So I have this Maglite 3D sitting around...

[Player]

I've been a long time lover of Maglites. While I was scoping the threads here I discovered that people spend a lot of time modifying them, yes I've been sitting in the dark far to long. So after reading many of the posts on this forum I've determined that I want to do something with my Maglite 3D. Several are a bit labor intensive and beyond my means but many seem easy enough to do. What worries me is some of the posts that I've read go back a few years, are they still current in terms of moding? Right now I'm thinking of going the easy way and ordering a Malkoff drop-in, glass lense because I was surprised at how scratched my stock plastic one is, and a fancy new bezel because we all know that a fancy new bezel is good for five lumens. So what I'm asking is, is this a good thing to do or should I be looking at something better? One post I've seen mentioned 1000+ lumens to modify a Maglite, I dont have the means to machine or otherwise I might consider something like this. The real question is what am I wanting. Well my great collection consists of various Maglites sitting around in places, a Fenix TK12R5 and a Fenix PD22, so naturally I want something that can light up the world. Any help would be appriciated.

[yazovyet]

which drop in?

do you plan to use NiMH or alkaline cells?
generally alkalines suck at large (over 1 amp) current.
http://www.candlepowerforums.com/vb/...tery-Shoot-Out

there are options to use AA cells (incase you could use NiMH AAs but not NiMH D cells) such as holders that do 2 AAs in parallel or other more solder requireing options. I have assumed you didnt want to get inot lithium cells.

generally i like an XM-L led on a h22a heatsink with a 7135 based driver with 3 NiMH cells. theres only a little bit of soldering and you can get in the 900 lumens (from the emiter) area without trouble (wont stay in regulation [fully bright] the WHOLE run-time, but it should be pretty good).

[Player]

You've sparked my interest in overindulgence again. If I went with the XM-L and H22A heat sink, would I still use the same driver if I wanted LiPo's? My knowledge of batteries is zero, I just follow the directions and put in what the item tells me to. I have played around with them a bit with my son's RC truck so I'm not a complete idiot contrary to what my wife says. Would I have to modify the switch doing this mod?

[yazovyet]

you'd have to use a different driver if you went with lithium. lithiums give 3.7 volts each (up to 4.2 when they are just charged for the first few min, as far as i understand) so 3 of them would be (3.7*3) 11.1 volts. a 7135 driver only works up to 6 volts and you really want to keep it in the high 3.x to 4 volt area (a little above the voltage teh LEd needs). you also need a lithium charger and should read about the dangers of lithiums cells ( http://www.candlepowerforums.com/vb/...e-of-batteries is the link to the cpf forum about battery explosions)

the mag switch seems to be able to handle the 2.8-3 amps to drive the XM-L hard, a mod might make it run a little better but doesn't seem to be required, i have seen a numnber of them done without switch mod. if you went with the lithiums it woudl be a higher voltage and thus a lower amperage so would definatly not need a switch mod.

incase you were unaware, NiMH is a normal rechargable cell. they give 1.2 ish volts where as an alkaline gives '1.5' volts (it starts at 1.5 and tapers off). NiCd is another type but no one seems to use those (they give in the 1.5 volts area too, i think those are the old style of rechargeables)

edit:
i don't know if that link is all horror stories or if there is precaustions and such too. but there is special handeling protocall for litiums above and beyond NiMH and alkaline

[Player]

So if I go with the NiMh what type and how many would I use? From running RC cars I know about the dangers of LiPo's but wow can they make an RC car go fast! Depending on the motor and gearing combination his truck went from 20+ mph to over 60 mph. Let's get back to the flashlight, so using what you've said I would need to make a carrier(s) for the batteries and install the parts listed, I think I can do that. Anything else I should know before I fall off the edge?

[yazovyet]

If you use 3D cells (NiMH) in a 3D light then you don't have to worry about carriers (but it may be hard to find NiMH D cells at a reasonable price?).

