Lumileds Luxeon V 4000K CRI70 LED color, CRI and output testing

The thermal interface is very good just like Lumileds claims. Here’s a thermal video which shows that the temperature difference between the emitter and MCPCB is very small. You can barely see any difference in color.

Current increases from 0 to 5 amps in 200mA increments every 5 seconds.

Difference between emitter and MCPCB

Also Fv up to 5A created during the automated test.

Not being a CRI connaisseur (well, I’m not connaisseur of any thing really :D) I do have to say that I have only 2 lights with this LED installed. I must say that all the nights I use them at home and I never thought the CRI was only 70 :open_mouth:

The beam, tint and the colour rendition of this led is super, even not being 90+ CRI!

Soon I will get more from led4power and I will not regret that!

Thanks for the test maukka :beer:

I use a cigarette lighter under the mcpcb and the led literally centers itself. Like sucks into place.

Technically its better to gradually heat it and cool it which is easy on my cook top

I wonder how would this graph look if LED is under frosted optics? "Average" seems to be very close to BBL.

You should take into account that most of the output is in the hotspot. The average is therefore still above the BBL.

Here’s a measurement with a Convoy diffuser for a very effective averaging (but not necessarily perfectly neutral as can be determined from the CCT change) and d-c-fix:

Duv value tells the distance from BBL. Still above in both.

Thank you for this and your other very informative test results. I hope to someday understand much of your reports.

Here is what may be a very dumb question. One of my favorite aspects of this emitter is how it picks up reds (my personal experience). I, however, do poorly in tests where I have to sort colors. So I don't put much stock in my experience. Anyway, my question is: Doesn't the -32 R9 mean the emitter has poor color rendition when it comes to reds?

Depending on the answer, can someone please help me reconcile my experience with the test results?

The R9 value of –32 means that compared to a reference blackbody radiator at the same color temperature the spectral intensity in the deep red wavelengths (above about 640nm) is very small. This results in duller red colors, but that doesn’t necessarily mean that the reds are anywhere close to invisible or grey.

Thanks maukka. Seems I need to get out tonight and do some side by side comparisons to help me reassess what I think I'm seeing. I'm thinking the tint is what caused me to assume I was seeing reds better.

It very well might be idk never damaged a led doing that way. That’s the way old lumens told me to do it. Instead of exposing the led long exposure to hundreds of degrees the solder melts and is cooled to the touch in 20 seconds

All the datasheets spend like 2 pages going over the proper temp and time to do it so i assume there is something to it. You might have damaged one just not enough to make it not work :slight_smile:

Lol very true. I have no way of measuring output. Only current draw at the tail or led I’ve used my iron a couple of times. But takes to long for me. I’m sure there is something to it.

Added output measurement to the original post.

Super efficient!
It would be interesting to see at what current the efficiency peak happens. Every LED has such a peak. Sometimes very small LED actually have a better maximum efficiency compared to larger ones .

Damn. It has the efficiency of the Samsung LH351D at all levels.

The low VF is appealing though compared to the LH351D.

That is great.

It’s very low, since it starts dropping already at above 10mA.

I need to get a precision source measurement unit to do and trust these, but just for the hell of it, let’s go lower. Wouldn’t hurt to test in a smaller sphere either, even though the meter still read 12.3 lux at 1 milliamp so it should be somewhat reliable (minus the 1 lux that seeps in through the heatsink gaps). There’s still some uncertainty which is apparent when you compare the 10/20/30mA results to the previous tests. Shouldn’t heat enough to create a 3% discrepancy at 10mA.

Except that at the lowest currents the

efficiency goes way over the theoretical max of white leds (~300lm/W) :party:

(did not check source for reliability O:) )

http://www.archlighting.com/products/leds-and-drivers/the-truth-about-how-efficient-a-white-led-can-be

That’s why I said I don’t trust the measurements. The PSU can’t control such low currents. I’ll check later with a bench DMM what’s the lowest possible.

S2+ mod, luxeon V 4000K on board with TIR wide optic. 4,18A 100% mode, with 2.0 extended driver

First photo shape of the beam:

BTW, maybe the topic is not 100% convergent but it is testing output in practice :smiley: