I also think you're seeing these results only for the Cree XR-E because it's an older LED which came around long before the PCT existed. Cree may have simply interpolated some the calculations based on its newer LEDs, and this will give incorrect results for the XR-E which is based on older LED technology.
J
jtr1962
Flashaholic 61 From Flushing, NY
- Joined
- Messages
- 7,505
- Reaction score
- 768
Profile posts Latest activity Postings Media About Post areas
-
Tobias,
It's pretty obvious to me what's going on. The Cree PCT gives too high a value for Vf. That in turn drops the efficiency by too much as current increases. And at low currents they overestimate the output. I don't think the model they're using accounts for the fact that quantum efficiency decreases once current density falls below a certain point. I think the specific mechanism involved is fewer electrons have the energy to cross the bandgap and generate photons. I'm not surprised this isn't accounted for in the PCT. For other thing, it makes the calculations a lot harder. For another, few people are going to run LEDs at the currents where decreasing efficiency begins to occur. As far as I'm concerned, the largest issue with the PCT is that it overestimates Vf at higher currents. They might be using a worst case value, which is actually appropriate. They would rather have customers get higher efficiency than the PCT tells them instead of lower efficiency.
At the moment I make measurements of light output of XR-E R2, XT-E R5, XP-G R5, XP-G2 R5, XM-L U2 and XM-L2 U2 in the current range 10mA to 350mA or 700mA respectivly. The results are compatible to Cree Product Characterization Tool (PCT) for all LED under test with one exception: XR-E R2.
My results compare nicely to your findings in 2008 as you can see in this Excel file .
But they don't compare to the values derived from PCT. The differences are so big, that I gess there is another bug in PCT.
(The first is a minor bug only: do not use the second (middle) column, it reads some parameters from the first column rather than its own (Tj <> TSP switsching)
First I thougt my LED was damaged and therefore I bought a new one from LED-Tech - but it delivers the same behaviour. As far as I know, there is no "new" XR-E R2 on the market with totally different behaviour.
Regards Tobias
Love the work you do, thanks a lot! -
Loading…
-
Loading…
-
Loading…
-
Loading…
-
Loading…
