Inrush currents and solutions......

[davesparks]

I think one fundamental issue here is that LED lighting is still being installed and controlled by a method designed for incandescent lamps.
Switching and/or dimming the supply to the electronic driver circuitry is never going to give the best results for inrush, lifespan or dimming performance.
Electronics like this performs much better with a constant supply and a separate control input.

New technology needs to have a new approach to the way it is installed and controlled otherwise these problems are never going to go away.

 

[Pete999]

How can he calculate this when the manufacturers haven’t supplied the required information?

Hi Dave I was in a hurry and was going to suggest to the OP that he calculate the circuit parameters to see if he could use a different type of CB

 

[Midwest]

I concur the advice re JCC, so Enlite might be of same assistance.

Although I recently fitted some JCC Hybrid 9's, where there is no such advice on maximum no's on B or C mcb's in the installation instructions. So I rung their technical, who said they don't provide that default info anymore. They did say to ring the local rep, who might provide more detailed information. Looks like their testing has taken some cutbacks.

 

[Octopus]

Initial thoughts is certainly a C6 or C10, will have to see what the Zs is first.

Another interesting strategy ................ if you 1st fixed it you should have a VERY good idea what the Zs is likely to be.................

 

[Dobes_88]

Another interesting strategy ................ if you 1st fixed it you should have a VERY good idea what the Zs is likely to be.................

As in #6 post I've taken over the job from a friend (family bereavement) and client wants some lighting on by end of next week, first time on site yesterday having a nose around and just preparing for worst case scenarios.

Kitchen lighting has around 25 fittings, couple of switches operating 10-12 lights.

Upstairs and downstairs lighting around 35 fittings with hall switch operating around 14 lights.

Upstairs lighting around 35 fittings with landing switches operating around 12 lights, rest of rooms have around 4-8 lights with single switch.

 

[Pete999]

I f as you say did the 1st fix, does he have any proposed Zs calculations from the circuit designs

 

[Dobes_88]

Not that I'm aware of! It's not a massive house so the cable runs aren't too long, will look like Terminal 5 when up and running though!

 

[Octopus]

As in #6 post I've taken over the job from a friend (family bereavement) and client wants some lighting on by end of next week, first time on site yesterday having a nose around and just preparing for worst case scenarios.

Kitchen lighting has around 25 fittings, couple of switches operating 10-12 lights.

Upstairs and downstairs lighting around 35 fittings with hall switch operating around 14 lights.

Upstairs lighting around 35 fittings with landing switches operating around 12 lights, rest of rooms have around 4-8 lights with single switch.

Are ALL these lights on 1 circuit?

 

[Dobes_88]

Are ALL these lights on 1 circuit?

Thankfully not, 3 x separate circuits -

1 - Kitchen lighting
2 - Ground floor lighting
3 - !st floor lighting

 

[Midwest]

As others have said, it would be the switching which might cause the issue. Consider splitting up the luminaires across more switches.

 

[redbarnman]

One cycle of mains current takes 1/50th second or 20milli seconds. In-rush currents may be very high but last for well under a milli second - that is much less than one mains cycle. There is very little power in a transient and heating effects can be ignored as far as cables are concerned so there should be no need to change any cabling. Traditional Z calc's don't work with transient currents because the effective frequency of the transient is 1000's of Hz so that is not much help and as electricity travels along wires at around 0.8 the speed of light or a quarter of a million kilometers per second the length of cable is not going to have much impact either. You also have to remember that tungsten, halogen and fluorescent lamps also have higher inrush currents - smaller peak but over a longer time. As LED lamps always need a driver (built in or external), it is the driver circuit that will determine the inrush current and these driver circuits contain far more sensitive components than a mains switch. Unless someone comes up with real measured transient data I don't see why you wouldn't work on the basis of 20 off 10W LED's per MCB circuit. By the time you have then broken down to 6 to 10 lamps per final switched circuit there is no problem. Changing from a B curve to C curve will only provide a factor of 2.5 over current capability. You would be better running off dimmers. I'm sure if LED's were a real problem they would be in scope for the 18th edition.

 

[Octopus]

Thankfully not, 3 x separate circuits -

1 - Kitchen lighting
2 - Ground floor lighting
3 - !st floor lighting

Then I would suggest that you have nothing to worry about .......... but based on the assumption you have low Zs readings then use of Type C breakers from install may prevent the possible call back later .....

 

[Dobes_88]

Just a little update, lights all up and running on a C10 MCB's (highest Zs being 1.1 Ohms) with no issues when turning multiple switches on at the same time.

Also had a response from Enlite for those interested -

"The inrush on the EN-DE8 is 0.6A PK measured in 100ms. We have tested 2-10 fittings on our AU-DPS400X dimmer module".

Thanks to all for your input!

 

[1Justin]

Good. - & just FYI to help understand the range of these figures and how the method of measurement greatly affects to answer.. The 30W LED drivers which I mentioned in a previous post (Meanwell) where I had fitted eleven into one circuit, are individually specified at 55 Amps /500us. Hence since Meanwell use 1/200th of the time slot specified by Enlite and come out with a peak current nearly 100 times that you got from Enlite. (OK, a more powerful driver, but not by that extent). It illustrates the potential for problems though.

 

[alasdairp]

My putting the cat amongst the pigeons again...
The IET have a 200 A4 page Code of Practice for the Application of LED Lighting Systems (2014).
They also have a CoP - wait for it - for Low and Extra Low Voltage Direct Current Power Distribution in Buildings (2015) (only 60 A4 pages) that predicts we will all soon be distributing LV DC for lighting and electronics alongside normal 230 mains for high power devices and appliances - compliant with BS7671 subject to suitable sleeving and labeling - installing a second DC CU alongside the normal one.
Should be fun, eh?
One advantage its that it would get rid of the in-rush current problem for LEDs.

 

[davesparks]

My putting the cat amongst the pigeons again...
The IET have a 200 A4 page Code of Practice for the Application of LED Lighting Systems (2014).
They also have a CoP - wait for it - for Low and Extra Low Voltage Direct Current Power Distribution in Buildings (2015) (only 60 A4 pages) that predicts we will all soon be distributing LV DC for lighting and electronics alongside normal 230 mains for high power devices and appliances - compliant with BS7671 subject to suitable sleeving and labeling - installing a second DC CU alongside the normal one.
Should be fun, eh?
One advantage its that it would get rid of the in-rush current problem for LEDs.

Using a driver which has a permanent feed and a control input will get rid of the inrush problem, without the need for an entire DC distribution.

 

[alasdairp]

Using a driver which has a permanent feed and a control input will get rid of the inrush problem, without the need for an entire DC distribution.

I agree, that would be much better. However, 20 or 30 mains-voltage switch-mode power supplies located in ceiling voids must be more of a fire hazard and reliability issue than having one high quality DC supply for LEDs and electronic devices in the house. I am not push for this, just pointing out the way IET thinking seems to be going.
Many LED 'bulbs' have inbuilt 230 V AC conversters built inside them and they are by far the main cause of failure. The LEDs themselves now very rarely fail (some early high power ones did - and the higher the power the more they still fail).

 

[davesparks]

I agree, that would be much better. However, 20 or 30 mains-voltage switch-mode power supplies located in ceiling voids must be more of a fire hazard and reliability issue than having one high quality DC supply for LEDs and electronic devices in the house.

I suspect you would still require an electronic power supply of some sort for each light fitting to drive the LEDs correctly as they will almost certainly need constant current rather than constant voltage supplies.