XM-L T6 advice please

richard8922

Newly Enlightened
Joined
Dec 28, 2009
Messages
3
I have just bought a XM-l T6 emmiter to upgrade my 4D mag from a single P7. i am running 4 NIMH D batteries and a 2.8amp driver. would i get away with just running a resistor along with my driver to bring my voltage down to safely run this diode. I understand this emitter needs around 3.5 volts. Does anyone know roughly what ohm resistor i would need.


Cheers in advance for any help
 

moderator007

Enlightened
Joined
Jan 1, 2010
Messages
990
If its a 2800ma 8-amc7135 driver the driver will use the excessive voltage up as heat. With all the parasitic resistance added up in the light there should not be that much excessive voltage to begin with. And you factory in voltage sag from the batteries. In my opinion the light will stay in regulation longer with 4 Nimh's than with 3 Nimh's. Making the light run longer before decreasing in brightness.
The P7 (depending on bin) and the XML T6 at 2.8 amps aren't that much different in forward voltage. 3.3v vs 3.5v depending on bin and luck.
 
Last edited:

richard8922

Newly Enlightened
Joined
Dec 28, 2009
Messages
3
If its a 2800ma 8-amc7135 driver the driver will use the excessive voltage up as heat. With all the parasitic resistance added up in the light there should not be that much excessive voltage to begin with. And you factory in voltage sag from the batteries. In my opinion the light will stay in regulation longer with 4 Nimh's than with 3 Nimh's. Making the light run longer before decreasing in brightness.

Thanks very much. i think i will get this all built up and chek the voltage and current draw at the emitter. just out of interest what current have people been pushing these to or is 3 amps about it. cheers
 

Hoop

Enlightened
Joined
Dec 18, 2011
Messages
316
Location
Spokane, WA
3.5v is probably much higher than what the LED will actually need at 2.8 amps. Also, as LED's get hotter they require less voltage to maintain the same current. The XML spec sheet is based on a 25C Tj and is not a very accurate reference for what voltages you are going to see. Refer to this thread: "Super Low XML VF's."

Edit: just be sure and heatsink the driver properly so it can dissipate excess voltage.
 
Last edited:

richard8922

Newly Enlightened
Joined
Dec 28, 2009
Messages
3
3.5v is probably much higher than what the LED will actually need at 2.8 amps. Also, as LED's get hotter they require less voltage to maintain the same current. The XML spec sheet is based on a 25C Tj and is not a very accurate reference for what voltages you are going to see. Refer to this thread: "Super Low XML VF's."

Edit: just be sure and heatsink the driver properly so it can dissipate excess voltage.

Thanks for that link, i think i will wait till the led arrives and have a play with it till its around 3 amps, being off ebay i suppose its a bit hit or miss as to how efficent the led will be.
 

alpg88

Flashlight Enthusiast
Joined
Apr 19, 2005
Messages
5,282
many 8amc drivers i used heat up from 4nimh cells, after sometime, and start blinking.

if you wanna go with resistor, use 3 cell, and use 0,1 ohm resistor.
 

moderator007

Enlightened
Joined
Jan 1, 2010
Messages
990
many 8amc drivers i used heat up from 4nimh cells, after sometime, and start blinking.

if you wanna go with resistor, use 3 cell, and use 0,1 ohm resistor.
Just curious alpg88, did you thermal epoxy the driver down to any kind of heatsink. I have not noticed the blinking from the driver with the driver heatsinked.

As far as the xml at 3 amps goes some have pushed it to 5 amps and more with a proper heatsink. One members test showed that their wasn't much to gain over about 4 amps from the xml with a lux meter. Then another member said he noticed quite a bit of difference between 4 amps and 5 by eye. The beam shots he provided also looked as though their was alot to gain at 5 amps. I have not pushed the xml over 4 amps and have noticed quite a increase (by eye) between 3 amps and 4. Some say the sweat spot is at 3.5 amps. You may just have to experiment with your xml and judge for yourself.

Some good reading on the amc7135 drivers and voltage sources vs current and led vf.
 
Last edited:

alpg88

Flashlight Enthusiast
Joined
Apr 19, 2005
Messages
5,282
yes, i did, but you can only heatsink 1 side, chips are on both sides. bottom chips are still have no way of getting rid of the heat, i got good results by soldering few pieces of braid (3 or 4 spread them at 120* or 90*) to the negative ring, (that is where thickest leg of the amc chip dumps the most heat. wedge those copper braids between heatsink and the tube, i use file to file a small channel in the heatsink so braid fits in between.it also doubles as negative contact.
it works fine with li ion, once it is below 4v, or 3x nimh, the driver runs cool. (but it only runs at 1,5-1,7A)
 

moderator007

Enlightened
Joined
Jan 1, 2010
Messages
990
I make my own heatsinks for the xml in a mag to fit the rebel reflector.
side.jpg


Then I make a quarter round the same radius as the hole in the back of the heatsink.

quarterround-1.jpg


Then thermal glue the driver to the flat spot on the quarter round and thermal glue the quater round to the side in the bottom of the heatsink. Seems to work well for me.

quarteroundinstalled-1.jpg


I agree that you can only heatsink one side of the board which doesn't make it a very good thermal path for the other side. All the 7135's are connected in parallel so some heat does have to be transfered.
 

