DMX connection suggestion from china

[roekeloos]

Hi,

 

Okay so everywhere I read that you cannot passively y-split a DMX connection. Im busy with a project now and this is the diagram I received from the supplier. It clearly shows the connections being y-splitted. The guy tells me that it works like this and they have done many lights like this. He says its how the t-split is connected. but our understanding is limited to each other. please have a look at this and tell me what you guys think. If this works then it works, why does everyone else then say its impossible ? wtf im confused

 

http://img442.imageshack.us/img442/8475/connectionguide.jpg

[timsabre]

It's not impossible to passively T-split DMX. The problem is that it can lead to instability/flickering in some of the fixtures - sometimes.

 

If the wire betwen the T and the fixture is relatively short, this helps it work OK. If all the wiring is relatively short, as it looks to be in this setup, that helps as well.

 

They have probably been lucky not to have any problems so far.

 

Or, the other possibility is that there is a powered DMX split in the T-junction.

[smalljoshua]

If you look inside a standard DMX Fixture, most do just this, link the DMX in to the out and then a 2nd connection from the DMX In to the DMX Receiver Chip, this can be 10 or 15cm without any issue. The problem comes when you have 2 long legs (as you would get if you Y Split a Run)

 

Provided the Cable Length between The T and the Fixture isn't too long (I can't give you a number on that though), there shouldn't be a problem.

 

Josh

[Brian]

...I can't give you a number on that though...

As long as the stub's electrical length is less than one quarter of the signal's transition time then the stub will still behave as a lumped load and not a separate transmission line.

 

For a typical DMX driver (75176B) that means you can have stubs of up to 1m in length.

 

You will however be sitting on the digital cliff edge.

 

A safer figure is a length of 10% so stubs around 400mm will be OK.

 

 

(Guess who's been modelling DMX signals over long cables in complex systems this afternoon.)

[bigclive]

Whoa! Now THAT was an interesting post. I've never really contemplated stub lengths inside equipment before. But I have been forced to "Y" into really crap "waterproof" fixtures that had a single "sealed" tail hanging out. I did keep the connection well within one metre. Not that it really mattered with that high profile piece of Italian junk, since they were a dog right out the box.

[musht]

They are `meteor` tubes or similar, have a travelling line of LEDs going down them.

 

DMX controller is a DMX decoder for the tubes own data stream , power, ground, data and clock.

 

Edited to add:

 

Part# are Ledwalker meteor tubes, you`ll find you can eat universes of DMX at full control or look at Madrix :-)

[Don Allen]

Yes, providing you keep the external stub short as explained by Brian , you can get away with it because it is the same as an internal Y split. I use a long run of 3 pin DMX connectors to connect 8 led strips, the cable loops in and out of 8 female 3 pin cannons. It saves a lot of setting up time and I have not had any problems with it. The cable is a lot cheaper than all of those connectors shown in their diagram.

[roekeloos]

The guy I'm ordering with tells me that it will still work even if the leads are up to 8m. He said they did a DMX installation like this and it worked perfect. so I don't know what the hell

[rich_lambe]

Simply this does not obey the USITT standard so IF there is problems guess where the finger will be pointed.

It might work but as the previous replys show you have been warned. I dont think anyone here will give you any guarentees if thats what you are looking for.

Best of luck :)

[roekeloos]

Im just new to the whole thing so. Thanks for all the replies. it helps a lot.

[musht]

People are obviously confused by the mention of DMX, the DMX `controller` in the diagram isn`t a master console it`s a slave fixture.

 

One of the clues is 4 pin connection, another is power entering the chain at both ends.This is important on long runs because voltage drop can cause subtle colour variation.

 

Biggest one is the part#, which google shows is a Ledwalker product, from this family:

 

http://www.ledwalker.com/product_details.aspx?pid=1&sid=24&id=18

 

In my humble experience Ledwalker make an effort to make innovative, well made products, site isn`t stocked with Avit Leopard desk knockoffs.

 

They are all based on LED pixel controllers, where each RGB LED is individually controlled by a chip beside it, the chips don`t talk DMX , they use a serial protocol with seperate clock, see above :-/ They regenerate the signal at each chip but trying to keep distances down between master controller and first in line isn`t bad idea.

 

Hence also the mention of eating DMX universes at a rate,only 170 RGB pix per universe.

 

Some of Ledwalkers controllers run `canned` effects triggered by DMX as well as direct control.

[timsabre]

People are obviously confused by the mention of DMX, the DMX `controller` in the diagram isn`t a master console it`s a slave fixture.

 

One of the clues is 4 pin connection, another is power entering the chain at both ends.This is important on long runs because voltage drop can cause subtle colour variation.

 

Ah, of course. It is a DMX to 2-wire data+clock interface.

Although the 4 pin isn't really a clue, if you were running DMX+power it would still be 4 pin.

[musht]

 

Ah, of course. It is a DMX to 2-wire data+clock interface.

Although the 4 pin isn't really a clue, if you were running DMX+power it would still be 4 pin.

 

Those 4 pin screw togethers are omnipotent on Chinese LEDs, though snowfalls are usually 2 pin;

 

 

Add more control :-)

 

 

Its really single line serial data and clock, its not balanced so line length can be an issue.

 

http://www.world-semi.com/en/