If you use AAs then you'd have to do some sort of carriers. If the carriers put the cells in parallel then you just have to get 3 carriers in there (with the 1,2, or 3 batteries each) and you're good. If the only carriers you can find are serise then you'd have to put non-conductive spaces in between them and wire all the +ve together and all the -ve together. you tend to be able to buy different kinds of carriers on lots of websites (google or CPF-google search for aa-d convers/carriers or for parts websites is/are your friend). In a 3D light you should have space to put 3 carriers.

A D cell may have 10 Ah or more while an AA has ~2 Ah. Assuming you can find 3 AAs for each D you still end up with less capacity when useing the AAs. For that reason and the simplicity I would say D cells are better. But it does depend on cost, if you have to buy a new charger for Ds but not AAs and how much it costs you to find the D cells versus AAs. I'd look in the batteries section for info on batteries (cost, websites to check), I don't know much about where to source them or what they should cost since I buy AAs locally and on sale and have no Ds.

I'd say your next step is to price things out, start with the D cell batteries and see if those fit your price range, then move on to the other items (maybe find a build thread/guide and use that as an example). before you start you could post a list of parts you intend to get to see if people see some glaring fault (or just get some different oppinions since everyone would do things a little different).

[yazovyet]

one quick thing to add. i mentioned that you didnt have to mod the switch. by that i ment that you don't have to do an electrical mod. you DO have to cut part of if off with a hacksaw (not hard though) and solder some wires to it, check out other people's build threads to get an idea of this. also what modle of mag light? is the first lette/number of the serial number a "D" ? real old ones are a different size and have no letter whie LED versions have an L and thus a different reflector in them.

[Player]

Let's say I go the easy way with D size NiMh batteries. Battery Junction carries a Titanium brand with 1200 mAh. I'm no good with batteries and all the numbers, I've fried my fair share of things. I think I've seen the mod where you need to cut part of the switch away so I should be able to find it again. There was another thread somewhere that had a mod with an LED that changed colors depending on the remaining charge, the LED was just above the switch button.

[Quest4fire]

Ebay seller marrsinc is selling blue Tenergy D cells four (4) cells for $16.99 (Yes they are back in stock, but who knows for how long), Item# 350560785640. That's $4.25 each with free shipping, you really can't beat the mAh for the price. Others have reported capacities in the 9000 mAh range after doing break in cycles. Check out this thred for more info. I bought some without really having a use for them just because the price was so good! Happy modding player.

[yazovyet]

I would expect the titanium brand to work well based on having seen a graph indicating their AA cells to be decent. I have seen reviews that the tenergy brand is good aswell. Another big thing to consider if how long you leave the light sitting between charge and use. NiMH tend to discharge over time and you'll find them with less power after a month or more. There is a newer kind of NiMH called low self discharge (LSD) that will do this at a much slower rate, so that you basically don't have to worry about it on a time frame of a year or more. So you have to decide if you want the higher mAh regulars or the slightly lower LSDs. The LSDs seem to be sold is 2 packs also so they will cost close to $32 for 4 where as you can get the regulars in singles and thus be around $22 for 3. Check the price on a charger if you don't already have one (but maybe your RC charges can do NiMH cells?)

Also those are 12 000 mAh (or 12 Ah), the LSD are 10Ah. Battereis in general tend to be a little over rated in general in their capacity, but I would still expect 3 or so hours on high, 10+ on medium, and 50+ on low (assuming you used the generic 2.8A driver I am thinking about). also drawing a high current from a battery will generally reduce the capacity it can give you (not in a damaging way but in a uses the power up faster way; although super high current can damage a battery but I bet you know that).

That colour changing LED mod would need: some electronics that sensed the voltage and ran the LED (or a more complicated driver that would also be more expensive), possibly a new hole to stick the LED through, a little extra wiring to run the LED. Chcecking the voltage of an NiMH isnt totally accurate, the voltage can move around depending on how much current is being used.

