[Review] E21A R9080 5000K

Alright, my last post about the E21A was more of a discussion into why the results sucked. The conclusion in the discussion seemed to suggest I needed to turn the power down to achieve perfection. So here is the official review, with a whole range of power levels (verified perfect by my Fluke), a fresh dark-calibration in the meter, and just a do-over in general for everyone’s sanity. Let’s see how this goes.

I am testing a:
Quad E21 “NVSWE21AT 5000K sm503-D240-M1-R9080” from Virence.com

For these tests I am not using any light modifiers, just a bare board as seen above. I hold the meter a measured 1-foot from the emitter, and at a slight 10-degree angle so I avoid the direct center of the emitter just a bit.

And the results are in:

—0.250A—
CCT: 5813
Duv: –0.0034
CRI: 95.9
R9: 82.9
TLCI: 98

—0.500A—
CCT: 5862
Duv: –0.0047
CRI: 95.3
R9: 82.9
TLCI: 98

—1.00A—
CCT: 6000
Duv: –0.0070
CRI: 94.2
R9: 80.5
TLCI: 97

—2.00A—
CCT: 6176
Duv: –0.0104
CRI: 92.8
R9: 75.8
TLCI: 96

—2.50A—
CCT: 6357
Duv: –0.0116
CRI: 92.5
R9: 74.0
TLCI: 96

So the CCT is horribly cold compared to the advertised 5000k. The Duv ranges from low to horribly low. And low Duv is not a rosy tint at this high of a CCT. It’s a blueish-purple tint, as seen below. The CRI (Ra) does hold spec across all tests. The R9 (the 80-minimum part of R9080) falls out of spec past 1 amp.

And this all makes sense with what I see, The meter is not wrong. Grass and plants appear cold and blue when lit by this light at night. The effect is mild when running 0.500 amp and extreme when running 2 amp.
So is this a fluke? Nope. I have two of these quads, and they are identical. It’s just the way they are when used bare. This is the true performance of the emitters.

Update: It turns out the advertised CCT and tint are rated at 700ma in a lab-grade integrating sphere, and that bare emitter performance is expected to be wildly different.
Adding a TIR lens to mine and turning the power down to 700ma was able to bring the CCT down to 5261. This is much closer to the advertised CCT, though still high. Unfortunately adding a TIR lens is not an option for my photo lighting needs. It simply doesn’t match my ambient light. This project is still a bust for me.
When I bought the ‘5000K’ I wrongly assumed that advertised output was mostly valid for any real world use. The Virence store makes no mention of needing to add optics to approach the advertised CCT.

Review Data:
E21A Review Data.zip

When I measure the new SST-70 from straight above the led, I measure 7500K. But when a beam is formed by a S2+ reflector, I measure 5950K in the hotspot. I can not measure integrated tint, but the average tint for all leds I have seen sofar will measure warmer than straight above the led (blue light escapes the led more difficult at an angle). It is a bit dependent on the application if a measurement from straight above the bare led is relevant or not, for a mule and for an aspheric build it is useful information, for use in a reflector light or with TIR it is quite a bit off from what you see in your final beam.

Well I can’t be using a reflector for my photography, I need wide, smooth, even lighting. But I do have some test data like that in the other thread. I will re-post.

CIE

Have you tried measuring an incandecent bulb with your meter?

I just tested one now for you:
CCT: 2638
Duv: 0.0001
CRI: 99.6
R9: 99
TLCI: 100

I just measured my TO50R that comes with the stock Samsung LH351D 5000K 90CRI last night with and without optics:

Without Optics:
CCT: 6053K
DUV: –0.0051
Ra: 92.2

With Optics:
CCT: 4868K
DUV: 0.0045
Ra: 91.2

Once you start measuring more emitters with you spectrometer, you will be less surprised with your measurements.

Thanks for posting the info. This thread really isn’t about ‘surprise’ for me. It’s just a bare emitter test. I wish more information was posted on BLF about the very significant differences that can happen without optics installed. I would have ordered what I wanted the first time.

