LuxLuthor's Most Powerful Maglite Mods List

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,657
Location
MS
This is my personal ranking, which I update from time to time. These do not yet include the Osram IRC 50W & 65W and other bulbs. I moved my post out of this topic location which was getting a bit long.

Please note: Based upon my destructive bulb testing, these Lumen claims are using an outdated, innacurate AWR spreadsheet that does not reflect real, verified lumen numbers...so at best use these rankings as relative comparisons to each other. In other words, if I say the "MaxBlaster" puts out 4400 (bulb) Lumens, that is not turning out to be a "real" lumen number. It is a case of everyone agreeing upon the spreadsheet predictive values, akin to the "Emporor Has No Clothes" fable.


  • 1)Sylathnie's Sublimator Osram 64663 36V 400W with 10s Emoli 18650 cells in Elephant-II with FM 2.5" Turbo Head.





  • 4) LuxLuthor's HyperBlaster. This uses FiveMega's Chrome Elephant with extender body, and holds 3 stacks of the new Elite 1700mAh AA cells welded into a single pack. 10A side switch replaces stock Mag switch. Overdriving the Osram 250W 64655 bulb. 21 cells of Elite's new AA raw power....Oh yeah. It's bright.:naughty:

  • 5) Mac's 250w Elephant using 20 x 2/3A Elite NiMH Osram 24v 250w 64655 bulb at 10K L (?)

  • 6) Osram 150W 64633 at 16.8V (ideally with new AlanB/JimmyM regulator). I just added this based solely upon my destructive testing, but have not made a working light with this due to the 50 Hr rated bulb life, and high liklihood of flashing the bulb with most battery combinations. If you can set a regulator to 16.8V, it should fit into this ranking.




  • 10) Ginseng's Legendary and beautiful Aurora/Aurora 2. One of the most beautifully made lights I have seen, made by one of the "Grandfather's of CPF" (I put Newbie, js, bwaites and a few others in that rarified category). Reading his post reminds us of how much love and attention can be put into projects like this. He even worked on the finish by buffing and had wet-sanding before having it nickel plated. This was made way back in 2003 using the 64623 & GP2000 4/5A cells, and a 10A rocker switch. It screams "custom made" but looks better than 95% of any lights out today. Likely puts out the same as Mac's Torch.

  • 11) The USL mod (search for it), and Mad MaxaBeam's Maglite like above MaxBlaster that uses a 13.2V battery pack with the Osram 64625 or 62138 100W bulbs. It is a whiter, and cleaner bulb than the 64623, but not as bright, not as hot, and not able to be driven with the higher voltage.

  • 12) The 64430 35W 6V bulb overdriven to 9.8 or 9.9V gives about 1800 bulb lumens. Best to have a regulated drive with soft starting. Wonderful solution to a smaller light. Because it needs about 9.5 to 10A of current you should look at powering this with either the new A123 or Emoli safe Lithium chemistries, or a NiMH high output pack using cells such as the Elite 2/3A or 1700mAh AA's. I have a large supply of the authentic Osram 64430 bulbs that are no longer made.


  • 14) The legendary Philips 5761 30W 6V bulb overdriven to about 7.2V Many have used this bulb in C size Mags with 2 x AW C Li-Ion cells, but it requires about 5.5A, so it is close to tripping his protection circuit. Again ideal for a regulated driver with soft start. The Emoli 26700 or 18650 are alternate battery sources. Works out to about 1400+ lumens.





  • 19) ROP using a potted Pelican 3854-H which is a 24W bulb (larger one that has pointed tip), powered by 2 x AW C Li-Ions in a 2C. Just tested this with freshly charged Li-Ions that measured 8.2V in bulb socket. No resistance fixes done. I did attach a lead to the tailcap spring and making contact with side gave me 4.1A. Recent 3853-H was also tested in Destructive thread.



Updated 3/1/13
 
Last edited:

DM51

Flashaholic
Joined
Oct 31, 2006
Messages
13,338
Location
Borg cube #51
A bump for this very useful thread - good to have all the info here, away from the other thread.

You show in your list:
  • Host configuration
  • Bulb type
  • Power pack type
  • Vbatt
  • Lumens (bulb & torch)
It would also be useful to show the current draw of these individual bulbs, to save people having to look this info up elsewhere or work it out with the hotrater etc.
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,657
Location
MS
Yeah, I agree.

I am supposed to get the Mastech HY3020-D late this week, and I'm thinking of using AW's new D (unregulated) Driver as a platform (out of the Maglite--to try and give a standard bulb holder that people can compare to) which will tell me at a particular voltage exactly how many amps the bulb is drawing....so "weak" batteries or various resistance of mixed Maglite setups will not be an issue.