Walterk

Enlightened
Joined
Jan 21, 2010
Messages
755
Location
Netherlands
If I have it right, with any amount of AMC7135 it would be for XM-L (3.7-3.3=0.4V) x 2.8A = 1,12 Watt of heat from the driver?
 

yazovyet

Enlightened
Joined
Apr 9, 2011
Messages
221
If I have it right, with any amount of AMC7135 it would be for XM-L (3.7-3.3=0.4V) x 2.8A = 1,12 Watt of heat from the driver?

yes and no. You assume 3.7 volts from the battery and 3.3 volts to the LED and 2.8 amps; if those are correct then the heat you calculated is right (if you had greater or fewer amc chips you'll have a different current). BUT battereis can sag lower or be running hot off the charger (depending on lion vs 3NiMHs) and in the case that the OP was talking about his 4 NiMH would give something like 4.8 volts and thus generate far more heat in the driver (around 4.5 watts).
 

alpg88

Flashlight Enthusiast
Joined
Apr 19, 2005
Messages
5,282
yes. which is why i recomend getting rid of 4th cell, even 1W makes the driver hot, and blink as heat protection kicks in. why use 4th cell and waste it into heat, not to mention, 4W is way too much for 8xamc driver to dissapate
 

yazovyet

Enlightened
Joined
Apr 9, 2011
Messages
221
I agree with alpg, runing 3 D cells and leave the 4th one out of it (get a blank cell sort of thing to take up space). Putting the 4th cell in will not increase run time significantly unless you get a different style of driver (althought with 3 cells you might find the last quarter or so of the run time has diminishing brightness, I'm not sure how it will go with Ds) . If you wanted to keep all 4 cells and use that driver then you might look into getting a .3 ohm resistor that can dissipate over 3 watts (to be safe, 4 or 5 watts) and put in in serise (mounting it in the tail cap would be better so the heat is away from the head). But again, that sounds like more work than getting a blank battery and useing 3Ds.
 

alpg88

Flashlight Enthusiast
Joined
Apr 19, 2005
Messages
5,282
you can have a resistor inside the spacer for the 4th cell, circuits with resistors or with linear regulators (aka amc boards\drivers), have same current running on either side of the circuit, by the led, or the battery. you don't have to put resistor in series with led after the switch, you can put it before the switch, in series with battery, first goes resistor, (or last) than the cells.
 

Walterk

Enlightened
Joined
Jan 21, 2010
Messages
755
Location
Netherlands
Not all off topic: I would glue the driver with thermal epoxy in place, and fill the remaining part of the void with thermal glue.
 

moderator007

Enlightened
Joined
Jan 1, 2010
Messages
990
yes and no. You assume 3.7 volts from the battery and 3.3 volts to the LED and 2.8 amps; if those are correct then the heat you calculated is right (if you had greater or fewer amc chips you'll have a different current). BUT battereis can sag lower or be running hot off the charger (depending on lion vs 3NiMHs) and in the case that the OP was talking about his 4 NiMH would give something like 4.8 volts and thus generate far more heat in the driver (around 4.5 watts).
Yazovyet brings up a good point here. The quality of the batteries used could play a big part in the wasted heat. More voltage sag less heat. Most all the graphs I have seen for newer type Nimh D cells show that they can hold above 1.2v for greater than half there discharge under a 2 amp load (3.6v with 3 cells). My cells are about 4 to 5 years old and are some no name brand green label cells. This could explain why I have not had any issues with the 4 D setup with the 2800mah driver. My cells may have alot less voltage under load than the newer type cells especially with a hundred or so cycles on them. When you factory this in and the parasitic resistance of the entire circuit path, all connections and the voltage drop of .12v of the amc7135 driver this could explain why mine works well? I'll try to test my Nimh D's and post back what I find.
 

wildstar87

Enlightened
Joined
Nov 12, 2007
Messages
369
I also haven't had any problems with 4D Ni-MH with this setup. The heatsinking actually doesn't need to be that much, a small slug of aluminum on the 7135 chips will do, one side can go to the body, the other side can just be in free air (w/slug).
 
Top