This is not common knowledge that just has not been posted, these things are fairly new to me too, I do not own a spectrometer for very long still, I’m still exploring all kinds of measurements that I can do with it.

maukka goes into great detail in his tests,
including power levels, and
reflector or optics arrangements, and their effects…

I admire your dedication to testing LEDs and sharing your results.
Your testing protocol may benefit from some standardized power levels and reflector or optics used, to be more comparable to actual use in a flashlight.

I have several lights that use the E21A led’s from Virence, ranging from 3000K to 4500K in both TIR lenses and Reflectors. In all my lights, the colour temperature seems spot on compared to all my other lights (Zebralights, 219C, XP-L Hi, etc).

In my experience, an LED that has a neutral or even negative DUV appears cooler in appearance than one that has a positive DUV at the same colour temperature.

Reading the entire thread below, I can see that the LED’s you ordered are indeed the correct CCT and as they are being measured bare as opposed to in an optic or reflector, the reading comes out cooler at that particular angle. Might be worth adding a *note to your original post for those who don’t read the entire thread.

To my eyes, even a 4000K LED can appear horribly cool if my eyes are adjusted to incandescent or halogen lighting.
I do think the E21A LED’s from Clemence are the best LED’s I have used to date.

BLF is an encyclopedia of wealth and knowledge, but the information is not always easy to find.

I do have a question for you. If you are using it as a photography light, why does the CCT matter so much? Just adjust the white balance in your RAW images at post processing, or set the camera White Balance to match that of the light.

I did post an update in the opening thread about the optics like you said. Good idea.

As for why it matters for photography lights, it would not if this was the only light, like a flash. But I need an ambient light from the side and top, and this light from the side to give a glisten/glare. If the 3 lights don’t match each other, there is 3 different tints on the same object. That can’t be edited out.

I see, multiple source lights do indeed need to match. But aren’t strobes and speedlights around 5600K +/-? I would imagine you will struggle to match any of the E21a CCT’s to a photography strobe/speedlight without the use of gels.

Thanks for adding your comment, although I do feel that Virence is merely selling the LED’s exactly as spec’d from Nichia which is what I would want from a retailer. I don’t think it’s fair to put the responsibility on the store for real world variances. I don’t think any of the other big LED suppliers online even list anywhere near as much information as Virence does with the Maukka charts. Wouldn’t you have the same experience no matter who you bought from?

Great discussion
Iirc combining the 5000k and 6500k optisolis could be a great daylght white

Maybe just test in the host such as a d4, not bare

(*1): LED CCT and tint shifting described clearly in the datasheet provided by Nichia and I also put the link to it in the product page. CCT/tint shifting occurs in any LED, the higher the working current range, the higher the swings. This explains why mid power and COB LED has minimal CCT/tint shift - their current range is limited compared to high power LED. COB and arrayed LED also has less shifting because they have more uniform color separation averaging over angle.
You already have the most tightly binned Nichia: 3 steps and R9080. The higher the CRI the less CCT/tint shifting it will be. Manufacturers always measured the binning CCT at certain current and then provide additional current and temperature shift curve in the data sheet.


To get the most consistent color you better use (lots of) mid power LED (within spec current) or fix the current and use PWM to alter brightness.

(*2): Manufacturers use big calibrated integrating sphere with at least 1 meter diameter. Most of us don’t have such equipment and I consistently put any LED I sell to Maukka since he’s the only person in this community who has the closest equipment and setup to manufacturer’s standard. I’m sure many flashlight manufacturers own such equipment but I have no access to them. LED’s output CCT/tint vary with viewing angle. And E21A is one of the most consistent LED about it.