The problem with AWR's Hotrater is for the most part he uses the formulas that were taken from the WA website, with a little adjustment for various Ohm assumptions.

My goal is to test the bulbs at what is supposed to be the ideal Voltage (this Mastech will supply however many Amps the bulb wants, and with its soft starter I should be able to get some useful real world readings. This is also important for specifying Voltages to set up once AW comes out with his REGULATED D Driver, for a given bulb.

Once I see how the bulb looks at the AWR Hotrater "Ideal" voltage, I am thinking of slowly climbing up the voltage until bulb flashes...which will be useful information. I may have my Meterman light meter in play at a fixed distance to get some objective comparison between bulbs in some sort of bounce reading.
 

DM51

Flashaholic
Joined
Oct 31, 2006
Messages
13,338
Location
Borg cube #51
That will be EXTREMELY useful info to add here. Lol I'm hopeless with the hotrater - it usually throws up weird results when I try to enter anything new.

I have found Mdocod's formulas very useful too.

EDIT: The hotwire info in this old thread needs updating too, although it might better just to incorporate the results here.
 
Last edited:

ExZeRoEx

Enlightened
Joined
Dec 11, 2006
Messages
227
Just wondering, would I have better luck with some ROP bulbs or WA 1111s in a 2C mag with 2 18650s in them? Output and runtime wise, because I don't really know the voltages on AWs protected cells.
 

Pyzon

Newly Enlightened
Joined
Oct 16, 2006
Messages
35
Thanks for the fine compilation of totally useful information, Lux !

I'm patiently waiting for AW's regulated switch for C hosts to see what the 5761 run time will be on 3 of his cells. What voltage setting would you suggest as a happy medium of bright and fair bulb life ?

Thanks again for your hard work that we all benefit from.
 

petrev

Flashlight Enthusiast
Joined
Jun 12, 2005
Messages
1,535
Location
SW England
Thanks for the fine compilation of totally useful information, Lux !

I'm patiently waiting for AW's regulated switch for C hosts to see what the 5761 run time will be on 3 of his cells. What voltage setting would you suggest as a happy medium of bright and fair bulb life ?

Thanks again for your hard work that we all benefit from.

Note: The AW regulated driver will be D Mag size ! ! !

Cheers Pete
 

cat

Enlightened
Joined
Jun 24, 2007
Messages
558
Location
South Africa
Another good Maglite mods thread. An interesting list. What would be really good would be one or two of those links leading to a lists of components and where to get them, and if they're still available.
I think I've been searching and reading for...well, I'm suffering from lack of sleep, so I don't really know, but it's about 10-14 days, and I still haven't got one organized.
 

Patriot

Flashaholic
Joined
Feb 13, 2007
Messages
11,254
Location
Arizona
Ah...cool! Thanks Lux for making this a seperate thread. I could never find the other one. :thumbsup:
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,657
Location
MS
I am going to try and find what I consider "the sweet spot" for a bulb which I believe means a life of 10-20 hrs. I want to clearly establish the flash point of various bulbs, so that is no longer a mystery, nor have it be related to initial spike current, from some level of overcharged batteries, or from a variable series of resistance setups, NTC, etc.

I'm also planning on using the temp probe which came with my Fluke 179 to see actual temps inside of a FiveMega enclosed aluminum head with Borofloat lens. It tests up to 750° F, so I may need to drill a hole just large enough for the probe in a head to get accurate sampling. This would also allow me to see how much a layer of FiberFrax would insulate the KIU, wires, and hotdriver below the bulb.

I recognize this will likely involve sacrificing 3-4 bulbs of any particular bulb to get some accurate data points, and may subject AW's 4-D to more heat/current stress than he anticipates. To minimize that, I will start out with the lower 6V and WA bulbs before considering moving to the 50W and higher bulbs.

I'm thinking it would be good to do the run times in 30 - 120 min segments which would be a more realistic representation of how these bulbs would be used. Not to mention the fact that I am not going to want to sit, transfixed on a bulb and readings for 20 hours straight.

I want to try and standardize and control as many aspects as I can think of when setting this up, so it could be used in conjunction with AWR's Hotrater spreadsheet....perhaps leading to some adjustments, and ultimately becoming a reliable bulb reference, similar to the excellent work SilverFox has done with the battery run time curves.
 

petrev

Flashlight Enthusiast
Joined
Jun 12, 2005
Messages
1,535
Location
SW England
Big Thanks to Lux for sacrificing his bulbs for the cause. Hard data is most appreciated for the upcoming Regulators and SoftStarters

:twothumbs
 

bxstylez

Enlightened
Joined
Aug 19, 2006
Messages
693
Location
NYC, US
LuxLuthor:

I have a large supply of the authentic Osram 64430 bulbs that are no longer made.


