Overdrive stats requested

[across the pond]

Does anyone have any figures/tables that indicate lumen output at different currents for luxeons under overdrive conditions (1W and 5W)?

And does anyone know a safe limit for the 5W (assuming adequate heatsinking).

Thanks.

 

[Doug S]

Originally posted by across the pond:
Does anyone have any figures/tables that indicate lumen output at different currents for luxeons under overdrive conditions (1W and 5W)?

And does anyone know a safe limit for the 5W (assuming adequate heatsinking).

Thanks.

<font size="2" face="Verdana, Arial">You really cannot consider just current alone since increased current results in increased die temperature. The efficiency effects of current and temperature are different for each color. Consult the datasheets. The amber, red-orange, and red ones are especially adversely affected by temperature. In the case of the amber, the output at 120C is only 20% of what it is at 25C for the same current! The luxeon datasheets show the effects of current and temperature seperately. Beyond a small fraction of rated current, efficiency drops with increasing current but less so for the amber, red-orange, and red ones. For all temperatures, efficiency drops with rising temperatures. Combining the two sets of graphical data will give a reasonable approximation of real performance. Unfortunately, the luxeon datasheets do not show data for currents beyond their "absolute maximum" ratings. I recently took some data on a 1W cyan configured with a Tj-Amb thermal resistance of 47C/W. I tested at various currents and normalized efficiency at a current of 320 mA. Here are the relative efficiences as a function of both I and P

If****Eff(I)****Eff(P)
20***1.15******1.46
40***1.27******1.54
80***1.29******1.47
160**1.17******1.26
320**1.00******1.00
480**0.86******0.83
640**0.76******0.71

Efficiency as a function of I, Eff(I), is most relevant to direct drive, resistive, or linear CC ballasts. Efficiency as a function of P, Eff(P), is most relevant to switching regulator ballasts.

Regarding safe limits for the 5W, I would go by the datasheet [that's what is it for] which states 700 mA.

 

[Rothrandir]

well, i don't have the stats, but you can pretty much drive a luxeon quite high. pyscho bob, i think it was, attached a (blue 1 watt) to a computer heatsink and fan, and was driving it at ?v and around 2350mA. i think that is about what it was.

 

[evan9162]

I have some test data to add to the mix

The other night, I ice-water cooled a 1W Royal Blue /O. The Al PCB is thermal epoxied to an Athlon heat sink. I poured a bowl-full of water, put about 5 ice cubes in, and dipped the HS in fin-first.

The heat sink stabilized to about 9 degrees C.

At ~700mA, it was appreciably brighter than rated current.
However, bumping the current to 1300mA gave no appreciable brightness gains. The heat sink never deviated in temperature (stayed at 9 C).

At 1300mA (~5W), with the heat sink at 9C, and an estimated thermal resistance of 20C/W (17 J-B, ~3 conservative estimate for the thermal epoxy), Tj was probably at 100C. Lumen/Temperature derating is spec'd at 80% for the RB..

So the output must simply level out due to other limitations..

It seems that about 600-700 is about the max reasonable overdrive before you reach diminishing returns.

-Darin

 

[Minjin]

I just measured my Lamda Ill Pill with 2 fresh lithiums at an even 1 amp tonight. My Inretech with 2 lithiums measured 150mA. Can you guess which was brighter?

That Ill Pill has to be putting out some serious lumens...

Mark

 

[Vikas Sontakke]

How do you measure the Ill Pill Luxeon current? You are only measuring the input to the Pill, not the output from it.

Obviously, even assuming 80% efficiency, IllPill probably is pumping out at least four times the current and at somewhat higher voltage!

Do you get lot more output from the IllPill with Lithium as compared to freshly charged NiMh? I recall Lambda saying that he has optimized his circuitry to work best on NiMh and Lithium did not add any significant brightness to the output.

Is there any risk in running Lithium with IllPill?

- Vikas

 

[remuen]

Originally posted by across the pond:
Does anyone have any figures/tables that indicate lumen output at different currents for luxeons under overdrive conditions (1W and 5W)?