(*3): could you please kindly take a close-up macro pictures of both LED modules? Your LED was from the same reel and batch tested by the “other” reviewer. I also used this 5000K LED for hundreds of cafe light, streetlight (no optics but PMMA cover), Jet-UE21A, Wizard Pro, and Tiara Pro mods. And I never heard anyone bought those modified lights from me complained about the 5000K E21A horribly colder than expected. So, if you think your LED is defective/different I’m also very curious about it. Just let me know if you’re interested and we can start an investigation openly. I always feel bad when people disappointed with my stuff/service.

(*4): I only provide manufacturer’s data (there’s link in the product page) and add some more missing data not covered in the data sheet which is individual CCT/CRI measurement for each CCT class. So far, I haven’t seen any web store providing this data. I also states each reel binning code, every time I get a new one.

[Clemence]

I don’t think they are defective, and I sure am not saying your store is deceitful. I’m quite happy your store makes these LEDs available to the average person.
But when offering stuff like this to the average person, the listing does need to state anything the average person would not know. Like the fact “CCT ratings are evaluated using integration equipment”. Or that “tint and CCT change significantly with power input”.
One way for a seller to have great customer satisfaction is through great customer service. You do good at this. The other way is to continuously improve the store pages with feedback. You’re store has truthful information from the manufacturer, but in my case, it led me to believe that the CCT and tint were stable properties of the LEDs. I know you did not mean to mislead, but you just made assumptions too, in assuming shoppers already knew CCT and tint were highly variable.

I’m not upset, just disappointed by the general situation that I paid good money and invested good time for a project that didn’t go as expected.

I respectfully disagree with your belief that Clemence is responsible for educating you about how CCT ratings are evaluated, and that he is responsible for educating you that tint changes with power.

And I think you are out of line in your implication that you wasted time and money to do a project that did not work as expected, due to your lack of experience and education with using and measuring LED specs.

I suggest you thank the community for helping you learn what you are doing wrong when attempting to determine CCT and tint values.

That would be like me buying a lathe, and then blaming the manufacturer for not teaching me how to use it.

you bought a tool to measure CCT and Tint, and you used it wrong. Then when people helped you learn the error of your ways, you acted a bit indignant, like you were entitled to an education on how to measure LED properties.

you are not entitled, I suggest you thank everyone for helping you learn why your measurements are not compatible with maukkas

enjoy your lights, and your growing education on how to test LEDs in a way that is consistent with maukkas tests.

Completely agree jon_slider, what Clemence sells are boutique, enthusiast products that require some prior knowledge of how they will be assembled and used.

Basically every single LED on the market has a (usually significant) shift in tint from different angles when run without any optics or reflectors, and when driven hard will often change tint and temp drastically as well (case in point, SST-20 which you have extensively tested).

Couldn’t agree more.
Clemence made countless posts since several years with pictures for tint and CRI comparisons and later send samples to maukka for extensive CRI measurements, some with bare LEDs + with reflector with measurements in the hotspot and spill, etc.
But it’s never too late to learn.

You say you respectfully disagree, but then you go on to straight-up do the opposite. You twist what I said about improving the store into some attack on me. You call me “indignant” when I straight-up said Clemence did nothing wrong or deceitful. My OP was very clear about the how and why I measured the emitter bare. You need to be very careful about stomping on people that post bare-emitter test data, this barking you do creates the situation where BLF has nobody that posts this stuff. You talk education, while barking at test results. These tests were done bare for a good reason, you are in the wrong thread if you can’t tolerate this type of test.

Nobody minds you posting bare LED tests, but claiming Clemence may have shipped you the wrong item or that it’s his responsibility to educate customers about basic characterististics of all high power LEDs is taking it way too far. Clemence provides significantly more information about the LEDs he stocks than literally any other seller in the entire world. He also goes to great lengths to personally assist many customers in selecting the correct product. If you had told him your requirements in advance he would almost certainly have advised you to purchase a different product or take an entirely different approach to your lighting setup design. Nobody is “barking” at you over your tests, they just think you should accept that no one else is to blame here. You didn’t fully understand what you were purchasing or testing, and now you do. Clemence has also very graciously offered to help you investigate further if you believe there is something wrong with the E21As you received.