RGB smart pixel mini trees

[mbarber21]

Having trouble running my smart pixel mini trees in a daisy chain. Each tree has 16 pixels on it, trying to daisy chain 8 together. I have power injection in the line, but will only light up the first tree and a couple of pixels of the next. Trees are 8' apart. Any suggestions.

[Ebuechner]

2 thoughts either you exceeded the 512 channels of DMX. Or you're having problems with the signal. If there's too much wire between pixels sometimes you have to put in a couple of null pixels. And power injection is great but you have to make sure you have enough voltage. At the pixels test to see what you have there. If you have long wires you might start off at 12 volts but you don't end there. But your problem might be the signal. Your controller sends out a really nice strong signal to the pixels. From there each re transmits the signal on to the next pixel. This signal isn't as strong in can't be sent as far. In a case like that you have to put dummy pixels in the line to make the distance.

[k6ccc]

What kind of pixels? 

[mbarber21]

Pixels are 2811 bullets. I should add that the first tree is about 30' away from controller.

[Ebuechner]

if you use heavier wire you can get a little more distance and less voltage drop. I buy the 3 prong green extension cords when they come on sale and use that wire most of them are 16 gauge. It beats the 18 or 20 gauge wires that most premade extensions are. I cut the ends off and solder on the three prong waterproof connectors for the pixels. I build my props with the extensions already built in. In your case my first tree would have a 30' cord coming off of it to plug into the controller. My second tree would have and 8' cord on it to plug into the first tree. Also if you tie the positive and negative together from the first pixel to the last pixel on the tree you will get less voltage drop. (But not the data wires)

Edited November 19, 2015 by Ebuechner

[Ron Boyd]

My first thought is power, but you said you are using power injection. Double check the channel config, network config.

 

Going by your stated 16 pixels times 8 trees, there are 128 pixels or 384 channels. (I have the exact setup but in a circle form instead of mini trees). That's less than one universe so that should not be the problem. Your channel config must have all 384 channels and the network config should match Universe # in the channel config.

 

One other thing to try, as stated above, you might need to add a null pixel in line about half way there. I don't really think that's the culprit, since the first few are working properly. It's something you can try though. Just be sure your pixel controller reflects that there is a null there. Otherwise, it will think the null pixel is your first pixel.

 

To recap,

1. Check channel config
2. check network config
3. check that controller matches up with the first two, and lastly,
4. try a null pixel at about 15 feet

Oh, it's preferable, in my experience to have the power as close to the element as possible. At 30', you may have power loss, whether it's 12v pixels or 5v pixels. The 5v pixels will have drop-off more so than the 12v, but if you're not using a heavy enough gauge wire, you could have some power degradation. Stranded wire is better than solid wire, in my opinion.

Edited November 19, 2015 by Ron Boyd

[Ebuechner]

He says his first tree is lighting up and working fine. The problem comes in with the second tree. I ran into the exact same thing with the distances being almost the same. To solve my problem I had to install 2 null pixels(about every 5') between the two elements. If the first tree is working fine you don't need null pixels there. Are the pixels on the second tree that are lighting up working perfectly fine or are they flickering and acting funny. If the pixels are acting up then you need null pixels between the two trees. But if they're working perfectly with your sequence and they just simply stop and it's consistent I would look at your setup

[tmeyermn]

Are they all in the same universe and do they have consecutive channel numbers?  

[mbarber21]

The pixels on the second tree are flickering. All channel and network configs are correct. Right now I am testing off the controller(alphapix4). I added a couple of null pixels between the controller and first tree now the first tree does not work except for a couple of pixels flickering. Going back to the way I had it and try null pixels between trees. Just not to sure how to configure board for that.

[Ebuechner]

As I said before in my post you do not need null pixels between the controller and the first tree. You would need them between the first and second tree. And this is assuming that your voltage is correct. You will also need null pixels between all of the trees after that.

[tmeyermn]

The pixels on the second tree are flickering. All channel and network configs are correct. Right now I am testing off the controller(alphapix4). I added a couple of null pixels between the controller and first tree now the first tree does not work except for a couple of pixels flickering. Going back to the way I had it and try null pixels between trees. Just not to sure how to configure board for that.

 

I had 3 minitrees with 100 pixels on each of them (power injected at the end of the 3rd tree) and the 3rd one was flickering.  I am using an AlphaPix 4 as well.  I spoke with HC and they had me try reversing the order of the trees (swap tree 1 with tree 3) and see if the problem follows the output or the tree.  I would try swapping tree 1 and tree 2 and see what happens.

 

In my case, it turned out to be some bad soldering because when I swapped them everything worked just fine.

[Ebuechner]

I'm going to add this. The problem you're having is due to the length of the wire between the trees. So when you're adding the null pixels do not increase the length of the wire. If it was me to be on the safe side I would add 2 null pixels in between each tree evenly spaced. I'll explain a little each pixel has to resend the signal to the next one. But it can't send it that far. The other part of this is if you let the signal go bad the pixels will be sending the signal on with corrupt data

[Ron Boyd]

I'm going to add this. The problem you're having is due to the length of the wire between the trees. So when you're adding the null pixels do not increase the length of the wire. If it was me to be on the safe side I would add 2 null pixels in between each tree evenly spaced. I'll explain a little each pixel has to resend the signal to the next one. But it can't send it that far. The other part of this is if you let the signal go bad the pixels will be sending the signal on with corrupt data

Sorry, but I have to disagree with this statement. The OP said 30' to the first tree and then evenly spaced 8' after that. If the controller will send the signal 30', without a null pixel, why would it not send it 8'? The signal is re-generated at each pixel in the string. The first pixel strips off it's 3 channels worth of data and sends the rest to the next one in line. The signal is re-generated 16 times at the first tree, so the data should be fresh, 8' later. 

 

Now, if the problem is, that the signal is weak at the first tree, maybe, but the null pixel needs to go somewhere in the  first 30'.

 

That said, In my setup last year I had about 20-25 feet from controller to the first pixel in a 136 pixel strip with 17 pixels, 2' of wire, 17 pixels, 2 feet of wire, etc. In the picture below, it's the 2 bursts up high against the house. The controller was at the foot of the center burst. I added V+, Data and ground from controller to a null pixel at about 12', carried the data and ground only to the first pixel in line, and power injected at the first pixel. It worked like a charm.

 

What gauge of wire are you using and is it stranded? If it's stranded, is it a lot of fine wires or a few thicker wires. A couple of years ago at our mini, we tested and discovered the power and data traveled further on fine, multi strand wire as opposed to a thicker, fewer strands.  

[Ebuechner]

I ran into the exact same thing with my display. It was driving me crazy until I talked to somebody who told me roughly the same thing. Now my setup has 2 null pixels between my mini arches which are roughly 10 to 13 feet apart. And I have 30 feet of wire from the controller to the first arch without any null pixels. Once I did that everything worked perfectly and I'm not using any power injection. I'm just working off of my own experience (your results may differ) but there so many other things we don't know like what type and gauge wire is he using. What sort of connector is he running this through has he tested the voltage at different points. And can I save 15% or more with Geico. Hopefully he figures this out soon he is running out of time. With me my elements were working fine until I put an extension wire in between them then all hell broke loose. I tried different gauges of wire power injection and a few other things. And just shortly before I resorted to selling my soul I tried the null pixels and life was good again.

Edited November 20, 2015 by Ebuechner

[mbarber21]

Gauge of wire is 18/4AWG shielded alarm wire. Power is constant @11.92v. Just ordered some uAmps instead of wasting pixels. Hopefully this will solve the problem. Thanks to everyone with all the suggestions. It is very much appreciated.

[Ebuechner]

Here I learned something. I didn't know about the uAmp. I like that idea then you don't have to accommodate it in your programming. I just read an article on it and it pretty much confirms everything I was telling you. http://falconchristmas.com/forum/index.php?topic=2115.0 https://youtu.be/lp8QYcV57bo. This is good info

Edited November 20, 2015 by Ebuechner

[wmilkie]

Here I learned something. I didn't know about the uAmp. I like that idea then you don't have to accommodate it in your programming. I just read an article on it and it pretty much confirms everything I was telling you. http://falconchristmas.com/forum/index.php?topic=2115.0 https://youtu.be/lp8QYcV57bo. This is good info

 cool device

That's sort of what a null pixel does, regenerates the data;

I'm using 5v ws2811 pixels with no signal generation(no null pixels, no extra power supply); for long lengths I've been able to use as much as 30 feet 18awg wire with only about 0.25v average drop; no visible loss of signal at actual lights

I set up some different wire lengths inside throughout the year to test this but I didn't put a scope on it to check what the signal looks like, next year

 

would be interesting , using an oscilloscope, to see the before and after signals using null pixels 

[k6ccc]

Gauge of wire is 18/4AWG shielded alarm wire. Power is constant @11.92v. Just ordered some uAmps instead of wasting pixels. Hopefully this will solve the problem. Thanks to everyone with all the suggestions. It is very much appreciated.

 

The only mention of a price on the uAmp in the referenced thread (and that they hadn't actually shipped) was $4 each.  So you are spending $4 to avoid using a 30 cent pixel.  Other than the uAmp not needing to be accounted for in programming like a null pixel would be in this case, what are you getting for the extra $3.70?

 

Am I missing something here?

[Ebuechner]

I want that scope. All I have is a really old B&K 1471 dual trace.

[wmilkie]

I want that scope. All I have is a really old B&K 1471 dual trace.

I had a b&k, but don't have any at home anymore. First scope was an old 19"rack Dumont, single channel, weighed 50lbs

The lab I work in, we have tons of Tektronix ones if I need one

 

Jim's got a point, a null pixel is a much cheaper solution; maybe those uAmp's have other uses?