2023-10-23
This LED appeared on Aliexpress a few months ago, and some lamps from Convoy are now also equipped with it. As with other Chinese models, there is no concrete information about this emitter (not even basic information like maximum current or power is known at all); this test shall try to gain at least some insights.
The emitter was provided to me by german flashlight forum user @Palladin, he purchased these emitters on Aliexpress. Many thanks for that at this point.
- Type: lateral, single die
- Bin: unknown
- Color group: unknown
- CRI: unknown
- Rated voltage: unknown
- Max. Forward current: unknown
- Max. Peak current: unknown
- Viewing angle: unknown
- Thermal resistance: unknown
- Max. Temperature Tj: max. unknown
Official data sheet: not known!
This LED looks amazingly similar to an older XM-L2 or SST-40. The only significant difference are the larger contact points of the bonding wires and the different arrangement of the dots on the LED chip. Furthermore, according to listings on Aliexpress, this LED is only available in 3000 and 4000 K to date.
This LED measures 4.75 x 4.75 mm. Depending on the accuracy of fit of accessories for 5 x 5 mm XM LEDs, this could already be a problem, although in my test all centering aids and boards were usable without problems.
The LED chip sits on a gray substrate, connected with bonding wires as with older emitters.
The black dot on one side indicates the side of the cathode. Noticeable is the slightly swollen solder underneath the LED chip on the cathode side.
Due to the standard 5050 footprint (usual size 5.00 x 5.00 mm), any accessory for the XM LED size can be used freely. Even centering rings made with lathe can be used freely thanks to the symmetrical housing.
The footprint reveals no surprises and is the same as other XM LEDs. The thermal pad is electrically isolated, so the use of DTP boards is possible without any problems.
The luminous surface is basically the same as that of the XM-L2 and SST-40 LEDs, with two major differences: the phosphor seems to be applied unevenly - recognizable by the clearly protruding blue areas - and there seems to be basically a larger amount of phosphor missing at the edge of the LED chip, which may have an impact on the light pattern in optics. In addition, the entire luminous surface is āblurredā, indicating less transparent silicone and thus minimal loss of luminous flux.
The luminous area is 9.25 mmĀ² in size. The LED chip is connected with bonding wires. The disadvantage here is the loss of luminous area due to the bonding wires, but the advantage here can also be a better light image, since the LED chip does not emit any lateral light, unlike many newer flip chip emitters.
Chemical dedoming is not easily possible. Due to the problematic optical properties (there will be more detailed info in the course of the test), the goal should be to completely remove the silicone, similar to what was implemented earlier with the old XP-G2 and XM-L2 LEDs. With shaving (cutting off the dome with a razor blade), there is a risk that the optical properties will not improve, but actually worsen. However, with high quality thinner (over 85% solvent), chemical dedoming did not work.
After about 3 hours, the dome came loose and floated around in the thinner - but with phosphor, which had also come off the chip. When the LED was removed, it turned out that the yellow phosphor had a gel-like consistency and was very easy to remove. This has not been seen in this form in any other LED tested so far. It is possible that the phosphor is always āgel-likeā and is also applied in this way during the manufacturing process, which could explain the irregularities and bluish spots on the previous picture of the LED chipā¦>
The LED chip is exactly 2.10 x 2.10 mm in size. Noticeable are the very large contact areas for bonding compared to chips of the Cree XM-L3, for example, which reduces the effective luminous area somewhat.
The wavelength maximum is 450 nm (also called āRoyal blueā), which is common for most LED chips.
Besides the large contact area, the arrangement and thickness of the dots on the luminous surface is also striking. This is not known from Cree, Luminus or Osram in this form. Possibly, this LED chip comes from Chinese production, but that remains speculation. The manufacturer of this LED and the LED chip is not known anyway, as is the case with most LEDs from China.
A data sheet or any other information at all is not available for this emitter. Concrete statements with regard to official data are therefore not possible.
Overcurrent:
- Maximum reached at 9.8 A, at this point 2600 lm @ 3.92 V
- Power at maximum 38.4 W
- Sweet spot at about 6 A (2062 lm @ 3.43 V)
- Power at sweet spot 20.6 W
- Efficiency at maximum 67.6 lm/W
- Efficiency in the sweet spot 100.1 lm/W
The LML2AW.DC can be overcurrented quite well. Almost 10 amps are possible. The massive increase in Vf from 8-8.5 A is striking, which indicates changing resistance of the bonding wires heating up. This is a common effect when the cross-section is small.
Note: Although the sample tested here was not destroyed by burning through the bonding, destruction in other samples in the range of 9 to 9.5 A is quite possible and must therefore always be taken into account for applications with currents greater than 8 A!
Based on the available data, I assume that this emitter, like the SST-40-W, is specified for a continuous current of max. 6 A and also has a similar thermal resistance as the SST-40-W (in the range 0.8-1.2 K/W).
The comparison with other LEDs is really interesting. Basically, the curves of the Luminus SST-40-W and this LMP are extremely similar. The LMP probably has a slightly higher thermal resistance, which is why the efficiency is always lower with increasing current. Otherwise, the characteristics are virtually identical. Even the Vf is extremely similar to the SST-40-W.
The already more than 10 years old XM-L2 of the first generation (ā1st genā) offers a higher luminous flux, whereas the Vf is extremely high. The new generation XM-L2 (ā2nd genā) offers extremely high luminous flux and is almost āindestructibleā thanks to its flip chip design.