[yazovyet]

I occurs to me, you are familiar with lithiums so you could use those with a different driver. The main draw back is I don't know too much about those so it is hard for me to help you if you encounter problems.

http://www.videofoundry.co.nz/ianman...driverlist.php

Is a list of drivers if you wanted to look for something that worked on 3 lithiums.

[alpg88]

generally i like an XM-L led on a h22a heatsink with a 7135 based driver with 3 NiMH cells. theres only a little bit of soldering and you can get in the 900 lumens (from the emiter) area without trouble (wont stay in regulation [fully bright] the WHOLE run-time, but it should be pretty good).

which driver did you use??
cuz i tried at least a dozen of 8x amc drivers, none would pull 2,8A, from 1 li ion, let alone 3xnimh. the most i got was 2a using imr 26500 cell. no 18650 ever made past 1,7A. tio get 2,8a you need at least 4x nimh cells.

[Player]

which driver did you use??
cuz i tried at least a dozen of 8x amc drivers, none would pull 2,8A, from 1 li ion, let alone 3xnimh. the most i got was 2a using imr 26500 cell. no 18650 ever made past 1,7A. tio get 2,8a you need at least 4x nimh cells.

If you're asking me I haven't modded my light yet, still trying to find a good alternative that's not out of my limited expertise.

[Player]

Ebay seller marrsinc is selling blue Tenergy D cells four (4) cells for $16.99 (Yes they are back in stock, but who knows for how long), Item# 350560785640. That's $4.25 each with free shipping, you really can't beat the mAh for the price. Others have reported capacities in the 9000 mAh range after doing break in cycles. Check out this thred for more info. I bought some without really having a use for them just because the price was so good! Happy modding player.

Thanks for letting me know about the deal on batteries. The more I save the more I can put into getting another light!

[Player]

I occurs to me, you are familiar with lithiums so you could use those with a different driver. The main draw back is I don't know too much about those so it is hard for me to help you if you encounter problems.

http://www.videofoundry.co.nz/ianman...driverlist.php

Is a list of drivers if you wanted to look for something that worked on 3 lithiums.

I'll give a look around the link you posted. I'm worried that I might be getting in over my head now. I saw somewhere that people were putting the drivers in a series, why would they do this?

[Player]

This is a link to a mod that really appeals to me but I'm not really sure if I'm up for the challenge. I must admit the beam and flood sure looks a lot better then stock!

http://laserpointerforums.com/f64/fs...ght-59581.html

[yazovyet]

Talking only about high mode running an XM-L:
I used the (2.8 A) one from SB with 3s2p AAs (eneloop). I did not measure the current (my amp meter capped at 2 amps) but I did find (useing the brightness meter on my camera) that the brightness did not change between 3 or 4 cells in serise. I also found that the brightness stayed constant for 40+ min then started to drop off. From this I concluded that the driver was running at it's full output for a while before it fell out of regulation. I also assumed that when the driver was running at full power that it would be putting out 2.8 amps.

I also tested it with 3s (duraloops) and found it started dropping off in brightness soon after starting, thus the extra current draw from the batteries brought it below the needed voltage to stay in regulation.

I think that 3 D cells should be able to supply current better than 3s2p AAs and thus it would hold a voltage decently relative to the 3s2p AAs I used. This conclution comes from looking at the NiMH battery shoot out and the voltages recorded int he eneloop AA at 2 and 3 amps vs the accupower D at 5 amp.

Two important things to note are that (a) I used an XM-L and that this would not work with an XP-G (I think a p7 is in the same boat) since they need a higher forward voltage and (b) that useing an amp meter tends to introduce a resistance that can push the voltage to the driver down past regulation (the often long leads can becomes important on high currents).

I have never used litiums so I don't know how it would work with them on that driver.

image of the graphs,
brightness of the hot spot (in units only useful as a relative comparison) vs time (in minutes) for 3s2p and 3s1p config.
http://imgur.com/f3wEQ

[alpg88]

no it wont, get real dmm and mesure current, you will not get 2,8A from 3nimh, 2p or 5p, from sb drivers, i've tried it more than once. and not just me.

i wish it wasn't the case, it would make things a lot easier, but it just doesn't happen.

