LED test / review - YLX N3535B cool white - round-die LED with high luminance for 3535 footprint, but very green

koef3 · 2023-11-11T17:46:43.838Z

LED Test / review EN

YLX N3535B cool white

2023-11-11

This LED appeared some time ago in the Wurkkos TS12. The special feature here is the round LED chip with a compact XP footprint (3535). This test should clarify the overcurrent capability and show whether this LED is suitable for throwers due to its luminance.

The ‘YLX N3535B’ tested here was taken from a Wurkkos TS12, which was made available to me by German flashlight forum member ‘LightintheNight’ for conversion to dedomed Nichia 519A. I am not aware of a source of supply for individual emitters.

Technical data

There are no official specifications for the LED, nor is an official data sheet available, so no maximum specifications can be given.

First appearance

In principle, this emitter is visually similar to the 5050 variant already tested, which is also known as the “Yinding 5050”. The LED chip is fitted with a glass cover and four bonding wires ensure the power supply. This LED does not have a dome, so a high luminance is to be expected.

The glass cover was not placed exactly in the center of the LED, it is noticeably slightly crooked, but this does not affect the function of the LED.

The symmetrical XP footprint (3535) allows the use of rotated centering aids and other accessories. Depending on the reflector used, it may be necessary to pay attention to the different heights of the centering aids.

The footprint has no special features. The thermal pad is electrically insulated from the anode and cathode so that DTP boards can be used without any problems.

LES

The illuminated surface is round, as already seen in the larger version. Only the Luminus SST-12 offers a similar luminous surface with a round shape.

There are initial speculations that these chips with a round LES all come from Sanan. Luminus is 100% owned by Sanan, so a relationship between the LED chip used in the SST-12 and this ‘N3535B’ is quite possible, especially as the LED chip of both emitters is the same size and the bonding wires are arranged in the same way (only the number of bonding wires in the SST-12 differs here).

The luminous area is 1.40 mm² in size.

The luminous surface is connected with 4 bonding wires. This should reduce the sharp increase in forward voltage with increasing current; this effect was particularly noticeable with older Cree or Osram emitters, caused by the reduced cross-section of the bonding wires, which resulted in increasing resistance.

Power and overcurrent capabilities

In the absence of official information, no statements can be made about performance at official maximum values. Based on the curve, however, it can be assumed that the maximum current from the manufacturer is probably 3-4 amps.

Overcurrent:

Maximum reached at 9.4 A, at this point 1421 lm @ 3.87 V
Power at maximum 37.2 W
Sweet spot at about 5 A (1155 lm @ 3.43 V)
Power at sweet spot 17.2 W
Efficiency at maximum 39.2 lm/W
Efficiency in the sweet spot 67.3 lm/W

For an LED in the XP footprint, the overcurrent is considerable. Experience has shown that the (too) small heat dissipation surface of the XP footprint is one of the limiting factors in the 10-12 A range. The forward voltage is pleasingly low, especially for an LED with a classic connection using bonding wires. The presence of four bonding wires is probably the reason for the low forward voltage.

Direct drive (connecting the LED directly to the battery without a driver) should be possible in principle, but this practice is highly risky when using batteries with a good voltage range because, on the one hand, the series dispersion of the forward voltage can be significant and 21700s in particular can sometimes reach over 3.9 to 4 V under load at 10 A discharge current, meaning that the LED would be energized far beyond the theoretically possible maximum.

The ‘YLX N3535B’ is very efficient for an LED with such a small light-emitting surface. Although the Nichia 519A used as a reference for LEDs with high color rendering achieves a slightly higher luminous flux, it also uses a considerably larger illuminated area, especially as it also has a dome.

The variant known as ‘Yinding 5050’ in the XM footprint has a significantly higher luminous flux, and the CSLPM1 with its larger illuminated area of just under 2.0 mm² also has a higher luminous flux. It is noticeable that the 3535 and 5050 LEDs have very similar characteristics, which suggests that the LED chips are very similar to each other, even if the Vf differs greatly and the 5050 LED has a very high Vf.

Luminance

Values at 25 °C Tsp, at 85 °C Tsp values are 13 % lower

The luminance is extremely high. Due to the small luminous surface, it is higher than most other emitters even at low luminous flux.

Although an older Osram Black Flat HWQP, for example, has around 10-20 % higher luminance at a lower operating current, the ‘YLX N3535B’ benefits greatly from the higher luminous flux and significantly better overcurrent capability.

Further advantages are the standard XP footprint and the simpler reflow process. In addition, heat development at maximum luminance is more manageable thanks to the lower electrical power. Basically, the ‘N3535B’ is therefore a good LED for throwers and applications where a small reflector is required for the longest possible range with low power consumption.

Light quality and use in optics

The light image is perfect. There are no tint shifts or other artifacts. Depending on the reflector and quality, rings may appear, although these do not occur in this form with commercially available lamps (Wurkkos TS12) and are therefore dependent on the reflector. The spot is sharply defined and merges into a narrow corona - the spill is relatively dark.

Color and light quality

Simply said, the light color is green. The green tint is extremely noticeable and annoying. The spectrum, on the other hand, is typical of an LED with reduced color rendering and high efficiency.

There are definitely ‘YLX N3535B’ which have a pleasant tint, but this is either a matter of luck or depends on whether this LED was purchased in a corresponding color binning, which is usually completely unclear to the buyer or user of a flashlight.

