Most Lumens from Amps and from Watts

Tobias Bossert

Newly Enlightened
Joined
Oct 7, 2007
Messages
146
Location
Germany, Munich
I made a mistake by comparing XT-E S2 instead of XT-E R5. S2 isn't availlable jet. I corrected the Excel sheet now.
Al statements in this thread which must be altered are marked red/bold now!


Which one is the best performing led at the moment?

This frequently asked question cannot be answered without specification of what is the meaning of "best".

For this post I will ignore things like die size, angular pattern, color rendition etc and focus on maximum flux at given electrical drive conditions.

If you drive a led with a linear driver, you are interested to get as much flux as possible from the drive current. The forward voltage of the led (and therefore the led power too) is of no interest, as long as it remains below supply voltage. In this application the quantity 'Lumens_per_Amps' counts.

If you drive a led with a buck or a boost driver, than you are not interested in the current directly, you would like to get as much flux as possible from power consumption. In that application the quantity 'Lumens_per_Watts' counts.

I used Cree PCT (product characterization tool) with the setting TSP = 25° to compare XR-E R2, XP-E R4, XB-D R5, XP-G S2, XT-E R5 and XM-L U2 with respect to lm/A and lm/W.

I put all data into an Excel sheet. The result is clear:

For linear driver applications up to 400 mA XT-E R5 or XB-D R5 is the winner. For currents above this level, XM-L U2 is the best performing led.

For buck and boost driver applications XT-E R5 or XB-D R5 is best up to led currents of 220 mA, above that, XM-L U2 is better.

The limits are different because XM-L has a lower forward voltage and thus a lower power consumption compared to XT-E at the same current.

XT-E R5 and XB-D R5 seam to have their best efficiency below 100 mA, because both values lm/A and lm/W still show a significant slope at the lower end of the range specified by Cree PCT. At 100 mA the luminous yields are 510 to 524 lm/A and 182 to 184 lm/W.

XM-L U2 seams to have its best efficacy 165 lm/W at about 150 mA or just below that value, because the slope of this quantity is nearly flat at the lower end of the range specified by PCT. The best yield in terms of flux per current is reached at about 200 mA (443 lm/A).

Unfortunately Cree PCT doesn't support the new led generation XP-E2, XP-G2 and XM-L2. Those should be better at least at high current levels. The datasheets don't allow to extract reliable values.

Does anyone can provide data for XP-E2, XP-G2 and XM-L2 as well?
 
Last edited:

Tobias Bossert

Newly Enlightened
Joined
Oct 7, 2007
Messages
146
Location
Germany, Munich
I have updated the Excel sheet (link in the first post) with diagrams for lm/A and lm/W.

Obviously there are two different technologies used by Cree:

XP-G and XM-L show a comparable behaviour: XM-L U2 (red curve) has just double die area of XP-G -S2 (dark red) and the peak efficiency is up to one luminous bin stepping ahead. This technology seems optimized to achieve reasonable efficiencies at high current levels on the expenses of peak efficiency at low current levels. Peak efficiency roughly occurs at about 1/20 of maximum spezified current.

XT-E and XB-D show comparable behaviour: XT-E R5 (blue curve) has just double die area of XB-D R5 (green blue). Both models may have comparable peak efficiencies propably, but these occur much below the specification window of PCT (100 mA). This technology seems optimized to achieve maximum peak efficiencies at low current levels on the expenses of decreased efficiencies at higher current levels. The peak efficiency seems to occur much bolow 1/15 of maximum specified current - but it's not proven until someone does some measurements in the range 10 to 100 mA.
 

Tobias Bossert

Newly Enlightened
Joined
Oct 7, 2007
Messages
146
Location
Germany, Munich
Another estonishing result:

Nowadays XR-E and XP-E seem to use quite different dies, their behaviour is completely different.
At the time when XP-E was introduced - and a long time afterwards - both used identical dies!
 

Bolek

Enlightened
Joined
Apr 1, 2007
Messages
246
Location
Paris
Which one is the best performing led at the moment?

XT-E R5 seams to have its best efficiency below 100 mA, because both values lm/A and lm/W still show a significant slope at the lower end of the range specified by Cree PCT. At 100 mA the efficiencies are 543 lm/A and 193 lm/W.

XM-L U2 seams to have its best efficiency in terms of flux per power (165 lm/W) at about 150 mA or just below that value, because the slope of this quantity is nearly flat at the lower end of the range specified by PCT. Unfortunately Cree PCT doesn't support the new led generation XP-E2, XP-G2 and XM-L2. Those should be better at least at high current levels. The datasheets don't allow to extract reliable values.

What effiency corresponds to 193 lm/W ?
 

Tobias Bossert

Newly Enlightened
Joined
Oct 7, 2007
Messages
146
Location
Germany, Munich
What effiency corresponds to 193 lm/W ?

Yes, you are right,:
Quantities with dimension 'lm/A' and 'lm/W' shouldn't be stated as 'efficiency'.
'luminous efficacy', 'light gain' or 'light yield' would be better.

And yes too:
There is no value of 193 lm/W, this was due to my mistake using the values of XT-E S2 instead of XT-E R5
 
Last edited:

Tobias Bossert

Newly Enlightened
Joined
Oct 7, 2007
Messages
146
Location
Germany, Munich
Attention
I made a mistake by comparing XT-E S2 instead of XT-E R5.
XT-E S2 isn't availlable jet.
I corrected the Excel sheet now.
All statements in this thread which must be altered are marked red/bold now!
See post #1 and Excel sheet
 
Top