Which is better, a Mag85 or a P7? That's kind of like an apples an oranges question. For the P7 I don't think you can do any better than the Malkoff dropin. So you've already got that covered. Here's a
Shootout where StefanFS made a direct compared between a 50 watt halogen and a
P7 MagLite. The classic Mag85 gets its name from the 30 watt Welch-Allyn 1185 medical bulb. Take a look at this
re-rating chart for the 1185. It only goes up to 10.56 volts. If you click on the little R for that voltage, you'll see the bulb only has a life span of 16 hours at 1140 lumens.
Nothing wrong with WA's site and its built-in rerater (after all, they should definitely know how to rerate their bulbs!), but AWR built a hotrater spreadsheet by deriving the formulas from their rerater (so it's equally valid) that can be used for WA
and other manufacturer's bulbs, and LuxLuthor has done actual testing of all the popular bulbs and some less-popular but promising ones. Best of all, he's presented his results-to-date along with useful data from AWR's hotrater spreadsheet, making
this post a one-stop information center on most bulbs of interest;
here is the data for the WA 1185.
So 3*18650's with a spacer tube and a reworked tailcap spring would be a little brighter than that but with less than 10 hours of bulb life.
True, but I'd go with 26700 molicells instead, for longer runtime.
By way of comparison the
50 watt halogen bulb StefanFS used puts out 1250 lumens and lasts 4,000 hours. So it's 50% brighter than a P7 and only needs 4 amps at 12 volts.
A point worth noticing -- that one's an IRC bulb, which can get much higher efficiency than normal bulbs. The coating reflects infrared back in, so the filament can keep the same temperature (and thus light output) at lower power levels. (At the nominal 12V rating, it's making the same efficiency as the 1185 (at nominal 9.6V), but that's only because the 1185 has traded off bulb life to get efficiency and whiteness. The 64440 has tons of life at nominal, so it's got
lots of room to overdrive, and the efficiency naturally increases. Check LuxLuthor's tests -- the 64440 can go up to
22V without popping, and gets comfortable bulb life around 30 hours at a more reasonable 18V.)
Until the recent introduction of high drain Li-Ions (
1 &
2), driving 100 watt bulbs required a ton of batteries. Lighthound has been selling the rest of AW's IMR line and will probably have the
26500's in shortly. Three of these C cells is about 30 mm shorter than 3 D's. A homemade wooden dowel with a bolt going through it will both make up the distance and help isolate the batteries from the heat. Then you could go with a
64623, 100 watt halogen. For less than $4 the Osram has a 2,000 hour life span and puts out 2,800 lumens. That's more than 3 times brighter than a P7. The downside is with AW's 2,300 mAh C cells, you'll only get a 15 minute run time at full power.
Sure, but you'll be underdriving it most of its life -- I'd use a fourth 26500 (it'll fit, with a tailcap-spring mod) for a brighter & whiter light, or at least use 2.9Ah 26700 cells instead to boost runtime by 25%.
But really, you can get more output than that 100W 64623 (2800 lum, 106W at 12V, 8.8A) from a lightly overdriven 65W IRC 64447 (60W at 12V, 5.0A, but run it off 4x26500 for 2965 lum, 92W at 15V, 5.6A), and boost runtime by 60% over the 64623 and 3x26500, or 25% over 64623 3x26700. And you'll still get about 275 hours bulb life -- that may not sound like much, but you'd have to change batteries about 700 times first. The IRCs really pay off in runtime, brightness, or both. (And if you
want something more like the 1185, with less output and a good runtime, Osram also makes 25W and 35W IRCs, and the previously discussed 50W IRC as well.) But they should all have at least 4 cells for good results, as should most 12V bulbs.
