Author Topic: wire size  (Read 10858 times)

Offline jnealand

  • Hero Member
  • *****
  • Posts: 1421
    • View Profile
Re: wire size
« Reply #30 on: January 18, 2016, 05:56:03 PM »
If I did not run data and power in the same wire, my display would not be working.  I have many props connected using cat5 and also many connected with 10ft or more of 3pin waterproof extension cables.    I am using hubs with my controllers (uSC and Lynx SSC) placed right at the prop.  6 wires total for power and 2 for data using the cat5.  Works like a champ.  That's my real life experience.   Been doing it this way for 4 years.  Been doing animated lighting for 8 years.
Jim Nealand
Kennesaw, GA

Offline babybear

  • Hero Member
  • *****
  • Posts: 540
  • Here to make everyones show better
    • View Profile
Re: wire size
« Reply #31 on: January 18, 2016, 06:11:40 PM »
I'm thinking a lot of my data issues were two things The length of my run with E682 and the amount of rain and snow. next season with F16v2 my issues will disappear. Im hoping... 
JimmyG
Rochester, New York

Offline pixelpuppy

  • Jr. Member
  • **
  • Posts: 73
    • View Profile
Re: wire size
« Reply #32 on: January 20, 2016, 06:50:24 PM »
There is a bit of speculation and hearsay bouncing around in this thread.   Here's my take on a few of the things I see...

Just realize cat5 has a low current limit, even if you double/triple up.
Not sure what you heard, but you might be confusing voltage with amperage.  Cat5 has no lower current limit than any other 24awg wire and it *CAN* carry more current when you double/triple up. 

What makes Cat-5 different when it comes to power is the conductors have very thin insulation so it has a low VOLTAGE limit - even if you double/triple up it is limited to about 100 volts max.  But each conductor can carry about 0.5A and if you bond two conductors together then it can carry 1A.  Three together can carry 1.5A ... as long as you stay under about 50 WATTS.


As for using Cat5 for data are you double that up to? I was always told thats a no no as when the same data goes down two different wires and one hits a little resistance the data would get there at different times and cause issues. 
This might be true it two independent data lines took two completely different paths.  But with a twisted pair, they are the same resistance and the same length and the same path.  Bonding the pair as one conductor will have less total resistance and less voltage loss.

I don't know what voltage runs down the line either.  I assume there has to be a threshold that is crossed to indicate a zero or one so I was assuming using double wires would lower total resistance and allow you to have a longer run before going over that threshold is not possible. 
The voltage on the data line is 5 volts.  This is true for both 5v AND 12v pixel strings.  Thats because the 2811 chip is a 5v chip and it generates the data out signal.  12v strings drop the 12v down to 5v to power the 2811 chip and the chip can only output a 5v data signal.

You are correct that double wires will lower total resistance.  However it will increase the capacitance.  Since data is a modulated signal, its a bit more complicated to determine if the extra capacitance will hurt the signal more than the lower resistance will help the signal.  This is where field testing becomes crucial.  However, at these speeds (~800Khz) and relatively short distances (<100m) its probably a wash to use one conductor or two for the data line. 

I prefer to use one conductor for the data signal and use the other conductor in the same twisted pair as ground for the data.  Its NOT because the ground "shields" the data as some people think.  I do it this way because for a single-ended data signal, ground is the reference.  Carrying ground in the same twisted pair as the data helps make ensure same ground reference at the receiving end as I do at the sending end.

But the bottom line is that in this application either way probably yields about the same results.

My brother who was a data expert I thought had told me at one time Data only runs on the outside of a wire thats why data lines are solid and not stranded. I have to find that out..
This is a big stretch of the truth.  Its true that higher frequency signals tend to run on the "skin" of the conductor rather than throught the "core" of the conductor.   But this doesn't really apply to our relatively slow data rates of less than 1Mhz.

Most people are running WS2811 chips at 800KHz so our cable lengths are not long enough for using 2 conductors to make enough of an error to matter.
Exactly!

I was assuming the data would be limited to 100m like you normally see on cat5.  I don't know either way, it was just an assumption.
Again, the 100m limit for Cat5 has to do with the high speed data signaling used for two-way ETHERNET.  The 100m limit does not apply for low speed one-way serial signaling that we are using.  But as already mentioned, resistance and capacitance will ultimately determine how far the signal can be carried before it can no longer be reconized by the receiver.  But its not necessarily limited to the 100m limit for other Cat5 applications.

Offline babybear

  • Hero Member
  • *****
  • Posts: 540
  • Here to make everyones show better
    • View Profile
Re: wire size
« Reply #33 on: January 20, 2016, 07:28:58 PM »
LOL.  How did we get so far off my original post......   All I was asking in the building of my snowflake would it be better to soider up strip light with 24gauge solid wire than 18 gauge stranded. No more than 6 nodes 18 LEDs for power and 40 nodes for data. 
  But still Pixelpuppy thank you for all the info. The more I learn the more I can help others.  Your response was not wasted.
JimmyG
Rochester, New York