[awenta]

Yaz your forgetting to take into account that the drivers are only around 80% efficient.

If you want to properly drive the xml then get a TaskLed Flex driver. (6 or 3)

Id be comfortable running it on 3 callies 3100 18650s. The better option would be 26650 or 32600 but those are few and far between in comparison. And with 18650 you can get a survival tube.

[yazovyet]

alpg:
hum, I would have expected to see (or for my camera to pick up) a difference in brightness between 3 and 4 cells if 3 cells can't put out enough voltage but 4 could. were you able to get the full 2.8 amps off 4 NiMH? I wonder if I got lucky and my XM-L has a lower forward voltage and thus works as I think it does of if there is something I am not seeing here. I'm quite suprised that you tested 3s5p AAs (or was it a larger type of cell?) and found it to be less current that 4 cells, do you remember what the current reading was in the 5p config?

player:
That mod is along the lines of what I had in mind but to use a simple driver and battery set up. My thoughts on the voltage moniter is that is has a higher cost than value and takes more effort for little gain. I think (but am not sure) that if you were to do the mod and found you wanted to add that in after that it would not be too much extra effort to do (provided you could get the heatsink out easily and didnt glue/epoxy anything ot the sidewalls). I made a similar mod with a 2D light:
http://www.candlepowerforums.com/vb/...ed-to-6AA-XM-L
I had envisioned you doing similar but with 3D cells instead of the AAs I used.

[awenta]

Alp, did you try the newer 3.04a (IIRC) from sb?

Also yaz get thicker or makeshift leads for more accurate high draw applications.

[Player]

player:
That mod is along the lines of what I had in mind but to use a simple driver and battery set up. My thoughts on the voltage moniter is that is has a higher cost than value and takes more effort for little gain. I think (but am not sure) that if you were to do the mod and found you wanted to add that in after that it would not be too much extra effort to do (provided you could get the heatsink out easily and didnt glue/epoxy anything ot the sidewalls). I made a similar mod with a 2D light:
http://www.candlepowerforums.com/vb/...ed-to-6AA-XM-L
I had envisioned you doing similar but with 3D cells instead of the AAs I used.

Thanks for the link to your post, the beam shots look great! So for the time being do you think I can just get the LED, heat sink, and driver and go with standard D cells for the time being? If so I'll try that and if the mod bug really bites me maybe I can push this light a little more.

[alpg88]

it made little difference between 2p and 4p AA only 150ma. the most i got was from off fresh imr 1,9A. forward V doens,t make any differnce, since i tried with different bins of xml. the drivers are cc it will adjust voltage, but all xml i tried were between 3 and 3,3V even at 2,8a i didn't see more than 3,3V.
those drivers aren't exreamly efficient, that is true, but it wont affect output current, or voltage, it will just cosume more from the cells.
the difference between 1,6 1,7A and 2,8a is about 250-300lm, your eye would not see difference, human eye preseption of lumens isn't leniar.

even with 4 AA i got only 2,2A but i credit voltage sag of energizers 2300 for that, with 4D nimh cells i got 2,8A.

like i said, i would love for that driver to supply 2,8a from 3nimh or 1 li ion, but it just doesn't work like that in real world.

not to say that 1,5-1,7a isn't bright, like i said you most likely wont miss those few hundred lummens anyway, also when supplied with enogh voltage to make 2,8a those drivers heat up a lot and if not heatsinked will start blinking soon no matter the mode (internal heat protection kicks in).

few 1D maglites i build with those drivers work at 1,5-1,7a but they don't even heat up much, they can run on high untill the cell drains, but driver stays cool all the time.

