Strange behaviour on a DMX chain

[Stuart91]

I've been asked to help troubleshoot some outdoor floodlights. There is a cluster of five, all running together on a single DMX chain from a controller. 

The main symptom is that the first two units in the chain misbehave, whilst the other three respond as they should. 

Unfortunately the connections are all hardwired so it's quite time-consuming to try different configurations. However what I've discovered so far is:

• Running the DMX signal direct into either of the first two fixtures, with nothing connected to the outputs, and they behave perfectly. 

• Attach the rest of the chain, and 1+2 stop behaving, but 3-5 work. 

• Bypass fixtures 1+2, and feed directly into 3-5, and they won't work. 

• Adding termination doesn't seem to make any difference. 

What's making it harder is that I have just trailing wires coming out of each of the fixtures, with no easy way to tell whether they are correctly wired internally. 

 

I am resigned to having to undo all of the wiring and check every fixture individually, before trying to get it all working together. However I thought I'd mention it here in case there is any obvious fault that anyone could think of that could be causing this misbehaviour?

[Stuart91]

A couple of other bits of information:

Control for the system is via a casDMX unit which takes bluetooth control from an app and produces DMX. 

The DMX connections to the fixtures are run in 4-core H07 (unscreened). Brown and black cores are the DMX positive and negative respectively. There is a blue core, and the earth (green/yellow). 

It looks like the earth is connected to mains earth.

The blue core seems to be connected back to a "return" terminal on the casDMX. 

The importer/distributor of the fixtures doesn't know what the blue core is for. However, there are three other identical clusters that are all wired the same way, and are operating without problems. 

The installation/wiring was done by an electrical contractor who is no longer involved in the job, so I'm second-guessing some of the choices that were made when it was being put in...

[timsabre]

I think the "Return" terminal is just a way of saying DMX ground.

I would check the continuity on the hot, cold and ground legs from the controller all the way through the 5 fixtures (using a long bit of wire to get back from the far end). Then I would check for stray voltages on the DMX hot, cold and ground relative to mains earth when the first 2 fixtures are connected, then when 3-5 are connected.

I suspect you will find that there is some mixup in the wiring, possibly more than one fault, on the way through. The fact that fixtures 3-5 don't work when directly connected make it sound like some sort of grounding fault.

You don't say what cabling lengths are involved but (although not the best plan) the HO7 cable shouldn't be an issue.

Edited September 29, 2023 by timsabre

[Ynot]

As far as I know, (and stand to be corrected if wrong) there's NO reason for DMX to ever need a connection to mains earth. That, to me, is just another potential source of noise that could be introduced into the system.

 

[Stuart91]

16 hours ago, timsabre said:

I think the "Return" terminal is just a way of saying DMX ground.

Looking through the casDMX manual, I think I agree. So it looks like we have the blue core on the 4-core cable connected to DMX ground and the earth coloured core connected to mains earth...

9 hours ago, Ynot said:

As far as I know, (and stand to be corrected if wrong) there's NO reason for DMX to ever need a connection to mains earth. That, to me, is just another potential source of noise that could be introduced into the system.

That makes sense - reminds me of the "pin 1 problems" that used to plague audio kit back in the day. 

16 hours ago, timsabre said:

I would check the continuity on the hot, cold and ground legs from the controller all the way through the 5 fixtures (using a long bit of wire to get back from the far end). Then I would check for stray voltages on the DMX hot, cold and ground relative to mains earth when the first 2 fixtures are connected, then when 3-5 are connected.

I suspect you will find that there is some mixup in the wiring, possibly more than one fault, on the way through. The fact that fixtures 3-5 don't work when directly connected make it sound like some sort of grounding fault.

That seems like a good plan of attack, and certainly less time-consuming than taking everything apart and starting from scratch. 

16 hours ago, timsabre said:

You don't say what cabling lengths are involved but (although not the best plan) the HO7 cable shouldn't be an issue.

The fixtures are all clustered together. They were supplied with 2m long tails, which have been trimmed to fit the short run into the junction box, with no extensions added. With all the in+out jumps we're still 10+m for the total run, but I agree that whilst poor practice it's probably not the root of the problem. 

[Ynot]

2 hours ago, Stuart91 said:

Looking through the casDMX manual, I think I agree. So it looks like we have the blue core on the 4-core cable connected to DMX ground and the earth coloured core connected to mains earth...

Hang on - you're saying that you have mains cable running the data to your fixtures?

If nothing else, that won't be near the correct impedance for DMX, and where you're having data issues, that would be the first thing to rip out and re-do properly...

Unless I misunderstand your OP

 

[Stuart91]

34 minutes ago, Ynot said:

Hang on - you're saying that you have mains cable running the data to your fixtures?

Yep - it's four core H07 rubberised mains flex. (0.5mm) 🙄

This is how they came from the importer/supplier, apparently. I think we all agree it's not ideal and may be exacerbating any problems, but all the other clusters have their fixtures wired exactly the same way, so I'd lose any argument to replace the whole lot. 

[sandall]

Bear in mind that over a 10m length DMX will quite happy run on wet string!!

[sunray]

I've only just come to this and I'm a tad confused.

 

Are you saying the 5 fittings each have a 2m long 4C HO7 exiting the fitting (presumably a compression gland or such similar) and open wire ended?

Is there a second similar cable for the output?

 

Do all 5 (or possibly 10) come together in a junction box and from there  a cable to the controller? 

