Carley 1499, T1.5 size bulb, smaller diameter glass envelope than the WA's. 7.2v/1.84a 385lumens. M*g2D with 8AA-2D adapter running a M*g6-cell xenon bulb will output even less lumens, but the hotspot will be slightly smaller and very white (note, I think I have need for better set of batteries, and/or way to reduce resistance for the test M*glite, and/or the 8AA-2D adapter, so YMMV, you may instaflash or get too short a lifespan for your liking). I love the white light I'm getting from my particular setup with the M*g6-cell xenon bulb.
Note also, something bwaites and Ginseng have not mentioned yet (to my knowledge) about the M*g bi-pin adapter using M*gChr socket. That socket is of a specific width, you will likely need to use needle-nose pliers to bend bulb leads of differing widths to fit into the M*gChr bi-pin socket, and you may end up with broken leads and wasted bulb...I have done it myself a number of times, the leads are somewhat brittle and difficult to bend. YMMV. Maybe the M*gChr (Welch-Allyn makes it?) bulb, rated at 175 lumens or so, overdriven with the 8AA-2D adapter will get you a tight spot?
MOP or LOP reflectors only smooth out the pattern and don't change the shape much, but it's the filament width that makes for a tighter hotspot (proper centering in the reflector also effects that uniformity of the hotspot).
Not sure about this, but there maybe a axial filament Carley 852 (500lumens, 6v/3.33a) in the works in the next month or so?---from fivemega, but don't quote me on that.
A more detailed specification of which bulbs you have tried, and what flashlight configuration, including type of batteries, would be helpful in recommending what you need. Ginseng has information on where to get an Osram projector bulb that also has an axial filament, which gets you that nice round beam you seek. 6v/3.33a 470lm?
