Luxeon Star (LS) - 3 D Cell conversion info / stats.

[Silviron]

Enough people have recommended that I start a new thread on this conversion for the sake of "posterity"; So here it is. A lot of this has been covered previously and in somehat more depth in several posts in the "What's up with Luxeon Star" thread at:
http://www.candlepowerforums.com/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=3&t=000625

Full details of the conversion including MANY photos that show how I did it, what problems were encountered and how they were solved are at:
http://az123.com/LED/3DLSconv/

Below is a compilation of all the statistics concerning battery life, voltage, current, light output etc. covering the three days + (74 hours) that the light has been "burning" constantly. I'll add more information as I can.

After THIS test using regular Eveready Alkaline D cells is done (which will be when NO light is coming out of the LED) I'll run another test with rechargable batteries, and let you all know the results on this thread too.

Of course anyone else doing a 3D conversion is more than welcome to add their info too.
____---------___________----------___________

With 3 "Fresh" Eveready alkaline D cell batteries (~ 4.9V) the light is " pulling " 370 mA (which is over the manufacturers specifications by 20 mA) Output is ~ 200 LUX at 1 meter between sensor and lens.

After one hour, the light is pulling ~350 mA and measures 3.14 V at the module terminals. (I neglected to measure LUX here, sorry-)

After 2 hours there is 3.11V at the module terminals, drawing 335 mA and putting out 290 LUX at 1 meter.

(You will note that LUX output after two hours is actually higher than it was at the beginning. This is likely due to inefficiency in converting current to light due to overdriving the LED {Perhaps generating extra IR OR UV which don't register on the meter???? That could also explain why the camera lens saw more blue than my eye did??} ).

At 4 1/2 hours there was 3.09V at the module terminals, using ~ 290 mA and emitting 220 LUX at 1 meter. (battery voltage is 4.28V)

At 7 hours terminal voltage is 3.06V and drawing 250 mA with ~ 190 LUX output. (Battery voltage is 4.2V)

At 9 1/2 hours terminal voltage is 3.07V drawing 250 mA and putting out ~ 190 LUX with battery voltage at 4.17

(Again a kind of discrepancy- there appears to be .01 V more at the module terminals with .03V less in the batteries at 9.5 hours than there was at 7 hours. My guess at the reason for this is that the module is running much cooler. {either that or meter error})

At 11 hours, still 3.07 V at terminals, at 240 mA (4.12 V on batteries at no load) 148 LUX at 1 meter.

At 17 hours , 3.06 V at terminals, drawing 220 mA (4.04 V on batteries at no load) 190 LUX at 1 meter

(Again, another discrepancy: after 6 hours the LUX reading has gone up by almost 25% at lower voltage / current. Most likely explanation? I didn't have the light aimed at the light meter sensor properly at the 11 hour measuring)

At 27 hours, it is 3.02 V on the module terminals, drawing 1.95 mA (3.94V on batteries with no load). 116 LUX at 1 meter.

At 48 hours it is 3.00 V at the module terminals drawing 163 mA (3.77V no load) 98 LUX at 1 meter.

At 65 hours: 2.95V at 118 mA - 112 LUX. (3.5V on batteries with no load)

After 74 hours voltage at module terminals is 2.91 V at 85 mA (3.42 volts on batteries with no load. 60 LUX at 1 meter.

 

[Silviron]

After 100 hours: Voltage at module terminals = 2.82, drawing 40 mA (no load voltage is 3.1) 41 LUX @ 1 meter

 

[X-CalBR8]

Silviron: Nice conversion. I just wonder how much of the heat problem would go away if this was an aluminum flashlight such as a 3D Mag-Lite instead of a plastic and rubber flashlight.

Has anyone tried a Luxeon Star mod on a 3D Mag-Lite yet? I would like to do this mod eventually when I get my finances straight.

 

[Silviron]

I thought about doing a mag light conversion, but didn't want to spend the extra $12.00 for the mag light originally- and don't want to "waste" the time to do one now.

But, to answer your question- I'm pretty sure that you would need to heat sink a LS in a Maglight, (but it would be easier to do than with the plastic one I did).

A simple circle of copper or aluminum just the diameter of the inside of the mag casing- You would want as much metal to metal contact as you could get, and slather heat conductive paste around the edges of the heatsink.

That would probably keep it cool enough even for "continuous" use without having to drill vent holes.

 

[Silviron]

At 120 hours ( 5 DAYS !) of continuous operation there is 2.78 V at the module, drawing 30 mA ( batteries have 2.99V no load) Light output is 26 LUX at 1 meter

 

[Silviron]

I found some great heat sinks that might be useful for just about any LS conversion or fabrication using C or D cells etc:

I ordered five earlier, they arrived today, and I'm going to order some more just to have on hand. They could be useful for Nichia projects too.

50 cents each. Only took 2 1/2 days to get here after ordering on the web.

http://www.allelectronics.com/cgi-bin/category.cgi?category=270&item=HS-59&t ype=store

 

[Silviron]

At 147 hours (6 days 3 hours) voltage at module is 2.75V, (2.91V no load) at approximately 20mA current draw, with 23 LUX light output at 1 meter.

The light is getting " *noticably*" dimmer, although there is still more than enough light to find your way around a dark house, change a flat tire or read etc.

 

[Silviron]

After 172 hours continuous, the 3D - LS conversion has 2.73 V at the module @ 11.3 mA (2.85V no load) Output of 13 LUX at 1 meter.

 

[Silviron]

At 196 hours, voltage at module terminals is 2.72V @ 10.7mA. (no load voltage is 2.82V), Light output is 11 LUX at 1 meter.

 

[Silviron]

The 3 D cell conversion after 223 hours of continmuous operation is at 2.71V @ 9.75mA
(2.81 V no load) with 13 LUX output at 1 meter. (I realize that this LUX reading is a bit higher than the previous one. I probably didn't have it "aimed" right last time.)
-------------------------------------------
Have also started longevity tests on a 3 C battery pack with LS (using Rayovac C cell alkaline 1.5V rechargables) Using a 3.6 ohm resistor initial readings on brand new, just charged batteries were:
3.14V at module terminals @ 340 mA current (4.88V no load) Light output was 202 LUX at 1 meter.

After 4 hours there were 157 LUX at 1 Meter, 165mA @ 2.98V (4.01V No Load)
-------------------------------------------

 

[Silviron]

After another 20 hours (243 hours total), the 3D - LS conversion has hardly changed at all: 2.71 V @9.6 mA, 12 LUX @ 1 meter.

The test with a LS on 3 C cells after 16 hours of continuous operation has 2.94V @ 125mA at the module, (3.74V no load) and light output at 1 meter is 106 LUX.

 

[Silviron]

At 267 1/2 hours it is 2.69V @ about 10 mA and 10 LUX @ 1 meter

At 291 1/2 hr it has 9 LUX @ 1 meter with 2.68V @ about 8mA.

After 12 days of continuous operation, the 3D-LS conversion is starting to get to the point that I would not want to have to find my way out from the bottom of the Mines of Moria without a change of batteries available. It is plenty bright enough for close up-tasks, but beyond say 12 ft. it is pretty weak.
--------------------------------------------------

ON the 3 C battery test, at 42 hours: 2.81V @ 35 mA and 28 LUX.

And the little 2 AA Rayovac rechargable alkalines are still running a LS module after 304 hours (12 2/3 days!) continuously. The light output is only bright enough to read with from 12 inches or so, pretty much useless for much more than that. 2.53V at module, and .48mA current draw.

Have a good weekend everyone- I'm outta here til Monday.

 

[Silviron]

Well, after 342 1/2 hours (14 days +), the 3D-LS conversion is putting out 7 LUX at 1 meter, running on 2.65V at the module with 8mA. Still enough light to get around the house in a blackout or change a flat tire on a cloudy night in the country.
_____________________________________________

The 3C-LS test after 65 hours shows 2.67V @ 4mA (2.73V no load) with 4 LUX output @ 1 meter.

This 3 C cell test is using Rayovac rechargable alkalines and the results are disappointing given the results on my other battery life tests with the LS module. It could be that they just are not "broken in" yet as they were brand new, and freshly "topped off" with a charge when I started the test. Will re-run this C cell test on "regular" alkalines pretty soon, then re-test with rechargables again.
_____________________________________________

After 357 hours (almost 15 days), the LS module being tested on the 2 AA Rayovac rechargable alkaline cells is still putting out enough light to read by from about 12 inches, although I wouldn't want to have to do much more than that with it. 2.52V @.43mA

 

[Coherence]

Thanks for keeping us updated Silviron. It really is amazing how the LS (and other LED's) self regulate to have outrageous runtimes.

I would love to have the test equipment you have - what kind of meter are you using to measure less than 1mA ? I also look forward to a runtime test of 2 AA lithium cells with the LS if you are willing to do it and post the results.

Again, thanks for all the info.

 

[Silviron]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Coherence:

I would love to have the test equipment you have - what kind of meter are you using to measure less than 1mA ? I also look forward to a runtime test of 2 AA lithium cells with the LS if you are willing to do it and post the results.<HR></BLOCKQUOTE>

I have to admit that my mA measurements may be off a bit when they fall below about 30mA- I have been using a $17.00 digital Multi-meter from Walmart, and I have noticed that it is registering about 30% lower than my two analog meters on the 3D-LS tests-

Lacking any kind of known current standard I'm just having to go with what the DMM says on the readings below about 5mA and am using the two analog meters (averaging and interpolating) on things between 30mA and 5mA. The analog meters are pretty useless below 5mA and completely useless below 1mA, so I am having to rely completely on the DMM for that range.

So, unfortunately, my current measurements are probably not precise- however, I'm pretty sure that they are useful for comparison purposes at least, because the error rate seems to be pretty consistant.

I WILL do a longevity test on a couple of AA lithiums- I kind of would like to run the alkaline rechargables to "death" first, but that could take another week the way things are going now.... I may get tired of waiting on it though and start the lithium test in a day or two...

