Wireless DMX HELP!!

[microstar]

I agree fully TuckerD. I have used stranded Cat5e successfully for light-duty (read "very seldom moved") circumstances where it is soldered to XLR's. I fully understand that this is not condoned by the DMX standard but it consistently works for me.
Regarding the use of Cat5 for installation use, here is a quote from the ETC knowledge base: (KnowledgeBase: DMX Over CAT 5 - Electronic Theatre Controls

"From the ESTA report:
Conclusion
Data obtained from all three of these test sessions confirms that, in most respects, UTP and STP Category 5 cable can be expected to perform at least as well as EIA-485 rated data cable for DMX512 applications."

The ETC statement “This is why we offer an alternative IDC header to support CAT5 wire on most of our DMX products” seems to lend some credibility to the use of DMX over Cat5, not just in an Ethernet setting. They also mention the use of STP Cat cable.

 

[Jay Ashworth]

Ok. You convinced me. Let's go to the video tape.

DMX cable has a characteristic impedance of 120 ohms, UTP Ethernet 100. STP isn't actually specified for any EIA Ethernet-ish use yet, so I can't find a spec for it (it's used on "Cat 6a, 7, and 8" runs, but those are not EIA standards).

STP has a capacitance around 50pf/ft. I wasn't actually able to find a capacitance spec for DMX cable, but it may be instructive to note that AES/EBU cable is apparently 10-12pf/ft.

So, *can* you use Cat5/6 for fixed DMX installs? Sure. Will you get as much distance out of it? Maybe not. Are you going to find devices that will misbehave more because of it? Quite possibly.

It all comes down, as it always does, to your specific installation. If you're a roadie, and often plugging together lots of different people's gear, in ad-hoc installations, then you probably ought to run with box-spec DMX 5-pin cabling.

If you're doing a permanent install, with your own fixture inventory, and it won't change much, and you have sufficient time to test and characterize the entire system to make sure it will be reliable? Knock yourself out.

It *is* probably* a good idea to minimize the cable-type transitions, though; if you're going to do one long backbone run of cat5 froom a booth to a splitter, you will tempt the SWR gods less than if you swap back and forth 5 times in a run.

And, of course, DMX over *Ethernet* (Artnet, etc), is a totally different and unrelated issue to all of this, having it's own issues (primarily: simulating an isochronous channel over a non-synchronous, but much faster, transport).

 

[ledstixx]

Hi, I am using Elation EWDMXR and EWDMXT and I saw some command delays on this devices. Tested distance was like 100m.

 

[Chris15]

So we need to remember that DMX is a signal running at something like 250 kilo baud. It's a relatively slow signal.
There is not a significant impedance mismatch between 100 and 120 ohm cables that will make SWR a terribly relevant factor. Include mic cable of random impedance, often hundreds of ohms, and things get less predictable...

Dr DMX has proven that DMX will happily run on barbed wire.

Think about this. ISDN BRI phone signals run around 130 kbps, and will happily run down kilometres of loosely twisted pair cabling from the exchange to a customer. Cat5 has a much higher twist rate, and thus a greater mutual coupling and rejection of noise.

Frankly, if you are soldering Cat5 into XLRs, the most likely cause of issues, BY FAR, is the mechanical failure of the joint. A data issue is orders of magnitude less likely.
Solid core cable into an XLR, good luck keeping a connection if it moves at all. Stranded, IF terminated well, will survive better, but if you're in something that will regularly move, use a flexible cable designed for soldering, OR, treat it as a consumable and keep spares on hand ready to swap out if failure is not going to affect life too much...

It would be interesting to see if some of the work being done wih AVB etc with PTP and such can help improve the syncronisation of DMX over Ethernet transport, but I wasn't aware that there were widespread issues with timing when DMXoE was involved...

 

[TuckerD]

@Jay Ashworth Here are some more numbers for you on impedance / capacitance for DMX cabling as specified in ANSI E1.27 - 1

Section 4.4 titled "Impedance" states

Portable DMX512 cables shall have a characteristic impedance in the range 100 to 120 ohms. Due to the
characteristic impedance of 120 ohms in EIA-485 systems, 120 ohms is preferred.

and section 4.5 titled "Capacitance" states

Capacitance between conductors within a shield shall not exceed 19.8 pF/ft (65 pF/m). Capacitance between
any conductor and the shield shall not exceed 35 pF/ft (115 pF/m).

It's worth noting that neither of the two parts of standard E1.27 specify maximum cable lengths and neither does ANSI E1.11. As a general rule, less is best. Every time I have had an issue with a DMX network it's usually because people are using 50' 3 pin mic cables between lights that are 12" apart.

Edit:

The values given for impedance / capacitance in E1.27 - 1 are the same in E1.27 - 2

Edit #2:
The values given on this page say that the capacitance specified for cat 6 cable is 5.6nF/m Max.

 

[Jay Ashworth]

Those DMX capacitance numbers ate substantially lower than Cat 5. Typo: 50pf/ft was UTP, not STP.

So that's 3 times the rated number; that'll shorten your runs...

Sent from my SPH-L720

 

[Jay Ashworth]

5.6nF/m is like 1800pf/ft, which seems really high...

Sent from my SPH-L720

 

[TuckerD]

I am not sure. I only read what was on the page I linked. I agree though, it does seem high.

 

[RC4Wireless]

Hello,

Has anyone, or does anyone have an opinion on the Made in China Wireless DMX Sticks? Specifically the ones you find on amazon? How reliable are these? And for a high school would you recommend them?

Thanks!

Yikes... seems like the cable capacitance police have commandeered this thread.

It appears there are two kinds of wireless dmx "stick" products: the ones design and built in Italy, which are quite good; and the copycat products out of China which are less expensive and may not perform as well. If you like that form-factor, I would go with the good quality Italian original.

Regarding the comment that only FHSS is reliable: This is incorrect. The SS part -- spread spectrum -- is certainly beneficial. But DSSS (direct sequence spread spectrum) has advantages over FHSS in some cases. And the opposite is also true. There is no signal encoding method that is ideal in all situations. WiFi uses DSSS, by the way, so we're all using it *all* the time.

Simply using spread spectrum is no guarantee of data security. Show Baby products all ship with the same digital system IDs, so anybody else can go buy one, get it transmitting, and flip through the 5 options until they find the one that messes up your show. And then you have interference on all the channels. No better than the suggested weakness of being on just one channel.

There are numerous alternative products to consider, including systems I design. Mine is not your only other option, not even the only good one, but Disney, Cirque, Blue Man, and numerous others have chosen our stuff for reasons noted above, among others.

I am happy to carry on a non-branded tech-only conversation on this topic with anyone interested. This is not intended to be a brand plug, my apologies if it appears to be one.

Thanks,
James Smith (Jim)
President and Product Designer,
RC4 Wireless