3-pin DMX declared 'ILLEGAL...'!

[Tomo]

It's all about data speed and DMX is relatively slow. It'll run on bell wire, mains cable, all sorts.

This is just outright wrong Tim, please don't keep repeating it.

 

In any comms installation, some of the time, some of the bytes are corrupted by the time the receiver processes them.

 

Following the standard is how you load the dice in your favour.

 

Disobeying the standard (eg by using the wrong cable) increases the chance you come up snake eyes.

 

Throw a DMX packet down the wire to a receiver:

• 
   
  
• If the bytes that get through perfectly happen to be the start code and all the bytes up to and including the ones your receiver needs:
  The receiver gets a good update with the right value. Hurray!
   
  
• If any of those bytes do not get through perfectly:
  The receiver either doesn't get an update at all, or even updates with the wrong value.
   
  

If the receiver gets any update - right or wrong - then it has a full second (possibly on the wrong value) to get another before it decides the DMX has gone away.

 

The receiver has no way of knowing whether the value it just got is correct.

- While they can tell if DMX packets were too short, or if they seem to arrive at odd intervals, few fixtures display this.

 

So in a bad installation, you will get steppy fades (missed updates), and/or flicker (wrong value sometimes) long before any receiving device will say "There's no DMX".

 

Also, the probability of failure increases as the DMX start address increases. A dimmer on 1 only needs to get a break and two correct bytes, while the one on 512 needs to get 513.

 

One fairly quick way to test a DMX wiring installation is to try using RDM across it. (Only between active components like splitters/boosters).

RDM receivers can detect both "no value" and "wrong value" situations, so give you a reasonable idea of how the dice are loaded.

 

If RDM doesn't work reliably, then sending DMX across it will also be unreliable.

 

PS:

As to the OP:

 

The ANSI standard explicitly prohibits the use of 3-pin XLR for DMX. Thus, any device using a 3-pin XLR doesn't meet the standard and is not DMX.

- But it might be compatible, and that's often sufficient.

 

However, as the standard prohibits it, anyone using it cannot have read the standard. What else did they get wrong?

[timsabre]

It's not wrong as in untrue, I have seen it in real life in a number of dodgy installs. However it is wrong as in "very bad practice" and of course we should be doing our best to build systems which are going to work reliably.

 

I think what I am trying to say is don't agonize for ages about getting exactly the right type of cable, because having the correct topology (correctly wired, no branches) and using splitters to break down the system into sections is much more important.

 

I've been called out to fix many non-functional DMX systems and have never seen a fault due to using the wrong type of cable. However I've seen many faults due to passive splitting, one side of the DMX getting disconnected, DC voltage on the DMX ground pin, fixtures with faulty DMX transceivers...

[gyro_gearloose]

What Tim said.

 

I've always thought of DMX as a data standard first and foremost, and the physical hardware far less important. I too have seen more problems caused by wiring faults, odd topology, and fixtures outputting spurious data than problems due to 'illegal' cable or connectors.

 

Some food for thought: Who uses terminators on the last fixture? I rarely see anyone use them, even though RS485 networks require them to avoid signal reflections.

[alistermorton]

Some food for thought: Who uses terminators on the last fixture? I rarely see anyone use them, even though RS485 networks require them to avoid signal reflections.

 

Always now, after (almost in desperation) adding a terminator stopped random flashes on some LED battens. I decided it wasn't worth the aggro of inviting problems when it's hardly arduous to plug a terminator into the last fixture.

 

 

 

[timsabre]

Some food for thought: Who uses terminators on the last fixture? I rarely see anyone use them, even though RS485 networks require them to avoid signal reflections.

 

If I was installing a permanently installed system I would fit a terminator. On temporary event rigs, only in times of trouble. Strategic use of splitters is much more important for a reliable system in my opinion.

[Don Allen]

+1 to Tim's practice of using splitters for reliability of temporary installation. I rarely use terminators, unless something does not appear rock solid, but like to have isolated DMX feeds to each section of the stage.

[empyfree]

20 years. Never owned or used a terminator... watch this space for egg on my face next gig!

[timsabre]

20 years. Never owned or used a terminator... watch this space for egg on my face next gig!

The best thing a terminator does is show up a missing leg on the DMX. If the hot or cold is missing the DMX will usually still work using the ground connection as the other side, but unreliably. When you plug in a terminator it links the broken side to the working side making them equal and the DMX stops working. By plugging the terminator at various places in the line you can usually find the fault.

[indyld]

The best thing a terminator does is show up a missing leg on the DMX. If the hot or cold is missing the DMX will usually still work using the ground connection as the other side, but unreliably. When you plug in a terminator it links the broken side to the working side making them equal and the DMX stops working. By plugging the terminator at various places in the line you can usually find the fault.

 

Agree with this method of fault tracing a one-legged signal line. The flip side is people saying "Well, my DMX worked fine until I plugged a terminator in - so I got rid of it again. Terminators are BS!"