im lost and need help

[notos&w]

i want to do a luxeon conversion into a maglite 3D, mainly for fun. i have almost no experience w/ electronics but can follow directions well.

i have read all the information at these two sites but am not sure this is the right info for my conversion.
http://lambda10.tripod.com/3dmag.htm
http://az123.com/LED/3DLSconv/

i plan to get my luxeon here: http://store.yahoo.com/flashlight/arcls.html
which is 1W.
should i get a 5W? is that the right type of luxeon?

can someone please give me simple instructions on what i need to do as far as resistors, heat sink, optics, etc.
im pretty sure i can follow those directions at those two websites for construction as long as im not overdriving the thing and blow it up.
of course id like this thing to be as bright as possible but i dont care about being the brightest.

i apologize if this subject has been beaten to death but i did a search and didnt turn up too much i could understand.
thanks

 

[Rothrandir]

the heatsinks are available through lambda...just drop him an e-mail and tell him you want a couple.

that page shows you almost exactly how to do it, and there is no resistance required as a luxeon can take the overdriving of 3d cells.

the only thing i would change: bother removing the anodyze, this does not effect thermal transfer.

you can use either a batwing or a lambertian 1w emitter, i don't know ofhand if the ls's from arc are emitters. www.futureelectronics.com sells them, but call them at their 1-800 number to get a better price.

you could also use a 5w, but you must use more batteries, 6 is optimal.

you need no optics, as the reflector focus the beam much nicer in my opinion.

that being said, i follow almost the exact same instructions, and i prefer the 1w ld to the 1w hd.

if you want to use a se (whiter, a little brighter, but worse beam) you have to do some things a little differently. if you would like more information on that, i will tell you later.

 

[Silviron]

I'll admit my first 3D-LS conversion is a bit crude by today's standards, but what do you mean it "goes fizzle"????

The one I made has well over two thousand hours (including many days of continuous operation) on it with no problems, nor has anyone ever mentioned to me any problems in following my design, as long as it was followed completely.

Could it be done better? Sure, but as far as I know it is the easiest, cheapest and fastest LS/O mod. You don't need any special parts or equipment; except for the LS, you can get everything you need at Walmart and Radio Shack for under $23.00, INCLUDING the cost of a soldering iron, solder and needle-nose pliers if you don't already have them.

Even though this light was just put together in a hurry as an experiment so to be the first LS mod to be posted here at CPF, it was a light my father used every night until he passed away last month, and I still occasionally use it myself.

Edited: simple mutual misunderstanding

 

[INRETECH]

The most important piece of gear that you can own, is a current-meter

LEDs are CURRENT devices, not voltage - the most current, the brighter they get; until the ineff of the device which generate heat are not able to be removed from the junction, and then it damages the device

The LS wants 350ma - but if you are running without a heatsink, you might want to lower that current

If you are using 3xD batteries - this will put out 4.5v; you need to get rid of approx 1v

Use some small value 1/2w resistors and get the current close to 350ma; my experience is that 2.7 ohms is very close; 3.3 would be a more common value and give you slightly less current

 

[Rothrandir]

please forgive me silviron, it was certainly not intended as a snide comment. think of it rather as ignorance on my part

. i looked at the first picture of the link and closed it out thinking i had already seen it, and that it was a light i had seen on another page that used 8 or 9 nichias, and i thought it was notos&w intention to duplicate the mod with a luxeon star instead of nichias. i was concerned with heat tranfer because i thought the instructions would only provide information on the nichias and bypass the important issue of heatsinking.

i would also like to, complement you on your creation. after reading the instructions and details you provide, i see that this mod is well thought out and took a lot of time and effort to write such a detailed instruction. again, please forgive my ignorance.

 

[Silviron]

Oh, OK !

Your posts here have always struck me as well informed and well thought out

.

Therefore I have always respected your opinions, so I didn't know where that earlier comment was coming from

....

I don't remember posting any instructions on multiple Nichia lights here or on my own web-spaces, so it must have been someone else's light you were thinking of....

