Lantern Torch LED Conversion

Picture of Lantern Torch LED Conversion
Harness the power of modern lighting technology by converting a standard lantern torch into an LED torch. LEDs provide the advantage of longer bulb life, reduced power consumption (originally 750mA, now 320mA) and reduced heat. This project uses 32 LEDs which provides slightly more light than stock while using under half the juice. I've made a video showing the basic steps involved in making the project. You'll be building a small LED light and installing it in the reflector. This light is then connected to the circuitry in the torch so it operates as per normal. If you're keen, take a few minutes to view then continue on to the detailed instructions. If you enjoyed the video, please consider subscribing to my youtube channel as this helps towards producing more videos and projects.


Remove these adsRemove these ads by Signing Up

Step 1: Equipment

- 1x 6v Lantern Torch / Flashlight (Amazon)
- 5mm White LEDs ( Amazon / DealExtreme ) Any multiple of two. I used 32 LEDs for this one.
- 1 Ohm 1/4W Resistors ( Amazon ) One resistor required for every two LEDs used.
- Disc shaped wooden or plastic object that fits inside the reflector to mount the LEDs. I used a jar lid that I trimmed down.
- Thin gauge wire.

*Affiliate links provided to suitable products on and

- Soldering equipment
- Drill, tape and stationary

Step 2: Disassemble Reflector

Remove the lens from your torch. You may need to remove the bulb before you can get the reflector out. Ensure nothings securing the reflector in as it can be a bit delicate and may break if removed with force. In my case, the lens was attached directly to the reflector and had to be removed also.

Find or make a circular disc like object made from a thin wood or plastic that will fit in the reflector. I used a jar lid with the threaded section trimmed off for some bonus kludge points.
LP21 year ago
This will definitely help me with my Boat spotlight.
Thanks !
but does it boast the same focus as it had previously....i don't think so....i would consider the same task being done with 26k cmd leds in a tighter arrangement so that it fits deeper down the reflector...also you have a plane reflector...not the ones with tiny individual rectangular reflectors arranged within the reflector surface ...i aave one and will modify it but with some design inputs from my side too...
krstn12 years ago
If I wanted to run a configuration of this off the 12 system on my truck, what kind of resistors should i use so not to blow the leds?
profpat2 years ago
nice idea,

i have an old rechargeable lantern that has busted 4.8volts halogen light bulb, i think i can convert it to this kind of design.. i want to use a smd hi intensity single led bulb the one used for gun flashlight!!
レレイ2 years ago
Is it still possible without the reflector? Cause when I opened my old flashlight, it was broken. Is there anything I can do? In exception of getting a new reflector or buying new one.
Dipankar2 years ago
Are you using a rechargeable 6volt battery or Alkaline Battery?


which uses a rechargeable 6 volts 4.5 Amps leak proof lead acid Battery. The Charging is done outside the light casing with a 6 volt charger. Published:Aug 8, 2010.
Its the same game with a different name?
x2Jiggy (author)  Dipankar2 years ago
Just a standard 6v spring terminal lantern battery. You weren't kidding, that flashlight is power packed! Plenty of LEDs there. Nice touch with the rechargeable battery.
smalcolm2 years ago
Nice work, looks like a project for me.... but I wouldn't be smug about owning a prius lol
lool True good project. wish one day I can make one I love LEDs things this for me as well look like a project specially because I dont know NOTHING about electrics lol. Good Work! May be one day I will try this.
radracer2 years ago
Now this is a sexy flashlight.
medionlvr2 years ago
I'm looking at and there are white led's that have a luminous Intensity of 18,000 millicandelas. They have a typical forward voltage of 3.4 volts with a max of 4 volts. Would a resistor still be necessary ? I don't pretend to understand this but it seems to me that a pair of led's in series that operate at 6.8 volts each and i'm only providing 6 volt to the pair i should be under the typical voltage and surely be under the max voltage of 4 volts.
You must ALWAYS use a resitor with LEDs, or it'll burn itself with time (or very fast).
I recommend you to use an online tool for the design, and try different (parallel and series) setups:

Hmmm, I am lead (led) to believe otherwise, by persons that know on various 12v and electronics forums,that is one DOESNT always need to use resistors...but hey ! tell me Im wrong , I like to learn !
x2Jiggy (author)  shocker1872 years ago
One of the big benefits of LED lighting is long life and LEDs are very sensitive to increases in current. Unless you can guarantee the current supplied (usually by building or purchasing a regulated driver) it's good practice to add a resistor or risk an early death for your LEDs.

