Having a larger area of phosphor is independent form the construction being flip chip or not, A dedomed or sliced 519A or LH351D barely have any tint shift even though they display a noticeable amount of it with the dome on, B35A is flat flip-chip and also have litle tint shift (except artefacts due to the quad dies) conversely Optisolis isn’t a flip chip construction yet displays large amount of tint shift.
Another example are the XHP 1st, 2nd and 3rd gen :
1st : lateral chip with bond wire, no large phosphor pour, domed, reasonable tint shift
2nd : flip-chip, phosphor pour on the whole package, domed, extreme tint shift
3rd : flip-chip,no large phosphor pour, domed, reasonable tint shift
3rd HI : flip-chip,no large phosphor pour, flat, lower tint shift (still some though)
The size of the dome seems to have an impact, though I’m not sure about it, small dome LEDs seems to have lower tint shift (SST-20, 219B)
It does seem to make sense that a larger dome would spread the tint shift over a larger area of separation, to make the tint shift more obvious, even w non flip chips like LH351d. I do see a green corona: .
it is also possible that the hotspot does not look green simply because of the overexposure, and in fact the whole beam is green
Interesting… so while dedoming can mitigate tint shift, it won’t necessarily get rid of it altogether. In the end, if the degree isn’t much then it becomes a wash in practical use.
XP-G3 has been an LED of choice for RovyVon and it took them a long while to start branching out to Nichia and others. It was the ‘leash’ that kept me from buying certain models. I didn’t like the CCT rendition… and the tint shift is likely the culprit. Noticeable on close-range usage with simple & smooth surfaces.
So with flat flip-chip style emitters, is tint shift less noticeable with lower CCT than higher CCT? I’m a bit confused, because I saw this:
Anyway, “flip chip” is a nomenclature that was unfamiliar to me. Fascinating,, the premise of the chip flipped inside the LED resulting in 20% greater efficiency and double the lifespan! What LED’s are made with flip-chip design aside from E17, E21, 519a and B35AM (E21x4)?
I didn’t realize the flip-chip design has been around that long, going back 4 years? I thought the whole 219 series was before that; so 219c has it as well. But in terms of efficiency, isn’t a 519a is still appreciably more efficient than 219c?
Very interesting question! I also notice more tint shift on my 4500K 519As compared to other CCTs. My conjecture is that this particular color temp is right between cool white and warm white, and tint shift easily puts the spill into the cool category, and the hotspot into the warm category, which makes the difference seem very apparent. 4000-4500K is also kind of an awkward CCT range that tends to make things look greener than they should, I'm not sure why.
In my experience the amount of tint shift does not vary monotonically with CCT, positively or negatively. I will say, however, that at extremely low CCTs (3000K and below), the same amount of tint shift becomes less noticeable than at, say, 4500K.
I believe there is no correlation OR causation between flip chip design and amount of tint shift. The dominant factors are generally 1) whether the LED is domed and 2) whether the phosphor pour extends past the underlying blue die. An external factor is the choice of optic/reflector--shallow reflectors like S2+ tend to bring out tint shift, while TIRs generally do a good job of smoothing out tint variations.
The 519A is much more efficient than 219C. On paper you don't see much gain in lumens, but keep in mind that the 519A is 95+ CRI and 80+ R9, while the 219C is not! Gaining R9 (deep red) at the expense of green makes the lumen count go down quite a bit, but makes the light so much nicer.
dont undestand what kind of Tint Shift you mean at Lower CCT?
imho, efficiency is the wrong question.
I have not seen any numbers… what percent do you consider meaningful?
Bear in mind the 219c has lower CRI R9.
The 519a is not more efficient than LH351d… I just like the higher R9 and lower DUV of 519a… efficiency is not a priority to me, I care about spectrum quality, not lumen quantity.
Efficiency favors greener tint and Lower CRI… neither of which I want.
It’s more efficient yeah, the die is larger and the thermal resistance lower, lower temperature means higher efficiency, the glass + silicone dome construction supposedly increases light extraction and there could be other improvements as well
Just to be super precise and pedantic: what you just described (presumably measured in lumens/watt) is actually efficacy, not efficiency. This appears to be a very common misuse of terminology within the flashlight world.
One nice way to distinguish them: efficacy measures how well a source produces visible light, and would thus have lumens (which is computed from the luminosity function of the human eye) in the numerator. Efficiency is always a unitless quantity: it measures the proportion of electrical power that has been converted into radiant power.
Given that the 519A still produces more lumens than the 219C despite having a R9580 spectrum (which is less efficacious than a 9050 spectrum), it is reasonable to conclude that the 519A is more efficient.
I can’t remember if it was Cree that introduced flip chip for emitters or not, but they were making them at least as far back as 2017 in their XD series, and in those they were able to mitigate wasted light and tint shift a little by putting tiny reflectors around the die, attempting to emulate a LES with one point instead of five (edges of the die). If memory serves the increase in output was around 15% or so, just in rescuing those lost wayward photons.
