Rechargeables in a hyper-cheap chinese flashlight

Some time ago, I bought for the car one of those cheapo chinese flashlights that would give the shivers to any self-respecting flashaholic.
Hey, it was €4.
Two pictures follow of the flashlight in its original state:



Its main problem, aside from looking like being made out of concentrated cheapness, is that it uses three AAAs in one of those triangular holders.
I hate AAAs. They are more expensive than AAs but hold less energy, and all for a size difference that is, to me, negligible.
So I took it apart with the intention of building a 3-AA holder and using rechargeable NiMH cells.

However, in the process of taking it apart, I discovered there doesn't seem to be any kind of resistance whatsoever between the cells and the LEDs.



My question is: can you use rechargeables in this light? I suspect it's using the internal resistance of alkaline cells to avoid its LEDs going into thermal runaway, but if that's the case the much lower resistance of rechargeables could turn all the LEDs into friodes... this means it'd probably go pop even if I used rechargeable AAAs...

I also checked the switch for resistance; it has none, it's a simple interruptable contact.

Usually there is atleast 5 springs between batteries and led, there might be enough resistance in those.
Good part whith these cheap lights is that if you fry the leds, it does't matter.
I have used mine with reghargeables, but I do measure the current draw and if it's around 30mA per led it's OK.

I also bought some 4$ for 4x 900mAh nimh's just for these

Juha

I have a 3AAA 8LED cheap light and measured approx 400 ma using nimh fresh off the charger which basically would fry all but perhaps nichia LEDs in awhile. I ended up replacing half the LEDs before I put a resistor in it and changed the LEDs to better ones.
If the LEDs are not matched well (Vf) then a few will get more than just 50ma and fry rather quickly thus diving the current among the remainer which would increase for each of them. I think I used a 2.5 or 3 ohm resistor in it.

It will probably be all right with NiMHs, but I doubt you will get 3XAA in there. Simpler soution (if you have them around) would be an 18500 li-ion. I use them anywhere a 3xAAA carrier lurks (cheapy chinese multi-leds and Dorcy Metal Gear).

Three rechargeables will give you 3.6 Volts, isn't that just about perfect?

Thanks for showing the pictures. I have two of these coming (for a total of $4.99 free shipping). If the LED's are really lame, I might throw something better in there. It's nice to know it's easy to disassemble and solder.

I won't get 3xAA in there. I'm planning to use the head and the switch and throw the rest away. The main body is going to be a 3xAA battery holder I'll fabricate out of this...



Its size is almost perfect to hold 3 AAs. A bit too large, but some duct tape will take care of that.
It's gonna look pretty ghetto, but I don't care.

The LEDs are of the cheap and nasty category all right, but they aren't quite as nasty as I expected them to be. I have another almost identical light that gives a noticeably bluer and dimmer light with the same cells.

Question: considering a voltage of 3.6 and 12 LEDs, what resistor should I put in series inbetween the head and the battery holder to prevent the LEDs from going pop?

Fallingwater said:

Question: considering a voltage of 3.6 and 12 LEDs, what resistor should I put in series inbetween the head and the battery holder to prevent the LEDs from going pop?

Click to expand...

Since nobody has answered, I'll tell you what I would do.

3.6 volts is less than the 4.5 volts the manufacturer designed the flashlight to, without a dropping resistor. So hooking up NiMH's shouldn't do anything bad.

Even so, hook up an ammeter in series with the battery to measure the average current, just long enough to get a stable reading. It should be around 240 mA on the NiMH, so use the 500 mA or higher scale. If you read more than 300 mA, you may want to consider a dropping resistor.

Nimh hot off the charger can measure 1.4v or more, approx 1.2v is the average voltage throughout their useful charge and with their lower internal resistance they can deliver more current than alkaline batteries in LED lights, you will need a resistor to keep all but nichia LEDs from frying off 3AAA nimh batteries over time.
The easiest way to determine current to the LEDs is to get a meter, put it on the 10A scale and measure current inline. Lower scales will affect your reading more than the higher amp scales. I sometimes use a variable resistor of about 25 ohms to figure out what resistance is needed for LEDs. I recommend approx 30ma max/LED with NIMH hot off the charger. This will give you the longest brightest runtime. If you want to increase runtime drop to 20-25ma/LED.

I would agree that you need to limit current to 20 ma per LED (0.24 amp total in this case). I have melted plastic reflectors even at only 25 ma per LED. Depending on the LEDs and their particular optimal drive voltage, even with NiMh you can exceed 20 ma/LED without a resistor.

Lynx_Arc said:

Nimh hot off the charger can measure 1.4v or more, approx 1.2v is the average voltage throughout their useful charge and with their lower internal resistance they can deliver more current than alkaline batteries in LED lights, you will need a resistor to keep all but nichia LEDs from frying off 3AAA nimh batteries over time.
The easiest way to determine current to the LEDs is to get a meter, put it on the 10A scale and measure current inline. Lower scales will affect your reading more than the higher amp scales. I sometimes use a variable resistor of about 25 ohms to figure out what resistance is needed for LEDs. I recommend approx 30ma max/LED with NIMH hot off the charger. This will give you the longest brightest runtime. If you want to increase runtime drop to 20-25ma/LED.

