E681 and pixel strips

[abigroot]

I have a few questions regarding my setup. First, let me explain my setup (I think) I have two sandevices E681's. one will be handling my pixel tree using 12 strings of 50 pixels attached to three universes of the E681 which will total 600 pixels being controlled and the second E681 having 4 pixel strips of 150 pixels each being connected to 2 universes totaling 600 pixels again (these I plan on using for arches). My questions come in when I look at things, Am I under utilizing my controllers? according to the manual, each E681 can handle 680 total pixels per unit (not universe). How do most users handle the arches. The strips are @ 16 feet total so either 8-8foot arches or 4-16 foot arches I guess. Also, I am not grouping the pixels at all. I think most of the ccr's that I have seen usually group into threes which would greatly increase how many strips I can use (by my great math skills, 12 strips) am I correct in saying that? I am using "power injection" as well and have not tried a null pixel as of yet, but I seem to get some flickering. Would the null pixel help? my leads from the controller are around 8 feet long and I am using the typical 4 core connectors.on the pixel strings for my pixel tree, how would I make it so I can use all 16 "channels" available? It seems, by count, that I cannot use them all. My tree will be using one string to make two "branches" of the tree (so start at the bottom and go up the angle and back down) As I said, it seems like I should be able to do more with these cards, but the "hard limits" read as such that I cannot. Please let me know what all of you are doing to make these work and work good.

[Shubb]

1st. When you get into these kinds of numbers you are not going to use every single channel.

Compared to the days of 32 or 48 channels you wanted to utilize all of them, but when you're talking about 1200.

2nd. If you use a 42 pixel string you will get the most "bang for your buck"

Scott

[BillyT]

I too have a 12 string x 50 RGB tree. I am using the first three channels of each universe as the fourth one only has 62 channels left. Programming 1800 channels has turned out to be quite interesting

[dougd]

My mega tree is 16 x 40 using a total of 640 pixels on one controller. I wanted to maximize the controller as much as possible. My arches poles and fans use 32 pixels each times 16 outputs for a total of 512 pixels. Didnt maximize that controller as much but wanted to keep things uniform.

[Crazydave]

I am using "power injection" as well and have not tried a null pixel as of yet, but I seem to get some flickering. Would the null pixel help? my leads from the controller are around 8 feet long and I am using the typical 4 core connectors.on the pixel strings for my pixel tree, how would I make it so I can use all 16 "channels" available?

Null pixels are needed when you have long runs from the controller to the 1st active pixel. I'm using 12v 2801 pixels and i have about 36 feet from Controller to 1st pixel.....I used 3 Null pixels at about 10' 20' and 30' for the 2801 they seem to be able to go about 15' at the most.at 18' - 20' with NO null pixel I was having problems with all the pixels staying on when they were supposed to turn off and the the 1st pixel would light up blue when it wasn't supposed to and do funny things.your mileage may vary depending on wires used.. I was using 2 strings of SPT2 to carry power(+), Power (-) , Data , and clock. Edited October 4, 2012 by Crazydave

[Tim Fischer]

I'm having trouble with your math. 12 strings of 50 pixels is 600 pixels, or 1800 channels. each universe on the 68x can handle 510 channels. So you'd need most of 4 universes to handle that, yet you said 3. Then for the second 600 you said 2 universes? Maybe you meant clusters, not universes?

I'm using 2 681-6's (the new upgrade which allows 6 universes) to control 1536 pixels. Total capacity would be 2040 pixels if I had it "fully loaded" so I'm only using about 3/4 capacity myself. If the 6-universe upgrade didn't come out I would have needed a 3rd 681 which would have meant I was barely over half-populated. What's even odder is that since I'm using 128-node strings, I'm only using 6 of the 16 sockets on each board, which makes it look pretty bare

[FBW4]

A universe may not cross a cluster on the E681. So, when you get to 450 (3 strings of 50), you have to go to the next cluster. That leaves one connector unused and the balance between the 450 and 512 unused as well.

