Any thoughts on fitting a Cree MC-E on an AW TH tower?

[Justin Case]

Good point on thermal path for the KT4 TH. I planned on using a KT1/2, which should have better thermal performance -- analogous to using a standard SF 6P bezel for an LED drop-in vs. a Z32 shock bezel in a weaponlight.

 

[Justin Case]

A few more indoor beam shots.

Control photo of a dark basement:

MC-E ceiling bounce. Distance to the windows is about 21 feet. EV -1 setting.

SSC P4 ceiling bounce. EV -1 again.

DX11836. EV -1.

MC-E down a hallway, 16 feet to the wall, EV set to -2 (note the wide beam pattern for the MC-E tower fills the full frame):

Same as previous, except SSC P4 tower (note the brighter hot spot for the P4 vs. the MC-E and the DX11836, as well as the wide beam pattern filling the full frame as before):

DX11836, EV-2 (you can see the polygonal outline of the SF anti-roll bezel's retaining ring where each apex is one of the notches in the ring):

 

[maxspeeds]

Thanks for the beamshots, Justin! This is an awesome thread! :twothumbs I read that you built two towers, one with a SOB1000 and one with a SOB1227. I wanted to know if the difference in brightness is worth the reduction in runtime between the two?

I know with XR-Es, the difference in brightness is not linearly proportional to the input current. In fact, I'd rather drive an led at 400mA over 750mA due to the marginal return on brightness versus shortness in battery life. However, when this is multiplied by 4, the extra brightness may be worth the reduction in battery life, especially when used in a large reflectored light (where distance lighting is it's main goal).

What are your thoughts?

 

[Justin Case]

I didn't measure any appreciable difference in hot spot or spill lux using my light meter for the SOB1000 vs SOB 1227. I did measure about a 30% increase in tailcap draw, in proportion to the increased drive current (plus some fuzzy, assumed value for the driver board efficiency) for the SOB1227 vs SOB1000.

My lux measurements are a bit rough, and I've found measurement variations for the same light tested say weeks apart. For measurements made on the same day, things look reasonably repeatable. I guess my hand is steady in the short term. I run the light for 2 min, then hold the light in my hand at 1m from the light meter, manually aim the beam at the light meter, try to keep the light square to the meter, and read off the lux value. I think that most of the measurement variation comes from the beam not being absolutely square to the meter.

But within that variation, the SOB1227 tower didn't measure any brighter than the SOb1000 tower. Only one data point though.

 

[Justin Case]

I must be hooked on epoxy.

The AW tower kit includes a small ring of clear sticky tape to be used as an insulating layer on the exposed, non-component side of the driver board for the assembled tower. I've never used this tape because it just doesn't stick well to the driver board.

So, I've decided to encapsulate the non-component face of the driver board with epoxy to insulate it from potential shorts, mainly from the center spring.