2023-10-24
The Luminus SFT-40 has been on the market for some time now. Available only in cool white light color at launch, a warm white variant with high color rendering was also available for a short time, which indeed makes this emitter special, as it combines high luminance with good light quality and warm light color at the same time.
The cool white SFT-40 was purchased from Mouser in November 2022, while the warm white SFT-40 was purchased from Kaidomain (as far as is known, it was also only available there).
Tj 85 °C, If 1,500 mA
- Type: single die
- Bin: N4 (min. 594, max. 634 lm)/L5
- Color group: 652/HB4 (6500/3000 K)
- CRI: typ. 65/95
- Rated voltage: typ 2.8 V V
- Max. Forward current: 8,000 mA
- Max. Peak current: 10,000 mA
- Viewing angle: 120 °
- Thermal resistance: 0.7 K/W
- Max. Temperature Tj: max. 150°C
Datasheet can be downloaded here: [Datasheet (newest version, Luminus)] (<URL fĂźr Datenblatt>)
The SFT-40 is an LED without silicone domes, similar to the Cree XP-L HI or Nichia 719A. The LED chip surrounded by white silicone sits on a gray ceramic substrate and is connected with a total of four bonding wires, making this emitter a classic lateral design.
A gold-colored metal frame serves as a perimeter for the white potting compound. This makes this LED very robust, making it easy to handle without having to take special protective precautions. However, care must still be taken to keep the emitter free of dust and particles!
On the left side of the following pictures the warm white version is shown, on the right side the cool white one.
The emitter is 5.00 x 5.00 mm. All accessories for XM LEDs (5050) can thus be used, and thanks to the symmetrical package, centering aids made by lathe can also be used.
The footprint is fully compatible with XM and 5050 boards, but the individual contact areas protrude visibly from the gray substrate, which also applies to the thermal pad. This makes for a supposedly âbadâ soldering with âgapsâ, although the LED is firmly seated on the LED board (see also the lateral image of the LED).
The luminous surface consists of a single LED chip. Thanks to the classic design, no or very little light is emitted laterally, which should significantly improve the optical properties. The missing corners should also further improve the light pattern in reflectors, as the LED chip thus more closely resembles a round surface.
The luminous area is 4.02 mm² in size. There are no differences between the two versions. The LED chip has connections in all four corners, which should improve the voltage characteristics compared to LEDs with only two bonding wires. Lateral radiation has been reduced to a minimum here; this should not have a significant impact on luminance.
Within official parameters, as far as known:
- at 8,000 mA (official maximum current): 2521/1384 lm @ 3.47/3.55 V
- Power at official maximum: 27.8/28.4 W
- Efficiency at 8,000 mA: 90.8/48.8 lm/W
Due to their design, LEDs without silicone domes have a 15-20% lower luminous flux (compared to the equivalent with silicone domes), while the luminance is increased at the same time.
The SFT-40 in 3000 K has a surprisingly low efficiency. Here it is clear that the achievable efficiency has been subordinated to the goal of achieving the most perfect light quality possible. The 3000 K version has an efficiency that is almost 80% lower. I did not expect such a clear difference here.
The Vf differs minimally in both variants, by 70-110 mV. This is not really much and could simply represent series dispersion, especially since Luminus specifies a relatively wide range for the forward voltage in the datasheet. The LED chip seems to be identical for both variants, the changed light quality of the warm white variant results exclusively from the changed phosphor mixture.
Overcurrent:
- Maximum reached at 14.2/13.8 A, at this point 3208/1717 lm lm @ 3.97/4.08 V
- Power at maximum 56.4/56.3 W
- Sweet spot at about 9 A (2703/1475 lm lm @ 3.55/3.64 V)
- Power at sweet spot 32.0/32.7 W
- Efficiency at maximum 56.9/30.5 lm/W
- Efficiency in the sweet spot 84.6/45.1 lm/W
Despite the already very high maximum current specification, an overcurrent of the SFT-40 is possible with both variants. The warm white variant has a slightly lower maximum current, whereby 400 mA is little deviation and this can simply be due to the soldering on the board or a minimally deviating thermal resistance. According to my tests, the performance of the Kaidomain Cu DTP board is on an identical level as the Noctigon XP20 specimens.
However, the clear difference in luminous flux and thus efficiency is again striking. In the worst case (the cold white SFT-40 tested here is in a slightly worse luminous flux binning), a halving of the luminous flux must be expected when using the warm white SFT-40!