We are OK then.

 

[evil.mooo]

Ummm, just about the whole thing with LED's being "Current devices".

By increasing the current, do you not have to increase the potential difference across it as well? Or does the LED's resistance change as more current is fed through them to keep the voltage about constant?

Sorry if this is a stupid question

 

[Doug S]

Originally posted by evil.mooo:
Ummm, just about the whole thing with LED's being "Current devices".

By increasing the current, do you not have to increase the potential difference across it as well? Or does the LED's resistance change as more current is fed through them to keep the voltage about constant?

Sorry if this is a stupid question

<font size="2" face="Verdana, Arial">Not a stupid question at all! The voltage does increase in response to the increased current but the increase tends to be very small compared to the increase in the current. In some special cases, it is actually possible for the voltage to drop as the current is increased! This can occur if there is poor heat transfer from the device and the device heats up. At a given current, voltage will drop with increasing temperature. This effect can overwhelm the effect of increasing current tending to make the voltage rise slightly. The fact that the voltage varies with temperature is one of the reasons that it is better to regulate the current rather than the voltage.

 

[INRETECH]

A semiconductor device, such as a diode or LED resistance does change depending on the current across it; it is not a linear device like a resistor - its more exponential

Each type of device will have a different Voltage/Current curve from another

Ohms law will get you close, since the datasheets for each LED will list its typical operating voltage

Take the input voltage, subtract off the typical operating voltage and then divide by the current - this will get you "close"

Then, using a current meter adjust the resistance for the desired current; I have a 7 ohm 10w variable resistor that I use and once I adjust it where I want it to be, I measure the resistance and use the equiv resistors, or sometimes I use a bunch of 2.7 ohm 1/2w and just try series/par arrangements

If you were to setup a bench power supply to a fixed voltage and then try several LEDs - even thou they show in the datasheet to be at the same typical operating voltage, you will get a different current for each one - since the curve of each one is different

Once you start to light up an LED, do not worry about the voltage; monitor and adjust the current

Each semiconductor has a "knee" or Threshold voltage - until you reach that voltage, the device will totally ignore you; then after that - the device will suddenly "ramp" up in current

 

[webley445]

Now you see this is where I start getting confused. You have two postings above, one saying you can drive it with 3 D cells, the other saying you'll need a resistor. Which is it? How do you determine if an led can or cannot take the 4.5volts?

 

[Rothrandir]

the led can be driven as high as thermal consideration allows.

if you are using a nice big metal (like a 3d mag) body, then you can drive it higher (like direct drive off of 3d batteries) if you plan on using a non heat-conducting body, or using a lesser heatsink, then it is a good idea to use resistance (like in silviron's example

)

it depends on what you want and can afford.

 

[webley445]

Originally posted by Rothrandir:
the led can be driven as high as thermal consideration allows.

if you are using a nice big metal (like a 3d mag) body, then you can drive it higher (like direct drive off of 3d batteries) if you plan on using a non heat-conducting body, or using a lesser heatsink, then it is a good idea to use resistance (like in silviron's example

)

it depends on what you want and can afford.

<font size="2" face="Verdana, Arial">And this is refering to using LS leds correct?

 

[Rothrandir]

yes it is.

other leds can be overdriven also, but not nearly as much as the only place for the heat to go is down the "legs" the reason the luxeon can be driven so hard is due to the slug at the bottom, it dissapates the heat much more effectively than the "legs" of a standard nichia.

 

[EMPOWERTORCH]

I agree...I've got round the problem of heat dissipation in 5mm LED technology by making sure that I can solder my cathodes to a large lump of metal. In my direct driven 2-Cell mods, I've managed to heatsink effectively so that the LED legs only bcome mildly warm... the key is to keep the cathode lead as short as possible. Most of my 5mm mods use plastic bodied torches but I always make sure that there's a good mass of metal to which I can attach my cathodes to. Metal bodied torches are essential for Luxeon conversions though! In most luxeon mod torches I've seen, the entire torch body becomes a heatsink and will warm up after extended usage.