5D Mag - SST-90 - 95mm Aspheric - 9+amp build - Finished on 03/17/15 and beam shots are in Post #1

Thanks OL. ChibiM, we must be very careful in analyzing OL’s results. The battery one and battery two results must be treated as separate sets.

  • In the “Battery One” testing the Wavien collar was 0.48% better than the bare LED. The Mag collar was worse than the bare LED.
  • In the “Battery two” testing the Wavien collar was only 0.09% better than the bare LED (best focus case). The best focus case for the Mag collar was 0.29% better in this case.

What’s also worthy of remark is that between the two battery tests we see that battery one gets the bare LED lux 1% higher than battery two - but the gains from the Wavien collar do not stay steady! They are much better with battery one than battery two for some reason.

All told, here’s what I see:

  • Yes, they both kind of suck. This hinges on what you are looking for though. Once all other avenues are exhausted, this becomes the only path towards increasing lux.
  • This testing has been done with a big-die emitter, maybe we can get closer to 1-5% improvements from smaller die emitters? Focus is certainly key and in the pictures comfychair showed I recall that the spot generated from his Mag collar was quite small. I think it was not enough to cover an SST-90 die. Take a look at Post #6 over here, comfychair - 64mm aspheric - which LED?
  • OL’s testing was quite granular. It shows what range we could expect benefits to be in with this setup. I’d guess probably no more than 1% best case for this setup.
  • OL’s testing also shows that focus is key. OL easily reduced the lux with the Mag collar and one must assume that if a person messes up badly enough they can also reduce the lux using a Waiven collar. (perhaps by using the wrong size?)
  • It’s possible that using alligator clips and magnets may have contributed to measurement error. This depends partly on Vf headroom and what current the light was regulated at. In DD I think it’s certain that the clips/magnets would mess with the results a little.

Anyway, thanks again OL. I’m definitely intrigued.