Clearcom thru Ethernet

Its an analog signal, DMX is a digital signal. Analog signals do not do well without shielded cable. 99% of the cat5 cable out there is unshielded. Will it work, probably, will you actually want to listen to whatever comes over the com, no. BTW, ESTA does have a few pages written on using cat5 in ethernet systems, http://www.alia.com.au/features/dmx.pdf
 
Actually it is not quite that way, analog balanced line level signals do work quite well over UTP, it is the twisting that rejects the noise. The problem is that many people when they attempt to wire up to UTP decide to send each pr the two lines down a different PAIR on the UTP and this does not work, instead send your two conductors down a single PAIR.

The Problem with the clearcom system is that it is not balanced audio
Pin 1: The pin 1 is common ground.
Pin 2: The Pin 2 carries the station power (24-30V DC up to 100 mA per station, typically 50 mA).
Pin 3: The audio is carried using pin 3.


you could TRY connecting it with pin 1 and 3 in a pair, but I don't think you will get the rejection you need from the twisting


Now if you have a RTS TW system which is two channels, you can use two pairs of UTP

http://www.epanorama.net/links/intercom.html has some general information

Sharyn
 
Last edited:
I now have my next project: Ethernet intercom!
 
So stepping into the middle of something I've never heard about... So are you saying I can take 20 feet RJ45 and throw two 5 pin XLR plugs on the ends and call it a DMX cable? With no loss in quality like when using audio cable? That's so much cheaper!
 
There still is this debate as to if you need the STP vs UTP, or if you need to run it thru conduit that is metal for shielding. In my experience, totally non scientific, it works fine. UTP is 100 ohms, use termination, and again make sure that you have the data carrying pair all on a single pair in the UTP cable.

It would be interesting to have other people try it and see their experience, when you look at how high a data rate we can run over this cable today with data, it would not be surprising as to it working. The twisting is very precise and the rejection that you get from the twisting is tjhe key

The only question is what sort of outside signal could you experience in a long run that would cause interference.

Sharyn

Sharyn
 
There still is this debate as to if you need the STP vs UTP, or if you need to run it thru conduit that is metal for shielding. In my experience, totally non scientific, it works fine. UTP is 100 ohms, use termination, and again make sure that you have the data carrying pair all on a single pair in the UTP cable.

It would be interesting to have other people try it and see their experience, when you look at how high a data rate we can run over this cable today with data, it would not be surprising as to it working. The twisting is very precise and the rejection that you get from the twisting is tjhe key

The only question is what sort of outside signal could you experience in a long run that would cause interference.

here are some links
http://www.belden.com/pdfs/Techpprs/DMX-512.pdf

http://www.esta.org/tsp/working_groups/CP/docs/DMXoverCat5_P1.pdf

Sharyn
 
So stepping into the middle of something I've never heard about... So are you saying I can take 20 feet RJ45 and throw two 5 pin XLR plugs on the ends and call it a DMX cable? With no loss in quality like when using audio cable? That's so much cheaper!

Yep, and you can run it a good distance as well. I have been using UTP cat5e for a long time and it works great, cat3 also works. Most places that are new installs are putting cat5 into the walls and electrics to run DMX through so when the switch to ACN/any given distro protocol happens you simply rip the 5 pins off and put on ethercons or whatever we finally settle on. The only thing to be aware of is off the shelf cat5 is a fairly brittle cable. Buy cable that is specifically made to be handled often, not the 1000' boxed rolls.
 
I'm running 16 universes of ethernet/DMX at the moment along with eight ethernet/ethernet, 5 BNC and one DMX 2-pair that's going to be dual channel clear com. Or at least I will be in about another 30 hours after I'm finished wiring the thing.

Attempt #4.5 at pulling my cable thru the conduit finally worked, "its thru" was the cheer of the morning - this after some very major getting stuck points on the way there that almost snapped the chain.. again and made the fork lift do little bounces of going over it's capacity as the chain/loom in conduit passed thru rough points or had to go up than down than up in finally freeing itself at time. Next time as potentially one hang up, I won't use a shackle at the Kellems grip - that was bad, sheered the screw pin head right off. Fork lift verses screw pin on a 5/8" shackle, the eyelet lost but gave quite the fight. Threw out the last attempts at pulling this cable loom - eight pair ethernet, two 16-pair digital grade XLR snake, a BNC snake and a DMX cable. Been thru the conduit too many times by way of pulling it in, than taking it out afterwards. Couldn't trust it even if it was a few hundred dollars worth of cable trashed. Also tossed the 800# safe working load chain that I broke yesterday after one final trip thru the conduit today - what was left of it was too stretched out in some areas and none of it could be trusted - lots of cracks at the weld and narrowing of the chain. Also threw out all quick links and shackles associated with it - rather throw out than trust it given they were associated with this nighmare of a cable pull. Finally even though it didn't show any real damage, I also tossed out the 5/8" spectrum braid rope used on the initial pull attempt - just don't want to trust it and it had in some areas wire pulling lube impregnated into it which would turn into a magnit for dirt.

Hope it all works as the conduit was fairly trashed and I think broke somewhere beneith the concrete so there is no telling how many circuits will work - this given the DMX works at all on ethernet. I used the solid conductor paired 1000' boxed version. Solders really easily to panel mount terminals. Both green for pin 1, both orange for pin 2, both blue for pin 3, striped brown for pin 4, solid brown for pin 5. If it don't work as there is a few shows designing starting Saturday and I have other stuff to do starting Saturday, it will be time for a major punt.

65' length, 27 cables of ethernet (3 spare) one 5 pair video, one DMX and one 5/16" spectrum braid rope in a 2" conduit. (Rope is there for pulling more stuff - doubt I would ever attempt.) By the time I was done with the first of the three imput/output panels, the cursed BNC and ethernet plugs were almost easy. Next challenge now that I have mastered the plug tool, using the punch down tool.

Fingers crossed in a major way that it's all going to work - no time for troubleshooting if I want sleep Friday night/Saturday morning.

If it all hopefully works, in a month or two I get to install the second part of the system, the right side is a duplicate of the left side I'm attempting now. Next time it's all quick links and at least 1,200# chain.
 
Ship, really hate to rain on your parade, but I think there might have been a bit of a boo boo. For UTP to work for DMX, you need to use a pair for the data lines, so say blue to pin 2 and blue / white for pin 3. The pair are twisted and so you get the cancellation effect that is the benefit of a balanced system. I don't know whether if you want double wires for reduced resistance, then you could join say blue and orange and blue / white and orange / white, they should keep their noise rejection I'd think - anyone else care to comment? Again sorry for raining on your parade.
 
Chris is correct, you need to use two pairs, but you need to split the two dmx lines across the two pairs, NOT use a pair for each line. This is how the twisting for rejection works.

Sharyn
 
Back to the Clear Com question of the day, (DMX response below) will it work on ethernet cable, at this point how to wire it if possible give the above twisted pairs concept.

Interesting... it's up and running in doing the in shop programming for a major tour. Tested it using the cable check on a DMXter and it passed the DMX signal 100%, and it's now running a light board.

I was kind of amazed that the pull didn't at least break or make unusable any of 16 lines of DMX/Ethernet lines pulled but it all worked given the possibly wrong wire pulled and wrong wires to the terminals... granted it's only 60 or so feet and I have passed DMX thru the brushes on a shop light cable reel but in this instance as with the other, it seems just fine. DMX code for me at least seems fairly rugged in passing thru all sorts of things including cheap crap microphone cable for short distances at least... this given I'm a barbarian when it comes to stuff like that. I do use the proper stuff when possible - I have the budget to do so I do it properly, but also don't at times. Interesting that it should not work - it works. Most likely not recommended for others to try or at least most reasonable to do so with the recommended shield / drain being tied to the second wire of a pair. Sounds reasonable but I'm not re-soldering what's done.

Perhaps with a longer length and or perhaps there is not enough datta going thru the pipe yet. Guess that's a back of the head thing for later should it come up and I hope it does not. In the mean time, the DMX and Clear Com are done, (the Clear Com was in this instance run thru a two pair DMX with pins 1-5 wired up so I could if necessary use it for DMX in addition to the unknown.) The Clear Com thru Ethernet question for this instance was more musing or academic. I back burnered the ethernet and BNC connections on the project - not enough time, not enough time... a tour left last night at 12:30 that needed by far too many custom fixtures from scratch built. Had to say bye bye to the wisiwig project given the amount time needed to complete the other tour's project. Since Thursday morning at 8:30 AM (and I was on time for once... or really early all week on all days, as of last night I just finished working a 37.1/2 hour day Friday/Saturday in going from one project to another which added up to a total as of last night of 101 hours in six days... Yep, so far at least taking the weekend off, and came in a half hour late this morning. I'll get back to the Wisiwig project perhaps in a week or two after I'm done converting 5K Mole light Fresnel into Pixel Pup mounting fixtures - this second half of the week project for a third tour. Told the various shop managers that the right side of the exact same panel won't be even thought of doing for at least the next two months. They can find someone else to do it if they want it before than. That said, no doubt next week both systems will no doubt be due. It's touring season and we have that big three bay room to potentially program three shows inside of. Gotta test it I suppose.

