Interesting article on LED bulbs in ET&T ...

Last year we dissassembled a GU10 LED bulb - I forget which manufacturer it was. I was curious to find out what kind of driver circuit was being used to "control" the current through the LEDs, which is quite a difficult thing to do (cheaply) given a 220-250V AC supply. I thought I might find a cheap switched-mode circuit or something like that.

Instead, what we found was that about 80% of the electrical energy was dumped into a simple resistor which was used as a crude "constant current" device. In this case, whether the LED is 60 or 99% efficient becomes almost irrelevant - its the efficiency and cost of the driver circuit which is important, and making more efficient, reliable drivers at such small power levels from 220-240VAC, within the cost constraints, is pretty hard.

Maybe a 24V DC low-power domestic ring, specifically for LED lighting (and nothing else) is the way to go ... it would make the driver circuits much easier to build?

Andrew

-------------------------
Dr. Andrew Roscoe

http://personal.strath.ac.uk/andrew.j.roscoe

Nothing new in that Andrew - look back far enough or deeply enough and you'll find plenty of dimming devices using power loss resistors - remember the tungsten or resistor ballasted fluorescent lamp circuits - you just simply dropped what voltage you didn't need for the fluorescent across a resistance - if it was a tungsten lamp you got some light (and heat) - if it was a resistor, you just got heat - sometimes the heat was useful, in many cases it wasn't - if only due to diurnal cycles

Is the economics of a 24V lighting ring in domestic houses worthwhile - I doubt it (at least for many years to come)

Regards

OMS

-------------------------
Failure is always an option

I've seen some cheap-and-cheerful AC step-down circuits, where electrical isolation was not required, using the reactance of a small film capacitor to drop the voltage. Doesn't help the power-factor but probably more energy efficient, overall, than using a simple resistor.

What about keeping the 240V ring main and unhooking the the upstairs and downstairs lighting circuits etc at the consumer unit and efficiently transforming and regulating to 24V DC there?

This could work if the total energy consumption for well designed domestic lighting is low enough when using LED lighting or other low current DC lighting fitted throughout.

(For example 5A at 24V (120W Max) or 5A at 48V (240W Max) per lighting circuit).

The DC power converter could be specified by the number of separate lighting circuits it can power and have current limiting and earth leakage detection fusing, and power saving mode when no lighting is required.

Are there any reasons (safety, fire risk etc) why this sort of conversion to an existing lighting circuit should not be allowed from a wiring regulation point of view?

You could piggy back the lighting control comms, signals on top of the DC supply, to give dimming and colour change controls for higher spec installations.

James Arathoon

LED lighting is something my son has been investigating for some time. His results from power consumption tests carried out on commercially available LED bulbs are similar to your own findings - in that it is common for more power to be dissipated in the driver circuit rather than the LED itself. He hasn't quite reached a conclusion on whether reducing power consumption of LED drivers will have a minimal impact on overall energy consumption from lighting compared to exchanging all less energy efficient bulbs for LED lighting, or whether it is something that manufacturers must take more seriously.

LEDs can be driven either from a constant current source, PWM, or both. The method used affects the colour of the LED during dimming. There is quite a good article in a magazine about LED drivers and if anybody is interested I will try and find it.





Low voltage domestic lighting using incandescent bulbs is commercially available but it has never been popular in Britain. Most systems work on 12V AC rather than 24V DC supplied from a high quality transformer. A low voltage LED lighting system with a driver circuit in each bulb would not require such a high quality transformer. Altervatively a switched mode PSU could be used. It would be quite easy to incorporate a battery back up into a low voltage lighting system to enable it to function during power cuts.



Are there any IET wiring regulations for low voltage domestic supplies? If existing wiring is used then the maximum current will have to be limited to 5A as that is the norm for domestic lighting circuit breakers . Assuming a 12V supply then 60W is the maximum power per lighting circuit.



The possibilities are endless. Over the years my son has designed and produced prototypes of a number of controllable lighting systems. Each LED driver circuit is allocated a code and data can be sent along power wires for switching and dimming. Security features could easily be incorporated such as particular lights that switch on between sunset and 11 PM or lights that switch on and off randomly to give the impression that the house is occupied.

From what I have found it appears perfectly safe to set a 60V DC standard for lighting circuits, (especially so if earth leakage detection trip are included as standard). That would allow up to 300W on a single 5A lighting circuit, more than enough for most houses, if the LED driver effficiency is as low as you suggest.


[Low Safety extra-low voltage
42.4V peak or
60V dc "Safe," user-touchable secondary circuit designed and protected to remain
under safe voltage levels in normal operation and under single fault;
double insulation

(IEC 60950-1 and other standards.)]

"New 60V LED Drivers from Infineon Improve Efficiency, Light Quality and Operating Lifetime of Solid State Lighting Systems"

http://www.infineon.com/cms/en...INFPMM201302-025.html

They are claiming Typical DC/DC conversion efficiency up to 98 percent and they give pricing levels.

Using DC/DC conversion techniques to DIM LED's might be better anyway, as I've heard that some people can see the modulated light given out by some of PWM dimmable LED lights currently on the market.

A switch mode power supply at the consumer unit giving out 60V can be over 80% efficient, and together with a DC/DC converter local to the LED lamp we get a power transmission efficiency in the 60-90% range, rather than the 20-30% range that you say is the norm at the moment.

This one seems like a no brainer to me.

James Arathoon

The Emerge Alliance is pushing for a 24DC standard.

http://www.emergealliance.org/...ace/StandardFAQs.aspx

This seems too low to me.

They say:

"Why is the standard set for 24 Volt DC?
Power over 30 Volts is not considered "safe" when conductors are exposed (without insulation) and must be enclosed in metal jackets, conduits and enclosures. Power less than 24 Volts (i.e. 12 Volt DC systems) can create more significant power losses in the wiring. The selection of 24 Volt DC provides the dual benefits of flexible, modular wiring and safe, efficient power distribution. Also, many digital devices used within commercial interiors today that inherently use DC power - such as occupancy and day lighting sensors - are already based on this voltage."

This appears to be guff and nonsense to me as no power system can be run without insulating and protecting conductors. In regards to safety the PoE standard allows for a voltage of 57V DC and BT runs 50V DC on its telephone lines. Has anyone ever be injured coming into contact with these sorts of voltages?

I haven't yet found a reason why DC power for lighting circuit upgrades should not be in the 50 to 60V DC range giving up to 300W of lighting on a standard 5A lighting circuit.

Such a conversion standard will be suitable for use in smaller industrial premises as well.

What are the possibilities for converting 50-60V DC to 30KHz high frequency (HF) AC for fluorescents?

Energy savings seem possible by using a 30KHz HF supply rather than a 50Hz supply. (with no flicker) Some DC to DC converters use intermediate 30 KHz switching frequency anyway, so are small low cost efficient 60V DC to 30KHz HF supplies feasible for fluorescents?

How do the ballast requirements and costs change when powering from 60 V DC?

See for example?
http://www.tlc-direct.co.uk/Te...heets/Lighting/HF.pdf

James Arathoon