 

[Silviron]

On the 3D-LS flashlight conversion:
At 373 hours I get 3 LUX at 1 meter. 2.61V @ 4mA

At 390 hours (16 1/4 days!) the 3D-LS conversion is putting out 3 LUX @ 1 meter- 2.60V @ 2.88mA.

Enough light to get around the house in a blackout if necessary, plenty of light to read by.
____------___________----------___________--------

On the 3 C cell - LS test- (using Rayovac Rechargable alkaline C cells) the light output fell to "reading level" only at around 80 hours.

Am now re-testing with "regular" "cheapie" alkaline C cells ( EverActive brand 2005 date code).

Here is the data on that so far:
(Going through a 3.6 ohm resistor)

Initial readings- Fresh batteries: 170 LUX @ 1 meter 4.85V on batteries- 3.14V @ 365mA at module. (running pretty hot- have to keep heat sinked module in front of a small fan to keep it reasonably cool)

1 hour: 139 LUX @ 1 meter, 3.08v@ 280mA at module

8 hours: 109 LUX, 2.99V @ 190 mA at module. (3.94V no load)

18 hours: 93 LUX @ 1 meter 2.97V @ 165mA at module (3.78V no load)

21 1/2 hours: 93 lux still---

27 hours: 82 LUX @ 1 meter 2.94V @ 130mA at module (3.62V no load)

31 hours: 73 LUX

40 hours: 46 LUX 2.87V @ 65mA at module (3.26V no load)

SIGNIFICANTLY better (so far) on these than on the rechargable alkalines previously tested. It COULD be that the Rayovacs just were not "broken in" yet as they were brand new??? I don't know if that type battery needs a few charge/discharge cycles to get to peak performance or not...

 

[Silviron]

On the 3D - LS at 420 hours: light output is 2 Lux at 1 meter -- 2.55V @ 1mA at module.

On the 3C - LS at 56 hours I get 23 LUX at 1 meter with 2.78V @ 24.5 mA (3.0V no load)

Got a new multi-meter today ( I've been wanting one of those from Radio Shack that interface with a PC for several weeks; Went in today for a new soldering iron tip, saw the meter on sale for $49.95, so decided to get it.) More on that later...

Anyway, It appears that some of my earlier mA readings are higher than "reality". The ones below about 5mA are probably 20 - 50% high. Hard to say for certain without running the whole test over.

 

[Silviron]

On the 3D-LS conversion-- after 442 hours of continuous operation, I Get a measurement of 1 LUX @ 1 meter- 3 LUX @ 2 ft. I'm not going to bother with volts / mA measurements this time.
---------------------------------------------
On the test with 3 alkaline C cells, - after 78 1/2 hours continuous, I get 13 LUX at 1 meter. 2.74V@ 15.1 mA at the module, 2.88V no load.
--------------------------------------------

On a side note, that probably has relevance on any conversion or design using a LS LED in a non-metallic body:

I have just started testing of a LS module built to be a reversible mod to plug in to a Koehler / Wheat miners / cavers lamp (these run off of a 4V lead acid battery, and the head is sort of a bakelite material with a glass "lens")....

Well, heat once again rears its ugly head, even though I built the module into one of those round heat sinks that I found at All Electronics and mentioned on an earlier post.

After about 20 minutes of operation, the temperature on the heat sink was up to 94 degrees C. TOO HOT!!!

So, even though it defeats the water and dust resistance of the light head, I drilled some holes for ventilation. 12 holes of .051 inch diameter does the trick- the heat sink is staying at a safe level of 63 degrees C. now.

 

[**DONOTDELETE**]

I am thinking of putting 3 LS into a Mag 3D.

Can you measure and give me the diameter of a LS? I know that heat will be a problem based on your reports, but I was thinking of heat-sinking to the aluminum case and probably removing the front lens. I just need to determine if they will fit and I will buy them.
BTW: I am asking you for the measurement because the specs page from the acflashlight.com link isn't working.

Thank you.

[edit] ok it looks like the LXHL-NW98 is the the /O model, spec sheet says 25mm for the base and 21.6 dia. for the optics. Hmm the Mag 3D has 48mm to work with. Time for a little CAD work... I will let you know what I find.

 

[Silviron]

It is about time to give up on this test of the 3D LS conversion:

At 493 hours we have 2.55V @ .6 mA 1 LUX @ 1 meter, 3 LUX @ 2ft. and 4 LUX @ 1 ft.

at 510 hours still 1 lux @ 1 meter.

Will probably do this test over using some Rayovac rechargable alkalines in the near future.
______---------___________-----------_________--------____________

On the 3 C LS test :

90 1/2 hours: 11 LUX @ 1 Meter 2.73V @ 14.5mA (2.83V no load)

97 1/2 hours : 12 LUX 2.73V @ 15.0 mA

103 1/2 hours : 11 LUX

127 1/2 hours: 2.71V @ 11.0 mA (2.81V no load) 8 LUX @ 1 meter.

137 1/2 hours : still 8 LUX

147 1/2 hours : 7 LUX @ 1 meter