Common Grounds on PSUs

[Seb]

Hi, I wonder if someone can shed some light on this quick question?

I understand from the Loxone docs that all the PSUs need to have a common -'ve (eg all the -'ve terminals linked together) [1]. I also understand this is the case if you have separate PSUs powering the DMX dimmers [2], however, the documentation on the Loxone site has me confused as they have not used previous notation [1] on linking the grounds of the PSU [2].

What I would expect to have seen is both PSU -'ve wires linked, but the are not. They have however added a wire from the middle terminal of the DMX module to the -'ve on the PSU that powers the DMX dimmers. Does this create the link for me / do I still need to link the -'ve / do I need to do both?

Any help greatly appreciated as I am about to connect a RGBW DMX dimmer to a different PSU and would rather no hear a crackle 'n pop followed by a few curses and the sound of my credit card flexing to some more Loxone kit :o)

[1] CONNECTING THE MINISERVER TO SEVERAL POWER SUPPLIES

[2] CONNECTING SEVERAL DMX DIMMERS

[Peter]

I'd like to know the answer also.
I'll take a wild guess though and say it won't be a problem.

[seb303]

If you have devices with separate power supplies that need to communicate, you generally need to connect the -ves so they have a common reference.  DMX wiring does this anyway - one of the 3 wires is a common/ground.

[Seb]

I have an answer from Loxone:

Question:
-------------------------------------
Does the middle wire from the DMX controller to the DMX dimmer:

1) create the link for the -'ve so I don't need to link the PSU -'ve?
2) do something else so I still need to link the PSU -'ve?

If 1 is correct, if I omit that wire, would linking the PSU -'ve be an alternate solution?

Answer from Loxone:
-------------------------------------
Yes linking the -ve at the power supply if it can be done is the best and simplest way to do this, it can also be done at the DMX extension as this is simply a common ground pin and in some cases it is more practical. In essence your answer is [1]. Your understanding of connecting any DC GND that has a wire between it's products.

[seb303]

If you use a shielded 2-cord cable for DMX, then the shield is your GND, so you don't need another wire.  I would suggest using shielded DMX if you have long cables running alongside other cables that may interfere with the DMX communication.

I used non-shielded (cat5) for DMX on the advice of our electrician, and sometimes the lights flicker when there are transient spikes (e.g. when a PSU for some drivers is powered up) on the power cables that run alongside the DMX wiring.

[James Mitchelmore]

It is perfectly acceptable to use cat5 cable for DMX in an installation environment however as it is a balanced transmission system it is important that the + and - are within the same twisted pair. See 7.3 table 4 http://tsp.esta.org/tsp/documents/docs/E1-11_2008R2013.pdf

If your link is wired properly I think it highly unlikely that any induced electrical interference could cause a flicker. It would be more likely that there is brief voltage drop on the supply of the affected lights when the other PSUs are powered on.

[seb303]

Hi James,

I am indeed using a cat5 twisted pair for DMX +/-.  In lighting rigs I've seen, DMX uses shielded twin core cable, hence my assumption that it may be needed in a noisy electrical environment.

For the mains wiring, I'm using 2x 1.5mm cable from the consumer unit to the Loxone cabinet for the 2 lighting circuits.  And then 1.0mm cable from the Loxone cab to each group of drivers that share a common relay output.  A voltage drop on the 1.0mm can't be the problem since all these drivers are turn on at the same time.

The run from consumer unit to Loxone cabinet is 17m of 1.5mm cable.  Do you really think that a few lighting drivers (or rather 1 PSU powering a few lighting drivers) would cause a enough voltage drop that another PSU/driver will flicker?

It's not something I've figured out how to test for.  Do you have any suggestions, or better still any suggestions to fix it?  Perhaps a ferrite core on the mains power inputs, or would this interact problematically with the switch mode PSUs?

[James Mitchelmore]

What size PSU are you switching that is causing the problem? I have seen substantial inrush currents on what are relatively small sized units causing MCB trips at power on.

James

[seb303]

On Tuesday, January 10, 2017 at 5:28:47 PM UTC, James Mitchelmore wrote:

What size PSU are you switching that is causing the problem? I have seen substantial inrush currents on what are relatively small sized units causing MCB trips at power on.

I happens with a Meanwell HLG-320H-24 for example.  This is a 320W PSU, so quite large.  It also happens when groups of eldoLED POWERdrive 562 drivers are switched on.  These are 50W each, but 6 or 7 of them switched together.

Looking at the HLG-320H-24 datasheet, it states this for Inrush Current:
COLD START 70A (twidth=1010μs measured at 50% Ipeak) at 230VAC

So quite a substantial current but for a very short time.  I guess the other PSUs are not coping with this momentary loss of power.

Any ideas how to solve this?  I'm reluctant to leave all the PSUs running all the time even when no lights are being used.  Seems a bit wasteful.

[Kops]

Hi Seb,

I am experiencing absolutely same behavior (flicker on the DMX controlled outputs when PSUs for other sections are powered up). I originally thought, that this is caused by some noise on the DMX wiring while the DMX dimmers are initiated (dedicated zone PSUs for certain DMX dimmers are controlled by relays), and I have not yet successfully solved that issue. I am using five MeanWell HLG 240 and I don't believe it is the inrush current causing this (in the sense that it would influence other PSUs 24V outputs), but I might be wrong. The flicker is only noticeable on lights behind DMX RGBW dimmer, it is almost like an error on the DMX bus, moreover it is not visible every time. There is some random factor that causes this being visible only with certain power up's.

I would appreciate any thoughts or findings on this topic. I am going to investigate on my end as well and will share any findings.