[yazovyet]

awenta:
It isnt a lead problem for me, the meter simply will not display a current above 2 amps. I was wondering if alpg had done this in his 3s5p test on an XM-L.


i think driver efficiency with the 7135 is variable based on votlage input and output:
http://www.candlepowerforums.com/vb/...ide-**UPDATE**
Efficiency appears to be based on the input and output votlage. Note that in those tests he appears to be useing an LED with a higher Vf (like an XP-G). It looks like the board drops around 0.1 to 0.2 volts, after you are applying more than that much over what the LED needs the extra voltage is stopped and "burned off" by the regulators. So from that page, with that emitter you'd be getting 80% eff with 4 cells and 63% eff with 5 cells.

If you looked at the efficiency only in terms of current then those drivers are very high. I understand this sounds like a strange thing to do but I find that when your batteries are rated in mAh and the efficiency is variable based on the input voltage and LED Vf it is nice to look at it as if it is 100% efficient in current and any extra voltage is wasted. It makes it nice for calculations of estimated run time by knowing when the batteries will drop out of regulation due to voltage sag at a given current but can NOT be used to refference it to other drivers that use a real efficiency calcultion.

So since current-in is effectivly equal to current-out for these drivers I don't think I need to worry about efficiency durring those calculations. That being said I was mostly basing my thoughts that 3 D NiMH would work on the idea that I was able to measure the same brightness form a light with a similar set up from 3 or 4 cells in serise for the initial portion of a brightness vs runtime test.

[alpg88]

i didn't try 5p i tried 4p max, but result with 5p is very predictable.
i never get into detales how it works, what voltage drop or how many watts it dessapates, efficiancy%,.....etc, i just tell what results i had by actually testing\using them in real world.

[yazovyet]

player:
please confirm: your light has a "D" as the first digit in the serial nuimber. (that shoudl be stamped near the switch).
You also need thermal adhesive (epoxy) and soldering equipment (and a few spare wires). And if you had thermal grease as well thats a plus, you use it between the heatsink and the light where you wouldnt want to use epoxy.

That should allow for a voltage sensor if you wanted one at a later date (it might be hard to pull the heatsink back out without scratching it/putting plier marks in it). If you want to change the batteries to lithums you'd need a new driver.

I think that it will run at 2.8 amps for a decent amount of the runtime. alpg says it will not. I think him and I are both in agreence that it will be a nice bright light* though. I am certainly impressed with mine.

edit:
* by normal person standards. Flashaholics might not consider a single XM-L (even over driven) to be that bright.

[LilKevin715]

Player:

To help make your decision easier on which route you want to go I have one simple question for you: are you comfortable/experienced enough with a soldering iron? If you know how to solder then the rest is fairly easy. If you don't know how to solder then you can pretty much forget the DIY route and go with a dropin.

If you wanted to go the DIY route with a XM-L take a look at my build thread in my sig (don't forget yazovyet's either). It shows for the most part step-by-step instructions. If you are comfortable enough with the overall process (and assuming you know how to solder) than it can be done.

Otherwise a dropin such as a Malkoff XP-G or a Terralux TLE-310M-EX (with 4 C cells) dropin would be easier.

[alpg88]

or get der witchel 3xm-l drop in, (2500lm) it will reqire some soldering, and switch modification, but it isn't hard, step by step instructions are on this site, more than one thread, you can run it from 4x26500 cells in 3d tube, or 3x26650, pbly the brightest mag drop in available.

[Player]

yazovyet
Yes it has a D at the beginning of the serial number.

LilKevin715
Yes I can solder, maybe not to any great standard but I've done my fair share over the years. Thanks for the help and pointing me to your link.


Why does it seem like most people are modding the 2D and not the 3D? Is it because the 2D is a cheaper light to buy?

[LilKevin715]

Why does it seem like most people are modding the 2D and not the 3D? Is it because the 2D is a cheaper light to by?

It's simply due to size most of the time and nothing more than that. I used a 2D with 3 C cells, while the same results with longer runtime can be achieved with a 3D and 3D cells.