kosPap
Flashlight Enthusiast
I meant to make a 2000 lumens thrower for sometime now, but I waited till I could find a good combination of Host flashlight, driver and LED.
Enter intl-outdoor with an 6V MT-G2 in P0 bin (very good)
http://intl-outdoor.com/noctigon-mtg20-mcpcb-cree-mtg2-p0-5000k-6v-led-p-727.html
and 2 choices for electronics
http://intl-outdoor.com/xml-multicell-circuit-board-3a-55126v-p-361.html
http://intl-outdoor.com/xml-multicell-circuit-board-35a-ouput-55126v-p-543.html
And fasttech, with the Convoy l2 host which is a bit of a beast as flashlights go.
The parts before assembly
Convoy L2 as a Host
The good thing about this host is that it has a LED opening the size of the MT-G2 and accepts a 21mm wide driver.
But internal room for the electronics is limited heightwise and the screws that hold the LED protrude into the pill. So you got to file down their length.
By and large it is a good host, especially for the price. Well machined, good enough anodization, AR coated glass, and press fit rings on the tube ends for better conductivity.
O-rings are a bit stiff and dense and not that high. Since the threading relief cuts are big enough I added silicon 26mm ones too.
I judge flashlights by the threads in the body and tailcap. In the Convoy they are anodized but thin for the size of the flashlight. I would expect a beefier design.
Did I tell you it is BIG?
Mothership BIG?
The Assembly in stages
Prepping the materials
The very first thing to do was cover the Huge 68mm reflector opening in shrink wrap. It makes for a clean and safe handling of it. Then the reflector base was isolated with self-sticking vinyl, the kind you find in car banners.
The reflector screws on the body and that makes for good heat transfering and good centering. On its edge there are two opposite tool cuts for the pliers so you got to be extra careful not to let them slip….
The LED socket
Preparing the LED
I added silver paste on its bottom, just enough to fill any metal-to-metal voids.
The LED MCPCB screwed on.
Reflector test fit…..It is a hair off-center, do not know why.
(the red blot is the vinyl part we removed from the LED opening)
Preparing the electronics
The driver came with a self-adhering thermal pad. But since space was limited I forego it. I positioned the wires that joined the two boards as best it could be done than filled the rest of the space with thermal glue (Fujik).
I also used some solid wiring to join the two boards tight against eachother.
What the driver socket looks like.
The LED wires soldered. It was a Pain in the A** cos the housing was so deep the tweezers and soldering iron were almost vertical.
I ended up soldering in 3 passes. A first join to just keep the wire in place, a second pass with extra flux to fix the wire properly and a third to "polish" the job…
End by screwing the brass retainer ring and adding a spring. While the original one was copper (?) colored I found it to be a bit weak for the calling and of too thin a wire.
I added an aftermarket thick and HARD one.
Measurements
I have a problem with them. They are too good to be true.
Either my lightbox is off or there is something fishy with the driver.
After the assembly I found out that the driver specs changed after I ordered it. It is supposed to be a 3.5A one but the new specs talk about 4.5A of output.
Practically I cannot know what I got, and I have no urge to desolder a LED wire to measure…
Reverse calculating it the driver must be feeding above 4amps to the LED
So…the numbers are.
4.2A current draw for the 18650 batteries
Lux at 1m: 48540
3350 Lumens on high!
1050 lumen on mid!
A runtime and temp test ended early cos the flashlight was scorching hot!
Min – Output – Temp C
0 – 100% - 23.7
1 - 94% - 32.7
2 - 92% - 37
3 - 91% - 39.7
6 - 87% - 51.7
9 - 81% - 58.6
12 - 80% - 62.9
15 - 77% - 67.5
18 - 75% - 69.8
I used protected Sanyo 2600s batteries. When the test stopped they measured 3.83volt and that makes for a 45% estimated capacity left.
Tempwise the battery had reached 65 C at the end of the test.
A bit latter I tried measuring the driver temp. The body temp had dropped some but the driver was cooler than it, at 45 C (at the time of measurement the body was at 54 C)
And that explains why the speced driver temp protection did not kick in during the test.
Sadly no outdoor pics are possible for now.