Some people choose not to use a resistor when their input voltage matches the LED forward voltage, however without a resistor a small change in voltage can produce a large change in current. Resistors are cheap, easy to install and when you add the resistor a change in voltage will produce a proportional change in current protecting your LEDs.
In-ter-esting ! As a digression , the situation I had in mind was a string of leds that were to be fitted to a car as day running lights . I inquired about resistors and was told that since the forward voltage etc ,blah,blah, (what you said above !) I didnt need resistance added.
I have been worried that the current change produced by the voltage range on a 1990s land rover, roughly 12 and a half to 14 and a half volts dc would indicate the need for a resistor and since I cannot find out what even using online calculators I havent fitted them.
If you have a source of power (like a D cell battery) that never exceeds the max forward voltage of the LED you can do without the resistor.

But a automotive electrical system has a constantly varying output as noted in other post here. If I wanted to use LEDs from a automotive electrical system I would buy or build a circuit that limited the voltage and as well preferably limited the current going to the LEDs and included spike protection.

Years ago I connected an oscilloscope to a car battery and noticed that the ignition coil was inserting high voltage spikes into the 12 volt systems of over 1K volt every time the spark plug fired. This was in the days where they used "points" before they switched to electronic ignition systems. So an older car could be really rough on LEDs and other electronics. Most car radios and common auto electronics have built in spike protection to prevent this problem.

Has anyone ever connected an oscilloscope to a modern electronic ignition car to check things like this?
Diode for spikes, cap across neg/pos supply for smoothing ? In my case Im talking diesel,traditional diesel,always diesel.Not common rail or electronic injection . And yeah , the current variation caused by that 2v range....well,doesnt sound a lot until you look at it as a percentage of total.
To respond to your situation, yes I think the diode and cap would work for a diesel.

But I "think" it would NOT be enough for a gasoline engine (of any type). Spikes and short over voltage situations often only "weaken" the transistor based components like IC chips and diodes which then fail after some unknown time period. That is why that after lightning damage to your home, your home electronics continue to fail for months afterwards.

The more times the electronics are subjected to surges the weaker they become. Since a "spark ignition" system, if in bad condition, "could" hammer the diode and other electronics in you car thousands of times a minute I think I would install 24 volt MOVs to reduce the strength of any surges before they reach your surge protection circuitry in this type of car.

MOVs are devices that short out above a specific voltage and revert to a high ohm resistor when the voltage drops. They look like large (even giant) ceramic disk capacitors. They also wear out after X number of high power spikes and have to be replaced. In the case of an ignition system I don't think the power would be high enough to make this a problem.

The time the MOV conducts is limited to the duration of the spike plus a few milliseconds. In the case of spark ignition caused spikes, these only last for a few milliseconds and the voltage during that spike is limited to 24 volts.

Normally 24 volts would "burn out" any 12 volt electronic component. But because this "burn out" is caused by heat generated relative to the duration of the over voltage--"usually" the heat generated by this over voltage during the few milliseconds or less that the spark induced voltage is over 24 volts is not enough to damage the device. But given the thousands of times a minute that your electronics could be subjected to this over voltage it could be possible that the heat build up over time could destroy the LED or other circuitry.

