AmbiLobe:
"Are white LEDs more or less electrically efficient than say blue?"
A simple question for a complicated industry. It is natural for people to want simple answers, but face the facts: eight thousands facts, of which, seven thousand facts are trade secrets of billion dollar engineering efforts worldwide today. The answer is available after deep study of the IEEE Transactions on Electron Devices and the conference proceedings of the ISSCC International Solid State Circuits Conference. I will research the answer for you if you meet my wage demands.
Wow ok a little bit patronizing there. I am aware that I'm asking a very general question, of which very specific variables are at play. This doesn't mean with some basic research we can't find at least a general solution. For example ("Blue LED's are more efficient than red and green, therefore the phosphor-coated blue LEDs are generally more efficient than the RGB combination") Not saying that information is accurate, just an example - without citations.
If you have any valuable information that I can't find myself through internet research then I would consider a collaberation of some kind, if you are interested. If not, stop patronizing me. I also know someone on the board for IEEE.
AmbiLobe:
"I read somewhere that they were more efficient for the actual amount of light they put out but then I read in another paper the opposite. "
You read press releases from marketing departments of two vast corporations.
Who said anything about marketing? This was university based research, nothing to do with marketing or corporations (except for the department, which I don't think is selling anything). Though I do understand your point, I am careful of the research I take into account.
AmbiLobe:
"Are the RGB whites or the phosphor coated better? "
I like phosphors better because the spectrum is broader the the narrow band illusion of whiteness. Better at selling units to an uneducated cheap public? That information is a trade secret.
I am really only interested in performance at this point. Trade secrets? I've come across a few, and selling cheap units to an uneducated cheap public isn't really a secret.
AmbiLobe:
The link from the RP corporation was approved by their marketing department. Marketing people are known to be biased in favor of products they are selling.
I know this as well. Which is why I try aim for scientific information rather than pseudo-marketing science. Yes this information is from the RP corporation and may be biased. However the information they present seems well researched and documented, and therefore is viable for analysis. Or are there secret links between universities and companies? Now thats information worth paying for.
AmbiLobe:
"Does anyone have any information on actual light output vs drive current?"
Yes.
Alright, do you have any specific information of photon efficency across a range of well known companies for a variety of spectrum? Have you compared luminous efficiency of various brands?
AmbiLobe:
"What would be the efficiency gain of using twice as many LEDs at half the current?"
The temperature will be lower, so the figures for the product you linked allow you to estimate the gain yourself. Or you can pay a highly educated expert to read your literature for you.
Well just looking at the table I see the "Deep red" at 350mA has a luminous efficacy of 46%, while the 700mA has 42%. So only 6% more efficient at half the current. The Relative Luminous Flux only varies by about 2.5% for temperatures between 20 and 40 degrees celcius for the "Deep red". The blue LEDs seem to vary even less with temperature. From an initial perspective its not enough to justify the 2x increase in distribution area due to twice as many LED's. I guess thats why some companies overdrive their LED's at the minimal expense of increased heat.
AmbiLobe:
"This is hard to justify as it effectively doubles the cost, it better improve the efficiency by at least 25%"
That 25% is only a guess.
"Also could my passive cooling idea work?"
Yes.
"Or is the water cooling part just unpractical?"
It is practical for the unknown purpose.
I worked out for myself that the efficiency gain isn't really substantial enough. There are undoubtedly more factors to consider than I took into account - hence the aim of the discussion.
The unknown purpose is probably out of the scope of this thread, and once practical applications start becoming a factor even more variables will need to be considered. For now I am interested in how to consolidate information from various brands, reading the datasheets is difficult because of the way they are ordered. I'm still learning how the codes work etc..