Ra: 66
R9: -68
CCT: 5928 K
duv: 0.0189

Conclusion

In principle, the ‘YLX N3535B’ corresponds to the previously tested 5050 LED with round LED chip (also called ‘Yinding 5050’).

The big problem, as with the vast majority of Chinese LEDs, is the complete lack of data sheets and even manufacturer/product names. It is very likely that ‘YLX N3535B’ is some sort of placeholder name.

This LED has a very high luminance, which makes it ideal for throwers. The Vf is pleasingly low for an LED with a classic lateral LED chip and the efficiency is high. If the light color is not important and only max. throw with low power consumption is essential, this LED may be worth a closer look, although this LED is not yet available to buy separately anywhere.

Pro

very high luminance
full compatibility thanks to XP footprint
low Vf

Neutral

no variants with higher color rendering or warm color temperature available

Contra

no official information (data sheet, manufacturer, type) available
no source of supply known
very green tint (tested sample)

Thank you for reading the test.

Greetings, Dominik

v1.0.1

Kazik · 2023-11-11T19:58:45.351Z

Thanks for Your test!
Wow, luminance higher than Osram NM1…
Wurkkos put those in TS12, first batch was greenish, but then they released second batch with more pleasent tint. So not all hope is lost

koef3 · 2023-11-12T01:07:14.057Z

Kazik:

Wurkkos put those in TS12, first batch was greenish, but then they released second batch with more pleasent tint. So not all hope is lost

Yes, I have also sample of the ‘YLX N3535B’ with much more pleasant tint ( more like rosy, it was salvaged from a newer Wurkkos TS12), but I don’t know if all LEDs of this ‘second batch’ have this better tint…

The higher luminance of the ‘YLX N3535B’ is btw not really relevant here, as this only comes into play at all at maximum current (which means maximum heat and shorter lifespan of the LED due to higher temperature Tj) and at normal current like in a flashlight suitable (such as 3 A) the CSLNM1.TG still has a significantly higher luminance.

Henk4U2 · 2023-11-12T09:26:28.361Z

Thanks for the review/test. It’s always good to have such info on hand in case you need it.

YBF650 · 2023-11-13T06:31:56.446Z

Thanks for the review, as always.

This is such an interesting emitter, looks like it could be an alternative to the CSLPM1 for throwy hotrods like the Emisar DT8K, especially if they sort out the tint.

Fingers crossed it becomes available somewhere else, instead of having to remove them from a TS12.

djozz · 2023-11-13T20:29:08.870Z

Thanks for the test! It is close to a very nice led for flashlights. The “Yinding5050” nowadays comes in a fairly nice 3000K tint, so maybe this one also. If you can find it of course

koef3 · 2023-11-13T20:34:04.633Z

djozz:

The “Yinding5050” nowadays comes in a fairly nice 3000K tint

The real manufacturer of “Yinding 5050” is still unknown, maybe it is ‘Lumenpioneer’, but who knows…

The 3000 K of the ‘LMP W5050SQ3’ I tested recently are only in low CRI 70, which is also visible on colors, especially if a comparison (high CRI) in same CCT is available.

I think the round-die LED chips are really nice, but since they are only available in Chinese LEDs with extreme disclosure and mysterious background I cannot really recommend them (but they are not easily available so it doesn’t matter)

Sirstinky · 2023-11-14T00:12:50.389Z

Thanks for sharing. Good LED for a compact thrower, and should be good for a buck or FET driver. This is my experience with the Chinese LEDs coming from Sanan-luck of the draw on tint. Some have been okay, some really -duv, some really +duv. One of the greenest I’ve seen was the round die Lumengin in the Nitecore MH40S. There’s no predicting what you get.

1stein · 2023-11-15T16:22:14.962Z

20231115_1700271920×2560 384 KB

My precious…

koef3 · 2023-11-15T16:48:04.885Z

High CRI? Tell me more

1stein · 2023-11-15T17:50:05.741Z

Yes
CRI 101

But I’ll surely check it against my Opple3 once installed

koef3 · 2023-11-15T18:01:05.462Z

1stein:

CRI 101

You forgot the minus

1stein · 2023-11-15T18:18:38.101Z

LED tester:

You forgot the minus

No, it’s binary

QReciprocity42 · 2024-01-13T21:39:15.892Z

Found this LED that looks a lot like the YLX with the glass covering, but with a square and smaller (eyeballing 1mm^2) LES. Super cheap too at $0.6 each, wonder if it might have similar performance.

koef3 · 2024-01-13T22:17:11.218Z

Looks like a Osram LE UW Q8WP… especially for the 15 W variant

QReciprocity42 · 2024-01-13T22:19:57.830Z

Nice catch! You have tested the Osram, and the luminance was good, with issues being off-center LES and low efficiency. Don’t know about efficiency on this one but at least the die is centered.

koef3 · 2024-01-13T23:22:46.047Z

efficacy ≠ efficiency:

Nice catch! You have tested the Osram, and the luminance was good, with issues being off-center LES and low efficiency. Don’t know about efficiency on this one but at least the die is centered.

I would test this emitter, but it cannot be sent to Germany unfortunately.

1stein · 2024-01-14T07:00:13.714Z

LED tester:

but it cannot be sent to Germany unfortunately

No worries, just ordered a few. I’ll share 2pcs with you

TUIG101 · 2024-01-14T14:45:03.005Z

Did you also get the 5W version? It also seems to handle a lot of current, based on the massive bond wires quantity.

1stein · 2024-01-14T15:24:20.369Z

Nope. Only 15W.