Help with a LS problem.

txwest

Flashlight Enthusiast
Joined
Oct 30, 2001
Messages
1,773
Location
Houston, TX
I already typed the long version of this & it dissappeared, so now I doing the short version. I did a run test on 2 heat sinked, direct drive LS LED's useing 3 C NiCads. The 1st test was 8 hrs. It started being about as bright as 10-12 good LED's. At the end of the test, the LS, even hooked to fresh alkalines, was only about as bright as my SL Batonlite. Figureing it was a bad LS, I ran it again with a new LS. This time, the test was only 4 hrs. Now the 2nd LS, with fresh alkalines, is about as bright as my Eternalight.

I know NiCads release current faster than alkalines, but since they are being dirven at their optimum voltage, 3.6V, should this matter. You don't have the overvoltage condition present to cause the over current. Am I missing something here on the relationship of LS/voltage/current??? TX
confused5.gif
 
TX,
I know you will get a more informed answer from someone who really understands this but in the mean time......

I have been messing with the LS's my self and although I haven't tried them with batteries, I have been using a regulated power supply. In my limited experience, what was shown as 3.1 to 3.2 volts would allow anywhere from 150 to 600 mA to pass through the LS. With my Regulated power supply, I could set the max current at 350 mA. It was clear to me that if you had 3.6 volts without any current control, you could easily drive the LS with too much current. I am guessing this may have been the case for you. I believe you need a resistor in your circuit. I understand that the batteries themselves can provide some type of current limit but I don't have a clue how that works.

Can you measure the current flowing through the LS?
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by txwest:
200 mA is as high as my meter will read. TX<HR></BLOCKQUOTE>

Tex, what you describe, when running that Luxeon on 3 "C" Nicads is running at least 700mA+ into that Luxeon for most of the 10 hours.
shocked.gif

My experience makes me believe that's too much for several hours sustained running.

I'm guessing from what you describe that you have some damage
frown.gif
, BUT,-to test the theory out, take 3 "AAA" Ncds/Nimh and hook up the Luxeon to those smaller batteries and see if you are getting what looks like full brightness. That is suggested simply because the smaller cells cannot overdrive the Luxeon nearly as much.

Hope you haven't hurt those Luxeons yet, IMO-3 "C" cells without any resistor can deliver way too much for some individual Luxeons

p.s.-you didn't mention anything about monitoring the Luxeon for HEAT build-up (?)
 
In the 1st hr of running, the head of the light got warm, but nothing I would call hot. So you think the long term run is what may have caused the problem? I have a direct drive 3 Blaster I'm concerned about running now. I bought some 9000mAh batteries for it. Think I would be better off useing alkalines because of the increased internal resistance? How about keeping run times to an hour or less? TX
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by txwest:
200 mA is as high as my meter will read. TX<HR></BLOCKQUOTE>

You've got to measure the current flow. Lots of meters only go up to 200 mA.

But there's a way around the problem.

V=IR, volts = current x resistance

Put a 1 ohm resistor in series with the LS and measure the voltage drop across the 1 ohm resistor. If R=1, then V=I(1). Or a delta V of .35 volts = .35 A or 350 mA.

A 1 ohm resistor is kind of hard to
come by but you can make your own.

In parallel, 1/Rtotal = 1/R1 + 1/R2 + ...

Put ten resistors in parallel that are 10 ohms each and all together they'll be 1 ohm
like the equation above.

Radio Shack has a 1 ohm wire wound resistor but someone was saying that the inductance of the coil gets in the way. So probably don't use one of those.

.35 volts is a lot to throw away. I guess .1 ohms would be better for measuring in this range. Once you get something going with some resistance in series, measure the voltage drop across the LS that makes it draw the amount of current you want it to draw and that'll be the supply voltage you want.

You could also use a variable resistor to adjust the current for testing. Just make sure it's rated for enough power. Power = volts x amps.

Let's see, for the one above, Power = .35 volts x .35 amps = .1225 watts disipated in the resistor.

Did I get all this right? What I'm getting at is to put some resistance in series and then measure the voltage across the resistor and calculate the current. Start with a lot of resistance, then back off until the current is right.
 
TX,

I have one Digital Multimeter that will go to 400 mA or you have a 10 amp scale that is useless. I normally use this meter in line with my power supply so I can monitor in a range of 1 mA scale. When I started messing with the Luxeons, I kept blowing fuses in this meter and it was effecting other things as well.

I haven't been able to find a readily available current meter that can read say 0 to 1000 mA. Seems to me that such a meter would be very useful in fine tuning a circuit for the Luxeons (1 & 5 W). As I have stated elsewhere, there seems to be a significant variation between individual Luxeon units themselves. I fine tuned my power supply for one LS and say the voltage shown was 3.1 and 350 mA was going through the LS. Without changing the power supply, I hooked up another LS and only 180 mA was passing through it. I went to crank the voltage up a bit with the fine adjust and before the display could switch from 3.1 to 3.2 volts, the current passed 350 mA and hit at about 500 mA. Not being able to read 100ths of a volt, I don't know what the required voltage would be but less than a 10th of a volt on a non protected LS can result in significant under or over driven conditions. At least this is the conclusion that I draw.

