DMX robust - RDM not?

[svanciel]

Not sure if this comment fits here or under user implementation but I have discovered emperically that DMX is much more robust on a compromised wiring system than the RDM traffic on the same line. A NEMA outdoor, weather rated electrical box, was found to be full of water. The wires and screw terminals on an ETC-pcb mounted on the 4-pin panel xlr were quite corroded. Yet the Aquaram power (pins 1,4) and the DMX data (pins 2,3) were able to get through and power three units and run cues correctly. However, the RDM traffic was occasionally corrupted enough to result in "Dropped Device" messages and no sensor data readings.
I'll now be going through all of the 10 year old terminal boxes to seek out other issues that may be present.

Maybe that others will make this discovery when they try to run RDM products on their "old" DMX systems that have worked well enough in the past? Might be worth noting in an implementation guide.

SteveV

[sblair]

Steve,

This is not surprising at all! The thing that makes DMX so forgiving is the fact that it is spitting the same packet out over and over up to 44 times/second. This masks all types of sins.

RDM by its nature is going to require a better quality infrastructure than what you were able to get away with in the past with DMX. This includes using termination and the right type of cable.

The good thing about this is that RDM will make it much easier to spot all identify and pinpoint all types of intermittent issues that previously would have resulted in random flaky behavior of the fixtures.

[svanciel]

Originally Posted by sblair

This includes using termination and the right type of cable.

True that! With the new InfoTrace Aquarams, we had to add terminators to the fourteen 4-pin lines (power supplies to heads) that never had them with the previous non-RDM units. At least all of the cable was/is the special Wybron four conductor cable, so it is still good for the power and data. Just have to go around to all of the outdoor connector boxes and look for water.

SteveV

[derekleffew]

Interesting. This just came up on another discussion board. I stated I'd never heard of an A4M with a 120Ω resistor between pins 2&3.

The board received the following answer:

"Ok, so the OFFICIAL response from the Wybron engineering dept is that you should terminate the power supply just like any other DMX fixture. No termination of the scrollers is necessary."
Dusty Hudgins
Sales Manager
Wybron, Inc.
719-884-6480

[sblair]

You should always terminate the end of the DMX line with 120 ohms across pins 2 & 3. This has been standard reccommened practice since the beginning with DMX512.

In some cases you could get away without doing it, but there are many times where not having proper termination will cause all kinds of nasty issues with erratic behavior.

With RDM, you MUST have the line terminated in order to work at all most of the time.

[svanciel]

So, I logged in, got a welcome message, and typed a reply only to get a message that I'm NOT logged in and my message is lost.

Is there a bug in this hosting system? This has happened before.

SteveV

[svanciel]

The prior message was an experiment. After typing it, I hit the "submit reply" button, got the "not logged in message", logged in again and the post appeared as sent. Very perplexing. Unfortunately, I still have to go back and reconstruct the original post I wanted to send.

SV

[svanciel]

quote: "Ok, so the OFFICIAL response from the Wybron engineering dept is..."

That IS interesting because it was the Wybron reps who told me to terminate the scroller lines as we were tracking down the causes of a lot of spurious InfoTrace RDM error messages. The volume of error messages did go down when I added the 4-pin terminators from Doug Fleenor. These are not listed as a product on their web site. Perhaps they custom built these for us?

The Wybron product documentation indicates that the data protocol from the power supply to the scrollers is DMX/RDM. So, unless they have made some undocumented modifications for the ps output, it would make sense to terminate these lines just like any other DMX line. Each power supply does have its own UID. Even though very little is reported back from these, I had supposed that it was to give a warning if one should lose power. With a UID, you can identify directly (rather than indirectly as scrollers start dropping out) which ps was lost.

SteveV

[Milton Davis]

Hello Steve,

The DFD 4 pin terminators are built by request. The design of the Wybron products is such that you are supposed to cable out to the first scroller, daisy chain to the nex, etc. until you get to the last one. The last open connector on the last scroller is supposed to cable back to the power supply. This serves two purposes:
1. The additional cable serves as "doubling up" of the power wires so voltage drop is better controlled.
2. There is a data terminator in the power supply so the act of connecting this loop also terminates the signal.

On those occasions when it is not practical to run the final cable from the last scroller to the power supply, a terminator can be plugged into the end of the line. While this addresses the termination issue, it leave the possability for low voltage particularly at the end of the string of scrollers. Also, just for reference, it is a 2 Watt, 120 ohm resistor connected between pins 2 and 3. We use a 2W part because a 1/4 or 1/2 Watt part might be destroyed in the event of certain connector mis-wiring configuration.

I have not gotten a good sense of the issues prompting this thread, but is is possible that the real cause of problems is one or more scrollers resetting due to a power "brown out"? Given that there is no return cable, the last units may be getting voltage "starved".

Regards,
Milton Davis, Engineer
Doug Fleenor Design, Inc.