LED test / review - Luminus SFT-70-X Specialty White (6500 K, typ. 70 CRI)

LED Test / review EN


Luminus SFT-70-X Specialty White (6500 K, typ. 70 CRI)


2023-08-07



The SFT-70-X is now a well-known cold white LED without a dome in the 6 V class and is currently frequently installed in flashlights with a long range and high luminous flux at the same time.

This emitter was purchased from Mouser in early 2023. A date code is not known here.


Technical data


Tj 85 °C, If 1,500 mA
  • Type: multi-die (6 V class)
  • Bin: KB, min. 1120 lm
  • Color group: 652 (6500 K)
  • CRI: typ. 70
  • Rated voltage: min. 5.25 V - typ. 5.75 V - max. 6.25 V V
  • Max. Forward current: 7,000 mA
  • Max. Peak current: — mA
  • Viewing angle: 120 °
  • Thermal resistance: 0.6 K/W
  • Max. Temperature Tj: max. 150°C

Datasheet can be downloaded here: Datasheet (newest version, Luminus)


First appearance


The SFT-70-X is an LED without silicone dome (called domeless). The 4 luminous surfaces are well recognizable. The bonding wires protruding from the white silicone, which presumably serve to connect the individual luminous surfaces, are also visible.

The gold colored frame serves as a border for the white silicone and gives this LED a relatively unique look. The dark gray substrate is similar to most other LEDs.



Thanks to the XM footprint and symmetrical housing, commercially available accessories can be used freely.



The footprint is designed a bit differently than usual for most LEDs. Although it corresponds to the XM footprint (5.00 x 5.00 mm or also called “5050”), so that conventional XM (DTP) boards can be used without any problems, the individual contact surfaces protrude clearly visible from the gray substrate, which also applies to the thermal pad. This makes for a supposedly “bad” soldering with “gaps”, although the LED is firmly seated on the LED board (see also the side view of the LED).

When using commercially available LED boards like Noctigon XM20, the SFT-70-X is in general operated in the 6 V configuration (2S-2P). It is also possible to operate the LED in a 12 V configuration (4S), but this requires a special LED board with a modified connection layout.


LES


The luminous surface consists of four individual LED chips. These are similar to the older XHP35 HI, they are arranged as closely as possible to prevent any dark stripes or gaps. This is necessary for homogeneous light pattern and avoids artifacts when using reflectors.



The luminous area is 6.51 mm² in size.

Due to the number of bonding wires, a too large increase of the Vf with increasing operating current is not to be expected (as seen as with other emitters with bonded LED chips). Nevertheless, the Vf could be somewhat higher than would be the case for an LED with more modern flip-chip design (connection from the bottom of the LED chip directly on the substrate).


Power and overcurrent capabilities



Within official parameters, as far as known:

  • at 7,000 mA (official maximum current): 3605 lm @ 7.10 V
  • Power at official maximum: 49.7 W
  • Efficiency at 7,000 mA: 74.2 lm/W

Luminus sets the official maximum current relatively high at 7 A. The data suggests that this is a 50 W LED. My sample has a slightly higher Vf than typical, although this is within the range of what the data sheet states.


Overcurrent:

  • Maximum reached at 9,600 A, at this point 3926 lm @ 7.58 V
  • Power at maximum 72.8 W
  • Sweet spot at about 7,000 A (3605 lm @ 7.10 V)
  • Power at sweet spot 49.7 W
  • Efficiency at maximum 54.0 lm/W
  • Efficiency in the sweet spot 74.2 lm/W

Overcurrenting is not recommended. The power, which increases by more than 30% compared to the official maximum specifications, does not justify the luminous flux, which is only about 10% higher. If necessary, it may even be advisable to reduce the operating current to 5.5 to 6 A in order to significantly increase efficiency.


Luminance


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


The luminance is high, but the maximum current is required for this. At the official maximum current, the luminance is about on a level of the XHP35 HI or the old XP-L HI V3, but with significantly higher luminous flux. This is the feature, which is why this LED is currently mainly installed in lamps with high range.

Light quality and use in optics



The light pattern is good. There are no disturbing artifacts like donut holes, and the spot is sharply defined. There are also no disturbing color fringes or other anomalies. For these reasons, the SFT-70-X is suitable for flashlights without any problems.

When using a zoom lens, the structure of the four illuminated areas becomes visible, but this does not bother much thanks to the almost non-existent gaps between the LED chips.

Color and light quality



Thanks to the narrower color grouping (652), which excludes the reddest or greenest tints, the light color is okay, at least at higher operating currents. As the current increases, the tint changes from slightly greenish to clear cool white without annoying color casts. The SFT-70-X tested here is actually one of the more pleasant samples and is thus suitable for use in lamps without any problems, although at low operating currents (up to about 500 mA) the green tint is noticeable and annoying depending on the sensitivity of the user to green colors.

The spectrum offers no special features, it corresponds to the spectrum of a cool white LED with a low CRI of typ. 70. Due to the low color rendering, colors appear distorted, for applications where high color rendering is important, the SFT-70-X is thus basically not recommended due to the lack of availability of high CRI variants.