Due to the larger heat dissipation area compared to the smaller Footprint 3535 (XP), a higher maximum current and thus significantly higher electrical output is possible. The new generation dedicated XM-L2 shown here in comparison - which is equipped with a flip chip die - reaches a similar maximum current, which shows that in the range 13-16 A the heat dissipation area of the 5050 (XM) footprint is limited.
The XP-L HI has no chance despite high bin V3. It has a generally higher thermal resistance, the smaller footprint, and associated lower maximum luminous flux. The Vf is also significantly higher. Here, the further development of the past years is particularly evident.
The SFT-40 is therefore a worthy successor to the XP-L HI. However, the larger footprint can be a disadvantage depending on the individual case; in any case, another (suitable) LED board in the XM footprint is required in this case.
15 minutes of continuous operation at maximum possible operating current did not cause any damage to the LED or phosphor. This applies to both variants; the performance is also still identical after the extreme test.
Values at 25 °C Tsp, at 85 °C Tsp values are 13 % lower
The enormous difference in performance of both CCT variants is also evident in the luminance. The basic optical radiation properties are identical between both variants, so that only the significantly lower luminous flux of the warm white variant comes into play here. As expected, the cool white variant has a significantly higher luminance than the XP-L HI and also exceeds the luminance of the CSLNM1.TG (however, this is only at absolute maximum current), so that it is thus very well suited for throwers, but at the same time also provides a higher luminous flux.
The light pattern is excellent (whitewall beamshot of an Aplos T02, which has an OP reflector). There are no artifacts in the light pattern, nor is there a strong tint shift. Depending on the quality and calculation of the reflector, rings could occur, but these can be avoided by using an OP reflector or a good TIR optic. This LED can be found in a lot of lamps lately and thus already shows that it can be used well in different secondary optics.
For the following outdoor beamshots of the warm white SFT-40, it was mounted on the 20 mm DTP board in an Aplos T02. The advantage of this lamp is the very easy to open head and the compatibility of the 20 mm board, where only minimal mechanical adjustments have to be made, which also makes it suitable for faster tests with LEDs on board.
The tower beamshot (about 90 meters away) shows the great advantage of such a warm white LED with high luminance at the same time. Intrinsic glare in hazy or foggy atmospheres is lower, and the light beam also penetrates better through rain. This is the reason why sometimes yellow headlights were installed in auxiliary headlights of cars.
In addition, a lamp equipped with such LED can be used as a mood light if necessary, for example, with diffuser as a replacement for candles.
The difference in CRI and especially saturation is particularly noticeable on the last beamshot, especially since both LEDs are extremely close to each other in CCT and duv. The brown of the foliage as well as the green comes out much better with the SFT-40.
The cool white variant is more in the greenish range. With increasing operating current, the green component decreases and the light color becomes significantly whiter, without color component. The color rendering is low, as is usual for cool white LEDs.
The tint of the warm white variant, on the other hand, is absolutely first class. (Please keep in mind, that the basically existing series dispersion, whereby other emitters can turn out differently).
The duv here is virtually zero, the location is perfectly on the blackbody emitter curve (BBL). The color rendering value is extremely high, especially the R9 with almost 100 stands out. The color saturation Rg is very pronounced, but might seem too intense to some - values above 100 indicate oversaturation. This is particularly noticeable in the outdoor beam shot - green and brown really stand out, while the saturation of the comparison LED (XM-L2 3000 K) is much weaker in almost identical color locations.
- Ra: 70/98
- R9: -36/99
- CCT: 6755/3200 K
- duv: 0.0059/0.0003
The SFT-40 is basically a very well suited LED for throwy lights. The light pattern is very homogeneous, and when using the cool white variant, a long range with simultaneous high luminous flux is possible. By using the 5050 footprint, a lower thermal resistance and a higher overcurrent capability is possible.
The warm white variant completely subordinates its efficiency to the quality of emitted light. The CRI and especially R9 are extremely high, the light color very nice, because it is exactly on BBL. Due to the lack of LEDs with high luminance and warm white light color at the same time, this variant offers a unique selling point. Unfortunately, the development and background of this variant is completely unclear, as it was only available from a Chinese dealer for a short time and this also does not appear in any official data sheets from Luminus itself.
Pro
- Compatibility with XM footprint
- very good light pattern in reflectors and lenses
- extremely high light quality with high CRI/R9 (warm white variant only)
- high luminance with high luminous flux (cool white variant only)
Contra
- for warm white variant unclear history and no official data sheet available
- partly extreme greenish tint (only cold white variant)
- strong fluctuations of the color/tint (rough color grouping)
Thank you for reading the test. 
