DALI lighting, need some basic info

[cwoern]

Hi all,

I am not a lighting designer, but I need to review a lighting design that was done by others. It is fairly big, and I lack profound knowledge to judge if (at least on paper) the design will work.

To keep it simple:

We have 213 DALI controllers DDBC300 utilising their (mostly) all three channels to control light circuits. Each channel/circuit has up to 25 lights. I read that typical limitations are:

• 64 DALI members
• 300m maximum length.

My questions are:

• Is it 64 DALI members per controller or 64 DALI members per channel? In other. words, can our DDBC300 have up to 3 x 64 lights connected to it?
• Is it 300m maximum cable length added between all lights of a channel/circuit? Some circuits (one channel) reach up to 500m; the designers suggest using 2.5 mm² DALI cables.
• Could a group of lights (i.e. three DALI lights in a small workshop) be "one member/one DALI address", or does each DALI member within a circuit require its own address?
• Is it correct that the "TunneLogic tunnel lighting control system" has a limitation of 200 controllers?

Edit: It was meant to say 200 (not 2000) controllers. Now corrected.


Regards, Chris

 

[plugsandsparks]

If it was my project , i would be getting the design approved by the manufacturer, Philips ?

The last one i did, the biggest decision i had to deal with, which put me at odds with the lighting designer was whether to use one Dali Driver per lamp or one Dali driver per light zone, i.e. your workshop example.
I opted for one driver per zone which massively reduced the number of controllers.
It did mean that i had to own the selection of lamp drivers but Tridonic have a huge range of drivers and the new ones can be programmed for multiple constant current LED lamps.
Looking at the spec of that controller, it is saying it has 3 channels at 64 addressable devices, i.e. 192 addressable devices, now whether that is 192 addressable lamps or lamp zones is what is dealt with in my comment above.
The bus cable is a daisy chain cct, its possible you could exceed 300m using 2.5 cable but assuming you can get 2.5 cable into the terminal blocks, again i would get Philips to comment as Dali relies not just on a good quality signal in terms of voltage differentials but there is also timings to consider. It may be that the designer has experience of going outside specs of Philips, after all its a pretty bomb proof system

 

[cwoern]

If it was my project , i would be getting the design approved by the manufacturer, Philips ?

The last one i did, the biggest decision i had to deal with, which put me at odds with the lighting designer was whether to use one Dali Driver per lamp or one Dali driver per light zone, i.e. your workshop example.
I opted for one driver per zone which massively reduced the number of controllers.
It did mean that i had to own the selection of lamp drivers but Tridonic have a huge range of drivers and the new ones can be programmed for multiple constant current LED lamps.
Looking at the spec of that controller, it is saying it has 3 channels at 64 addressable devices, i.e. 192 addressable devices, now whether that is 192 addressable lamps or lamp zones is what is dealt with in my comment above.
The bus cable is a daisy chain cct, its possible you could exceed 300m using 2.5 cable but assuming you can get 2.5 cable into the terminal blocks, again i would get Philips to comment as Dali relies not just on a good quality signal in terms of voltage differentials but there is also timings to consider. It may be that the designer has experience of going outside specs of Philips, after all its a pretty bomb proof system

If it was my project , i would be getting the design approved by the manufacturer, Philips ?

The last one i did, the biggest decision i had to deal with, which put me at odds with the lighting designer was whether to use one Dali Driver per lamp or one Dali driver per light zone, i.e. your workshop example.
I opted for one driver per zone which massively reduced the number of controllers.
It did mean that i had to own the selection of lamp drivers but Tridonic have a huge range of drivers and the new ones can be programmed for multiple constant current LED lamps.
Looking at the spec of that controller, it is saying it has 3 channels at 64 addressable devices, i.e. 192 addressable devices, now whether that is 192 addressable lamps or lamp zones is what is dealt with in my comment above.
The bus cable is a daisy chain cct, its possible you could exceed 300m using 2.5 cable but assuming you can get 2.5 cable into the terminal blocks, again i would get Philips to comment as Dali relies not just on a good quality signal in terms of voltage differentials but there is also timings to consider. It may be that the designer has experience of going outside specs of Philips, after all its a pretty bomb proof system

Thanks for your thoughts and responses, much appreciated.

1.) "If it was my project , i would be getting the design approved by the manufacturer, Philips ?":
I am in contact with the nominated supplier. It is an odd situation. They, understandably, won't get roped into a design role, as it is claimed to be complete and functional by the design company. At the same time, I believe that the supplier just started to look into details and get back to me (as they have technically no relation to the external design company). I am also a bit unsure if the PM I am dealing with at the supplier end is the right technical person. In between the lines I can sense that they don't agree with the design in all parts

2.) Mentioning that, I now learned something. I probably assumed incorrectly that all lamps must have their own driver (and I think that is likely our case, but I will check and confirm that).

3.) I also just got off the phone with the PM of the supplier. He is almost sure (rightfully?) that the limitation of 64 members applies to the entire controller. That contradicts what you stated and what I also found on the www after digging a bit. However, the PM is awaiting further details from the pros within his company and will get back to me.