Hopefully someone can give us some idea of the animal we're trying to train here.
smile.gif
 
It seems to me, with these parameters, it would be impossible to get a LS at the right amperage. How can you even determine what resistance to use. You don't have control over the battery voltage enough to set your other values. TX
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by txwest:
It seems to me, with these parameters, it would be impossible to get a LS at the right amperage. How can you even determine what resistance to use. You don't have control over the battery voltage enough to set your other values. TX<HR></BLOCKQUOTE>

One thing is for sure-espec. for those Luxeon LEDs, you need a DMM that measures at least up to 1 Amp:p
 
If you look at page 8 of the Luxeon 1W- Star PDF file. It shows a very steep curve for current ranging from 100 mA to 500 mA as the forward voltage goes from 3.1 to 3.4 at 25C.

There is no indication of how much this steep curve may shift from one LS to the other or how it may shift or change with other temperatures. I don't know how accurate my $250 powersupply's digital meters are but I experienced even steeper Vf versus If curves.

I just tested two Luxeon emitters I have in heat sync'd fixtures (I measured the voltage across the LE's with a DMM):

LE Volts Current (mA)

#1 2.96 150
#1 3.20 430

#2 2.96 150
#2 3.17 430
#2 3.20 490

I don't know what the temperature was at the framastat but I venture to say that if either of these LE's werre taken to 3.4 volts as the LS data sheet suggests, there would be a heck of a lot more than 500 mA going through them.
shocked.gif


Perhaps a resistor is needed just to stabilize these guys?
 
You guys are scaring me re my hope to run the 5W LS off of 6R cells w/o regulation. Sounds like Tx may be pulling ahead of me in the fried LED sweepstakes.
 
McGizmo wrote:

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR> If you look at page 8 of the Luxeon 1W- Star PDF file. It shows a very steep curve for current ranging from 100 mA to 500 mA as the forward voltage goes from 3.1 to 3.4 at 25C.
There is no indication of how much this steep curve may shift from one LS to the other or how it may shift or change with other temperatures. I don't know how accurate my $250 powersupply's digital meters are but I experienced even steeper Vf versus If curves
<HR></BLOCKQUOTE>

Actually the datasheet is quite informative about the range of possible variation among devices. The I-V curve can be shifted to the left and right so that it passes through 350mA anywhere in the range of 2.79 to 3.99V [ref: page 6 of DS23]. As to variation with temperature, the tempco is -2mV/C [ref: page 6 of the datesheet]. It is this tempco that causes the apparent steeper I-V curve in McGizmo's testing since as higher current is applied the die heats up which tends to lower the Vf from what it would have been if there were not a negative tempco.

Regarding the various advice to txwest on installing a series sense resistor to measure current: For his specific application, the sense resistor must be very low to avoid significantly lowering the current to the LED. Since it appears that he only needs an approximate measure, I would suggest that the sense resistor be only 10 milliohms. Assuming that his voltmeter has a minimum resolution of 0.1mV on the 200mV scale, this will provide a minimum resolution of 10mA. Conveniently, 1 foot of 20 gauge copper wire at 20C will provide a resistance of 10 milliohms. Consult a wire table for different gauges.
 
10 milliohms...WOW, I should be able to stare at it real hard & get that much resistance.
astrosmiley.gif
TX
 
Doug,
Thanks for the info, illumination and suggestions.

Any of my Luxeon Stars and emitters will be driven by Xitanium or Dat2zip drivers
smile.gif


I'm typically not concerned about getting in over my head a bit (simple resistive circuits) but IMHO, the Luxeons need to be tamed and driven by professionals.

TX,

Let us know how you fair.
 
txwest,
About the only help I can give you is that I've never gone below about 1.5 ohms resistance in series with my LS. Using 3 fresh alkaline AA's I usually start out at 500+/-10ma. I've not done any real time tests on brightness, but even at 250ma my LS puts out a very nice usable light.

What gauge wire are you hooking everything up with? My battery pack had 26 gauge wire and I never got above 240ma, switched to 22 and things perked up to that described above.

I would definitely recommended using a resistor in series with the LS to stabilize things. I've been getting my small value resistors out of old tv sets.
 
a quick test with 3 fully charged 4/5a nimh cells direct to a ls showed over 1.6 amps!
this was just long enough to let the dvm sample and stabilise.approx 2 sec.so your direct drive is beating the crap out of them.
1 easy way to tell you are pushing too hard is to watch for a shift in color towards cyan.
 
I've got a couple 3c direct drive luxeons, where the internal circuitry of the flashlight switch adds 0.5 ohms of resistance running at 900-950mah. With 4.8 volts from 4 AA nimh, they were at 1.1-1.2 amps I think, one luxeon was drawing more power than the other though.
 
For every man, there's a woman. So it follows that:

There's a resistor that will make each LS happy. You just have to start out with a higher value and then try smaller ones until you get the right amount of current while the batteries are fresh.

But some men will search for a perfect woman. So it follows that:

For every LS, there's a perfect driver circuit. There are constant voltage circuits and constant current circuits. But y'know, there's always something about each one that's just not quite right.
 
I wonder if any extensive runtime tests have been performed on any of Elektrolumens' lights? I know they're overdriven...

Wayne?
Notice any loss in light output over time?
 
Additional input. I just found out my digital V/O meter has a 10A setting. I won't vouch for accurecy. The single LS is showing to be pulling 70mA. This could be correct, as it's not very bright. Now the 3 Blaster on alkalins is showing to be pulling 1.3A. On rechargeables, it's showing .9A. Does this sound right? TX
 
Top