Apparently, the 682 solves this problem but I do not have those yet.

[Lee Lunday]

I am using the ECG-P12R from Joshua 1 systems. It handles 12 DMX universes for a total of 6144 channels or 2048 RGB pixels. I setup all my leads with Cat 5e wire used the orange pair for the data and clock then used the other wires for power. twisted the solid color together for the + and the white strip for the -. All my runs are 35 feet and i did not have to use any null pixels.

I am only using 50 pixels on each universe right now 1/2 are 12v and 1/2 are 5v. I know when I expand next year I will have to inject power on anything past 50.

No problems so far.

Lee

[Tim Fischer]

A universe may not cross a cluster on the E681. So, when you get to 450 (3 strings of 50), you have to go to the next cluster. That leaves one connector unused and the balance between the 450 and 512 unused as well.

Apparently, the 682 solves this problem but I do not have those yet.

That's not true. I have 3 strings of 128 pixels on each of 3 clusters on my 681-6. That's 768 channels per cluster, so it definitely spills across universes. Some clusters actually have 3 universes due to starting toward the end of one, using up a second, and going onto a third.

[FBW4]

Hey guys- hate to argue- but no, on my 681, the universe is assigned by cluster. Only one per cluster. What am I missing here?

R

[Crazydave]

I am new to this too so Please correct me if I am wrong but this is my understanding.

The universes that exist on a cluster will be determined by two factors:

1) the universe# and channel starting number

2) the number of pixels to be controlled

So if you have 170 pixels on 1 string on cluster 1-1 then that cluster will use 480 dmx channels on universe1 and channels 481 - 512 will be unused

but if you have 200 pixels on cluster 1-1 then it will use u1-1 to u1-512 and u2-1 to u2-88

[edvas69]

Can a cluster have different length strings on each output within the same cluster as i have had some issues with my E680 controllers with mixing string lengths within the same cluster. It seems not to like an unbalanced load within a cluster. has anyone else experienced this because I have 2 E680s doing the exact same thing

[k6ccc]

In your first example your math is wrong.. If you have 170 pixels on cluster 1-1, then that string will use 510 DMX channels. Maybe you intended to say if you used 160 pixels they it would use 480 DMX channels. You could start cluster 1-2 at universe 1 address 481 and it would roll over to the 2nd universe for the remainder. At least that's the way it works on the E682 and I understand that is the same on the E681.

OK, to be technical, it is actually looking at the first and second socket which in most cases would be assigned to universes 1 and 2 (but don't have to be).

[FBW4]

DUH----- OK. I see what you are saying. Still limited to total number it will handle but the cluster will go more than one universe.

Mea Culpa

[FBW4]

In the beginning almost a year ago, I made the arbitrary decision to use 3 - 50pixel strings per cluster to stay under the 680 pixel limit each S681 had. With 50-pixel strings, the 681 (without what I now am told is an upgrade) can only handle 2040 DMX channels. So if I filled out each cluster with 4 strings of 50, I would exceed that. If I load up the first three clusters and not the last, I have an unbalanced load on my power side which actually doesn't matter to me any longer because of the way I am hitting it with power.

So, I am hearing there is an upgrade to the 681 that lets it handle more pixels?

[Tim Fischer]

Hey guys- hate to argue- but no, on my 681, the universe is assigned by cluster. Only one per cluster. What am I missing here?

R

What you're missing is that you're confusing "socket" and "cluster". Each "socket" is one universe. A cluster may use parts (or all) of one or more sockets.

For example, the E682 (or the E681 with the upgrade) has 6 sockets which are 6 different universes. But there are 4 clusters. Like I said above, I'm using only 3 of the clusters but 5 universes among them all.

The E681 (without upgrade) has 4 sockets and 4 clusters. There's a misconception out there where people think there's a one to one mapping between them, but there's not - as Jeff said above you could have all 4 sockets used up by the first cluster if you wanted (with 4 strings of 170 pixels each).