It seems that SST-40 and LMP have some similarities. Despite the different design and the slightly different looking LED chip, both LED chips and also the general design of the LED seem to be extremely similar. Although this is not necessarily an indication that we are dealing with a Luminus LED chip, which is built into an LED in a joint venture (?) together with a Chinese manufacturer, the similarities are simply too clear to simply ignore them, especially since such characteristics are hardly ārebuildableā and every LED manufacturer and manufacturer of LED chips leaves his āfingerprintā somewhere in these characteristics. On the other hand, I have only tested one LMP here, so there could be strong deviations in other samples.
Regarding the real manufacturer and more informations about this emitter, I have found interesting information. These probably concern not only the LMP LML2AW.DC, but also other LEDs with the same abbreviation āLMPā.
Maybe it could be that āLMPā simply stands for āLuminus Manufacturing Partnerā. In 2013, Luminus Devices was taken over by Sanan Optoelectronics (the newer SFQ43 or SFN55 LEDs come from them).
Basically, Luminus is a Chinese company. Sanan holds (as far as I know) 100% of the shares of Luminus Devices. The fabs (manufacturing factories) for the LED chips are located in China, including Xiamen and Nanan, also in China. Sananās main production facility and corporate headquarters are also located in Xiamen, by the way. Some types of individual LED chips are manufactured in the Philippines. The locations of the Chinese fabs, for example, come from this P2CN.
Luminus is silent on their own website about the location of the fabs and also about the acquisition of Luminus in 2013. It is also not mentioned in their list of milestones. In general, I have not found any publicly available information about the acquisition by Sanan on Luminusā website. There are only articles from other magazines about it.
The address of the Luminus headquarters given on their own page seems to be odd. Under the given address there is a relatively small, one-story building with an extensive cooling plant (?) on it. I donāt think that extensive operational activities are handled from there. I actually could not find the Operations Office in China, the address allegedly does not exist (or I am not able to find it on Google Maps and Bing). All very strange.
Looking at the trade-political situation of the last years, the sudden wide-spread use of Luminus LEDs in flashlights developed by Chinese companies (Nitecore, Fenix, Sofirn and Co.) seems very understandable: Independence from the West.
Either way, I now finally see Luminus for what it is: a Sanan Optoelectronics brand that exists only to be able to ship to the West. For me, Luminus is a Chinese LED manufacturer, even if the companyās own presence should suggest otherwise. Maybe I am also dead wrong with this, and I would be very happy to get more informations about this LED, the background of the development and production or additional documents like datasheets.
Values at 25 Ā°C Tsp, at 85 Ā°C Tsp values are 13 % lower
The luminance corresponds to that of a Luminus SST-40, which is to be expected due to the close resemblance. There are no other surprises.
The LML2AW.DC has a big problem with secondary optics. It is not clear to me what causes the extreme ink shift depending on the beam angle, but it is already very noticeable without optics. Possibly this is due to the silicone used (refractive index?), but I have never seen it so extreme with any LED with lateral chip (Lambertian radiation behavior).
The outer area is very blue, only the spot is in the light color according to which the LED is classified. Although the light is mixed well in a diffuse environment (e.g. integrating sphere), the difference is even greater when reflectors are used. Here you can already see how extreme the ink shift is in the reflector.
The use of slightly thicker centering rings can reduce the extent of this tint shift, but this has to be tested individually for each reflector and LED.
When using a smaller OP reflector with a lower height like the one from a Convoy S2+, the blue tint shift in the spill is much more visible. But even in the SMO reflector it is clearly visible and annoying. This LED is only really usable in very diffuse optics that mix all of the light, such as heavily textured (beaded) TIR optics.
The light color of the sample measured here is excellent. With a duv close to or below 0, there are no green or yellowish components. However, the low CRI of only 72 is a disadvantage here, which is definitely noticeable. For those who donāt need high color rendering and just want a nice light color, this LED (considering the problems in optics) is definitely a good alternative.
Interesting is the kink in the 530 to 580 nm range, which is probably intentional. It is possible that the phosphor mix was deliberately designed to reduce the green component to the point where the color localization is as close to BBL as possible. Such an obvious adjustment of the green component is relatively rare in neutral white LEDs (normally a simple curved curve is to be expected) and is an advantage for all those who want a good light color.
For this reason, among others, I suspect that these LEDs are specifically designed for use in streetlights and downlights without secondary optics, where the tint must be usable and reproducible, but high color rendering is not required. This seems all the more likely since there are also reports of very good tint with the 3000 K variant around.
- Ra: 72
- R9: -20
- CCT: 4125 K
- duv: -0.0008
The LML2AW.DC LED does not have it easy. As a ācloneā of the XM-L2 or SST-40, it has the potential to be a worthy successor to these older models, but it does not even begin to exploit this potential.
The biggest problem are the optical properties. The extreme tint shift with increasing beam angle is extremely problematic for the use of secondary optics and creates a not so beautiful light pattern with blue spill and yellowish spot. In addition, chemical dedoming is impossible, at least in my test, and variants with high color rendering are not available. Finally, the complete lack of any official data as well as the manufacturer should also be noted, which makes DIY use much more difficult.
Pro
- very pleasant tint (duv below 0)
- XM footprint, full compatibility with XM accessoires
Neutral
- low color rendition index (CRI)
Contra
- extremely high tint shift when using in conjunction with secondary optics
- absolutely no information to this LED available in public
- only available at chinese stores
Thank you for reading the test. 
