I definitely want to do this the safe way
The problem is, people are conflating the signal integrity on the data line (which has nothing to do with protecting the controller) with fuse protection (which does). I literally run my data lines for 12 feet together on a shared ribbon cable, with no issues. If your run length gets longer than that, data can get dropped unless you do some stuff, but that has nothing to do with frying the controller.
P. Short wrote:Option 2 - is much better - run the V- from the controller to the first pixel. The wire from the controller to the distribution board is not necessary.Neither option is totally to my liking, though. If it were my setup I'd prefer to have complete electrical isolation between the controller and the first pixel for both the data and v- lines between the two. The reason that I say this is to guarantee that current for the pixel strings never flows through the ground wire between the controller and the pixel string. Normally this wouldn't happen because the wire gauges are sufficiently different that the current 'wants' to flow through the thick power leads; however, if one of those leads becomes broken or cut, the heavy ground current to the pixels might take the alternate path through the controller->pixel ground wire. In this situation the pixels would still appear to mostly work, but would be flaky, and you wouldn't know to look for cut/broken main ground leads. You would have a much easier time tracking down the problem if that sneak path did not exist, because entire string(s) would fail, rather than just being flaky.
Phil,
I agree Option 2 without the wire from the distro board to the controller is preferred of the 2 options he showed here, and I follow Martin's recommendation of a separate PS for the controller, but I don't fully understand how you are proposing he do his connections?
tbm878 wrote:OK, I guess I had misunderstood a couple things.So what would be the best way to do this with the least amount of cables needed? My control box is going to be approximately 20' from the closest pixel. The control box will be mounted in the closet then connects to wire trough (red box in pic). With all of the matrix's and lights I'm planning to use, this had me running 53 18-3 cables for V+ & V- & D, then another 51 18-3 cables for power injection only connected at the end of the 200 pixel strings, not using the data wire in that pi cable. If i need to run a separate V- wire from the controller i would need to run an additional 53 cables to the 1st pixel.
Believe me, i want to do this right the first time, i am not opposed to making changes now to make for a better installation. I am an electrician, so i can install whatever setup i need to. I just dont have the electronics backgrounds that some of you guys have. I understand the theory, but some of the stuff goes over my head.
Another option. Since the concensus is that i should run a Data AND V- from the controller, can i use the unused #18 wire in one of the 18-3 power injection only cables and use this for that controller V- wire? Or with it being run in the same multi-conductor cable as the power injection V+ and V- would that cause too much or any interference to not work?
I am changing course a bit to try to do it this way....not using V+ from controller, because that's what some of the more knowledgeable users have suggested. I was originally going to power everything from the controller and power inject at the end. And cut the V+ wire after pixel 50 or 100 so the controller doesn't have too many amps running thru it. I could then power inject the rest. Is that any better of a setup?
I'm starting to get a headache!
You really want the data and V- from the controller to be in a thinner wire that is twisted around itself if you are going any distance over 15'. I suggest you use a CAT5 or CAT6 cable for your data / V- pairs. You can run 4 pairs in a CATx cable. That means your cable can make a few stops along the way to drop signals to multiple ports. A CATx cable can easily get you 40' from a controller and even further with a data buffer. If you are using a controller that supports remotes then you can use a single cable to a distro point and then shorter cables to nearby props. I personally have multiple controllers around the yard to keep the pixel data runs short and I have a few up on the house where I use the smart receivers.
1pet2_9 wrote:The problem is, people are conflating the signal integrity on the data line (which has nothing to do with protecting the controller) with fuse protection (which does). I literally run my data lines for 12 feet together on a shared ribbon cable, with no issues. If your run length gets longer than that, data can get dropped unless you do some stuff, but that has nothing to do with frying the controller.
Agreed. We are discussing two topics here. I was describing why I went with a separate PSU when I mentioned fusing. The separate PSU is an offshoot of the signal integrity discussion in which you want to limit the current on the controller V-. Using a separate PSU further reduces the chances that there is any current going through the signal reference wire. Having a dedicated controller PSU might be overkill, but I like the setup.
As for safety, If you really want to be safe when you are doing power injection, you make sure every current path (V+ and V-) has a fuse that is rated to protect the smallest wire that is directly attached to the fuse. If you have a distro board then you are mostly doing just that. However, the distro boards I have seen ASSUME that if you have V+ fused then V- is protected on the same circuit. If you attach multiple outputs from the distro board to the same string of pixels and you forget to cut the V+ between the injection points then it is still possible (not very probable) to overload the V-.
tbm878 wrote:I'm starting to get a headache!
Me, too.
Without knowing what each those 53 18/3 cables are for, it's hard to know what to say.
