This LED is less than half the price of a SBT90.2, with the same LES and a 7070 footprint, single 3V die. If true, the claimed 5500lm would put its output on par with the SBT90.2, possibly above–the SBT90.2 sample tested by djozz did not crack 5300lm @ 20A.
Interesting offering, though I worry that the lack of a glass cover would cause long-term reliability issues: given how large the LES is, a mote of dust is bound to eventually land on it, burn up, and cause cascading failure…
Looks like a new design and a new package. The electrical configuration looks the same as the sbt90.2, but the bond wires have been potted to the die like ither SFT leds. In all the sft leds, this has increased the robustness and lowered the Vf and thermal resistance vs the older designs (sst40 vs sft-40, sft-25r vs sst-20) for more output. Maybe this one will be the same? The smaller footprint may be a detriment though as it will make it less able to dissipate heat from the die to the mcpcb.
As with other Luminus LEDs I expect heavy tintshift with rising current (so visibly green at low current), luminance similar to SFT-40 and 5000 to 7000 lm, maybe 8000 lm, similar to SBT-90.2 - I expect the light flux to be lower as for the SBT-90.2 since the footprint is smaller. Unfortunately I don’t have SBT-90.2 for testing.
The KD samples are J7 BA, so relatively low bin and supposedly on BBL. But I am pretty sure that these are over BBL, duv 0.003 or something like that.
Because of the low price and smaller footprint it is likely that this LED will replace the SBT-90.2 in most lights. Just a matter of time (and availability) that Simon and Hank take notice of these, hopefully they have higher binnings and/or better color groups at this point.
I will get a SBT90.2 in the next weeks (maybe I will write a short review about it), so I can compare it directly with the SFT-90.
Very interested to see how these LED will compare to each other in terms of light flux and Vf. The optical properties should be more or less identical.
This is something that could make a small difference. But since we don’t know of which material the bonding wires are and how thick these are, this could be negligible.
Despite not being the biggest fan of Luminus LEDs, this could be very interesting since the SFT-90 is more like a replacement of the SBT-90.2 than a completely new LED. Wouldn’t be surprised if the SBT-90.2 is EOL after few months of introducing the SFT-90, since most customers with bigger quantities will likely switch to SFT-90 due to much lower purchasing prices.
I ordered 2 as well, will order a L21A host + FET driver + Ampace JP40 to go with it, I think. I was never interested enough in the “Big ass thrower” category to fork over 25-30 bucks for a LED, but 10 is fine by me
I’d been hoping Luminus would make an SMD version of the CFT-90, looks like this might finally be it! Hopefully the smaller thermal pad isn’t a problem.
Fingers crossed for an SFT-90-WDH (6000k Ra92) at some point.
It seems the SFT-90 uses a slightly larger chip as the CFT-90. At least the chips seem to be different (maybe due to the age, the CFT-90 was released several years ago, at least before 2017).
In my experience the sbt90.2 offers similar characteristics to other Luminus LEDs like DUV dropping as current increases and abnormally good power handling characteristics.
I’m VERY interested in the results. Im expecting performance around the top end of xhp70.3 range lumens-wise, but we are talking about a smaller LES for more throw, correct?
Screw it, I’m ordering on a 16mm PCB and slapping it in a 3v 8 amp Convoy.
I don’t have any host I could test it in yet… I have one M21B with a 3V 5A buck, and another with a 9mm hole OP reflector, but I don’t really want to Frankenstein 2 of my M21Bs into a single light just for testing… Also, 5A is very underdriven.
For some strange reason the Amazon delivery service picked up Kaidomains shipping at the Italian airport and brought it to me… This was crazy fast!
I have no host that can take a 7070 LED and has a 3V driver (yet) so I can’t really test them… Did a short test with the lab PSU, at 200 mA. I measured a forward voltage of 2.59 V during that, and took an opple screenshot (see below). Looks pretty much as expected.
I can’t wait to get my L21A! I bet the Ampace JP40 with FET driver will be crazy with this LED.
So, I did a first test of my Kaidomain samples. The SFT-90 is somehow quite… underwhelming. Tbh I don’t know if it is worth to write a full test about this LED, at least for these pre-production samples.
It barely manages the specified max. 20 A at 4956 lm and 3.87 V. The specified luminous fluxes for stated flux bin J7 are definitely not met. Vf is also quite high.
The J7 binning at 10 A 25 °C is not complied with at 3348 lm, but is still within the tolerance of ± 6 % specified by Luminus.
I got 2.584 V at 200 mA, so my sample could be some mV lower in Vf.
It is not worth overcurrenting the SFT-90, since it maxes out at around 21 Amps. I suspect that the heat transfer from the chip to the ceramic substrate is not perfect, because the 7070 footprint is not fully utilized here, since it can deal with over 150 W power. Incidentally, my second sample performs similarly, so I rule out problems with the board or reflow.
Pretty sure we will have to wait for mass-production samples with regular binnings (maybe I will re-testing it in 6 to 12 months).
However, the tintshift is hardly pronounced as the current increases - even at low levels the LED is barely green. The tint becomes pure white at just a few amps. Has Luminus finally managed to adjust the phosphor mixture?
