Andúril 2 coming to Sofirn - The general Sofirn development thread

Yes, they tend to wash out colors with a similar effect to a low-pressure sodium vapor lamp, but not quite to that extreme.

It just depends on your use case. For example, I’m a technician and I specifically choose high-CRI lights for work so I can tell different colored wires apart or read the bands on a resistor. I’ve had problems doing this with the cool white low-CRI lights they provide for us.

Yeah, I use my headlamp for auto/motorcycle work and also need to identify wire colors accurately. I tried a few leds and settled on a 80+ CRI Nichia 319AT in 5000K. It’s about the same as a 219c in 5000K.
I tried a 90+ CRI 4000K 219c, an 80+ CRI LG351D 5000K and the 90+ CRI LH351D 5000K daogf4rts but didn’t think the colors were as good. Just because the CRI was higher didn’t mean it was more accurate for my eyes.

In my normal walking around flashlights I’m not so picky. I just try to stay away from green.

Some people are high CRI snobs, but I guess it depends on the individual and therefore is not automatically better.

I have some NW and CW emitters which are pretty decent, those “angery blue” tints are just annoying (though I can tolerate them in a 5mm in certain applications).

That said I just don’t see any reason to tolerate low cri anymore. You have the sst-20 which offers high current and output in a small size, LH351d and even the xpl hi is offered in 80+ cri flavors. I guess if you are just trying to build a lumens monster that’s one thing but for real world use why wouldn’t you want a high cri led? I get with the 219 leds you were sacrificing some max output but with newer options you aren’t, really. At least nothing that would be missed.

For example I just reflowed a 4000k sst20 into my lumintop worm. It’s still every bit as bright as it was with the xpg but now in a beautifully neutral tint with great color rendition. Still the same excellent beam shape, too!

The SST-20 is quite low output for it’s amperage. The XP-G2 blows it away. It’s close to the 219c output. It’s just a small die, so output is not so great.

Note that it’s lower Vf, so at the same amperage it consumes less power. But at what amperageis it blown away?

IDK what tests you’ve been looking at but 5000K and 6500K SST-20s are certainly not far behind XP-G2 by much, if at all.

I thought this was quite well known. Here is a direct Djozz comparison from this thread. This is a 3,000K, maybe this particular version has lower output? The G2 in this test is a 2B tint so around 5700K. Maybe it is around CRI 70? That might explain some of the big gap.

Maukka’s test above shows higher output, but I can’t find any tests from him on the XP-G2 or 3 for a comparison.

The general characteristics I remember from the SST-20 is it’s cheap, looks good, throws well due to it’s small die size, but output is lower and it runs hotter.

The 3000K and 4000K are high-CRI only and therefore significantly less efficient. 5000 and 6500K are 70-CRI only and still compete with much larger die emitters in output/efficiency.

On Emisar lights for example the max output of the former is listed at ~3000lm, and 4200+ lm for the latter.

I think all the SST-20 tests I’ve seen were a while ago and they were all high CRI because that is what people seemed interested in. Like with FW3A for instance, the SST-20 were only for the WW tints then they switched to XPL-HI for NW and CW. You could see the big difference in output and heat. Maybe I have not given the cooler tinted SST-20 enough research. I haven’t fully read Maukka’s 6500K test. Maybe they are not as bad as I thought (efficiency wise).

Yeah for sheer output while having very high illuminance the cooler SST-20s are pretty solid. Only issue is how green the beam is compared to other options like X-PL HI.


SST-20 6500K:

Who’s numbers are these?
It’s not an apples to apples comparison due the dome difference, but still pretty close. Especially considering the XPL has a 2mm2 die and the SST-20 is closer to 1.20-1.25mm2.

Looks like the XP-L HI data points are from Texas_Ace and the SST-20 ones are from maukka

IDK how it’s “not apples to apples”. The X-PL HI is dedomed but the effective die size of the SST-20 is still smaller, that’s why its throwier. The point was that it’s comparable to emitters with a larger physical die in output and efficiency (but unfortunately has a pretty gross beam in comparison).

Gchart is right on the sources

Ooof, that TA chart is from Jan 2017. I think TA’s integrating sphere was reading a bit too high back then. We may need to adjust those results down a bit. When we do that the 2 LEDs are actually pretty close to each other even with different die sizes.

It would be a better comparison if it were the domed XPL compared to the domed SST-20.

Yes and no. :wink:
With optics, XP-L HI and SST-20 have quite comparable beam widths; the latter being a bit narrower despite having a dome.
For most uses where one would consider XP-L HI, SST-20 is worth thinking about as well.

BTW I wish there were intermediate CRI variants of SST-20. And a factory domeless option.
CRI95 is indeed inefficient, the price one has to pay for pushing CRI this high is very significant while CRI80-90 is not so bad in other modern LEDs (about 5-10% for 80, 15-20% for 90). CRI95 in SST-20 costs very roughly 45% output. :person_facepalming:

I just got an idea...

Looking at HaikeLite's new HT70 I wonder if Sofirn could consider upgrading their SP33 V2 (boost-driven 6V XHP50.2) with the new Luminus SST-70 including SMO reflector. According to Djozz' test it should output around 3.500lm at 5A (that's about the max. current SP33 V2 used). They could call it "SP33 V3 Pro".

SP33 V3 = 3V XHP50.2, FET driver, OP reflector, proprietary Sofirn UI

SP33 V3 Pro = 6V SST-70, boost driver, SMO reflector, Andúril UI, desert tan color

There would be no increase in output. Lumens might actually be less. The only advantage would be more throw distance since it’s a smaller led.

I really hope the majority is not always looking for more and more lumens, only. There are so many advantages to be expected such as a more stable runtime (even on turbo), less tintshift compared to the 3V XHP50.2, better throw and - most important - less heat generation at 6V. You need 10A at 3V what you can achieve with 5A at 6V.

Sofirn is currently working on a new corporate design, including a new slogan they may put on their cardboard boxes. The slogan should ideally reflect following content: high quality, reliable products, great customer service and affordable prices. Your input will be deeply appreciated. :THUMBS-UP: