• Here is a link to the full explanation: https://rollitup.org/t/welcome-back-did-you-try-turning-it-off-and-on-again.1104810/

Cree sphere test from chill led

shadow_moose

shadow_moose

Well-Known Member
Interesting that the CXB3590 seems to perform about 5% worse than Cree datasheets would suggest. In every independent sphere test I've seen they've fallen short of their claims. It seems like their plus or minus 5% is mostly on the minus end of things.
 
Stephenj37826

Stephenj37826

Well-Known Member
CobKits said:
data is about what cree pct would predict at Tj of 55-70C
Click to expand...
Yep pretty much and it goes to show that the tools are actually not too far off. With Crees binning they leave that min max not so much because of variances in the cobs but rather different ccts fall into the same bin meaning a 5000k CD bin will most likely have more "lumenous flux" than a 3500k that barely snuck into that bin. I always felt like the 3500k was at the bottom of the bin and 3000k was at the upper part of its bin CB. I hope people get where I'm going with this if not sorry for the confusion :)
 
Last edited:
CobKits

CobKits

Well-Known Member
whats odd is the 1400 would indicate a higher Tj vs 2400 mA which doesnt make sense to me

the curve must be "flatter" than pct would indicate

Stephenj37826 said:
Yep pretty much and it guess to show that the tools are actually not too far off. Renee with Crees binning they leave that min max not so much because of variances in the cobs but rather different ccts fall into the same bin meaning a 5000k CD bin will most likely have more "lumenous flux" than a 3500k that barely snuck into that bin. I always felt like the 3500k was at the bottom of the bin and 3000k was at the upper part of its bin CB. I hope people get where I'm going with this if not sorry for the confusion :)
Click to expand...
totally. a lot of their products are near bin cutoffs. a DD bin will prob never be a full 8% higher than a DB in the same color, as that is usually a low-DD vs a high DB, the difference could be 1-3%
 
wietefras

wietefras

Well-Known Member
CobKits said:
whats odd is the 1400 would indicate a higher Tj vs 2400 mA which doesnt make sense to me

the curve must be "flatter" than pct would indicate
Click to expand...
Perhaps that's a result of the cooling applied by the testers? If they ramped up the cooling for the higher wattage.

It's too bad these sphere tests seem to always neglect COB temperatures though. Only the ambient temperature is mentioned, but for COBs the case temp (or junction temp) is a pretty important parameter too. You lose about 2% in efficacy for every 10 degree Celcius the junction temperature rises.

:edit:
The LER seems to be around 326 (327 for 0,7A and 325 for 2.4A)
The QER is around 4.61
Lumen-PPF divider is around 70,7 (multiplier 0,014152)

The QER (and therefore Lumen-PPF conversion) seems to come out slightly worse than expected from the datasheets. Malocan observed the same figures with his spectrometer. Perhaps the datasheets are based on lower junction temperatures and/or lower wattages?

It's not a lot though (1% difference).
 
Last edited:
DankaDank

DankaDank

Well-Known Member
It would be great if we could get some 1212's tested. Judging by the results in the test it seems a 1212 3k 80CRI @ 700ma would be around the same efficiency as CXB3590 3.5K CD @700ma.
 
You must log in or register to reply here.
Top