4.) DALI cable length: Your comment regarding the terminal block is a good one. I should confirm that on the DDC end. Consumers run either through associated JBs (one per light) or daisy chain using plugs and sockets in the lights. I hope/assume that those plugs are capable of accommodating the OD of the cable and core size. I'd better check and confirm that as well.
Whether the designer has experience in exceeding the length or not, I can't answer. I am awaiting a response from the lighting system supplier regarding that. Some circuits include repeaters, which likely mitigate the problem; others don't. However, even this source doesn't say it is recommended to exceed 300m when using 2.5mm2 cable: Practical Issues Regarding DALI System Wiring - https://tillumelight.com/blogs/home-automation-lighting-application/practical-issues-while-wire-a-dali-system?srsltid=AfmBOoqf40iOUHp_w5aiHN9UweeX4c_ryZutfYBDwc8XEm-QDoOOiVQY


Regards, Chris

 

[plugsandsparks]

Sort of confirms why someone needs to contact Philips. Below is a basic spec PDF

https://dynalite.com/wp-content/uploads/2024/09/DYN_DDBC320-DALI_SPEC-R17.pdf

At the end of it, i sort of read it as 64 / Dali Line with 3 lines ?????

Does not surprise me that you end up with one driver per lamp and it does have advantages wrt fault finding as each lamp can be monitored, however if there are only 64 lamps per controller you would not get very far, lol
Again if its one driver per lamp then i would also check out the maximum number of groups you can create, e.g. if one room has 8 lamps, then you would want to group that lot together and call it Boardroom or something and then you can ensure they all work together.
Hopefully the whole schedule has already been created so you know which lamp is in which group and what part of the building it is in.......
Oh and the spec indicates it can take 2.5sqmm cable

 

[cwoern]

Sort of confirms why someone needs to contact Philips. Below is a basic spec PDF

https://dynalite.com/wp-content/uploads/2024/09/DYN_DDBC320-DALI_SPEC-R17.pdf

At the end of it, i sort of read it as 64 / Dali Line with 3 lines ?????

Does not surprise me that you end up with one driver per lamp and it does have advantages wrt fault finding as each lamp can be monitored, however if there are only 64 lamps per controller you would not get very far, lol
Again if its one driver per lamp then i would also check out the maximum number of groups you can create, e.g. if one room has 8 lamps, then you would want to group that lot together and call it Boardroom or something and then you can ensure they all work together.
Hopefully the whole schedule has already been created so you know which lamp is in which group and what part of the building it is in.......
Oh and the spec indicates it can take 2.5sqmm cable

Ouch, that's maybe where the supplier got his 64 from. If all our lights have a driver (and that's already confirmed by the supplier) and we read that spec sheet right, I will have a real problem, as some controllers serve more than 64 drivers. The max. cable length is now another concern as some channels/circuits have a total length of up to 500m connected. If the 300m limitation is a total per controller, my problem is even bigger, as some combined runs are 330m + 343m.

I shall see what the experts of the suppliers have to say and if there is an interest here in the forum, I will post their view.

Regards, Chris

 

[plugsandsparks]

Another question, even though the controller spec indicates 3 X 64 drivers, could you have a group of lamps, say boardroom fed from say Dali line 1, driver 64 and Dali line 2 driver 1...... if you cannot combine lamp groups across the lines, that could mean you dont get all the benefit from the capacity. lots of question as usual with this tech. Good Luck

 

[cwoern]

Another question, even though the controller spec indicates 3 X 64 drivers, could you have a group of lamps, say boardroom fed from say Dali line 1, driver 64 and Dali line 2 driver 1...... if you cannot combine lamp groups across the lines, that could mean you dont get all the benefit from the capacity. lots of question as usual with this tech. Good Luck

Honestly, I am unsure why the designers assumed the need for a driver/light (could be a project requirement). If the grouping is an option (at least for several of the almost 5,000 lights), it would also mitigate other problems (i.e., the lighting supplier has already informed me that the MCU software can only handle 200 controllers; currently, we have more in the design). However, it may also introduce new problems, such as even more devices per controller, which could further exceed cable lengths. I will see what the supplier is telling me when they assessed the complete design.

 

[cwoern]

Having spoken to the integrator, I learned the following: Some lights are composed of up to three modules, meaning that those lights (although technically only one lamp with one asset tag) have three DALI drivers, which makes things even more complicated (in terms of adding devices to a circuit). Something that the designers didn't take into account. Grouping or re-grouping will not be a quick or straightforward task as there are redundancy requirements and other conditions that need to be met (i.e. only certain areas/sections can fail at a time). Given that we are discussing more than 6,000 luminaires over a distance of 6.5 km, there appears to be a considerable amount to consider. The integrator also recommended not exceeding 300m, even when using a 2.5 mm² cable. Although 3x 64 drivers can be connected to the chosen DALI controller, the integrator suggests not exceeding 60 to have some spare options if needed. To make things even worse, the control logic can handle only 2 x 100 DALI controllers, meaning there is a lot to consider when redoing or correcting the design. We shall see how that all pans out.