Help needed: Adding resistance to a mag tailcap

About 4 months ago, I took my parents 15 year old Maglite (pre-D serial) 4 Cell D and updated it. I had someone make me a custom heatsync and a custom driver. I installed a chopped rebel reflector and put an XM-L T6 in it.

The problem with the light is, with 4 freshly charged cells NiMh cells, the driver goes in to thermal step-down mode. It will run at 3.5A for about 10 seconds (what it's programmed to do) but then ramps down to 2A as it overheats for the remainder of the time the light is on (since it's too much excess voltage to burn off for the driver). The driver is only designed to accept 4.5V, anything over that, it heats up quickly no matter how it's potted (even under load, my 4 NiMh D batteries don't run at 1.2V for a while).

If I power it with 3 freshly charged cells, it will run at about 3.1A for the first 30 minutes or so, then it ramps down accordingly over the next two hours because the batteries can't keep up with the Vf needed to run the light at 3.5A.

I'm looking to have the light's output be regulated at the correct 3.5A for the entire time it's on (or as close to 3.5A as possible). The only way I can think of doing this with 4 cells is to add a resistor to the circuit. The problem I'm having isn't with finding the right resistor, it's figuring out how to get the resistor installed in the tailcap correctly...

I'd like to bury the resistor in the tailcap of the light since I have no room at the top (near the switch) for it. I want to put the resistor underneath the spring, attach one side of the resistor to the middle of the spring and the other side to the base of the tailcap, then prevent connection between the base of the spring and the tailcap.

I cannot, for the life of me, get solder to stick to the inside base of the tailcap, no matter what torch I heat it up with. I've tried my soldering iron, a plumbers torch, flux, drilling a small hole in the tailcap, etc. Does anyone have any suggestion as to how the heck I can get one end of the resistor to stick to the inside bottom section of the tailcap? Also, a suggestion on the best way to shield the base of the spring from the base of the tailcap? I've tried heat shrink but it just gets all torn up when I put the spring back in the tailcap.

I think a resistor is an inefficient way to loose voltage. You might be better with a 1 D sized spacer and run the light on 3 D cells.

Also, you can't solder to Al, steel, or some other metals (well, it's hard to and/or need a different type of solder). You can solder to bronze and copper with the usual tin-lead solders.

Since Maglites are made of Al, you won't be able to solder to it.

I have a 1D sized spacer. When I use 3 cells, the light wont reach 3.5A with only 3 cells and it doesn't run regulated. I'm saying I need to add the 4th cell for the light to regulate correctly and have something burn off the excess voltage (besides the driver) to keep it in regulation.

Gunner12 said:

Also, you can't solder to Al, steel, or some other metals (well, it's hard to and/or need a different type of solder). You can solder to bronze and copper with the usual tin-lead solders.

Since Maglites are made of Al, you won't be able to solder to it.

Click to expand...

I was really hoping for suggestions from someone that's installed a resistor in a Mag tailcap before. Maybe it's been done, maybe it hasn't, I was hoping someone else had done it before and could offer advice. Perhaps soldering isn't the best way and someone has done it another way. As far as not being able to solder to Aluminum, I found this on another forum and I'm willing to give it a shot...

"no need for an oil "bath" a heavy spray with WD40 is all you need, then give it a vigorous brushing with a WD40 wetted brass wire brush and solder through the thin liquid film (any light oil will do, I use WD40 because it is convenient and the residue isn't messy).

While you can use ordinary 60/40 & 50/50 tin/lead solders it is easier if you use 95/5 tin/zinc"

I think you can use resistor or Schottky diode (one for 5-10A with required Vf, about 0.3-0.4V may be enough). Diode is non linear and will give you voltage drop closer to designed across wider current range, regular diode will work too but it has higher Vf.

I would try to make some cap for narrow end of spring, insulating spring from battery bottom; metal something (brass screw in the middle) for battery contact but not touching spring. Maybe one sided copper PCB. Then with thick wire (AWG 18-14) solder diode (in right direction) or resistor between this contact and spring (it is easy to solder to Maglite spring, rough it a bit with sandpaper of file and use hot iron). Use thermal potting compound to couple diode/resistor with tailcap for cooling. Depending on shape you may be able to use thermal tape or soemthing.

arek98 said:

I would try to make some cap for narrow end of spring, insulating spring from battery bottom; metal something (brass screw in the middle) for battery contact but not touching spring. Maybe one sided copper PCB. Then with thick wire (AWG 18-14) solder diode (in right direction) or resistor between this contact and spring (it is easy to solder to Maglite spring, rough it a bit with sandpaper of file and use hot iron). Use thermal potting compound to couple diode/resistor with tailcap for cooling. Depending on shape you may be able to use thermal tape or soemthing.

Click to expand...

Ahh, I see what you are saying, almost like a thin-resistive-spacer. That's a good idea and I can still hide the resistor inside the spring area. I like it! Now we're getting somewhere!! Even if i solder one leg of the resitor to the base of the spring, it will transfer it's heat into the spring and then to the body. Worst case, I'll just have to remove some of the spring to accommodate the spacer.

See this post.

Thanks Justin. What holds the spring in the tailcap? Just friction? No solder joint needed to keep the bottom portion of the spring connected, just a friction fit with the anodizing removed?

Post deleted.

Rezolution said:

Ahh, I see what you are saying, almost like a thin-resistive-spacer. That's a good idea and I can still hide the resistor inside the spring area. I like it! Now we're getting somewhere!! Even if i solder one leg of the resitor to the base of the spring, it will transfer it's heat into the spring and then to the body. Worst case, I'll just have to remove some of the spring to accommodate the spacer.

Click to expand...

One method you can use is to get some sort of insulator that fits on top of the tailcap spring that will allow a contact on top of it that you can attach the lead of a resistor to and the other end of the resistor to the spring itself. You could use even a piece of a two sided circuit board.

fivemega said:

Wouldn't using a resistor on tail side cause less forwarded voltage to driver and runs nonregulated direct drive?
What about a custom made beefy head with large fins to remove the heat from driver so it will run much longer than 10 seconds at 3.5 Amps?

Click to expand...

3.6V is not enough, 4.8V is too much. 3 cells makes the light run at about 3.1A. 4 Cells will let it run regulated but it overheats the driver. I need a small amount of resistance to drop it like 0.6V or 0.8V for the light to run in regulation with 4 cells. That's why I'm doing this. With 4 cells and a resistor pad, it will run at 3.5A from start to finish.

If the driver is wired up to the switch in the head or the top of the body above the switch, why couldn't you install the resistor between the switch and the driver input.

moderator007 said:

If the driver is wired up to the switch in the head or the top of the body above the switch, why couldn't you install the resistor between the switch and the driver input.

Click to expand...

The only place I have space for a 5W ceramic resistor, sadly, is in the tailcap (under the spring).

After messing around with the light meter, I'm not sure how much the extra 0.4A is going to get me anyway... With the 3D cells and the light running at 3.1A I'm getting 46 kCd. I don't think it will do much better at 3.5A.

Maybe it's best to just leave it alone

My factory 2D rebel Mag is 22 kCd and "2" on ceiling bounce. My modded Mag is 46 kCd and "13" on ceiling bounce. That's twice the throw and 6 times the lumen output. Time to stop the modding and just enjoy it I suppose lol.

If you still want to tinker with the idea of adding some resistors to the tailcap... Techjunkie built a quad XM-L with 4C Nimh cells and ran into the same problem as you; too much Vin/Vbatt. You can take a look at his build thread here.

Another option would be to run 3 cells and do a some tailspring and switch resistance fixes to squeeze out as much Vin/Vbatt as possible.