Or is the controller in the junction box?

How are these then wired?

How to the fittings get powered, is it an additional cable?

 

Is there a chance the fittings can be run in master slave combination?

 

 

The manual for the casDMX does indicate the return wire is for ground

 

The reason I ask is I worked on some Flood lamps with a fixed 4c cable and the connexion were +V power, Ground/0V, DMX+ & DMX-  but cant recall colours used. The whole system simply ran on a 4c cable looped between the fittings and all wired in parallel.

[Stuart91]

3 hours ago, sunray said:

Are you saying the 5 fittings each have a 2m long 4C HO7 exiting the fitting (presumably a compression gland or such similar) and open wire ended?

Is there a second similar cable for the output?

 

Do all 5 (or possibly 10) come together in a junction box and from there  a cable to the controller? 

Or is the controller in the junction box?

How are these then wired?

How to the fittings get powered, is it an additional cable?

 

Each fitting has a mains cable (wired in 1.5mm H07) and two cables for DMX. Each of the DMX cables is 4 core 1.0mm H07. 

They have been labelled "in" and "out" by the spark that installed the system, presumably there was some labelling on them from the factory but that's been removed when the cables were trimmed to size. 

As things stand, they have been arranged in a daisy-chain configuration, but rather than jumping between fixtures, all the (in and out) cables run into a junction box, where they are linked together using chocky block. So "Out 1" is connected to "In 2" and so on. 

There is a separate junction box for the mains supplies, which I presume are just wired together on terminal block of some kind. 

3 hours ago, sunray said:

The reason I ask is I worked on some Flood lamps with a fixed 4c cable and the connexion were +V power, Ground/0V, DMX+ & DMX-  but cant recall colours used. The whole system simply ran on a 4c cable looped between the fittings and all wired in parallel.

That's interesting. 

Whilst I haven't opened up any of the fixtures themselves, I can see that there's a separate power supply unit, and an LED driver unit. Both about similar size and style to the LV transformers that you would use for 12v ceiling downlighters. I'm guessing this arrangement was specified so they could be run from the DMX supplied by the casDMX unit, but I wonder if there's also some capacity for this 4c parallel system, that shouldn't be wired up...

Another potential clue is that I did notice a bit of a tingle from the blue core, nowhere near 230v, but enough to give me a fright. That might square with the +v power being supplied that way, even if it's not needed. 

 

[sunray]

I've just looked up my notes from 2016, I didn't note the product, only wiring and colours. I show

Red = +24V

Black = 0V

Blue = DMX-

Yellow = DMX+

So straightaway there doesn't seem to be any similarity.

 

As Tim has suggested the next thing is to check for continuity between in and out on each fitting and round the whole circuit, hoping that each fitting doesn't contain a buffer amplifier.

I reckon you have a flip or two but I can't workout where. I'm probably not the best person to be  doing this though.

 

[timsabre]

On 9/30/2023 at 9:00 PM, Stuart91 said:

Another potential clue is that I did notice a bit of a tingle from the blue core, nowhere near 230v, but enough to give me a fright. That might square with the +v power being supplied that way, even if it's not needed. 

 

Well that would certainly do it... Get that voltage metered to earth then work out which fixture it's coming from. 

[sunray]

On 9/30/2023 at 9:00 PM, Stuart91 said:

... 

Whilst I haven't opened up any of the fixtures themselves, I can see that there's a separate power supply unit, and an LED driver unit. Both about similar size and style to the LV transformers that you would use for 12v ceiling downlighters. I'm guessing this arrangement was specified so they could be run from the DMX supplied by the casDMX unit, but I wonder if there's also some capacity for this 4c parallel system, that shouldn't be wired up...

2?

When you can some pictures or a decent sketch would be helpful.

From your description I see this:

I only envisage one power supply unless the fittings run on a reduced (S)ELV in which case may be in either of the areas circled pink.

There is also the possibility it could be connected to the blue and green/yellow of the 4core cables.

 

Is there a makers name for the light fittings?

 

Start with Tims measurements first. Continuity through the DMX chain light by light, shorts, funny voltages.

Edited October 2, 2023 by sunray

[Stuart91]

I was back at this job today (for my sins...)

What we've managed to establish is that fixture number four in the chain is the one that seems to be causing the problems. Omitting it from the chain means that the other fixtures all work as expected. 

I did try wiring fixture four into the end of the chain, but it still caused the same problem (the first two fixtures in the chain not working). 

The fixture has now been removed entirely, and I'm going to go through it on the bench and see if there's anything wrong with it. So far I've only been able to ascertain that it will respond happily to a DMX signal by itself. 

It looks like, in the meantime, another four or five lights have stopped working in similar fashion in the other clusters, so there'll be more to be done no doubt. 

[Stuart91]

I've taken the casing off the faulty fixture today. 

Inside is a PSU and fairly generic looking LED driver.

I'm getting continuity on the brown, blue, and black cores, but not the green. If as I suspect, it's some sort of power supply, I suppose that would make sense. But the driver unit has a separate power supply, which leaves me scratching my head as to why the DMX signal requires an additional (presumably DC?) feed along with it. 

However, when I traced the wiring back in-situ, these connections are wired back to mains earth. Again, there's no obvious reason why, although I'd expect the spark who did the install looked at the colour-coding and figured that they needed to go to earth. 

 

 

Edited November 1, 2023 by Stuart91