<font size="2" face="Verdana, Arial">IMHO it will be difficult to get such figures in lumen as nobody has a integrating sphere photometer (maybe McGizmo soon will have one

). I recently measured an old 1W white LS with my brandnew luxmeter. Here some figures which will give you an indication of the relationship between current and brightness (lux measured in 1 meter distance from the LS)

mA | Lux
---------
200 | 151
250 | 178
300 | 197
350 | 215
400 | 233
450 | 245
500 | 262
550 | 273
600 | 285

 

[Doug S]

Originally posted by remuen:
I recently measured an old 1W white LS with my brandnew luxmeter. Here some figures which will give you an indication of the relationship between current and brightness (lux measured in 1 meter distance from the LS)

mA | Lux
---------
200 | 151
250 | 178
300 | 197
350 | 215
400 | 233
450 | 245
500 | 262
550 | 273
600 | 285

<font size="2" face="Verdana, Arial">Rene, what amount of heaksinking was used for this test and were the measurements taken at thermal steady state? I am begining to like this thread. People are coming forward with real data.
Those that are inclined to drive the **** out of these devices should note the relatively small incremental increase in light per unit increase of current at the higher levels. Driving 3 devices at 200 mA each would produce 60% more light than one driven at 600 mA.

 

[remuen]

Originally posted by Doug S:

Rene, what amount of heaksinking was used for this test and were the measurements taken at thermal steady state?

<font size="2" face="Verdana, Arial">Doug
I'm using two LS only for testing boosters. They both are mounted on quite a big heat sink formerly used in an sound amplifier. Of course when measuring such data I first let these LS get warm for a few minutes until the current stays on a steady level.

Originally posted by Doug S:

Driving 3 devices at 200 mA each would produce 60% more light than one driven at 600 mA.

<font size="2" face="Verdana, Arial">I can provide you some more data proving your statement as I also have measured the brightness with two LS. This will give you a picture of what you could win when you use two LS instead of only one driven with a higher current.

The two LS are connected in series therefore a current of eg. 300mA through these two LS is about the same power consumption as one LS driven with 600mA. These figures below show the brightness in Lux at a certain current, measured in 1 meter distance:

mA | 1LS | 2LS (in series)
---------------
200 | 151 | 283
250 | 178 | 335
300 | 197 | 381
350 | 215 | 426
400 | 233 | 477
450 | 245 | 520
500 | 262 | 548
550 | 273 | 577
600 | 285 | 619

So if you compare these figures you will note:
Doubling the current with 1 LS from eg. 300mA to 600mA will give you an increase in brightness of +69%. Using two LS driven with 300mA (which is as already said before the same power consumption as 1 LS with 600mA) will give you an increase in brightness of +93%.

Hope these figures help to better understand all this.

 

[Jonathan]

I bet that it gets even better than this if you measure voltage in order to determine power consumption.

Rene states 'The two LS are connected in series therefore a current of eg. 300mA through these two LS is about the same power consumption as one LS driven with 600mA.'

This statement is approximately true in that the voltage doesn't change too much, but I would expect the Vf of the single Luxeon driven at 600mA to be about 300mV higher than the Vf when driven at 300mA, meaning that the two LS in series at 300mA would be using about 180mW less power than the single LS driven at 600mA...not much, but about 5% of the total power being used.

-Jon

 

[across the pond]

Some good facts and figures guys. Thanks.
What I was looking for is the relationship between current increase and lumen increase.

You have to try and take some things for granted - like middle of the road luxeon performance, and assuming the heat sinking keeps pace with the led - if you want easy to interpret data.

 

[Doug S]

Originally posted by remuen:
Using two LS driven with 300mA (which is as already said before the same power consumption as 1 LS with 600mA)

<font size="2" face="Verdana, Arial">I did not say this. The power with the 2 at 300 mA each will be a bit lower due to the lower Vf.
Thank you for sharing your data.