MagLED 2D Performance in Heatsinked Flashlight - Test Results

sguyer

Newly Enlightened
Joined
Dec 8, 2006
Messages
12
Maybe this should be posted in the incredible "MagLite 2C/2D 4D 3W LED drop-in Technical Review". This is a relatively minor project in comparison.

So that you do not have to read all this, the end results are as follows:[
The MegLed 2D in a heatsinked flashlight dropped 10% - 15% total with current stabilizing at about 1 Amp. The light output and current was mostly stable in 2.5 minutes. This compares with the MagLed in the Mag Lite flashlight which dropped 40% (tests by others show up to 50%) with the current stabilizing at 540ma. The light output and current was mostly stable in 15 minutes.

It is unfortunate the design of the star flashlight for hosting all these new bulbs is locked into a poor heatsink configuration. As a result, all the new bulbs must be turned down 30%-40% to avoid overheat. Maybe Mag Instruments will redesign their flashlights before someone else comes along and stills the show. I believe they did just that for the AA Mag.

The Project – Use a heatsinked flashlight to measurement the MagLed bulb performance.

Heatsinked Flashlight
For this project, I used the often mention on this forum $3.95 2D Mag clone (http://www.improvementscatalog.com/product.asp?product=265943bluzz&dept_id=301). The flashlight consists of an aluminum plate, which holds the bulb. Unfortunately, the bulb is held to the metal plate with a plastic holder. Heatsinking the MagLed 2D bulb consists of only replacing the plastic bulb holder with a metal holder. The quality of this flashlight is not up to that of a Mag Lite. The springs are not quite as strong and the I do not believe the switch is the same quality

The Modification
Armed with only a drill press and hack saw, I purchased from Home Depot a brass pipe fitting Watts P/N A-116. Brass has about ½ the heat transfer as aluminum. The threads were not an exact match, but with a little force the first time with a wrench and the piece fit and then could but put in by hand. Next a hack saw to the correct length and then the drill press to enlarge the opening to allow the MagLed to fit. The 7/16 drill bit was not quite large enough and some reaming was necessary. Also, a slight bit of reaming of the aluminum disk was required for the MagLed bulb to fit. I used computer heatsink compound on the components and assembled. A piece of rubber tubing was glued in to prevent the spring from contacting the side of the brass fitting. The picture below shows the completed project. Left is the original disk with plastic part. The right disk is the completed disk with Magled bulb and the far right is the unmodified brass pipefitting. This is definitely not the world's best heatsink, but it is all I could do with a hacksaw and drill press.

144_4421_small.jpg


SETUP.
Unlike the previous tests, I decided to drive the flashlights using an external supply. This eliminates battery voltage changes from the picture. I elected to use 2.4 –2.5 volts, which is close to actual worse case battery performance under load (150-300mohms) and matches NiMH performance. I took data using the unmodified 2 D Mag flashlight and with the heatsinked Mag Clone. I took data points at start, 2.5 min, 5 min, 10 min, 15 min, 20min 30 min, 1 hr, 1.5 hr, and 2 hr. Lux was measured at 1 meter center of beam. Both the Mag and clone focussed to a tight beam.

Measurement Accuracy Problems
I could not get accurate starting numbers. The data changed very quickly before I could get setup to take readings. Maybe this had to do with the ambient temperature of 50 deg F which in theory gives 5% more light at the start. I give this data about 80% accuracy with the most error occurring at turn on.

Results and Conclusions
Mag Flashlight-
The Mag data follows closely data measured by others. The data shows the typical 40% drop in light output stabilizing in about 15 minutes at about ½ amp, similar to what others measured. Starting current was 1.1 amps and ending current was 540ma.

Heatsinked Clone
The modified clone showed only about a 10%-15% drop in light output stabilizing in about 2.5 minutes at about 1 amp. Starting current was 1.3 amps and ending current was 990ma.The Mag people have created a bulb, which will work well in different environments.

Ambient Temperature 50 Deg F



MyHTML1.gif



MyHTML2.gif
 
Top