Across the square the light is not that good a thrower as acroos the distance the hotpsot thaws out, but it is resistance-is-futile stunning!
thanks for reading
Enter intl-outdoor with an 6V MT-G2 in P0 bin (very good)
http://intl-outdoor.com/noctigon-mtg20-mcpcb-cree-mtg2-p0-5000k-6v-led-p-727.html
and 2 choices for electronics
http://intl-outdoor.com/xml-multicell-circuit-board-3a-55126v-p-361.html
http://intl-outdoor.com/xml-multicell-circuit-board-35a-ouput-55126v-p-543.html
And fasttech, with the Convoy l2 host which is a bit of a beast as flashlights go.
The parts before assembly
Convoy L2 as a Host
The good thing about this host is that it has a LED opening the size of the MT-G2 and accepts a 21mm wide driver.
But internal room for the electronics is limited heightwise and the screws that hold the LED protrude into the pill. So you got to file down their length.
By and large it is a good host, especially for the price. Well machined, good enough anodization, AR coated glass, and press fit rings on the tube ends for better conductivity.
O-rings are a bit stiff and dense and not that high. Since the threading relief cuts are big enough I added silicon 26mm ones too.
I judge flashlights by the threads in the body and tailcap. In the Convoy they are anodized but thin for the size of the flashlight. I would expect a beefier design.
Did I tell you it is BIG?
Mothership BIG?
The Assembly in stages
Prepping the materials
The very first thing to do was cover the Huge 68mm reflector opening in shrink wrap. It makes for a clean and safe handling of it. Then the reflector base was isolated with self-sticking vinyl, the kind you find in car banners.
The reflector screws on the body and that makes for good heat transfering and good centering. On its edge there are two opposite tool cuts for the pliers so you got to be extra careful not to let them slip….
The LED socket
Preparing the LED
I added silver paste on its bottom, just enough to fill any metal-to-metal voids.
The LED MCPCB screwed on.
Reflector test fit…..It is a hair off-center, do not know why.
(the red blot is the vinyl part we removed from the LED opening)
Preparing the electronics
The driver came with a self-adhering thermal pad. But since space was limited I forego it. I positioned the wires that joined the two boards as best it could be done than filled the rest of the space with thermal glue (Fujik).
I also used some solid wiring to join the two boards tight against eachother.
What the driver socket looks like.
The LED wires soldered. It was a Pain in the A** cos the housing was so deep the tweezers and soldering iron were almost vertical.
I ended up soldering in 3 passes. A first join to just keep the wire in place, a second pass with extra flux to fix the wire properly and a third to "polish" the job…
End by screwing the brass retainer ring and adding a spring. While the original one was copper (?) colored I found it to be a bit weak for the calling and of too thin a wire.
I added an aftermarket thick and HARD one.
Measurements
I have a problem with them. They are too good to be true.
Either my lightbox is off or there is something fishy with the driver.
After the assembly I found out that the driver specs changed after I ordered it. It is supposed to be a 3.5A one but the new specs talk about 4.5A of output.
Practically I cannot know what I got, and I have no urge to desolder a LED wire to measure…
Reverse calculating it the driver must be feeding above 4amps to the LED
So…the numbers are.
4.2A current draw for the 18650 batteries
Lux at 1m: 48540
3350 Lumens on high!
1050 lumen on mid!
A runtime and temp test ended early cos the flashlight was scorching hot!
Min – Output – Temp C
0 – 100% - 23.7
1 - 94% - 32.7
2 - 92% - 37
3 - 91% - 39.7
6 - 87% - 51.7
9 - 81% - 58.6
12 - 80% - 62.9
15 - 77% - 67.5
18 - 75% - 69.8
I used protected Sanyo 2600s batteries. When the test stopped they measured 3.83volt and that makes for a 45% estimated capacity left.
Tempwise the battery had reached 65 C at the end of the test.
A bit latter I tried measuring the driver temp. The body temp had dropped some but the driver was cooler than it, at 45 C (at the time of measurement the body was at 54 C)
And that explains why the speced driver temp protection did not kick in during the test.
Sadly no outdoor pics are possible for now.
Across the square the light is not that good a thrower as acroos the distance the hotpsot thaws out, but it is resistance-is-futile stunning!
thanks for reading
Last edited: