Flickering Strips - Please Help!

I

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
Hi all

I have spent hours trying to get my strips to behave and I'm still none the wiser. Please see the image for how I have them wired. Everything works fine except for strips 4 and 5. They have a severe intermittent flicker when they should be on and displaying colours or patterns. There is no flickering when they should be off. Strip 4 is almost new. The lengths specified are the lengths of the strips. The 6m strips have an extra PCB trace so they go further without voltage drop mattering. All data wires between strips are no more than 1 metre long and all strips are 30 LEDs per metre. So far I have tried
  • Swapping strips 3 and 6
  • Connecting strip 3 to strip 5 and removing strip 4
  • Adding a 5V WS2811 node between strips 3 and 4 (the null node lights up properly)
  • Replacing the wires between strips 3 and 4
  • Replacing the wires between strips 4 and 5 and the DC-DC convertor
  • Connecting the 12V power supply where the DC-DC convertor is so that the supply powers all 12V strips
  • Removing strip 5
  • Checking the voltage output from the convertor
Nothing has made any difference to the flickering problem. Tonight I will try disconnecting either the negative between the BeagleBone and strip 1 or strip 6 because this is a ground loop but I figured I would post this here in case anyone has any ideas.

Thanks in advanceFlicker1.png
 
Last edited:
okay
you have swapped 3 and 6 which means it is not the output of the last pixel in 3
you have removed 4 which means it is not the first pixel in 4
you have replaced the cable between 3 and 4 so it is not the cable

similar to what you have suggested, just disconnect 6 from 7 for a few minutes and see watch (yes 6, 7 and 8 will stop working) and see if the flickering continues with the ground loop removed,
The other thing is that the maybe the 12v strip does not behave nicely on the end of 5v strip. Is there any way you can move the 5v string into position 5 ?
 
Thanks JPB

I disconnected the negative between strip 6 and the BeagleBone and hence I removed the ground loop which made no difference. I did, however, notice voltage drop (pink) at the end of strip 6 which didn't make sense to me. I re-connected the negative and it didn't seem to make much difference so the voltage drop must have always been there. I then realised that strips 4 and 5 only misbehave when strip 3 is on. Strip 3 also shows voltage drop which I hadn't noticed before. Fortunately the physical layout is such that I can easily inject between strips 8 and 3. I'll do some more testing tomorrow but I think the problem is due to the 12V strip not understanding the signal from the 5V strip or null pixel when it is underpowered. I'll post an update then.
 
Hey mate, i had some flickering myself.. would only do it at the same point during sequences which was weird..

While going over some other things i discovered it came from 2 issues, which once fixed resolved the flickering..

Issue 1 was my main earth return feed connection had suffered some shrinkage leading to a crap connection..

Issue 2 was a dodgey pigtail which worked fine when squeezed/gently bent a bit, but would drop out connection when moved a different way.

You mention your 6m strips have an extra pcb trace to reduce voltage drop.. unless they have equal postive and negative sized traces then you will be limited by the current handling of the smallest trace..
 
Also without seeing the layout you are trying to cover.. would it be possible to seperate your 5v and 12v strip runs? Ie have strip 1, 2, 4, 5 on one output. And strip 6, 7, 8, 3 on the other?
 
I have 2812 5V nodes and 2811 12v strip mixed together for walls/windows, sharing a common data line.. however, there seems to be an impedance mismatch between the 2 types of chips. Strips feeding nodes works fine (mostly), but he other way around causes flickering at full brightness, even with a proper ground reference at each junction. Along with cable impedance/coupling even over short distances, it caused a lot of frustration (and there may be some minor pulse differences between the two adding to the problem)

Perhaps a buffer/repeater type board would help.
 
I had similar issue where 12v strip did not work after 12v modules (which is essentially strip cut into pixel sections). I separated the strip (arches) onto a different controller output, issue solved. When I tried to add modules after the arches on the same output, modules would not work correctly, so they are on their own output also.
 
I had flickering caused by a flaky connection on the controller board. One of the screws had actually needed a bit more tightening on the +12v line
Noticed it right away.
 
I seem to have solved the problems now. Firstly, I want to add that there is another 5V set of pixels not shown in the diagram. It uses the same 5V supply and PixelBone and hasn't had any problems.

I started by removing the ground loop by connecting strip 3 directly to the 5V power supply. This caused strips 1, 2, 3, 4 and 5 to flicker when on full brightness white. Strips 4, 5 and the second half of strip 3 kept getting stuck on white. I then tried replacing strip 3 with a 12V strip. Strips 1,2,3,4 and 5 then completely stopped working.

It turns out that either the connector or the start of strip 1 has a bad negative connection to the PixelBone. Having a 5V strip in the line connected the negative reference through the 5V power supply due to the working 5V pixels not in the diagram. This is why it didn't work when I changed strip 3 to a 12V strip. The negative reference introduced through the 5V strip still wasn't enough because it has multiple long thin wires and it is connected at strip 3 whereas the signal comes in strip 1. As a workaround I have now connected the negative output of the PixelBone directly to the 12V power supply and the flicker is gone!

Tonight I will get a better look at how well it works. I think there will be a little bit of voltage drop in the 12V strip 3 because it is powered through strip 2. There is currently too much daylight for me to tell if it is pink.

Roosta, you are correct that I will be limited by the current handling of the thinnest PCB trace but voltage drop is different and it tends to be a problem before the maximum current handling capacity is reached. Voltage drop happens in both directions; doubling one trace halves the voltage drop in that direction, reducing it to 3/4 of what it would have been if the trace wasn't doubled.

marmalade and djgra79, I like the idea of a repeater board which isn't something I'd thought of. Judging by last year, I know that I can successfully have these 5V strips followed by these 5V nodes followed by these 12V strip. This is why I added the 5V node as a null pixel. This isn't really relevant now that I've changed that run to all 12V strip.

bluzervic, I believe this was caused by a bad connection but unlike yours, it was the negative and it was at the start of the strip, not the controller. It's recommended that you don't run the positive connection to the pixels through the controller and the PixelBone doesn't have the ability to do this. I haven't actually checked my connection that I think is bad but judging by the behaviour, I'm sure this was the problem.

Thanks to everyone who replied and tried to help me. I will get a better look at this tonight but I think it is alll working now.
 
You must log in or register to reply here.
Back
Top