henkvis wrote:Thanks, this worked out for me.... don't know if twisting ground with data made THE difference, but it also couldn't hurt, so I used it....
Good to know. Still don't understand why...
I've ordered a bunch of different Ferrite Clamp On Cores from Mouser to arrive on Friday. I'll post here about whether it makes any difference.
I don't know anything about how to select a good Ferrite Clamp On Core, so my major selection criteria was: "Does it fit around the wire?" 🙂
Sandshark wrote:While my experience with pixels is short lived (started very basic last year), I have a lot of experience in electronics. I think it may be some kind of interference between strands and one reason it could be getting worse at cold is a bad connection that is adding resistance, making for higher susceptibility. A chip's drive could be getting weaker, but most electronics work better at cold, so that seems unlikely. Grabbing the cables is providing some capacitance between them, and possibly to local ground, depending on your footwear, and that can be a good way to reduce interference. Changing resistor values in the controller could help by increasing the signal-to-noise ratio. Putting ferrite cores over the wires (possibly also wrapping them in a wire to earth ground) may provide a similar situation to you holding them. Even just driving a metal stake in the ground and wrapping a wire from it around the bundle several times in a coil may have some impact.I've seen voltage drop issues cause all kinds of odd issues, but I can't figure how that would be affected by you grabbing the wire bundle.
I spoke with a friend who is super smart when it comes to electronics. He pretty much said the same thing you did. It's interference. He said it happens when it gets cold because the electronics work better in the cold and actually get more sensitive to the interference. His suggestion was to use Ferrite beads, so that is my plan.
-=Steve
Not so sure ferrite beads will solve your problem since you are transmitting an 800k signal I think it is.. But I would be curious how that turns out.
You may also have an issue with the wire itself. Perhaps try a jacketed cable with twisted pair for the data. At least the jacket would give you some moisture barrier. Just a thought.
plasmadrive wrote:Not so sure ferrite beads will solve your problem since you are transmitting an 800k signal I think it is.. But I would be curious how that turns out.You may also have an issue with the wire itself. Perhaps try a jacketed cable with twisted pair for the data. At least the jacket would give you some moisture barrier. Just a thought.
Just make sure you only use one of the pair for data. Doubling up the data line is a known issue.
szaske wrote:I spoke with a friend who is super smart when it comes to electronics. He pretty much said the same thing you did. It's interference. He said it happens when it gets cold because the electronics work better in the cold and actually get more sensitive to the interference. His suggestion was to use Ferrite beads, so that is my plan.-=Steve
The threshold of the receiver may be changing with temperature, making it more susceptible to the noise, or the strength of the interference may be increasing. Are you using beads on each wire individually or cores around each set? I would think cores around each set would have more impact.
deonb wrote:I've ordered a bunch of different Ferrite Clamp On Cores from Mouser to arrive on Friday. I'll post here about whether it makes any difference.I don't know anything about how to select a good Ferrite Clamp On Core, so my major selection criteria was: "Does it fit around the wire?" 🙂
I found the solution!
Order some Ferrite Clamps and once they arrive, take the strand off the tree and get ready to install them. This strand that has never worked properly in its life, since the first time I turned it on for testing, and that I've had up and down the tree a few times, plugged into different controllers, rewired from scratch, and having NEVER EVER NOT flickered while being off and not being touched... now that I have the Ferrite Clamps, started working and worked flawlessly throughout the night without so much as a hint of a flicker on any pixel.
The Ferrite Clamps are still unopened in the box...
So it seems like you don't actually need Ferrite Clamps, you just need the threat of Ferrite Clamps.
Sigh.
Yes, you separated the noise receiver from the transmitter, which pretty much confirms it's an interference issue. So, now to get them to play nice when they are in close proximity. Rewiring with shielded cable would likely do it, but try the clamps first. The idea of the clamps is to reduce the amplitude of the aggressor while shielding or space separate it from the receiver, reducing it's influance.
I managed to make mine act up again. Tried the Ferrite Clamps, and they make no difference. Finally just swapped it out.
Here is an update on this. The weather has been fairly warm in Seattle, so my problem has not returned. I ordered Ferrite clamps anyway, and wrapped 2 of my 8 3-core wires around them. I figured it would be a good test anyway. My smart friend (see my previous posts) actually instructed me to only thread the data and ground wires through the ferrite clamps. I'd already clamped all 3 wires, and was too lazy to change it so they stayed as is.
Well the weather got cold, and I came home yesterday and one of the strips that I had clamped was no longer responding. The lights were on full white and NOT changing to the sequences. So I figured it was a good chance to test my friends advice. I went outside, pulled the clamp off, pulled the Voltage/Red wire off and wrapped only the data and ground wires through. One time though, looped around the outside of the clamp and then through a second time. Well after about 10 seconds the lights started working again.
I'm not sure if it's the Ferrite clamps, or magic but didn't have another problem for the last 2 days.
-=Steve
I'm sorry to bounce this old thread, but I was having the EXACT same issue. The weather turned cold (around 0F) and suddenly I had flicker almost EVERYWHERE! It was the worst on the strings I have attached to my pixel extenders. These run off of independent E682s with a balanced output over CAT5. Pretty much every pixel on the extenders was flickering constantly with random colors.
I recently returned from an amateur radio DXpedition so have a load of ferrite clamps - adding numerous mix 31 ferrite clamps to the CAT5, pixel wires, and power wires did the trick and now I only get occasional flickering from random pixels (more ferrite would probably fix this, but I mostly just hope it warms up).
So while the ferrite has addressed the symptoms, I'd prefer to know and fix the root problem, but I'm not exactly sure what the problem is other than it seems to be RFI related and mostly affects the stranded wire.
Any ideas?
