2023-07-23
Some time ago, Nichia introduced the 719A, an LED with a so-called ‘stacked die’. This refers to the “stacking” of the luminous surfaces on top of each other, which is supposed to significantly increase the luminance. With the 719A, this technology was introduced for the first time in series production.
This test clarifies to what extent the 719A really makes a leap in luminance and whether this LED can still be used well in optics for flashlights.
This emitter, already soldered to a DTP board, was purchased from the LHT Store on Aliexpress.
Tj 25 °C, If 1,050 mA
-
Type: multi-die (stacked)
-
Bin: T550f26 (min. 550 lm @ 1,050 mA)
-
Color group: sm405 (5-step MacAdam ellipse)
-
CRI: R9050 (typ. Ra 93, R9 50)
-
Rated voltage: min. 6.4 V, max. 7.4 V V
-
Max. Forward current: 1,500 mA
-
Max. Peak current: 1,800 mA
-
Viewing angle: unknown °
-
Thermal resistance: 2.5 (max. 3.5) K/W
-
Max. Temperature Tj: max. 150°C
Datasheet can be downloaded here: Nichia, PDF
At first glance, the 719A could be mistaken for an XP-L HI (new generation) or a dedicated Nichia 519 A. The stacked luminous surfaces are surrounded by a white silicone layer. On one corner there is a label and a barcode, most likely for quality monitoring and automatic batch assignment during or after the production process.
In the side view, it is visible how the luminous surface ‘protrudes’ a bit from the white coating. This is probably primarily due to the special design of the stacked luminous surfaces.
Due to the symmetrical design and the standard XP format (3.5x 3.5 mm), the use of common accessories (centering aids) for XP footprint is possible without any problems.
The footprint reveals no surprises, although I can’t show a picture of it. The center thermal pad is electrically isolated, so as in this case, the use of DTP boards is no problem here.
The special feature of this LED is the so-called ‘stacked die’, i.e. stacked luminous surfaces. According to Nichia, this is supposed to increase luminance and general efficiency. In principle, this approach is similar to AMD’s ‘stacked’ 3D V-cache, which is partly installed on the newer Ryzen processors in mass production. Despite the obvious multiple luminous surfaces (6 V operating voltage = 2S configuration is likely), only a single LED chip is visible from above, which means there are no gaps or other artifacts that could influence the light image in optics.
The luminous area is 4.69 mm² in size. In principle, the luminous surface is quite similar to the luminous surface of a dedicated 519A. The only noticeable difference is the lower luminance and a darker area in the four corners.
Within official parameters, as far as known:
-
at 1,800 mA (official maximum current): 804 lm @ 6.72 V
-
Power at official maximum: 12.1 W
-
Efficiency at 1,800 mA: 66.4 lm/W
The specified binning was missed with 521 lm @ 1050 mA 25 °C Tj, but due to the specified tolerance of ± 7 % it is within the range of the specified bin T550f26.
The generally low basic efficiency is interesting, already at 200 mA only 103.4 lm/W is achieved, a value that is reduced by another 10-13 % at 85 °C Tj. Particularly worse is the fact that even the 519A (tested sample R9080 5000K) with a power of 10 W has a luminous flux that is about 15% higher (!) than the 719A tested here. Of course, the efficiency is higher with increasing CCT and it has a dome, but such a performance is still disappointing.
Overcurrent:
-
Maximum reached at 3,2 A, at this point 1104 lm @ 6.89 V
-
Power at maximum 22.1 W
-
Sweet spot at about 2.2 A (925 lm @ 6.78 V)
-
Power at sweet spot 14.9 W
-
Efficiency at maximum 50.1 lm/W
-
Efficiency in the sweet spot 62.0 lm/W
The performance of this LED is surprisingly low, despite the nominally low thermal resistance. The possible maximum is already reached at 3.2 A with very good cooling. I would not operate this LED with more than 2.2 to 2.5 A in continuous operation.
There were reports in the BLF that the LED died a sudden heat death at 3.2 to 3.5 A (burnt phosphor), which I could not confirm. The LED withstood a short ten-second burst to 4.0 A without problems, although the changing CCT clearly showed that the current was way too high.
This LED reminds me of the first Ryzen X3D. These also feature stacked dies (3D V-Cache), and the issue of maximum power and heat dissipation plays a crucial role in these as well. At the time, this was (presumably) one of the reasons why the temperatures were significantly higher than on the 5800X and voltage increases also risked a CPU defect. This also seems logical, since the heat dissipation is hindered by the stacked dies and the thermal resistance increases significantly. In the case of the 719A, the uppermost (externally visible) light area should get the most heat and thus die the quickest heat death when overheated.
In comparison, the low performance and especially the low maximum current are particularly noticeable. Pretty much all LEDs offer a higher performance (although not necessarily in Ra 90), and thus also offer a higher luminance, but this is depending on light flux and size of the LES.
Values at 25 °C Tsp, at 85 °C Tsp values are 10 % lower
Nichia reports a luminance that is increased by a factor of 3 for the 719A compared to the 519A. This is for example shown on the official marketing page for the “Super High Luminance LED”. However, there are several problems with this presentation:
-
the 519A is a 3 V class LED. The 719A is in the 6 V class.
-
the performance of the compared LEDs is therefore different.
In short: due to their different Vf, both LEDs do not offer the same power in comparison (3 W vs 6 W) and are thus not directly comparable. This falsifies the statement enormously, since in this case neither luminous flux nor electrical power are standardized. For me, this makes this advertisement simply meaningless.
At least at 1050 mA, the advertising statement is roughly adhered to (factor 2.6), whereby this value shrinks to only a factor of 1.4 when normalized to 10 W. I don’t know why Nichia does such misleading, probably the lower increase in luminance at the same electrical power might play a certain role here…
The light image is just perfect. (There is really nothing more to say about it).
There are neither disturbing differently colored areas, nor any artifacts or other unsightly light images when using SMO as well as OP reflectors. This LED can be used without restrictions in both zoom and reflector optics. Diffuser foils or other aids are not necessary to get a pleasant light image.
Subjectively, the light color is pleasant. Despite the duv being slightly above the BBL (0.0021), there is no annoying tint. The color rendition is very pleasant, there is no oversaturation or distortion as observed with a deviation of the tint.
However, the R9 of “only” 65 is somewhat unpleasant, whereby this is lower than expected from Nichia (with tolerance R9 min. 73.5). However, I do not see any problems here, since this difference is only relevant for measurement and does not play any role for the normal user.
-
Ra: 93
-
R9: 65
-
CCT: 4041 K
-
duv: 0.0021
The 719A is an interesting LED. As the first LED with stacked LED chips, the performance is particularly interesting here. Unfortunately, the overall performance is not really high. The maximum current to be achieved and thus the power is relatively low and so is the efficiency.
In return, the light quality is fantastic, there are no artifacts like color distortions or rings in reflectors and the color rendering is very high as typical for Nichia.
If maximum power is not necessary, but an extremely good light pattern and tint is desired and a direct replacement of a 6 V LED is necessary, this emitter is absolutely recommendable.
Pro
-
extremely good light pattern with lenses and reflectors
-
very high CRI, pleasant tint
-
common XP size and footprint, so the 719A can be used widely
Neutral
-
luminance only average
-
sensitive to heat due to stacked dies
Contra
-
low performance (maximum current/light flux)
-
hardly any advantages over a 519A in terms of efficiency and luminous flux
Thank you for reading the test. 