I don't think this would happen in a car where the ignition system was in good shape, but if it had problems and was feeding large spikes back into the 12 volt system it could. Also late model cars (which are dripping with electronics) must have installed something to prevent this or designed it in such a way that this was not a problem.
thanks for all the replies. as this project uses a 6 volt DC battery i guess my question was answered.
If the power supply outputs less than the LED's maximum voltage, you don't need a resistor. But if the voltage drops too much you will lose a lot of light. So it's best to use a higher voltage supply and drop the excess voltage through resistors. The more voltage goes through the resistors compared to the LED(s), the wider voltage range it will tolerate without changing brightness too much. But you waste more power and dissipate more heat that way.
pcooper2 stuuf2 years ago
Resistors or a constant-current source are ALWAYS needed with LEDs. They are current-controlled devices and a have a fairly sharp turn-on "knee" in their V-I curve. Too little voltage means they don't turn on. Slightly more voltage means they turn on and allow too much current to flow, destroying the LED in a matter of seconds, if that long. Relying on the internal resistance of the supply is a bone-headed approach and would get your fired if you were designing such things for a company.
stuuf pcooper22 years ago
there seem to be plenty of "bone heads" building $4 LED flashlights with a PCB full of white LEDs running directly from 3 alkaline cells with no resistor :P
Thanks Stuuf - good stuff ! Makes sense.
To lead on from your post, I have always thought that the main point of using leds (to me at least) was the lower power the need to use resistors to up the load suggests that one would be in a situation where the power consumption was similar to incandescent luminaires , so making it a pointless exercise.
However, with the emergence of better spectrum and brightness of leds maybe the cost/benefit scales begin to swing back to favour them. Maybe.
So I think with my limited understanding, anyway !
It might work. But LEDs are constant CURRENT devices, and working with under-voltages may not generate enough current to drive the LED into conduction. In which case, it won't heat up and burn out. But that's not much use since it also won't light up. And the discharge curve on your battery is such that as soon as you hook it up, the voltage starts to droop, so it's touchy how much time it would be at 6 Volts, anyway.

Since there seems to be a fundamental lack of understanding about this, Voltage into a load (the LEDs and any resistors) generates a current THROUGH the load proportional to the voltage applied. And as long as the voltage is somewhere near (say, no more than three times the rated voltage of the load) but higher than the rated minimum voltage of the load, internal battery resistance doesn't affect the equation, so the formula for current becomes Voltage (in Volts) divided by Resistance (in Ohms) equals Current (in Amperes) (or Amps). And for those of you who are accustomed to the metric system, milli-amps is, as you would guess, thousandths of amps. So one milli-amp is one-thousandth of an amp. You can figure the rest. And, yes, milli- is also applied to volts and ohms.
Your statement about LEDs being constant-current devices is correct. Excess current will kill an LED far short of its expected 100,000 hour life, like Chinese-made LED night lights that fail in just one year, when they should last 7+ years. Most LEDs run on 20 mA to 25 mA, and the current needs to be properly controlled by use of a regulated DC power supply and current-limiting resistor or a constant-current power supply.
I have found through testing that all my LEDs (superbright and standard) work off of 2 standard D cells or alkaline batteries at approximately 3 volts--but will not work off of most rechargeable batteries which usually output less than 1.2 volts each for a max voltage of 2.4 volts

Some LEDs have a max of 4.5 volts like mine. But when I tried 3 D cells per superbright LEDs they were brighter--but not that much (to the naked eye anyway). Therfore I decided to use 3 volts because that was as bright as I needed and power consumption and heat generated was a lot less.

A second advantage of using 3 volts and alkaline batteries was that the lights dimmed noticeably BEFORE it drained the alkaline to the point that it could not be recharged with my "ReZap Battery Engineer" alkaline battery charger.
I've noticed that allot of the tail lights in cars are LED. how hard would it be to us one of them in this manner?
x2Jiggy (author)  robertblacksmith2 years ago
Yep, micahd02 makes valid points. Anything designed for automotive use is designed to handle input voltages of 12v and up as a cars electrical system doesn't provide a stable source of power.

If you wanted to salvage LEDs from something and use it with this project, it would have to be designed for an input voltage of 6v as this is the voltage of a standard lantern battery.
The one advantage of the lantern battery is that its voltage will never be higher than when it is fresh and fully charged. If the current limiting resistor value is chosen on this basis, the life of the LEDs will be maximized.

