LED test / review - Nichia Optisolis 3000 / 5000 K - ultra high light quality, even with high performance in mid-power class

LED test / review
EN



Nichia Optisolis 3000 / 5000 K



NF2W757G-F1 / NF2L757G-F1


Some months ago Nichia announced a press release fort he new Optisolis LED series, emitters with ultra-high CRI „that demonstrates the closest match to that of the standard illuminant“, Nichia says.

This is achieved by a new developed LED chip and a mix of unique red and green phosphor.

I got some samples from BLF member clemence (‚Virence‘) for testing. Thanks a lot for these! :)

Technical data

Tj 25 °C unless noted otherwise

Order code: NF2W757G-F1 (white)

Official Nichia datasheet

Type: lateral, multi die
Binning: ---
Rated voltage: typ. 2.9 V

Forward current (Tj 150 °C): 100 mA

Pulse forward current: 130 mA
Viewing angle: 120 °
CCT: 5000 K typ.

Thermal resistance: typ. 24 °C/W, max 30 °C/W
Junction temp.: max 120 °C

Order code: NF2L757G-F1 (warm white)

Ofiicial Nichia datasheet

Type: lateral, multi die
Binning: ---
Rated voltage: typ. 2.76 V

Forward current (Tj 150 °C): 150 mA

Pulse forward current: 200 mA
Viewing angle: 120 °
CCT: 3000 K typ.

Thermal resistance: typ. 21 °C/W, max 27 °C/W
Junction temp.: max 120 °C

Interesting are the different electric characteristics of both Optisolis types. Overall the warm white version is more powerful and has also less thermal resistance. With 120 °C the maximum Tj is much lower than other LEDs (135 or 150 °C) which could limits the maximum overcurrent in a negative manner.

Appearance and die

Unless classic high power LED the Optisolis is different in design and footprint. Like most other low- and mid-power LEDs in this class the case is made of white epoxy resin and the two non-symmetric dies are embedded in a silicone layer. Everyone of them is connected with two very small bonding wires which could limits the maximum overcurrent possible.

Comparison with classic 3535 emitter (XP-G2, right)

A single Optisolis LED fits easily on a XP- or XM-board, with the bigger contact on thermal pad. But there is also more possible, like shown here with four Optisolis emitters on a single XM-board:

There are some specles on the die which I have also seen on Cree ML-E and some other LEDs in this power rating class. The four bonding wires can be easily discerned, because of the a-centric dies embedded in the LED itself the beam uniformity might be very worse.

Performance

25 °C Tsp, unless noted otherwise. In ‚real life applications‘ this values may vary.

Due to the higher thermal resistance and highest CRI these emitters do not deliver as much luminous flux as ‚classic‘ high power LEDs.

  • The 3000 K variant reaches 184.3 lm @ 4.22 V at 1,500 mA (6.33 W, 29.1 lm/W)
  • The 5000 K variant reaches 281.0 lm @ 4.37 V at 1,500 mA (6.56 W, 42.9 lm/W)

Despite the design and higher thermal resistance, they are quite powerful, with better overcurrent capabilities than other emitters in same design.

  • I would recommend a maximum current of 800 to 1,000 mAmps.

Despite the slightly lower thermal resistance of 3000K Optisolis, the reached maximum current at the same test setup is almost the same.

The 5000 K variant has a slightly higher Vf, but overall the efficiency of the 5000 K Optisolis is way better than of the 3000 K variant. If these LEDs are soldered in 4P config on a board together the warm white emitters lit up earlier, like here:

Light quality and use in optics

Because of the two non-symmetric dies this LED cannot be easily used in standard flashlight optics like reflectors, because the beam uniformity is not as uniform you would expect from a flashlight.

Using some type of diffusor (dc-fix or similar) is necessary if you prefer an absolutely smooth light distribution.

I can describe the tint in one simple word: awesome. The color rendition even without measure equipment is very, very high and very natural.

Our BLF member Maukka did extensive color testing of Optisolis in both CCT, shown here (with his permission).

Optisolis 3000 K test results

Optisolis 5000 K test results

You can find more results and information to the tint also in maukka's thread!


Conclusion

This might be the LED with highest light quality to date.

The Optisolis series are also pretty decent in light flux, and could be overpowered very well, especially for it's design. I did not like the difference in Vf between both CCTs, and the unusual footprint which makes flashlight modding much more complicated. For interior and accent lighting these LEDs are very interesting, especially the 5000 K variant!

Pro

  • Very high light quality
  • Pretty high maximum current possible for it’s design/type

Contra

  • Non-Symmetric dies which has impact in beam uniformity if used in conjunction with optics
  • Vf difference in CCTs
  • Unusual footprint

Thanks a lot for reading! :)

Greetings, Dominik (aka BLF member koef3)

Mistakes, suggestions or offers / sponsoring of LEDs which I should test next are best sent via PM.

Thanks for the testing koef3. Lots of useful information here.
I wont say its invaluable as djozz will harm me. :slight_smile:

Thanks for the data. Not sure if this is a good choice over the brighter E21A. I’m surprised the 5000k version has RG of only 99 whereas on the E21A, RG values are 100+. Also 5000k version has a DUV of 0.0024, whereas the E21A are on or below the BBL so the tint on the 5000k seems not as good as on the E21A.

Thanks for testing. Can I place a request for another performance comparison testing using supplied VR16S1 board vs your DTP board? Just curious.

Thanks,

Sorry, I just saw your post now.

For this testing series, my Optisolis samples are reflowed several times and are now built in a flashlight. From my experience with other Nichia emitters reflowing too often has a negative impact on the maximum performance possible. I have no more Optisolis on stock, so actually unfortunately the answer has to be 'no'... :(

Very nice test as always.
But there may be a problem with your Current/Lumens diagram, the current starts at 0.1A with 0 lumens.

OK, no problem Koeff. Thanks

- Clemence