How do fixtures split the DMX signal internally, electrically?

[IAmLumenator]

It's always been a big no-no to split a DMX signal with a passive splitter, but every fixture I repair always has direct continuity between the in and out ports. I understand reflections are an issue, is this possible because the split happens so close to the fixture electronics?

Also, this is purely anecdotal, I have never terminated a DMX network with a 110Ohm resistor in my life. I do have one soldered to an XLR connector in my toolbox, just in case, but never used it. My fixtures are pretty cheap and low-end (see my FreeStyler post to see exactly what I use) and none of them seem to have an internal terminator. I use exclusively 3-pin microphone cable too.

This stuff isn't in great shape, and I have run up to 200' total cable before the first fixture, 10-50' between fixtures for this, longer if I used house audio multicore for a DMX universe at a venue.

Why haven't I ever had problems?

 

[BillConnerFASTC]

Try running on barbed wire but no guarantee you'll have problems. Tolerant protocol.

 

[abbyt]

Regarding the use of "microphone cable"...it may work just fine...until it doesn't. I've found, when it doesn't work, it's at the most inopportune times.

Check out these articles:
http://www.controlbooth.com/threads/dmx-cables.1216/
http://www.controlbooth.com/threads/the-ongoing-debate-of-mic-cable-for-dmx.27615/
http://blog.apollodesign.net/techni...re-the-same-as-3-pin-microphone-cables-right/

Bill is right. DMX may well run over barbed wire without any issues. It's tolerant. One question to ask yourself, though, is if you're willing to risk unexpected issues during show critical moments...

 

[starksk]

@IAmLumenator,

To answer your original question, yes, in most fixtures with a passive pass-through, the DMX actually forms a "T" with the fixture being the third wheel. The very short distance and high tolerance of the protocol allows this to work.

As for running without termination, it will work until it doesn't. You may have never seen an issue because each setup may not have had the optimal conditions for reflection. It is good that you carry a terminator just in case that situation changes, but as a rule, I recommend always terminating fixtures that do not provide their own termination.

I won't comment on the "microphone cable"/3-pin debate as it has been covered in the forum ad nauseam. Suffice it to say that Bill and Abby are right.

 

[JD]

True on the passive split. Problems get real bad if there is any length between the passive split and the chip. In most fixtures, that is only a couple of inches. When you start getting into feet, a resonance trap is formed and noise from this degrades the DMX signal. I saw a rig a few years back where the DMX was "T"ed on the truss at each fixture location and a short jumper looped down to the fixture. It worked for him, but I would never recommend it! You can get away with a lot on DMX runs, but then the night comes when you pay the piper, usually an hour before a very large show opens.

 

[IAmLumenator]

Thanks for the info!

 

[RickR]

Another common no-no is the "touch-n-go" input.
An input jack in booth runs to a jack in the seats for a tech table and then on to the dimmers. Against the standard but done in countless houses for decades. The "right" way is with one of my favorite gizmos:

 

[epimetheus]

Another common no-no is the "touch-n-go" input.
An input jack in booth runs to a jack in the seats for a tech table and then on to the dimmers. Against the standard but done in countless houses for decades. The "right" way is with one of my favorite gizmos:
View attachment 10480View attachment 10480

That's pretty cool. Link?

Sent from my Nexus 5 using Tapatalk

 

[MikeJ]

I have gone years never using a terminator, then I've have stints whee every show had termination issues.

I think a lot of it probably boils down to what DMX IC is used and circuit layout. I've also seen specific cable length and fixture combination create DMX problems. Segregating the fixture types tends to solve the problem.

 

[RickR]

That's pretty cool. Link?
Sent from my Nexus 5 using Tapatalk

http://www.pathwayconnect.com/content/view/45/29/

 

[porkchop]

Why haven't I ever had problems?

What's your definition of a problem? Is it an issue if your lights vary a degree or two every time you bring them to a specific focus? Do you require very smooth slow 16-bit movement? What is you acceptable variance in color between fixtures set at the same value?

DMX-512 is a digital signal so it is a very stable protocol regardless of the transmission method. At the same time, the shows you've done with no problems might my hair pullingly buggy to a different designer with more stringent expectations of the fixtures.

 

[JD]

DMX has no error check. No way to know how well the fixtures are "hearing" it until they can't. Imagine you're an LD talking to 12 spot ops. You can't hear them, and you assume they can hear you. You would not know how well they are hearing you until they started missing cues. Same thing with DMX. In a quiet room with simple cues, it may looks like everything is going great. What you don't know is that the spot ops can barely hear you. You find that out later as things get more complex and the room gets noisy!

 

[BillConnerFASTC]

In my experience, when DMX goes bad, some fixture either doesn't work or strobes or a hole rack is dead. Not good. Never a case of a slightly different hue or intensity value.

 

[jcslighting]

Over the years I too have broken some rules and used unterminated lines, mike cables, etc. But all it took was a few issues, thankfully during a rehearsal, and not only did I have a mess to figure out but I had to correct it all. Now I never run unterminated DMX - why would I, a terminator can either be made for a couple bucks myself or even bought online for almost the same $. I also try to never use anything except real DMX cables and any issues I have with those cables are noted and repaired.

Last summer we did a musical with 20 movers over 5 DMX drops from an opto-splitter and had a couple issues of either fixtures fluttering or some lamps not striking. The issue was a couple bad cables in the middle of a run but the problem fixtures were no where near the bad cable. Why put a show you spend hours hanging, focusing, and programming at risk because of blatantly not following the proper protocols for the data runs? In my book its just not worth it.