What sort of lumens do you expect from the sst-50 drop in?.
Ive read over 500+ but doubt that.
It seems no one has tested a top-binned, properly installed drop in with clear lens, proper heatsinking, a good cell etc. etc., so all we have is theoretical for now. Here's what I conjured up using infallible mathematics. Skeptical though I am, Cree rates their numbers at +-7%, and Seoul +-10%. So variables like heat aside, my results are likely at least within ballpark range.
Flux binning specs:
Assuming I'm getting a W57S GJ bin, I can expect a minimum of 500 emitter lummens at 1.75A
Source: PhatLight SST-50 Rev. 2 Data Sheet
Based on this chart, 2.5A would give us roughly 135% rated output. Therefore, emitter lummens for a SST-50-W57S-F21-WJ-G4 bulb would be a minimum of 675. Assuming a 25% loss of efficency due to optics, we end up with a minimum of 506 OTF lummens - a spec that should be reasonably accurate, as all but the reflector waste come from factory specs, and reflector waste seems to be generally accepted at 25% for high-quality, 30%-35% for low quality. If we assume that our LED falls at the higher end of the binning spectrum (as Nailbender stated he tried to provide upon request), ideally we would see (600 * 1.35) * .75 = 608 lummens. In other words, the amperage boost over spec should result in rated emitter lummens closely approximating OTF lummens after reflector loss.
In short: A SST-50 should give anywhere from 221 (worst case WE bin) to 608 (best case GJ bin) OTF lummens when driven at 2.5A.
Now, if we tried to feed it 5A, ignoring the obvious overheating problems, we would get a 235% increase, giving us a best-case (600 * 2.35) * .75 = 1058 OTF lummens.
Again, all of this is forgetting about heat and the loss of lummens that result as well as overstatments in the data sheets. But all of it makes me somewhat optimistic that we'd see at least 500 instant OTF from a good WJ bin.
That being said, we could easily get <200 from another LED of the same color bin but inferior flux. So I think binning matters much more then the type of LED - we're talking about a 2.8x variation in brightness from the exact same LED.