color coded them below, and attached videos of each problem. I used pink instead of yellow to talk about the top level since it's hard to read: Right?

Strand One (cyan): This one is easiest to troubleshoot. No lights, nothing at all, plan on troubleshooting this as basic continuity issue that developed after it was already working.

My plan: Zero lights lit indicates no power/circuit (red/black) or no signal (blue/green). Check for voltage at the beginning with a multimeter. Suspect cable/soldering leading to first pixel if zero volts on any pin at pixel 1.

Update: I found one loose wire (looked to be ground) but a quick fix didn't repair it. Still need to tackle this.

Strand Two:
This strand is much improved, but still flickering. I suspect the first pixel is problematic/shorting which causes a cascade on the rest of the strand. This happens rarely (once every 30 seconds or so) but I can't imagine it's good to leave it that way.

My plan:
Update: 1) Cut out the first pixel, and connect the strip to the second pixel, maybe the third
Update: 2) For the last few inches where a second to last pixel is not lit, leave the strip alone and just turn off the last couple pixels programatically, which won't be noticeable if the last couple are not lit.

Strand Three: This one has a few lights at the end of the strip that don't flicker like the others.

Update: This strand seems to be much improved. For some reason the last couple feet are not lighting at times.

My plan:
Update: This one is greatly improved. Only the last couple feet are not lit. I suspect it's a short on the signal wire, because in my original video, these were lit constant red when both signal lines were connected. So I know the pixels are good, I just need to ensure connectivity of the signal all the way through. So a cut/re-splice of the strip may be required.

Strand Four (arch/peak yellow strand on the left): This one is also greatly improved but now it color washes, but with a flicker from time to time.

It may be hard to see, but it seems to flicker towards all white as it goes through the color wash cycle.

My plan: Double check each solder/connection point up to the beginning of the strip.

Strand Five (straight yellow bit on the right)
Update: As anticipated, not only did most of the strip begin working when I removed one of the signal lines, the rest of the strip lit up as well.

In summary
Apologies for the length of this post; I erred on detail in the hopes that you kind souls will save me a ton of time on my roof!
Thanks in advance for your assistance!


There's a good chance you find the cause of the flickering in here:

Errors above are made for entertaining readers only

English is not my native language.


@kockie great video, thank you!

My summary takeaways of things to try:

1) Double up on the ground wire to the pixels to give the signal more voltage differential to trigger +/-
2) Replace the cable with something that can handle interference (capacitance specifically) better
3) Since the cable impedance is proportional to the length of the cable, shortening the cable run could also seem to have a positive effect (say I had 10 feet of extra cable that I hadn't trimmed yet since I'm still in testing and have not yet mounted the controller ;) )
4) (Although the video didn't state this) I suppose power injection at the end of the strand (+12v and 0v) could clean up the overall triggering of the signal on the wire...
5) Install an F-amp before the flickering strand(s)
6) Replace the socketed 100ohm resistors on the F16v3 with 30-33ohm variants
7) Install a null pixel between the controller and the strand.

So I have a few questions :)

#2: I used 18AWG "CL2 Power Limited Circuit Cable", but I have 24AWG 4 pair CAT6 that I could replace the cables with, if that would be better?
#4: Am I onto something with power injection to introduce less voltage drop and cleaner ground over the length of the strand?
#5: Recommended suppliers of F-amps?
#6: Recommended suppliers for 30-33Ω SIP-8 in small quantities? I've found a few sites that sell them in 2000 qty. but I need 99% less than that ;) ) Also, are they bussed or individual resistors? Couldn't find any reference to them in the F16V3 manual.
#7 Recommended suppliers for null pixels or can I achieve the same thing by merely turning off the first addressable pixel in the strand and tell it not to light, converting it, effectively, into a null pixel? If not, is there some guidance on where the best place to put the null pixel on the run leading to the strand?

Of all of these solutions, keeping the first pixel off (if my hypothesis for #7 is correct) seems like the easiest one (no roof work, no rewiring), or introducing an F-amp/null pixel or output resistor seem easiest (some rewiring but no roof work) assuming the F-amp/null pixel can be installed near the controller.


Following up in the hopes it helps others and to thank @kockie for his pointer to the video!

I've remedied the flickering problem for 60% of my strands by simply replacing the socketed 100 ohm/100Ω resistors with 33 ohm/33Ω resistors (aka output resistors). Given these are cheap, and the process to replace them is easy, I recommend it as a first step before any others if you are -> <- close to having a functional set of strands. 

Some quick Q&A for others in the same spot as me:

Q: What resistors/part numbers should I be using?
A: The first thing I did was remove the existing resistors and look up the part number that was imprinted on the resistor.
On mine, it said B101J, and an internet search resulted in products described as "RES NET 8I 100R 4RES SIP".

Decoding this into English, this product is a "Resistor Network, 8-inline pins, 100 ohm resistance, in a 4 resistor network array, Single Inline Package" (as opposed to DIP which would have two rows of pins instead of a single row of 8).

This gives some clues about what to order for replacement. My research showed that there are three main types of resistor networks like this. One is a common bus layout (common ground pin, the remaining pins are V+), another is a cascading layout (each resistor is connected inline to another resistor in the package, called dual terminator), and the third is what the F16v3's resistors appear to be, four distinct independent resistors, called "isolated" networks.

Using the nomenclature of that internet search, you'd want a RES NET 8i 33R 4RES SIP or a Resistor Network with 8 pins in a SIP package, with 33Ohm 33Ω resistance.

Q: OK, but what part number did you get?
A: At the time I ordered, there were four product matches. One of them had yellow resin packaging (and in retrospect I might have bought that to distinguish the old from the new), one of them had rectangular packaging (ideal for using them in two rows for DIP sockets, apparently, as opposed to the bulbous/rounded packaging of the factory resistors), so to keep in line with what was already there, I excluded those two.

There were two manufacturers, and not knowing between them, I just picked one of them. The price different wasn't much, the temperature range and environmental specs (RoHS) were the same, so what I ordered was the "Resistor Networks & Arrays 33ohms 8Pin 2% Isolated"  from "CTS Electronic Components" with part number "77083330P".

Q: What supplier can I use?
A: I tried Amazon but was unable to find one. I also tried Digikey, but couldn't find it (not to say they don't have it or won't when you look). In the end, I bought from Mouser electronics. I'd link to the product page, but it's likely to change over time.

Reasons I chose Mouser:
1) Product images. Not having bought electronic components like this, having a representative image was a little piece of mind
2) Sub-bulk ordering. Some products/packaging comes in minimum quantities of 1000 or more.
3) US stock and shipping (vs. Alibaba or other sources that have cheaper prices but lengthier shipping and delivery costs)
4) Good search engine to select in-stock, small quantity, and right product criteria.

Q: How do I replace the exit resistors?
A: Chances are most folks won't be afraid of this, but just in case, I used needle nose pliers to remove the old ones, then gently aligned the replacements and pushed them into place with my finger. The key to removing without damage (as might be obvious) is to pull firmly, but with good control straight up, and don't apply any pressure until all eight pins are in the proper holes.

The pins can manage a little bit of bending abuse (just gently straighten out the pins if they got bent in shipping or handling). If the pins get too far bent, just toss it and use another one. They're so cheap, the hassle of fishing out a broken pin from the socket isn't worth the trouble.

Q: Any other tips?
A: If you want to dig into far more detail than you'll need to understand resistor networks, looking into datasheets can be instructive. Here's the data sheet for the product I bought, it explains a lot of what I touched on above.

Support FPP

+- Recent Topics

strang text behavior with all modell by Poporacer
Today at 07:13:42 AM

Playing video to Monitor on RPi4 by TomsLEDs
February 29, 2024, 10:08:20 PM

Problem with images in playlist. by CaptainMurdoch
February 29, 2024, 08:56:43 AM

Choosing P4 Panels / Colorlight by CaptainMurdoch
February 29, 2024, 08:52:41 AM

Struggling with media playback sync by MichaelD
February 27, 2024, 03:40:10 AM

Live video with xLights and Falcon Controllers by Laser Falcon
February 24, 2024, 01:20:52 PM

Waveshare 7" Screen by skulltronix
February 23, 2024, 03:36:17 PM

Need to expand by darylc
February 22, 2024, 11:41:01 PM

reviving a bbb on an octoscroller by starsman20
February 22, 2024, 04:34:23 PM

FPP After Hours Music Plugin has been updated by MikeKrebs
February 21, 2024, 10:24:39 PM

Powered by EzPortal
Powered by SMFPacks Menu Editor Mod