  • Ra: 70
  • R9: -49
  • CCT: 6370 K
  • duv: 0.0134

Conclusion


The SFT-70-X is an LED with high luminous flux and relatively high luminance.

I like the luminous surface without disturbing gaps between the LED chips and the associated high quality in the light image. In addition, there are no disturbing color fringes around the spot. By selecting a narrower color grouping, the tint of this LED is usable, with increasing operating current the slightly greenish color cast disappears.

I did not like the non-availability of variants with high color rendering and the almost not worthwhile overcurrent, which results from the already very high official specification of the maximum current.


Pro

  • very good light pattern
  • LES with 4 chips without gaps
  • high light flux for domeless LED
  • greenish tint disappear with higher forward current

Neutral

  • greenish tint

Contra

  • no high CRI variant available
  • overcurrenting not worthwhile

Thank you for reading the test. :slight_smile:

Greetings, Dominik


v1.0.1
9 Thanks

Thank you Dominik for this review.
This LED is inside Olight Warrior 3S. I am not a big fan of spot round optic but in this flashlight is quite impressive. With level 2 you have 12 hours of decent light without hot flashlight

Nice test! In your opinion does this led do something better than 50.3HI?

Not really. If you need a 3v LED then the 50.3 has a decided advantage, but the only other way is efficiency and output when overdriven becsuse the Cree is a flip chip/csp so it comes down to design limits. I will say the sft70 is a superior LED to the older sst70 though and if you have a light with a 6v sst70, this one will drop in easily. I’ve tested several with the sst70 vs sft70 and i really like the sft70 better. Cleaner beam and equivalent output.

50.3HI is usually 6V or 12V led as is SFT-70, so it could be used instead of SST-70 also.

Thanks for the test, good work.
I’m surprised you didn’t see a donut hole with this emitter. Iirc Simon and/or Hank stated the donut hole was the reason he/they don’t use the sft70.

Here’s mine

Close to the wall for emphasis, but the donut hole is still visible at most distances. Did they fix this on the new ones?

Please make a picture with more distance.
In this short distance a donut-hole could even occur with other emitters with only one LED chip, depending on the reflector. I took the picture with a distance of 1.5 meters or so.

I understand. It’s not as close as it looks, but I took it that way because my camera doesn’t capture the effect well; bright light all just blurs together. But I measured out 1.5m and gave it a shot anyways. There’s more of a donut hole in person, but I think you can still see it here.

1 Thank

Thanks for the review, better throw and output when compared to a XHP35Hi is good.

Isn’t that a problem with the design of this type of LEDs, using 4 smaller LEDs you get a center gap, the XHP50 use to have the same issue but Cree some what fixed it with the newer LEDs. You can’t get each LED close enough together and it creates this problem, That is where the flip chip design helps. Also being domeless makes it more obvious.

Yes, it is defintely there.
Weird… with my old Convoy C8 SMO reflector I use for testing the light pattern I don’t get a donut-hole, even if I change the position of the LED in the reflector a little bit…

Maybe your reflector is simply not designed for use with these LEDs, I don’t know. I could imagine that the LED is not focussed right, it is too high or low in the reflector. What reflector from which light you are using?

Not necessarily. The SFT-70-X is the best proof, that even with classic lateral chip design a decent multi-die emitter can be designed. The gaps depending on the distance of mounting the LED chips and if the LES covers the whole LED chip up to all edges. The most flip-chip LED chips are emitting light also to the side, which also helps a lot for these multi-die emitters.

About donut hole: I remember Convoy’s Simon mentioning that he doesn’t want to use SFT-70 because of that problem. Also Pioneman has(had?) a warning of a donut hole with lights that were sold with SFT-70. So it might need very careful tinkering for no donut hole in the hotspot.

This is just an off the shelf thrunite catapult pro, so it was literally built for this emitter, you’d think it’d be good. But thrunite does have some issues. For reference the reflector is similar to an L21b but not quite as deep. Giant centering ring.

I have another sft70 on order from KD right now though I’m going to try with the 6v5a convoy driver in something else. Maybe an m21a. See if it’s any better

There is a 3000 K 95CRI variant of SFT-70 available (I think color quality similar to SFT-40 3000 K). Ordered a few :slight_smile:

7 Thanks

Very exciting, I wonder how it compares to an XHP50.3 HI

1 Thank

Just saw a LM3 test result and some beamshots over at Reddit (not sure if links are allowed so I didn’t give one). Seems to be rosier than SFT40 3000K, I wonder about the output…

I ordered some SFT-70 3000K at KD. Since they ship extremely slow sometimes, I don’t know when I get them for testing. Could be in two weeks, or in four months… :roll_eyes:

I will post a light flux chart and luminance values as soon I tested them. I am pretty sure that the efficiency is a lot lower as with cool white, in similar range we seen with SFT-40.

4 Thanks

I ordered things from KD a few years ago with untracked shipment, still waiting for it to this day. :joy:

Looking forward to your results!

1 Thank