greenforce halogen to LED conversion

jspeybro

Enlightened
Joined
Aug 13, 2009
Messages
586
Location
Belgium
cutter.com.au has a 24mm quad xpg that is individually addressable so that would be suitable. It's a bit unclear though for the optic since the optic they link to is out of stock...
They also have some MR11 quad optics for 25mm boards so they may fit the 24mm boards as well. the optic would need some grinding though to remove the outer edge that is not part of the lenses.

I guess I'll send them an email when they are back in the office.

once I know what optic can be used, I can check if it fits in the head with a new heatsink.
 

jspeybro

Enlightened
Joined
Aug 13, 2009
Messages
586
Location
Belgium
a bit overdue for an update on this project...

meanwhile, the 24mm quad XPG module with 24mm optic arrived from cutter. I made a new heatsink, similar to the triple xpg board.
I used a AMC driver that drives 1 of the LEDs. The other LEDs are in series with the driver itself to drop the voltage.
Since the battery has a full charge of a little over 14V, this should be a nice fit. 3x 3.25V drop=9.75V drop. Add to that the input range of the driver of about 3-5.5V, this setup should be able to cope with the voltage of the battery pack. It will however start dropping in intensity a bit sooner than I wanted, but have yet to do some tests.
I measured a FWHM angle of 14-15°. That's a little more than I would really want, but this is still impressive for such small optics. I'll post a comparison of the intensity profiles and some info on the angle measurement in one of my next posts in this thread.
The AMC driver is one of 1400mA with a microcontroller and stars where I connected star 2 to the ground giving me a high and a low mode, which is 100% and a little over 10% intensity.

Currenlty I just need to get a new o-ring since one of the o-rings shows some damage and I don't want to risk a flood on my first dive with it.
pictures will follow.

Johan
 

jspeybro

Enlightened
Joined
Aug 13, 2009
Messages
586
Location
Belgium
Here's a montage of beamshots I made yesterday:

Montage.jpg


from left to right, top to bottom:
The 4x xpg mod I mentionned in my previous post, high and low;
the next one is another project I'm working on, using a cree xml and a 3A DX driver with a 35mm optic;
then the other greenforce mod shown in pictures above, using also a cree xml and a 35mm optic and a 2.8A DX driver;
the last one is a greenforce original tristar, however, this is one of the first of these heads ever made by greenforce and the intensity is really disapointing (even lower than the low mode of the quad xpg as you can see in the intensity comparison below.

IntensityComparison.png


This graph gives you a nice intensity profile comparison. I thought initially that the 2 XML lights used the same optic, but apparently the 3W Hollis lights use a slightly different optic than the DX optics. The DX 35mm TIR optics are a near perfect match for cree XML leds in my opinion.
While the quad XPG head on high has a slightly lower maximum intensity, there is clearly a lot more light and more evenly distributed in the cone of light. This graph can also be used to determine the position of the half intensity. Currenlty I used pixels of the beamshot pictures on the X-axis. I'll see if I can convert that to angles so you can more easily read the graph.

1 difference between the 2 XML lights, is the battery. The 3A driver runs off 2 18650's, while the 2.8A driver runs off the greenforce 12V Arrow NiMh 3.3Ah battery.

enjoy :)
Johan
 

jspeybro

Enlightened
Joined
Aug 13, 2009
Messages
586
Location
Belgium
Hi Guys,

I have some underwater beam shots of some of my greenforce mods :)
Sorry for the crappy quality, it's the first time I actually took a picture underwater...
The first is the cree MCE @ 2.8A with 35mm optic:
PICT0004.jpg


the second is the quad XPG @ 1.4A per LED on 24mm board with matching 24mm optic
PICT0003.jpg


The 3rd is a mod that I didn't show before. It's a greenforce PRO head in which I placed a cree XML at about 2.5A and an aspheric. I'm currently using this as my primary light:
PICT0002.jpg

My buddy called it my lightsaber ;-)

Johan
 

Pizeer

Newly Enlightened
Joined
Sep 26, 2012
Messages
21
Location
Montreal
The triple XPG setup doesn't seem to work very well... after a few seconds, the LEDs start to strobe fast and after 10-15 seconds or so, it starts to flicker. This of course indicates that the AMC chips are overheating and the thermal protection is kicking in.
Initially it started to flicker immediately, but the driver wasn't heatsinked at all. I then used thermal glue to glue the AMC chips to the back of the heatsink, which helped for a few seconds, so this doesn't seem to be enough.

I suppose the battery voltage when fully charged is too high (10x1.2V NiMh in series) compared to the forward voltage of the 3 LEDs, where as it might work fine with a partially discharged battery.
Any idea if there is another way to keep the AMC chips cooler, other than sticking the package to the heatsink?
would it help to cool the board somehow?

Maybe I need to start looking for another driver if this AMC setup doesn't work out.

any suggestions?

What kind of temperature are you seeing? Deg F or Deg C?
 

dexter9

Newly Enlightened
Joined
Dec 19, 2011
Messages
27
The 3rd is a mod that I didn't show before. It's a greenforce PRO head in which I placed a cree XML at about 2.5A and an aspheric. I'm currently using this as my primary light:
....
My buddy called it my lightsaber ;-)

Johan


Hello Johan,
what kind of aspheric are u using with the xml, i like the beam
 

jspeybro

Enlightened
Joined
Aug 13, 2009
Messages
586
Location
Belgium
What kind of temperature are you seeing? Deg F or Deg C?

I have no idea. I since replaced the setup with 4 XPG's so there are now 3 XPG's in series with the driver instead of 2. The voltage drop when fully charged is now OK, and although I haven't reached this point yet, I'm guessing it will go a bit out of regulation when the battery is nearly empty.
Basically the AMC chips had to burn too much voltage into heat, causing the temp protection to kick in. I think the set point for this system is somewhere in the datasheet, so that is probably the temperature that I was getting...

Are you trying someting similar?
 
Top