Pixels and power injection questions/problems

KenKol

New elf
Joined
Dec 28, 2011
Messages
4
I have some confusion that I hope can be cleared up by the collective heads on here. I will try to detail what I have going on as best as I can, so please bear with me.


I have my pixels set up for power injection three different ways, but all similar in hook up. the only difference being the pixel form factor.


First off, and one of the two ways I have no issues with, I start with a Sandevices E682. I am only using one output of one grouping. I am not using power to the pixels through the E682, but rather just the signal hooked up. I am only going out around three feet to a "signal In" outdoor three conductor weatherproof connector which goes in, by signal only, to a terminal block. The power comes from a Phoenix contact 12V dc 10 AMP din rail mounted power supply (we have a glut of these from retrofits to CNC machinery and the such). There are three of these power supplies in the box, all coming from the same 120 Volt plug (through a terminal block) Output joins the +volt in, negative and signal into a three pin weatherproof screw plug. From there it goes out on around a 4 foot pigtail to another three pin connector which is connected to the lights. coming back in through another three pin connector where the +volt is "interrupted" but the signal continues to the terminal block to continue the chain and the negative gets collectively connected to a terminal block. This continues 6 times using three power supplies and I have no problems.


Second setup I have is where I use another E682 and each power supply is in its own enclosure but basically the same setup; e682 - signal in - +volt, negative and signal out - light strip - +volt interrupted - Negative and signal continue on to the next power supply and then on to the next strip and so on. Again, no issue with this at all, works like a dream.


The third way, the only thing different is the actual lights. I have everything hooked up the same way as the first example, but when I hook up the lights, I get all on and flickering, some spastic flickering of various lights and generally not behaving when no signal for a test pattern is given. Once a test pattern is directed (all red for instance) it behaves exactly like it should, all red, and no problems. but once you direct no test pattern, it goes back to glitching. I have tried tying all 12 volt negatives together and that does nothing for the issue, I tried disconnecting everything and just using a single power supply and I get the same glitching. I hooked up the pixel strip to see if that would work and it did just fine, so I know it isn't the board. I still have yet to check the power supplies being as the connectors to the light strings are different, but I cannot see how that would affect this problem. The phoenix contact line of trio power supplies are supposed to be filtered and quite nice from what I hear. The first trouble shooting issue I would like to see is if I do indeed have the lights hooked up correctly. I cannot seem to find out any information on these lights, so I traced what I could and, since it lights up red when directed to (and blue and green respectively) I cannot see that my hook up is incorrect at all. I tried shorter leads (only a foot or so) and no change could the 120 volt wires passing by the 12 volt and signal wires cause actually that much disruption and noise to cause this? I will try to re-route them in the mean time to eliminate that possibility, but what else can be causing this?
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
If you have tested other lights in the same situation that use the same chip type and they work well then i would be looking further at the lights because not all lights are the same and in DIY consistancy is sometimes something that suffers.

But a picture says a 1000 words and if you want the best help then drawing a quick picture of your 3 injection methods will help people visualize your setup better to confirm you are injecting power correctly.
 

AAH

I love blinky lights :)
Community project designer
Joined
Dec 27, 2010
Messages
4,190
Location
Eaglehawk
If you have a known working output configuration then I would swap the non-working 1 to that and see if the fault follows the lights or the output. I assume that the 3 methods have all been on different outputs or controllers.
 

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
Could it be a speed/clock rate issue? If everything is the same except for the lights then they may require a different speed.
 

Kitman

Full time elf
Joined
Oct 25, 2014
Messages
395
Location
Munno Para
The E682 has a setting to specify the exact string being used, based on that setting it automatically sets all variables for that exact string based on the chip that you are using.


First thing to check would be to make sure that you have the correct pixel chip selected in the e682 configuration screen as choosing the wrong string could generate weird effects on the lights. I have just assembled one of my e682's and an e6804 and done some testing on my lights and they all come on nicely but I made sure that I had the 2811 selected from the drop down as that is the chip that is on my pixels.


Other than that yes some pictures off your setup and the configuration screen of the e682 or a little diagram would help.
 

KenKol

New elf
Joined
Dec 28, 2011
Messages
4
OK. Progress or maybe just more concise issues. I am leaning more and more on the lights being the culprit or the power supply box I made. I am trying to add pictures...we will see how that goes. below is the dropbox link to my power box.


20150805_105005.jpg

https://www.dropbox.com/s/uk7nfb2ufje1hti/20150805_105005.jpg?dl=0


and here are the pixels and their dropbox link


pixels.jpg

https://www.dropbox.com/s/1g9fztnjm78dyw3/pixels.jpg?dl=0


I swapped out another e682 and am getting the same glitching, so that eliminates the e682 itself, but maybe not the pixel type settings. Can anyone tell from looking at the picture either what these are or where I may have purchased them from (I purchased from the various sources including holiday coro, diyled) If the pictures do not come through, I will try to upload them another way.
 

KenKol

New elf
Joined
Dec 28, 2011
Messages
4
This is the power supply injection box for my pixel strip tree. Three Phoenix Contact Quint 12 volt 10 amp Power Supplies. The strip tree has 12 pixel strips. four strips per power supply. In the picture you can see: incoming 120 volt power distributed through the terminal blocks. In this instance, I combined all +power and -power at the power supply. The signal routes as a signal in through the terminal block - out to strip 1 adding +power - zigzag back down strip 2 - in from strip 2 through the terminal blocks minus the +power - and so on. This way works perfect for the strips. I tried this power supply setup with the problem strips and still am getting the random glitching. This is leading me to believe that the problem lies in the strings. Lead length? could 8 feet of combined pigtail and lead on the strings pose a problem? It almost relies on a signal cooruption issue. I can send test patterns through the e682 and the strings respond and light up accordingly but still has minor glitches with random lights. This points me to the fact that the lights are set up correctly in the e682 (right?). It almost responds similarly to when I had a ground not connected with the pixel strips but when I did that, the strip didn't respond well at all.


20150805_111433.jpg

https://www.dropbox.com/s/lmdfdyfxec5qviz/20150805_111433.jpg?dl=0


Same type of idea, I use this one for my arches where I can put the power at the beginning of every other arch. Same concept. Signal in - +power "interrupted" - signal and ground soldered to each respective wire adding +power and on its way. Again, this works flawlessly for the strips.


20150805_111654.jpg

https://www.dropbox.com/s/4iadlujze8z5uss/20150805_111654.jpg?dl=0


In the last example, where it does not work on the pixels, I have the same basic principle at work on it with the only exception being all the +power and -power being run through the terminal blocks as well as the signal
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
Confirm the lights are good by connecting to another output that you know works with other lights, if it works then you know you have the right chip. If it doesnt then i would then try the different types available as it wont damage the string by using the wrong protocol. This way you can confirm if your lights are good or not.
If it was a power supply then other lights should alos be effected in most cases, again you can eliminate this by just swapping it out.

Once you have done the above you should have a good idea on where to focus
 

KenKol

New elf
Joined
Dec 28, 2011
Messages
4
Well, after ALOT of swapping things around and testing until I have tried every possible combination possible, I have determined it is indeed the lights themselves. I had purchased these a number of years ago and just now got around to making something of them and they are way too unstable to do anything with. They do better straight off the board with the power coming from the board, but the amount I was going to use would have surpassed the amount allowed by the board thus requiring power injecting. If I was going to inject power at any point, I wanted the consistent color and power along the whole pixel tree so I went straight for power injection and did not utilize the board for anything else besides processing. The power supplies all work as I intended them to when I put "other" 2811 pixels on them and they are very stable. In the end, it was all about the firepix.
 
Top