Flip chip as a concept goes much further back, in computer chips and maybe even transistors…certainly was happening in the 80s but probably years before that.
Thanks. Especially helpful to learn about how tint shift isn’t necessarily progressive by CCT, that it can be more prominent in certain ranges and depending upon domed vs. dedomed. Yes, I’d seen an ambiguous mention about efficiency, and I’m getting a better idea of it now. Of course, output efficiency has to be qualified because it’s conditional on duration and comparative CCT, given equivalent host/driver/amperage. “Luminous Efficacy” helps explain a lot too
Thanks. Larger die and comparatively lower thermal resistance sounds like a good prescription for better efficiency.
Tint shift across the beam or tint shift manifesting at different levels of output. Perhaps both are worth considering. I had read someone seeing tint shift manifesting in a dedomed 519a emitter rated 4000k~4500k, but not seeing it in 5000k~5500k. They didn’t qualify it by brightness setting, but that is of course an important consideration.
Wrong? Didn’t seem to be taken wrong by others here. Does luminous efficacy sound more appropriate? I don’t think efficiency is wrong generally speaking, but it should be qualified. The presumption is, given same power level being driven, does one LED deliver more lumens or nearly equal lumens but with less thermal buildup (and thus able to sustain the output longer before needing to step down)? From what I’m seeing in responses, it sounds like the 519a will achieve same lumens output as a 219c while drawing less on the battery and producing less thermal buildup, given the same host.
I just got a Convoy M1 with this led in 4k. While I don’t like the cool tint, the beam and output are amazing. It is a huge step up vs my other M1 with a 6k xm-l.
I don’t see any green; even indoors. Maybe the OP reflector is mixing the light well. However, I only see green in my MF01 with *18 SST-20’s because people say they are green.
wrong because
lower CRI and greener tint, are more efficient than
High CRI and neutral tint
for example LH351d is more efficient than 519a, but has Lower R9 CRI and Greener Tint. I dont choose for efficiency, I choose for best CRI and Tint.
If I was choosing for efficiency I would be using LH351d 9050, instead of 519a and 219b 9080 LEDs
see what I mean? efficiency ignores spectrum quality, and promotes lumen quantity, even if the lumens are cool white, Low CRI, and Green, they can be more efficient…
the lumens race favors output over CRI… thats the result of chasing efficiency instead of Quality
Yes, lower CRI and higher CCT will be more intrinsically efficient. But I wasn’t proposing comparing higher CRI to lower CRI. 219c and 519a can be compared at same CCT. I get that it’s not in your wheelhouse to be concerned about efficiency, but some of us are. The more runtime minutes per battery charge means longer battery life. It’s a 3rd place interest for me, with beam characteristics and CCT being above it.
Notice that while dedome brings down the CCT, which can be more appealing, it also introduces a change to the beam pattern. Not as smooth as domed. But that may not matter to some based on application.
Again, what you describe is efficacy, not efficiency! Just wanted to make sure that everyone in the conversation is on the same page and talking about the same thing.
they do not have the same CRI R9
219c, and LH351d are 9050 LEDs
519a is a 9080 LED, it shows red colors better
I also care about battery life, I control for that by using enough light, but not too much. I generally dont use the maximum, nor the high modes. I use medium and less. This gives me much longer battery life.
Also, I use a trick to double the Lux on target, if my medium mode seems too dim…
I move twice as close to the target, based on my own Lux measurements, this doubles the brightness on target.
thanks… how do you measure how efficient the 519a is?
do you know output of 219c compared to 519a, at same CCT?
fwiw, my personal measurements comparing LH351d to 519a showed just a 3% lower outpu for the 519a… That is totally fine with me, to gain the R9 CRI, and more neutral tint.
and thanks for trying to point out the differences between efficiency and efficacy… I dont really have my head wrapped around those distinctions yet. To me, efficiency, in the way xevious is using it, refers to runtime or output… iow, more runtime from same output is more efficient
or
more output with same runtime is more efficient
Excellent question, and not an easy one to answer! I don't think there's a device that can measure efficiency directly. In theory, you could measure the input power (P watts), output (L lumens), measure the spectrum, compute the luminous efficacy of the spectrum (E lumens/watt). The ratio L/(PE) is the efficiency. The difficult part for me is computing the luminous efficacy of any given spectrum, I'm not completely clear on how to do that.
The usage of "output" is ambiguous; it may refer to the luminous output (measured in lumens) or radiometric output (measured in watts), which are essentially different quantities.
It might help if I give an extreme example: consider a UV light that converts 100% of its input power into UV light. Since UV light is invisible, it does not generate any lumens, so the efficacy is 0 lumens/watt. However, the efficiency is 1 (100%) by construction.
.