50 Watt LED's Not Behaving with Detail

[MSB]

We've been using these style of lights for years and have had our issues. Some were self inflicted but we really want to make an effort this off season to figure out the problem. Here's a link to the video that illustrates our issue. When they work, it is one of our most beautiful displays. They are very effective lighting up a stage that is very wooded and even looks better when there is snow.

The stage is called Northern Lights. We have 32-50w LED's on the Drivers Side and 34-50w LED's on the Passenger side of the road spaced approximately 20 feet apart. They are wired light to light with TRU-cable Cat5 direct burial cable eliminating the connections exposed to the elements. We use the T568B wiring method and each jumper cable is tested with a FLUKE Cable IQ tester before installation.

We have an Aux A network  set to 500 Enhanced feeding the driver's side starting at the top of the hill and a Separate Aux B network set to 500 enhanced mode feeding the passenger side also starting at the top of the hill. Both networks are terminated with a 120 ohm resistor at the bottom of the hill at the last LED on each network.

The distances from the PC that runs the show is in the center of the park but the park is quite large and we have no other issues with flaky controllers and running much longer distances. This year we did an infrastructure upgrade with direct burial cat 6 cable. We tested every segment after termination with the Cable IQ tester. We also have an APC ProtectNet (PNET1GB) at the end of every network before entering the LOR UART controller. The distance from the show PC to the start of the driver's side LED's (Aux A) is 1408 feet and the distance from the show PC to the start of the passenger side LED's is 985 feet. We are using LOR S5 version 5.6.8 pro

The sequence programmed for the video is alternating Red 0.75 sec, Green 0.75 sec, Blue 0.75 sec with no fades, strobe, nor motion effects duplicated on both sides of the road. The video shows there is better performance of the lights on the driver's side whereas the passenger side is very flaky. We did replace some of the control boards this year and the hardware scan from the show PC did register every light on both networks. They are assigned unique sequential UID's starting at 1-20 on the driver's side and 1-22 on the passenger side.

We did purchase a Siglent SDS1104X-E 100Mhz digital oscilloscope 4 channels standard decoder and our plan is to get into the weeds to try to figure out how we can fix this problem during the off season and looking for help. I did make a breakout board that has test points on pins 4 and 5. I have the scope setup using 2 channels with one channel inverted and not using the ground on the probes. I can see the heartbeat on the scope. I will add that picture in a subsequent post as I left the USB drive in the scope. 

[MSB]

This is a scan of the heatbeat on the scope.  This is the schematic of the break out board I made to measure the signals. I connected the channel 1 probe to Test Point 1 for Pin 4 and channel 2 probe to Test Point 3 for Pin 5.

Edited January 19, 2022 by MSB

[TheDucks]

26 minutes ago, MSB said:

This is a scan of the heatbeat on the scope. 

That looks fine . The waveform is Squares. with 32 devices (not even a loaded Pixie2) , you could try 128K

Is this done at the Far end (near terminator)? What about Mid way? Speaking of Termination: are you positive you are not double(or more) terminated?

Be absolutely sure  you do not have a 568A at one end and 568B at the other end of a section (AKA a Crossover Ethernet) . the main pair (4+5) is the same for either, BUT LOR Aux power is on the Green (568B) pair. That could get ugly

Is the direct burial CAT5/6 near the AC run (assumes direct burial or PVC). Speaking of AC run. Is it clean (is the circuit shared with  sodium or Mercury vapor lighting)?

Have you exchanged the Blinky nodes with good ones?

What happens if you break the run in half (move the terminator to the current last flood)? Does it behave?

What is odd, is you have 2 isolated runs (networks), that both act up. They are way under 4K ft. To me that says bad Power or damaged floods (water ingress)

BTW the spec for the APC says Ethernet Only which does not use 4+5 (the one you want protected). You should look at the Elk Products 592 which is for Alarm telephone line RJ31  (Alarm service override the tel line use). It has a 8pin jack

[k6ccc]

26 minutes ago, TheDucks said:

with 32 devices (not even a loaded Pixie2) , you could try 128K

If those floods are the only things on those two networks, drop them down to 56K or even 19K.  That is only 96 and 102 channels being controlled - and I assume it's not going to be massive fast transitions.  Speed is not going to be an issue.

 

[TheDucks]

22 minutes ago, k6ccc said:

If those floods are the only things on those two networks, drop them down to 56K or even 19K.  That is only 96 and 102 channels being controlled - and I assume it's not going to be massive fast transitions.  Speed is not going to be an issue.

 

@jim Did I miss any other  WAG's ?

Got to love it when most of the facts are in the first post 👍

[MSB]

38 minutes ago, k6ccc said:

If those floods are the only things on those two networks, drop them down to 56K or even 19K.  That is only 96 and 102 channels being controlled - and I assume it's not going to be massive fast transitions.  Speed is not going to be an issue.

 

I do have a question regarding setting up the serial ports on the PC. Is it necessary to also change the port speeds using the device manager from the default 9600-N-8-1 if the network speeds are changed using the hardware manager? 

Also, to be clear, the wave form I posted earlier is only with 4 lights connected. This is a whole new level of debugging so we haven't hooked up all of the CR50's yet. I made the breakout board as an easy and safe way to test the CR50's. The LED's on the breakout board do blink every 1/2 second with the heartbeat signal so I should be able to use that to check for signal in the field.  I did check the resistance across the high speed UART and it is 120 ohms across pins 4 & 5 which according to the RS485 spec there needs to be a termination on each end of the data line and yes, the termination on the far end is made using a 1/4 watt resistor across pins 4&5 plugged into the last light in the chain. Without the resistor and the new CAT6 cabling that was installed we were experiencing very significant problems with the signal. Some A/C controllers on other networks would even quit responding. The terminating resistor was the cure.

Many of our CR50's do not have a replaceable UART chip but we did purchase some replacement boards. That seems to help most of the time but not 100%.

I will update this thread as I get more data. My plan to to start plugging in the lights in order and checking the signal as we move down the line also moving the terminating resistor as we go. it will take some time to get this done. Just a question...was everyone able to click on the links successfully? I've had problems in the past with my sharing settings on google drive.

Thanks for all your comments and the heads up with the APC!

[TheDucks]

No! The network utility setting prevails. BTW HU does its own thing and does not run at the set speed.

You are using LOR HS USB dongles?  Is there a large number of Pixels that you need Enhanced mode (also requires latest 1.05 FW in the floods AND any other device must be enhanced capable)?

You may have a non-enhanced capable device on the network. (Normal is Bi-directional, so some of the blinking may be unwanted network chatter) IMHO unless you have thousands of channels on a network, speed should be enough  (not keeping up is the symptom )

[MSB]

52 minutes ago, TheDucks said:

No! The network utility setting prevails. BTW HU does its own thing and does not run at the set speed.

You are using LOR HS USB dongles?  Is there a large number of Pixels that you need Enhanced mode (also requires latest 1.05 FW in the floods AND any other device must be enhanced capable)?

You may have a non-enhanced capable device on the network. (Normal is Bi-directional, so some of the blinking may be unwanted network chatter) IMHO unless you have thousands of channels on a network, speed should be enough  (not keeping up is the symptom )

We need the enhanced network to take advantage of the motion control programming available in S5. It's a very helpful programming tool for many of our displays, even the dumb A/C controllers.  We typically use the motion control programming for the CR50's, but for this test I wanted to be sure that wasn't causing the problem and why we manually programed the lights for the video. If you look at our show (link in signature) we have a very unique setup and displays like our Whoville and Elf Choir conductor work well with vertical stack motion effects. We are using all HS dongles, even for our regular network that does not use any motion effects and mainly used for our animation loop. We don't have any non-enhanced devices on any enhanced networks. Good thought, though. 

Since RS485 is not a protocol, I've tried looking for documentation on the LOR protocol that runs on top of the RS485 network but have not been able to find much but have learned much about RS485. We did introduce an E1.31 network this year specifically for the 6 Pixiecon16 controllers. We had to add a SACN device on that network to run 2 of the Pixie2 controllers that are on top of our 48 x 100 360 mega tree. This is by far the best performance we've ever had with the mega tree but that is a totally different topic.

[dibblejr]

38 minutes ago, MSB said:

We need the enhanced network to take advantage of the motion control programming available in S5. It's a very helpful programming tool for many of our displays, even the dumb A/C controllers.  We typically use the motion control programming for the CR50's, but for this test I wanted to be sure that wasn't causing the problem and why we manually programed the lights for the video. If you look at our show (link in signature) we have a very unique setup and displays like our Whoville and Elf Choir conductor work well with vertical stack motion effects. We are using all HS dongles, even for our regular network that does not use any motion effects and mainly used for our animation loop. We don't have any non-enhanced devices on any enhanced networks. Good thought, though. 

Since RS485 is not a protocol, I've tried looking for documentation on the LOR protocol that runs on top of the RS485 network but have not been able to find much but have learned much about RS485. We did introduce an E1.31 network this year specifically for the 6 Pixiecon16 controllers. We had to add a SACN device on that network to run 2 of the Pixie2 controllers that are on top of our 48 x 100 360 mega tree. This is by far the best performance we've ever had with the mega tree but that is a totally different topic.

What's a pixiecon16?  

RS485 is a protocol.

JR

[k6ccc]

14 minutes ago, dibblejr said:

RS485 is a protocol.

RS-485 is an electrical standard.  LOR and DMX are protocols that are commonly carried over RS-485.  There are others.

 

[TheDucks]

12 hours ago, MSB said:

We need the enhanced network to take advantage of the motion control programming available in S5. It's a very helpful programming tool for many of our displays, even the dumb A/C controllers.  We typically use the motion control programming for the CR50's, but for this test I wanted to be sure that wasn't causing the problem and why we manually programed the lights for the video. If you look at our show (link in signature) we have a very unique setup and displays like our Whoville and Elf Choir conductor work well with vertical stack motion effects. We are using all HS dongles, even for our regular network that does not use any motion effects and mainly used for our animation loop. We don't have any non-enhanced devices on any enhanced networks. Good thought, though. 

Since RS485 is not a protocol, I've tried looking for documentation on the LOR protocol that runs on top of the RS485 network but have not been able to find much but have learned much about RS485. We did introduce an E1.31 network this year specifically for the 6 Pixiecon16 controllers. We had to add a SACN device on that network to run 2 of the Pixie2 controllers that are on top of our 48 x 100 360 mega tree. This is by far the best performance we've ever had with the mega tree but that is a totally different topic.

As others have said: RS 485 is an electrical standard. It's claim to fame is Mult-drop and distance. It is a VERY common protocol for Industrial and security controls.

Each Network can be set to Be 'Enhanced" independent of the others.  Also, you CAN NOT use the Trigger inputs (on most controllers, including your floods) on enhanced networks (they need to talk back). They must be on a normal network.

[dibblejr]

11 hours ago, k6ccc said:

RS-485 is an electrical standard.  LOR and DMX are protocols that are commonly carried over RS-485.  There are others.

 

Someone must have forgotten to inform BP, Exxon and Texaco that their POS systems do not run on the RS485 protocol that the manuals and the large hub mounted on the wall says. My in-laws service station has used the same system through the 3 brands.

JR