Author Topic: Any limitations with RPIWS281X?  (Read 1302 times)

Offline rcowan

  • Jr. Member
  • **
  • Join Date: Jan 2016
  • Location:
  • Posts: 82
  • Kudos: 1
Any limitations with RPIWS281X?
« on: April 13, 2016, 12:38:31 PM »
Since my lights and other hardware have started to trickle in, I of course wanted to see some blinky-blinky. So I configured the RPIWS281X output type, connected my lights using only about 3' of 3-core waterproof connectors to the GPIO, powered them separately and connected the grounds together. Everything worked as expected. So then I decided to start testing out the max distance I could achieve using coax. RG59 to be exact. That's where the problems started. I had read that some people were getting up to 300' using coax. So I thought surely my measly 20' of coax shouldn't be a problem. However, my tests are as follows:
Fill White at 255 brightness: Erratic behavior of the pixels at the start and then it tapers off after about 20 seconds. After that the lights stay solid as expected.
Fill White at 128 brightness: Works as expected. No erratic behavior.
Chase RGB at 255 brightness: Works as expected. No erratic behavior.

A little bit more about my setup. While the data was travelling along the 20' of coax, the power was only running about 4.5'. I did this to eliminate any issues of power drop over a longer cable. I wanted to test the components individually. Also, I tried increasing and decreasing the voltage level of the PSU just to make sure there wasn't an issue there as I read somewhere that people got better results out of some of their pixels at 11.8 volts.

So my question is, is the RPi just ill suited for longer data runs? Will there be a difference when I try this with a real controller like the F16v2?

I should also mention that the coax hookup was not direct from the Pi to the lights. There were a couple of 1.5' pigtails to actually connect the coax to the Pi and the lights. They were the same pigtails that I used when doing my initial test without the coax in between. Also, because I am swapping out connections for testing purposes, the coax wasn't actually soldered to anything. The connections were made temporarily using wire nuts. The pigtail to the lights though is soldered.
Rick Cowan

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,238
  • Kudos: 126
    • Granbury Christmas Lights
Re: Any limitations with RPIWS281X?
« Reply #1 on: April 13, 2016, 01:36:58 PM »
The Pi's output isn't great for long distance to first pixel.
First, it is 3.3v output instead of the expected 5v data.
Second, it isn't specifically impedance matched for distance.

Best bet is keep it short or use a Pixel Extender (like the uAmp). The uAmp will level shift the signal up to 5vdc and optimize the signal for very long distance.

Offline rcowan

  • Jr. Member
  • **
  • Join Date: Jan 2016
  • Location:
  • Posts: 82
  • Kudos: 1
Re: Any limitations with RPIWS281X?
« Reply #2 on: April 13, 2016, 02:00:18 PM »
Thanks Jon. I don't plan on using the Pi's outputs for my show. I have 3 F16v2's on order. I was just using it to test my lights with until my F16v2's come in. Glad to hear though that there will be a difference between the Pi's output and the big boy controller. Gives me hope that what I am seeing is just a limitation on the Pi and not indicative of how things are going to be with the F16v2.

Offline mararunr

  • Hero Member
  • *****
  • Join Date: Dec 2013
  • Location:
  • Posts: 606
  • Kudos: 14
    • Bentonville Heart Lites
Re: Any limitations with RPIWS281X?
« Reply #3 on: April 13, 2016, 02:13:23 PM »
I can vouch for the F16v2 running 60 feet to first pixel over Cat5 wiring without a pixel extender/uAmp.  I've never done coax and don't have a need to go farther than 60 feet since I own multiple F16v2s.
Bentonville Heart Lites (www.facebook.com/bheartlites)
 This is just my opinion/suggestion/viewpoint.  Others with other viewpoints/experiences may have different advice.  I am a hobbyist with a couple years real world experience, not an expert.

Offline rcowan

  • Jr. Member
  • **
  • Join Date: Jan 2016
  • Location:
  • Posts: 82
  • Kudos: 1
Re: Any limitations with RPIWS281X?
« Reply #4 on: April 13, 2016, 02:40:29 PM »
Thanks guys. You've put my mind at ease that my F16v2 will perform better than what I am seeing from the GPIO of the Pi.

I do have a follow up question. Using standard 3 core extensions I've read that 10' is the safe bet. Some people have managed to get a little farther. My question is, does that 10' include the pigtails? If I have 1.5' pigtail coming off of the controller and another 1.5' pigtail coming off of the lights then that will add 3' to the 10' extension giving 13' total distance between the controller and first pixel. I ask because there are some elements that are well within that 10' that I would like to use the already made up 10' extensions rather than creating up my own. What experience do people have in this scenario? My gut tells me that the extra 3' shouldn't be a problem. Just want to make sure before I order 16 of the 10' extensions for my mega tree. At $2.50 a pop I'd rather just buy them already made instead of sitting there and soldering all 16 of them.

Offline AAH

  • Hero Member
  • *****
  • Join Date: Jul 2014
  • Location: Australia
  • Posts: 727
  • Kudos: 19
  • Blinky blinky blinky
    • I love blinky lights
Re: Any limitations with RPIWS281X?
« Reply #5 on: April 14, 2016, 04:45:48 AM »
I personally have had 10m (33ft for those in backwards parts of the world) from a J1 Sys P2 that I swapped the output resistors from 300 ohm to 33 ohm which is what it really should be based on WS2811 specs. Providing you drive the 2811 line with 5V logic and preferably with a 20-50 ohm resistor in series you should have no issue getting 10m. The potential problems are that the output direct from a Pi is 3.3V logic, there is no output resistor and voltage drop can be a problem. If you are using ready made 3 or 4 core leads from Ray Wu or similar that use 0.75mm2 cable you shouldn't have a voltage drop problem.

 

Back to top