Hope I never have to fix the DMX/Ethernet and it's a bit more rugged than as stated.... I really hope that it continues working not as per the spec on concept but in reality in it working and continuing to work no matter how much datta in the future is run. that's going to be the main question of the future. Passes datta so far as the only line doing so, once you get 31 lines of various types of datta running thru that 2" conduit... don't know yet given the above.

On the right side, I'll give a shot to doing as a compermise the white/stripes to the drain and the solid colors to the pins. This at least for pins 1-3. Given the nature of a four pair cable of what ever type, it doesn't add up to a five pin DMX, this much less anyone ever try to fit a lot of wires into a five pin Neutrik panel mount? For pins 4&5 I'll probably just dual pair them green/green-white pin 4 and brown/brown-white pin five. This double wire concept for pins 4&5 would work well with a Hog remote given it's sending power down those conductors I would think. Might not work as a concept so well with a time stamp type data but given the robust nature of the signal theorized above, it should be fine. All stuff to look into by way of Ethernet/DMX I suppose.

thanks for the info, hope it's wrong... as it were.
 
Last edited:
the way you have it wired "will" work, BUT will not be taking advantage of the cancelation effects of the twisted pairs. I will be interested to see what it does in the field.

Out of curiosity, did you use the stranded (flexible) or solid ethernet cable? Not that it should make a difference but if solid I too am surprised that none of the conductors are broken, or if they are it is not showing up because of the wiring style used.
 
For reference, it was solid red/green/blue used off the shelf cable in 1,000' boxes. While my supplier might want to push that his is better in grade, I'm sure it was of a standard grade.

Used three colors to designate as per eight channel DMX/37 pin Socapex lines 1-8 A verses B, than Ethernet/Ethercon lines 1-8, than paint markered a wide color band on each of the 9 cables pulled of each type at the end. ROYGBV color code plus Black/White/Gold for each of the circuits plus the spare on each.


I'll also be interested in the cancellation effect between lines of datta as it were. It's a pipe full of 45 cables that is so far only running one DMX/Ethernet.

In broken conductors... Most of the circuits at this point are either not connected yet or not tested. Fact was I tested the first 10 DMX circuits. #A-1 had a short between pins 1&2. Two of them had problems in the test. Think I found a micro wire on pin 2 touching pin #1 but it could have been an errent strand of plastic - didn't get to re-test it yet. Circuit B-2 didn't work either in passing datta. That was ok, gee I forgot to finish wiring that circuit at the time of testing... I was about half way done or at least started in wiring it. So In theory, circuits A-2 thru B-1 so far work, the rest in theory work but have not been tested. Also the DMX that runs the light board is connected to A-1 so while I didn't test it, it's currently a solved problem as a theory by way of seeing somethign plugged into it.

Yep, one of today's projects - testing the DMX lines, but I never got time, instead I cleaned up my work area and started to get caught up with about two weeks worth of purchase orders and E-Mails I had to set asside. Half day more at doing so and I'll consider correcting my PO to the idiot at Litton/ITT/Veam that needs me to remove two lines of text from my PO about the specification of what I'm ordering before he will put thru the order that takes 10 weeks to complete.... written OK place the order already and I'll get back to you, I'm busy now don't work instead I get phone calls and further E-Mails... every day you wait in removing some text from the purchase order for a few thousand dollars worth of plugs you are buying delays the completion date... No sense of "go with it", industry standard with them... but at least they are not Socapex brand. Urr, waiting... waiting for my plugs they sold me on than can't seem to keep up with in getting to me once the order is placed. Waiting, hay, isn't it spring, does my 10+ week time frame take into account the French month long vacation - is it 14+_weeks until I get my plugs?

ah how I ramble on with other stuff to work on than ethernet concepts... Just ordered 600 four pin Neutrik plugs today for various LED fixtures - one week out to get them given also the French vacation doesn't come up within the next week as a consideration. As you can see, while academic and concept for me, I'm already two projects beyond the ethernet/dmx one.
 
Last edited:

Users who are viewing this thread

Back