LEDs, real world efficiency

[fredfish01]

I thought I would post this here as there are a few people here who seem to know a lot about LEDs.

 

There has recently been a lot of fud posted on 'that other site' by both the lovers and haters camps about the efficiencies of LEDs.

 

I've seen a lot posted about the efficiencies of individual diodes, but not a lot on finished units. Recently there was a thread 'over there' that looked at the total radiated power of 3 lights: a Kesil 150, an Ushio 14K 250W, a Radion XR30W.

 

I was surprised to see that both the Kessil and the Ushio were at about 16% efficiency with the Radion being at around 26%. It was the Kessil that surprised me. I would have expected it, as a new unit with quality LEDs, to be more along the lines of the Radion, yet it was not.

 

That leaves me with a few questions:

 

Just how efficient or inefficient are those Chinese units using Epileds LEDs?

 

Is the whole efficiency story being told just by output? (I think not)

 

To make thins more interesting, I see that Reefll has posted both watts consumed and total watts output numbers for their new pendants that yield efficiency number in the range of 35%. For instance, for the Vega 60 they give the max watts consumed as 58w and total radiated power as 21w. Unless I've misunderstood these numbers, that gives us an output efficiency of 36%, more than double that of the Ushio or Kessil.

 

FWIW, I understand total radiated power to mean a measurement of the light output in a 360% sphere around the light ie: everything. this is how the Kessil, Ushio and Radion were measured.

 

I'm curious to hear from the LED smart guys here.

[Arkayology]

I saw that thread. Lots of misinformation. I left it very quickly.

[DNK]

Hi guys!

 

fredfish01, thank you for raising this important issue! The question of real efficiency of LED lights is rarely discussed and, as you noted, this figure can vary significantly.

 

At the design stage of our pendants we applied huge efforts to make our lights as efficient as possible with the best existing technologies. This efficiency is important not only from the standpoint of saving consumers money: each percent of loss is converted to heat that needs to be dissipated. With less heat generation we were able to make the lights small and lightweight, and for the VEGA line we were able to achieve a completely fanless solution while keeping the LED crystals at comfortable temperature.

 

I would like to point out the three important reasons which let us build LED luminaries with as high efficiency.

 

1. In our VEGA line pendants, at luminarys max output, we operate the LEDs at about 50% of maximum current specified by the manufacturers datasheet. This improves the efficiency of each LED by roughly 20% compared with operation at max power.

 

2. In our budget-oriented VEGA60 the LEDs are not the top highest efficiency that exists, as in our other products (you can look up all the LEDs used in our lights on the TECH SPECS tab of our product pages, with links to corresponding datasheets). But dont mix our LEDs from Semileds with cheap Epileds that are commonly used in Chinese fixtures. Semileds is an advanced manufacturer (and is capable of making best UV LEDs so far), whereas Epileds uses outdated technology and its LEDs are not that efficient.

 

As for our VEGA 100 and SIRIUS 180, in both pendants we only use top quality LEDs, best bins from the best brands that are commercially available at the time of manufacture. In VEGA 100 too, the LEDs are operating at about 50% of current at most, therefore VEGA 100 sports highest conversion efficiency: 43%

 

3. As you pointed out, the question of total efficiency, i.e. the ratio of emitted radiometric watts to watts consumed from the socket. Note that our Total consumed power parameter, besides LEDs, includes power consumption of all electronic components, such as drivers, controllers, etc. To achieve such outstanding efficiency, we spent a lot of time and resources on each component of our luminaries. Our drivers sport highest efficiency, we are using high quality power supplies, etc.

 

As a final note, after recent rebranding, Semileds is now called TSLC.

[fredfish01]

Thanks for the reply DNK. Its funny how both sides get all excited about LED efficiency, yet not many really know what that means or what it might be for a given light.

 

43% represents a HUGE efficiency gain over MH or T5.

 

I have a question. Is 20% a typical gain from running diodes at some optimal point lower than 100% output or does it vary depending on the diode?

 

It would be interesting to have someone who has access to the proper equipment do some efficiency testing on some of these cheap Chinese LEDs.

[DNK]

I believe that 20% improvement of efficiency is indeed, quite typical for most LEDs. For some LED types, such as red-orange and amber (not PC-Amber that we are using) this figure will be even higher, around 30%.

 

Actually, the older is the technology by which the LEDs are manufactured, the higher will be efficiency gain when operated at lower current.

 

In this sense, Chinese LEDs will gain more when operated at low currents, but then they loose their only advantage of low price