If you’re comparing Texas Ace (219B) and maukka (sst20) numbers, on the one LED they both tested (E21A) TA was 7~8% above IIRC so that 219B number might be a bit overestimated. Plus the sst20 is 500K less.
What’s the die size on the SST-20 ? Might be a bit smaller, although the 219b is quite small already.
One other thing to consider is that less green makes the LED less efficient, so it’s difficult to compare sw45k. Non k bins would be better to compare to but they haven’t been tested.
I was using those two datasets as reference. I didn’t realize TAs numbers were that much higher… I have tested multiple SST20s and my results line up with maukkas very well. I’m ordering some 219Bs from Simon and will test those in the future apples to apples.
Effective die size on the SST20 is 1.75mm² according to my spreadsheet. Natively it’s 2mm², but the bond wire pads block some. I have some sliced 219Bs in my stash I could measure later but I’m fairly certain it’s just under 2mm² from memory. Nothing recorded in my notes. I’m getting about 1.5mm * 1.5mm so 2.25mm²
Also, my thoughts on use of 219B is as follows. If per emitter current is below ~2A the majority of the time (aka triples quads etc) and especially if there is a buck or boost (for AA lights) driver then I’ll take it over SST20 all day. By the time you tint correct or pwm a FET on the SST20, efficacy is at or below a 219B.
SST20w is a throw king though and wins out in single emitter lights with anything larger than 1*14500
Personnaly I like sliced 90CRI LH351D, especially 3500K : virtually no angle tint shift, neutral tint/slightly under the BBL at low/high output, R9 close to 80, while keeping a higher efficiency than an SST-20.
Also B35AM 9080 : neutral/slightly rosy at low/high current, virtually no angle tint shift, very high CRI, higher power 3535 sized LED.
For the 219B I would prefer a sw40k or sw35k since I prefer warmer CCTs, still like the sw45k though.
Oh man, this is a great thread question! I haven't been around long enough to know obscure awesome leds like many folks here, but I have a couple I really like, and I also have my first sw35 and sw30 219b's coming in on order from Simon, so I can't say either of those yet though I'm hoping the sw30 makes it to the top of the list, especially since I wasn't overly impressed with the sw45k from emisar (I just like warmer tints too much and it felt too cool was all really, the color was nice of course).
xpl-hi 5D (the rosy one from emisar 8 months ago when he did NOT have them in stock or as an option on his site but still had them in his store and sent me a KR4 with them and they are perfect - rosy, bright and awesome, thought a little more like 4200k).
xhp35 HI 4000k (like in my skilhunt from mbplitt and/or id30209 :), or my k1 from Noctigon- perfect)
osram CULPM1 (bright, throws very well, and nice tint for a throwing led)
sbt90.2 (just impressive in every way, including a decent tint)
sft40 (just, wow. And coming out in 5000k?! Awesome)
xpl-hi 2850k 8A (rosy, bright, nice tint and decent cri. It's a great all-arounder)
I have my first gt fc40 light coming in, so that is HIGHLY anticipated as well as the sw30 219b. I have a feeling one of or both of them may be on this list.
Thanks. That’s encouraging to know. Maybe at some point down the road, we’ll have enough of this info culminated into a chart of some kind. Would be great to understand the relative characteristics of emitters across the gamut, their features and limitations.
@xevious
There are a few scattered places to find a lot of this info, but it’s not perfect. Different equipment and methods and sample to sample variation keep it from being a precise science. When I get to my pc, I’ll link a few databases that are helpful. But there is often legwork to get the comparison results you are looking for.
In my book, tint (absolute and homogeneity) and CRI (for me R9 and Rg) are most important in anything but a dedicated thrower. Unless output or throw differences exceed 20%, then lumens are a distant second in the calculus. Combine smaller differences with ramping UIs with no reference points and linear or pwm based drivers, lm/w can be next to meaningless in practicality. Unless you are caving, climbing, or S&R with limited resources once you set out, then many of the details us nerds like to stress about really have no discernable effect on casual use.