I sent you a PM a few days ago.

Care to sell any of your 'overstock' Osram 64430 bulbs?
 

KevinL

Flashlight Enthusiast
Joined
Jun 10, 2004
Messages
5,866
Location
At World's End
Awesome, I love your summary.

I have been away for 18 months, so it's hard to catch up. Glad to see my ROP made the ranking at all. Thought it would have been too wimpy as you guys pushed the frontier back for all light-kind ;)

Not intending to compete for ranking position, but just to confirm - the six-SubC NiMH ROPs using Sanyo 3600 high-rate SCs is much brighter than the ROP on lithium. Eyeballing it shows a significant loss in output. Of course the ROP/LE is a much smaller 2C light compared to the 4D monster with the SC's. Instaflash risk is very real with the SCs, whereas the Li configuration seems to have virtually no instaflash risk due to significant voltage sag. My original ROPs have no resistance fixes or anything special. I've got a 4D/6xSC and 2C/2x18650 of my own.

4.1 amp is around there. I've taken measurements around the 4.15A mark too, close enough.

I'm considering building something in the 1.5K lumen range. What would you guys recommend? Planning to use A123 cells, soft start driver and preferably a very easy to obtain bulb. I hate my lights being disabled because of popped bulbs.
 
Last edited:

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,657
Location
MS
BxStylez, I just did reply to your PM. I was kind of hanging back because I wanted to test these for optimum voltage, and see what the Osram is actually doing.

Kevin, I'm sure higher voltage/amp NiMH cells would push the ROP higher....so it was definitely a great discovery for its simplicity.

I have yet to add the Osram IRC (Infrared Reflective Coating) bulbs, so I'll revise this list and what I think the best regulated voltage (& amps needed) for a number of lights.
 

petrev

Flashlight Enthusiast
Joined
Jun 12, 2005
Messages
1,535
Location
SW England
I was trying to figure out some version of QuickBeam's light box, but the wax coated milk carton won't cut it for these torch lights. (I mean "torch" in the American lexicon, since much of the world calls flashlights "torches.")

Hi Lux

Some thoughts on Light Metering ? and no cheap/expensive integrating spheres ?

Firstly some of the Light Meters generally available

CEM DT-1309 with Backlight + USB graphing to boot.

CEM DT-1308 and here Backlight but no USB but cheaper and more like it for me.

Meterman LM631 / Tecpel 531 Lightmeter With Backlight (Google it for prices)​

My idea is that these and even the cheaper Chinese ones with no backlight LX1010B or DT1301 greater range could all be used to give fairly consistent readings for comparison purposes.

First Incan Bulb-Lumens - Take off the Torch head to reveal the naked bulb in Candle Mode - measure Lux at say 1metre - right angles to filament - horizontal separation (same level) same for axial filament lamps just avoid the return wire if external, although it probably won't change anything much.

The filament acts a point source so the Lux meter reading is just a small part of the total 1m sphere surface

Lux level should be proportionate (?) to Bulb Lumens so take a known trusted value like the ones for the WA1185 from the WA re-rater page at a set Voltage or two and check the ratios - that is the first interesting experiment. Should be a simple ratio between Bulb Lumens and Lux at 1m ( I think :thinking: )
Might have to check how much the room, if not a Black-Box room contributes to the Lux value or experiment outdoors at night - but not as easy to use a constant Voltage supply outdoors !

Another value of Lux could be taken for all bulbs at 1m (or suitable distance) for Centre Spot value from say 3" FM MOP or other semi standardisable reflector of choice ( or lots of mirrors and heads ! ! ! ) using standard/optimum Voltages.

This method should be fairly standardised and repeatable by other members ? ? ?

I'm going to post these thoughts in the LuxLuthor Most Powerful Thread to see what everyone thinks and they can point out the glaring errors

:whistle: Let me know . . .

Cheers Pete
 

LuxLuthor

Flashaholic
Joined
Nov 5, 2005
Messages
10,657
Location
MS
Pete, thanks for your information. I bought the Meterman LM631 a while back, so I'll use that one. I can position the Meterman at a distance of 1 meter, and enclose the sensor head inside a 15 inch black tube sealed around it to eliminate reflected light. With the tube horizontal to the table, and so the attached sensor is the same height as the bulb, I can line it up so it is aiming at the side of the bulb. I'll check first with brighter bulb like 64458 to see if that maxes out the meter, and move it back farther if it does.

Figuring out the head size and zoom will be a more variable issue....but it would be interesting to see how a particular voltage/amps/lux @ 1m translates to a farther distance spot from various Mag head beams.
 
Last edited:

Latest posts

Top