Xagor
I think I read that those chips become more blueish with degeneration. But emphasis on "cheaper" white.
But heat and driving parameters are an issue. Id say: never run a chip near its max rated current. But the opposite is beeing done e.g. with those "3W led" Panels -they usually run near their max at about 2,2W for white/blue and 2W for red.
Almost all LED is run near its stated output, we run around 12% under and that's borderline affecting the output of the chip, very few manufacturers run anything like this amount under capacity.
For cree cobs/leds you can find this, their lab results for long term lumenmaintenance and spectrum shift:
http://www.cree.com/~/media/Files/Cree/LED Components and Modules/XLamp/XLamp Application Notes/LM80_Results.pdf
Im no expert, but those numbers look good to me. And while they give their result for 105°C or 85°C and pretty high currents-put on a nice heatsink, with reasonable currents, with active cooling Tj gets about 55°C -there is no need to worry at all.
In order to drive the LED correctly you need to give a high current, a higher current is much more important than the voltage. No matter what anyone says, you quickly get to some serious temperatures at the thermal junction, without a well thought out heatsink and fan combination LED quickly dies...
But they are significantly worse in photon output per W (regarding common blue + phosphor) and maybe even too much gets wasted as FR.
A high CRI has little or no effect on spectrum, they are usually slightly higher in either voltage or current requirements, but this is the big myth when it comes to LED, in that its hugely power efficient...it simply isn't, you are still going to have to put in somewhere around 50-60% of the power to get a similar light output to halogen or HiD...
But for MJ, neutral or cool white tends to prolong the flowering time, compared to ww.
Because most modern white is made from doped blue!
If I could make a suggestion for your RnD atleast for MJ grow lights with no additional sunlight, test some CXB3070/ CXB3590 3000K or better 3500K high bin 80CRI at about 50W. (this ww also looks totally different to what you have on ur hp as ww) I think you will be very impressed. Or just one chip for ur light measuring device at about 1,4Amps or less.
Or more budget oriented with Vero29 V2 chips.
Color mix doesnt get better then with white/ 1 type of chip that does it all, does it?
Is there any white cob available thats better then a 3590? I mean in a reasonable way -not some secret lab 99% efficiency 100K$ way.
Our other associated company manufactures commercial "white" LED fixtures, we've tried quite literally hundreds of whites, most make very little difference...the fact of the matter is that no matter what "version" of white you run, a neutral white is marginally (less than 2%) worse than warm or cold at either end of the cycle, and by adding in slightly more red to get more flowering, you tinge the white pink...which is very similar to what Illumitex use...So white has it's uses, BUT at the expense of not targeting wavelengths a plant wants and therefore wasting energy, which is what the argument for red blue is all about.
So for us, we have a dope which produces pretty much a very light pink to the human eye, but does have most of the wavelengths in that make up a full white, so there are greens, yellows and oranges just in very small quantities, as well as UV and FR into IR...
But we are working on a suitable white, but it takes time, and even in our research facility we can't grow mj because we don't independently hold a licence to, so any research we do carry out has to be in response to a clients wishes. What our client wants, a client gets, and that research belongs to them...although we do take some of that learned information away in order to produce lighting recipes for that client and tweak our own offering.
