Wire Type for the Data/Power

[jamestnoe]

I just am getting my permatrack up on the house and have a bit of issue. I ran some of the Pixel runs (250 pixels) close to my FalconF16V4 controller with 18 Gauge stranded wire with no issue. On a longer run of 165 pixels about 50 feet from Controller I was advised to use 18 G 3 wire shielded wire. NOT twisted pair just shielded. When I run tests on the lights they all work BUT the shielded wire flickers and is just different. It does not act the same as the unshielded wire. 
I did the same thing on the carport runs. 1st run of 229 pixels closest to the controller I used the unshielded then the 2nd run further away I use the Shield wire and get the same issues. 

Are we not supposed to EVER use Shielded. I reasoned that the shielded wire would get less interference for the data stream. 
Jim

[algerdes]

Is your problem the shield, or is it the distance to the first pixel (of the second sets) from the controller?  Would an F-Amp or equivalent be in order?

[tbone321]

You simply don't need shielded wire and I would avoid using it for this purpose.  Unless the shielded wires shield is properly terminated, the shielding can actually make things worse.  Instead of protecting the signal wire from outside interference, it can actually reflect and intensify the noise coming from the power and ground conductors into the signal line which may be what you are seeing with your "flicker".  You don't really need twisted pair either as the single data line does not benefit from it.  Just a high quality 18 gauge 3 wire cable is all you should need for 50 feet.

[jamestnoe]

I am not sure what the issue. I don't use the shield wire but it may be that I should try connecting it to the common. What I see is when I send like red fade to the lights the unshielded lights work fine but the shielded lights will be brighter and will have red/white... blinking lights. Always the shielded lights are brighter and have different colors showing. I will build a test string with a short shield wire and see if it is the length or just the shield causing the issue. 
Thanks for the thought process. 
Where would I find the F-Amp (data booster I assume)?

[jamestnoe]

Thanks Tbone321, I am coming to that conclusion. Crap that will be alot of rewire. I made the mistake of thinking I could only put 250 pixels per port. Now I know I can get 1024 pixels. If I have to rewire I will just add the 2nd 162 pixel run to the end of the 1st 250 run. They were just extensions of the 250. Many lessons learned. I would guess I can leave the shielded run there and convert it to power injection ONLY. 

[tbone321]

Not using the shield is exactly the problem as it is not properly terminated and is acting like a giant reflector.  Grounding one side of the shield wire  may clear up some of the issue.  A short shield wire would not have the same effect as a long one so you might not see anything on that one.  You can get the F-Amp from the Falcon store if you wind up needing one.  Keep in mind that power injection also requires a ground wire as well as current flows on both so just using the shield for power does not help with the ground and the shield was not set up for power conduction and may not provide the current carrying capability you may need.

[rayster]

Shield wire is called a drain wire and is not made to carry high current and is not normally considered a good negative wire. In the RS-485 specification you do not use a shielded cable which is what the data we use is considered. So the best wire to use for what we do is just 3 conductor cable. I normally use 18/3 for short runs, under 20 feet to 1st pixel then go up to 16/3 for 50 foot then 14/3 to 70 foot and that is max from a controller. When you add a power injection point to make your string longer make sure you tie the grounds together so you have a good path for the data ground the whole way. The problems I have had is the ground signal has too much loss so that pixels start doing strange things so I add a better ground wire to the board.

[jamestnoe]

Thanks Everyone for the input. I tried the following and finally had to remove the shielded wire completely. 
I tried to ground one end and it did not help at all. 
I moved the data wire to the end of the first segment and used the shield wire as power injection but it stilled caused havoc. 

I finally was able to just connect all wire (data, 12V + and Ground) to the end of the previous runs and all is well now. 

Conclusion: DO NOT EVER USE SHIELDED WIRE. 

Thanks all.