Anyone ever run into this? LEDs/DMX burst... Flourescent EMF?

Okay, I don't even know where to begin to describe this problem or even if this is the right thread to post in, but basically we have an Express 48/96 console on a Net2 system (three 4-port nodes, and two 2-port nodes). Everything was working fine throughout tech a week ago, then one day I came in and the LED's that are plugged into the nodes were randomly bursting (ie a channel blitzes on for a milisecond, say reds from SL box boom, then maybe 10 seconds nothing, and a burst from the SL dance boom, green, then maybe 30 seconds nothing, and a whole other set will burst). It has gotten to the point of flashing all over the place, like a rock show, at times, but more often there is lag time between the bursts.

Now, the oddest part is that there are times when the theatre is dark, nothing is on, just power to these errant LED's and nodes..... and everything is calm, no bursting. When we switch on the flourescent worklights, the LEDS then started to burst, turn off the works, and they settle down and can go minutes with nothing. 

We have tried, maybe too many things to list, trying to figure out this possible correlation. The thing is that the LED's had also done this during shows, when the work lights were off. 

The one thing is that when the console was plugged at the tech table in the house we had no problems. It really only started when it got moved back to the booth, and the LED's were left on overnight. 

We've since discovered that the DMX line from the booth seems to run to the back of the theatre in the rack, then ties into to the DMX that goes to the mid-house and back. 

The DMX from SR isn't even connected (is this really an ETC approved job?)

So, we've disconnected the mid-house DMX from the system thinking that the DMX run back and forth was too long and degrading the signal. It's seems to be better, though every so often one of the LED's gives a little "shout". 

The questions in all of this foremost are; would the fluorescent works light really be interfering with the DMX, is it EMF causing such interference? And how on earth does this DMX tying together make sense (shouldn't each of the three DMX lines come to the block in the rack and then gone into the 4-port rack node? Only one line goes to the node), and how did this installer get certified by ETC ----> A SR DMX line that wasn't ever connected to anything in the terminal block (D2), and we have a "LS Power" duplex circuit on our catwalk that has the Cat5 line, but there isn't even cable run to the outlet box (some random 16 guage wires we don't know what for though!).

Any ideas on how to troubleshoot this further would be much appreciated! (We have slowed down the DMX on the ports and at the console)

Couple of photos for reference....

 

    

 

 

 

 

  • Are your LED fixtures self terminating? If not, are those DMX runs terminated? Do you have any "Y" cables in play anywhere? Any optos in play?
  • Hi, all the runs have terminators on them (they are some cheap LED's "TechnoLEDgy MaxPAR's" and two Irradiant movers). Termination was our initial suspect too, but double checked the runs.

    No optos (wish we did!)

    As for "Y's" not entirely certain. Not on purpose, but in the install there might. There is a DMX-relay switch that can be assigned to control power outlets, running lights, the flouro work lights. Not entirely sure how that is circuited, since it doesn't seem to go the the Terminal block in the rack (it's DMX is labeled D7 which traces into conduit somewhere), perhaps that's "Y'd" off the main DMX?
    Thanks
  • Looking at your pics and chart you've got some weirdness with the wiring in this box. The "R" lines you can probably ignore as that is most likely the Remote Focus Unit wiring and this explains the 16ga wires, since the RFU also need 24v power. The "D" lines should be your DMX Input loop from the console location (D1), which then goes out to the (mid-house?) location on the R/B pair of D3 and then returns from that location to what is probably your issue - you indicate that the D3 line into this box is from the "D3 and Rack Node" and yet there is a corresponding cable on the other side of the terminal block to "D4". Inferring some info here, I'm betting that the problem is that this termination point has created the Y that Jason was asking about. The left side of the terminal block has been landed in a "touch-and-go" from the D3 plug-in location to hit an Input on the Rack Node (which would be fine if that was all on its own), but then you've also got the remainder of the run to "D4" which is creating the Y to whatever is on the other end of that wire (Dimmer Racks?, Other gear?). How does DMX get out to the fixtures? from what port? Are the Net2 gateways in the field that go out to the fixtures or hard-line DMX runs from some other splitter? The other possibility is that the run from the "Rack Node" is actually an output from the node to this point (connector sex is irrelevant if there is one - you can still make the port be either in or out), in which case you potentially have two "consoles" on the DMX stream at the same time (the Express and the Node), but I would expect that to manifest as a serious chase through dims and fixtures rather than just fixtures. Try disconnecting the "D4" wires and see what you lose - if you lose the dimmer racks, you can be pretty sure you've got a DMX Y. If nothing, then that may also be an abandoned run or it may be an input from another location that hasn't been identified. Also check the configuration on the rack node port - I believe red is output, green is input. If the fixtures get their DMX from a Node in the field, then those DMX artifacts are being broadcast on the network, and the Y here is probably the issue. What is the Job location and (ideally) Job Number? If properly documented at the time of the original startup, ETC Tech Services probably has copies of the drawings that could give you insight into how the wire runs were intended and if the installer deviated from intent to "make it work". Doubt the fluorescents are actually a culprit unless they are DMX controlled or possibly if they are controlled from the dimmer racks and the traffic from starting them is firing up the DMX gack - just like any other traffic would. I've also seen this behavior from a damaged splitter down-stream.
  • Yeah, if you've got D7 floating out there too, you got some wires to find... The Y in the box in the picture may not be your only one.
  • A little update on this problem.....

    Here's a diagram of basically how the system was when the LED's flickered; we've even disconnected some of the Y's (so that the D1 light went directly to the terminal and the LightMaster DMX (which we discovered wasn't installed with any termination), but didn't seem to fix the problem - especially the correlation with our fluorescent work lights (you can trigger some LED's to flash when switching on/off the wall switch).

    Before:

     

     

    And guess what.... when we insert a 4-port node at the booth and don't have any traffic through the rack node, everything becomes stable. Switching any of the four florescent switches doesn't cause the LED fixtures to flicker at all.

    Seems like some trouble with the Rack Node, 

    After:

     

     

  • Thanks for this in depth look at the issue! I've posted a diagram for some clarity and answers many of those questions you posed. We have shows this week, but plan to see if the issue can be recreated with a tour node in place of the rack node, and then we can also compare the settings on the rack node vs tour node via NCE. It's really the weirdest thing. Thanks again for looking at this thread!
  • Problem seems to have been resolved!!!

    Get this..... installers never ran a ground wire for the rack node. When we grounded, the flicker isn't replicated. When we disconnect the ground the problem starts.
    (It's only been 15+ years!).