Using LEDs salvaged from an automotive lighting system carries no disadvantages whatsoever. One must simply adhere to their maximum current rating. Finding out what that rating is could be problematic, though, if the part number and manufacturer can't be identified, which one would need to get the specification sheet.
Those are designed with built-in current regulation to run on the output of a rectified alternator, where the input voltage CAN'T go below 12 Volts for more than a couple of minutes, and, depending on the car, routinely go anywhere from 13.8 volts up to around 27 volts, with spikes into the 10s of thousands of volts. The lower "running" voltages alone make it expensive to design and build them, and make them cumbersome for a light-weight application like this, but the fact that they can't run on less than 12 volts means you would have to use TWO 6 volt batteries to drive them, and those would be heavy for a hand-held torch.
In addition, the voltage spike protection further increases the cost, and isn't necessary at all for this application. I doubt that it would be worth even considering using one of those for this.
Yes I see car voltages from 11. to 14.5 at least. So I am thinking a voltage regulating system using a 555? chip, they were used to regulate V. in several circuts I believe. It seems this would be perfect and maybe eleminate the resisror? ....
HELP please! I need to put in lots of lights in my camper; and cant handle the heat and huge power drain of regular lamps. Thanks.
pcooper22 years ago
x2Jiggy, LEDs come in all kinds of body styles and lens configurations. Did you choose narrow-angle LEDs, wide-angle LEDs, or just-any-old hobby-grade LEDs?
acastonguay2 years ago
isn't there something that can create a movement or a reaction, when it gets heated, using stronger led's, heat's up, the heat could generate power, that power in turn recharges the battery.. hummm...
Nice idea :-D, but (quite apart from the law of diminishing returns) I doubt there's anything mechanical that could provide enough power from such heat to make it worthwhile attaching (to say nothing of the initial effort of creating (and/or designing) it in the first place!) - But I'll leave that as an open challenge.

I do have a vague memory of radios being powered by electricity generated by two (bonded?) strips of metal heated by a candle, but the fact that such things aren't common (and they weren't ever common - just created to illustrate some point or other) suggests they're not considered that efficient. But Trevor Bayliss ignored such concerns when he created the first wind-up radio, so maybe someone reading this will take up that challenge...
That was called a Thermo-couple. They were very inefficient, begged to be overheated (after all, if I could get 1 watt out with a fire, maybe I could get 2 by burying it in the fire-place, and 10 with a welding torch, right?), and even if not over-heated, they were very fragile, and the bonds tended to break in ANY contact, vibration, or any other mechanical shock.
Peltier Junctions work better, and stand more heat, but the heat-sink required to make them efficient is large, and would have to get very hot to push much current.
Mechanical work still does the best when it comes to portable power.
But solar cells might work, if you have enough of them, and enough space, and the sun shines long enough where you would use them.
Trade-offs make life difficult, no?
Indeed they do!

Thanks for the info on the Thermo-couple - I never paid more than passing attention as a kid.

I don't know anything about Peltier Junctions, other than what you've written, above, but whether I'll investigate any further is highly debatable...

As far as solar cells go, if you're building an array, it might be worth incorporating a dedicated charging circuit for something like Dipankar's SLA-powered LED lantern (see my answer to Kiteman, below). But, as you say, you're unlikely to get enough cells, space AND light for a lantern-mounted (or, even, portable) system that will reliably charge it well enough for anything other than occasional use.
Peltier Junctions are really neat but they are horribly inefficient.

The simplest description of them is a "heat pump" which can also generate electricity if heat is applied.

Apply electricity and heat is pumped from one side of the device to the other which requires lots of cooling fins to keep it from melting the junctions.

Apply heat to one side and current appears across the wires but for this to produce very much electricity you have to "cool" the other side in some manner.

They are often used in car and RV refrigeration situations where their power consumption is not a problem. An example is one of those plastic beer coolers that plug into the cigarette lighter.

Using "pump" to describe this is not technically correct but pretty well describes the results.
Hmm - not entirely surprising - Micahd02's mention of the large heatsink required sort of implied the basis of what you say.

It really is a pity there isn't some modern equivalent of a steam turbine that would be small enough to mount on this lantern, to re-convert (some/most of) the heat into electricity to (partially) recharge the battery. I mean, I remember how fascinated kids used to be by holding a spiral (cut from a circle of paper) on a piece of cotton, over a radiator, and watching it turn with convected heat... :-)

I also remember realising almost immediately that the slightest friction would bring it to a stop... :-(
That's the concept of a Perpetual Motion Machine. Unfortunately, those go against the laws of thermodynamics, and have been proven impossible outside of self-enriching nuclear reactors (which don't last forever, either).

Or, if you may understand it better in economical terms, is the equivalent of a never-collapsing Poncy scheme. Impossible.
Richone2 years ago
I have also used tihis method of replacing the bulb in lanterns.One thing I did different was to point the outside bulbs toward the reflector so I got a better light spread.It works great!
keithwd Richone2 years ago
That works...