Thanks a lot.

[seb303]

Having done some further testing I've come to the conclusion that the flicker is definitely caused by interference on the DMX bus, rather than voltage drop.  James Mitchelmore had convinced me otherwise, but my tests are fairly conclusive:

(1) I have two main lighting circuits coming from the consumer unit, each running over a separate 1.5mm cable (17m long) to the Loxone cabinet.  I am finding that switching a PSU on one lighting circuits will cause a flicker on lights on the other circuit.  The cable to the consumer unit itself is 25mm so shouldn't be much voltage drop over this.  Furthermore, the inrush current period is very short and so I find it unlikely that such a brief voltage drop would affect the other PSUs anyway.

(2) Further proof is that sometimes turning on a PSU will cause lights that aren't on to flash momentarily.

In retrospect I think it would have been a good idea to used shielded cable for the DMX bus.  But there are some other aspects of my DMX wiring that I have to check: I don't remember if I added a terminating resistor to the bus for one thing, and I also can't remember if I wired it as a true daisy chain, or if I have a partial star topology (there are 3 groups of drivers in different parts of the building).  If I have wired it as a 3 chain star, then a DMX splitter may help.  These are things I have to check and I will report back ....

[Duncan]

a dmx splitter also cleans up and retransmits the signal, so this option may well be worth it anyway, and a star wiring with no termination wouldnt help, so these are well worth sorting out anyway.

i have had issues on larger installs relating to dimming on lighting circuits causing interference with other adjacent lighting circuits, including induced currents in adjacent circuits causing RCD breakers to trip so without a oscilloscope or something similar it can be difficult to see whats going on - i bought a £25 usb oscilloscope from ebay which for the money is fantastic to see what spikes etc are happening on the wiring, - you can see the waveform on the mains to see if its voltage droop, or interference spikes for instance

[Peter]

Happy New Year everyone.

Hopefully someone will be able to resolve this DMX wiring problem.

I have 3 x Loxone RGBW dimmers followed by a LTECH 512 dimmer. The use of RJ45 connections on the latter isn’t entirely helpful but i believe that DMX+ is the orange / white of the cat cable and the DMX- is the orange. 1 & 2 of the T568B standard as per the documentation . I’m also summarising that 7 & 8 is wired to the N (GND) of the last Loxone dimmer? The LTECH is powered separately to the Loxone dimmers).

Hope that makes sense. 

[David Wallis]

please don't confuse N and Gnd! 

On Thu, 2 Jan 2020, 19:44 Peter, <s69...@gmail.com> wrote:

Happy New Year everyone.

Hopefully someone will be able to resolve this DMX wiring problem.

I have 3 x Loxone RGBW dimmers followed by a LTECH 512 dimmer. The use of RJ45 connections on the latter isn’t entirely helpful but i believe that DMX+ is the orange / white of the cat cable and the DMX- is the orange. 1 & 2 of the T568B standard as per the documentation . I’m also summarising that 7 & 8 is wired to the N (GND) of the last Loxone dimmer? Hope that makes sense.


--
You received this message because you are subscribed to the Google Groups "Loxone English" group.
To unsubscribe from this group and stop receiving emails from it, send an email to loxone-englis...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/loxone-english/1f0999ef-6f19-412a-bd4a-0f7325e2424d%40googlegroups.com.

[Peter]

Ok, so Pin 7&8 (brown/white & brown) from the LTECH dimmer RJ45 connection wired back to the GND terminal on the Loxone RGBW dimmer.

Hope that’s clearer.

[Seb]

That is an odd diagram. Putting it to one side, DMX has two cables in and two out, with each unit daisy changing off the one before. Don’t forget the last unit having the resistor bridging +/-

Therefore, based on T568B wiring:

DMX in:
+ is white/orange
- is orange

DMX out:
+ is brown/white
- is brown

[Peter]

Hi Seb, thank you for the reply. You’re right the diagram is odd and also confusing. Initially I thought the same as you however the orange pair is both DMX in and out and the brown pair is GND. Which is how I’ve now wired the drivers and everything works as it should. Having said that I’m still a little uncertain re the GND wire. Loxone drivers (the old type) do not have a third connection for a GND however all my other drivers do. I think I am right in saying that the GND of a LTECH driver should be wired back to the GND of the Loxone driver which precedes it in the daisy chain? On the same theme, the GND of a different LTECH driver should also be connected to a Loxone driver which follows it in the chain. If that doesn’t make sense I’ll do a rough drawing.

[Seb]

Do you have the full diagram of that driver?

I take it the DMX in/out is the same terminal like the loxone and unlike my crappy 302 drivers that have separate terminals for in/out.

As for GND, what are the other terminals on unit? It could just be an earth?

[Duncan]

the orange and orange white are the signals and brown/ brown white are the zero/common. They do not have in and out on separate wires, but most dmx devices using rj45 have 2 connectors for daisy chaining, either one can be in or out

the remaining connectors within the rj45 are unused and not connected internally.

the dmx signals are differential so dont technicall need the common/zero/ground to work, however it does ensure a common ground between different bits of equipment (with separate power supplies) to ensure that there are not large voltages/currents flowing along the signal wire

[Peter]

Seb, that was the full diagram, well, actually the script on the driver itself. No other documentation except how to set the dip switches and a very short troubleshooting guide. Even more confusing, if the dip switches are set incorrectly the guide states the driver will illuminate red. What it fails to point out is that the driver had a red internal LED which illuminates when its powered up. The trouble shooter refers to a different red LED which is mounted externally and not immediately obvious to an untrained eye.

Duncan has summed the rest up quite nicely thank you