Click to expand...

Soooo... considering I'll run the light from AA cells (and not AAA) runtime isn't a problem.
If I want to give it 30ma per LED, what resistor do I need inbetween the battery pack (3x NiMH AA) and the LED head considering it's got 12 leds?

I've got exactly the same lights.

3AAA nimh's work fine.

And there is some resistance through the springs.

Fallingwater;

I've converted several lights from 3 AAA carriers to a single #18650 lithium ion. The length is just a little longer, but usually the negative spring post can be reformed.

The voltage (4.2) is just under the original 4.5 VDC, so no resistor should be required.

Larry Cobb

Using 3AA nimh will give you slightly lower internal resistance than 3AAA, as LED Vf varies my estimate is about a 1.5 ohm resistor on 12 LEDs based on my using a 3 ohm resistor on 8 LEDs for approx 25ma/LED. It is possible like someone else said that springs and wiring can make up some of that resistance.
I have a 32 LED light that uses 3AAA and can draw nearly an amp of nimh

Some cheap LEDs can take 40ma but the main problem is when you are driving all the LEDs at an average of 40ma some may get closer to 45-50ma as the Vf varies. One way I bin cheap LEDs quicky is to hook up a few of them slowly turning up the power to see which ones are the brightest and match them up together You should also look to see if when batteries get low to some LEDs go dark when others are still brigh, this will show you differing Vf of the LEDs in your light and can be more of a concern when overdriving them as the brighter ones will most likely fry first when you get around 40-50ma/LED or so.

In rather un-scientific testing, my Dollar Store 3AAA 3LED light pulls about 145Ma on 3 Heavy Duty AAAs. I just tried a 17500 I have and got 18XMa.

I don't use that light much anyhow, so I won't worry about it.

But it does show that higher current CAN hurt!

I just received my 9-LED direct drive 3AA LCKEBYs today from Fifth Unit. One is more blue than the other, so I pulled it apart.

Using 3 yesterday-charged Sanyo eneloop AAA 2000mA NiMH, I checked open circuit voltage at 4.14.

Hooking up directly to the stock board, the voltage dropped to 3.70V and the current ran 700 mA. That's 78 mA per LED. 2.6 Watts total.

On a breadboard I prepared 9 JELED 50,000 mcd LED's in parallel, just like the board is wired. Applying the same voltage, it pulled 460 mA at 3.80V. That's 51 mA per LED. 1.7 Watts total.

I only ran these tests for a minute or two.

The JELED's are ugly by themselves, but packed together in a row they throw nice white light. In the circle arrangement they should be awesome.

I'll build this light up with the JELED's.

When no dropping resistor is used, the power used by the LEDs only depends on the internal resistance of the batteries (low for NiMH) and the LED themselves. You almost have to test to find out what you get.

I wouldn't dare put lithium primaries in!!!

I've read a bit about LEDs and current and resistance, and I've also managed to find a few online calculators for resistor values.
They all gave me this.

Since I have 12 LEDs connected in parallel, I just multiplied the desired current.
Now, if I wanted to feed them 30ma that would be the resistor to use.
But I understand driving a standard white LED at 30ma might actually be a significant overdrive.

My question is: without having the original data on the specific LEDs I'm using, and knowing they are of the cheap chinese variety, what is their optimal running current? The one they are meant to run at without being overdriven?
20ma? 25ma? 30ma? Or what?

I don't want to overdrive them because they have no heat sinking to speak of.

Edit: I measured the current drain with fresh AAA cells, and it's around 500ma.
That's a lot! It means more than 40ma per led!
Are they being seriously overdriven, or have I got something wrong here?

Bump

Another question: which voltage should I give the LEDs? 3.3 or 3.5/3.6? The resistor value changes in that calculator with the voltage change...

The voltage depends on the LED and the current you want to use it at (see "white LED testing" thread). Do you see the LED you want to use there?

With 3 AA batteries, V(batt) is so close to V(f) that its hard to calculate exactly what you'll get. Especially at higher currents where the battery voltage drops.

The best thing you can do take the LED you want to use and hook it up in a circuit where you can measure current and voltage (having two DMM's helps!) Use resistors with progressively less ohms and plot the current vs. voltage.

Before you stuff all those LED's in the flashlight, hopefully you can test them outside using a breadboard and some test leads with alligator clips. Be sure to connect the tailcap in your circuit, since some resistance is present there.

For one of my LCKEBY flashlights, I ended up using 20 ohms in series with each LED. If I used one resistor for all, it would have been much lower value and a higher wattage.

EngrPaul said:

If I used one resistor for all, it would have been much lower value and a higher wattage.

Click to expand...

Why is this bad? (sorry, I'm kind of a noob at this. )

I think -- if you use just one resistor ---- then if one LED fails then all the electrons flow through the remaining LEDs, more than you want, which makes each of them more likely to fail, and so it blows one after another.

But I'm a noob too (grin). Hoping for correction.

I assume a resistor is just a waste of power turning it to heat, but there's no other way easily to drop the voltage until you get into the electronics-regulated devices.

I don't know if five little LEDs with five little resistors wastes any more or less than five little LEDs with one bigger resistor.