Another solution is to use larger diameter LEDs. I have been experimenting with 10m,m superbright LEDs which have the same internal sized and power as the smaller diameter ones. (Bought from American Science and Surplus)

The difference in the size allows for a larger lens which spreads the light more giving you a wider and more diffuse pattern.
x2Jiggy (author)  Richone2 years ago
That's an interesting idea. Will have to give it a test in future LED projects. Although 5mm LEDs can provide a strong light source, it's more focused compared to the stock bulb and reflector.
x2Jiggy (author) 2 years ago
Really pleased with the positive responses and suggestions for improvements to this project. Looks like a version 2.0 with some additional features and refinements could be the go in the near future. Thanks and happy tinkering!
Richone2 years ago
Also,Go to your local recycling places and look for the Walk and dont walk signs.They will sell them to you,because they are plastic and can't recycle the plastic.
-max-2 years ago
to make the torch super useful, stick a power led in it. i recommend cree t6! that thing is super bright! (its 100 lumens per watt @ 1000 lumens [max] )
Kiteman2 years ago
Nice job.

Next challenge: replace the lantern battery with a re-chargeable battery and a winding charger.
x2Jiggy (author)  Kiteman2 years ago
Cheers. That's an excellent idea. A friend wanted a fishing torch that provided good light with long battery life and a stand so I converted this one. Not having to carry a spare battery as well would be a plus.
Kiteman x2Jiggy2 years ago
Another thought I just had - in another iteration, paint the drilled disc a shiny silver before you fit the LEDs, to reflect that little bit more light forwards (the original reflector won't help much, since it's shaped to focus light from the original location of the bulb, well behind the LEDs).
Reflectance is useless with LEDs. They are, by nature, a point-source and highly directional emitter. No light comes from the back, unlike the hemi-spheric propagation of conventional incandescent, or even CFL bulbs. It all comes out one side of the emitter, and is all focused, by its very nature, in one direction. In fact, if you look at the lense of a conventional flash-light housing, it is flat, to avoid spreading a naturally spreading beam. But the lense of any commercially assembled LED flashlight is already domed, specifically so that the beam WILL spread. Otherwise, you get a tightly focused beam that is all but useless in the semi-flood-lamp pattern usage of a flashlight.
Er, by definition, a point-source is a spherical emitter.

Look at the side of a lit LED - you can still see the emitted light.

Point the LED away from you - you can still see emitted light.

Yes, the shape of the casing focuses most of the beam forwards, but a non-reflective backing to an LED will waste some of the light.

(We have a number of LED torches - all but the very, very cheapest (the cardboard one) have reflective, beam-focussing linings to the cavity containing the LED(s).)
Most of the sets of rechargeable tools (drills, saws, etc. from Royobi, Hitachi, Bosch, etc.) also come with a flashlight, that runs off the same rechargeable battery pack (usually 18V). Add to that, the 18V bulbs are hard to find, expensive when you do, and apparently have a lifetime of only a few hours. This makes them a natural for conversion to LED's. With the led's used here, you'd probably make them into strings of six.
As far as rechargeable batteries goes, Dipankar used a 6-volt, 4.5Ah Sealed Lead-Acid battery in his Instructable, here:

Or you could buy this: from Amazon (UK), and perform thi Instructable on it...

OR, if you can find one with a cage for 4xD cells, you could simply buy rechargeable D cell batteries. Of course, this means you'd still have to carry spares.

And as for a winding charger, a look at Yewfort's recent Instructable here:
may help (I'm just not sure enough about voltages, etc to know if it'd work as-is, or need beefing up).
ijakidi2 years ago
Muy bueno el proyecto, lo tendré presente para cuando reúna los materiales, creo que podría mejorarse la luminosidad si se tapara la parte trasera de los led, para que no escapara luz para el interior de la linterna. y como ya se dijo mejorar el reflejo la base de los led con pintura o papel aluminio.
Nice work! I think you could use less resistors, in fact, just one could be enough but It would dissipate a lot of heat.
For this array, I'd rather use 4x22 Ohm 1W resitors, saving a lot of soldering work and space. The schematics sholud look like 4 arrays of 8 LEDs in parallel. What do you think?

Spokehedz2 years ago
One of the driving circuits on DX would have let you run these all in series making the build a heckuva lot easier.
carlos66ba2 years ago
Nice job, but you will get more bang for the buck if you use a smaller number of high-power LED's (from dealextreme too).
x2Jiggy (author)  carlos66ba2 years ago
True. I wanted to keep this one simple and not have to worry about heat dissipation and building a decent driver.