Why are we trying to power 1.5V lights with Li-ons

[LEDninja]

It seems in every thread of 1.5V Cree -XRE lights, someone wants to know if they can use a Li-on battery in the light. Why are we trying to power 1.5V lights with Li-ons? Should we not just buy a light designed for a 3V Lithium/3.6V Li-on?

All this discussion about 10440/14500 batteries in AAA/AA lights prompted me to dig up Newbie's research into the Cree XR-E again.
http://candlepowerforums.com/vb/showthread.php?t=138503
The 1st chart shows Input voltage vs. current.
3.1V = 350mA nominal 1W
3.29V = 700mA nominal 3W
3.38V = 1000mA max current Cree has qualified the XR-E
3.6V = 1850mA almost double the max allowable current of the XR-E
4.2V is way off the chart into magic smoke territory

I believe the 1st 'production' XR-E light the Aleph A19-XRE has a boost/buck circuit. It will limit the current to 850mA even if the input voltage is 4.2V.

More common are boost circuits that shut themselves down and let the battery direct drive the LED. So at 3.6V we are driving the LED at double what it is rated for and at 4.2V.....

Worse still are dumb boost circuits. They keep boosting so 3.6V becomes 7.2V and 4.2V becomes 8.4V. Yikes!!!

-----

It is a little better with the higher Vf of the Luxeons. At 3.6V the Lux is only driven to 350mA and 4.2V 1300mA.
Under load the voltage may drop a bit and current goes downto 700-1000mA. Can someone with knowledge of Li-ons under load comment on this.

 

[mudman cj]

Well, as you probably know we all want more runtime! Li-ions have about 70% more energy compared to Nimh when drawn at single LED currents. And don't forget the ultra low self discharge! Yeah, hybrids like Eneloops are great, but they take an additional hit in capacity (about 20% less than standard Nimh) and still can't beat the self discharge of Li-ion.

I fully agree that it is not a good idea to run a Cree XR-E at around 2 Amps, or even 1.85. I would think that direct driving a cree off of a 10440 would put it in the 1.85 Amp range due to voltage sag, because smaller cells typically sag more at a given current, but Li-ions tend to hold at 3.6V. You wouldn't want to do that though, because that poor little 320 mAh cell would be running at almost 6C. Even the 14500 should not be used at more than 1.5 Amps.

I have driven Lux IIIs directly with RCR123A batteries, and with a fresh cell (4.2 V open circuit) it only drops to about 4.0 V at a current of about 1.2 Amps. Can anyone offer data for the smaller cells?

Provided the cell wouldn't run over 2C, manufacturers should make single cell AA and AAA lights work with Li-ion cells. Of course, that only increases sales among the likes of CPF, your average Joe doesn't know Li-ion batteries even exist. But, they could bundle them with a protected Li-ion cell and cheap charger to brag on the long runtime to the general public (but still offer it alone for the rest of us).

 

[chevrofreak]

Well, I don't see any possible way a 10440 Li-Ion will provide over 3.6v under the load of an L0D CE on high, so I really doubt it will be near the danger zone for the LED. It might get very hot very quickly, but you can always kick it down to a lower level. That's the one good thing about PWM, you'll get all the brightness levels.

The great thing is that lights light the L0D CE and L1D CE will run on any cell you put in them, be it Ni-MH, alkaline, Lithium or Lithium-Ion.

 

[cheapo]

the mah to the led depends entirely on the circuitry.. i think the numbers you have might be direct drive numbers, but the circuitry will limit the current to the led.

 

[HEY HEY ITS HENDO]

Remember there is also a weight issue, li-ions are so much lighter 🙂

 

[paulr]

The reason is they want very small lights running at overdrive or near-overdrive levels and there are no very small lights designed for li ions. So they use 1.5 volt lights instead.

By my calculation though, li ions have about the same energy density as nimh cells. 10440 cell = 320 mAH @ 3.6 V equiv to 960 mAH @ 1.2V.

 

[2xTrinity]

By my calculation though, li ions have about the same energy density as nimh cells. 10440 cell = 320 mAH @ 3.6 V equiv to 960 mAH @ 1.2V.

One difference here though is in a high-drain application like an overdriven flashlight, the LiIon will lose FAR less of that energy due to internal resistance. Let's say you want to power a flashlight that consumes 3.6 watts, that would be 1amp on the lithium and over 3 amps on the NiMH (taking into account voltage sag). IF the battery could even sustain such a current, the combination of the higher current and the higher internal resistance of the NiMH chemistry means it will not be able to deliver nearly as much as the lithium.

 

[Lynx_Arc]

All that and the ability to use normal 1.5v cells when lithium ion are not available too.
A light that can only use 3v+ cells requires you to have those cells around to use it while those using standard alkaline and nimh cells you can find those batteries easily around the house in quantity. If you made a light that could use AA, AAA and A sized cells (123A) in alkaline, lithium and lithium ion you would have many choices available to use in an emergency.

 

[WildChild]

The reason is they want very small lights running at overdrive or near-overdrive levels and there are no very small lights designed for li ions. So they use 1.5 volt lights instead.

By my calculation though, li ions have about the same energy density as nimh cells. 10440 cell = 320 mAH @ 3.6 V equiv to 960 mAH @ 1.2V.

That's what I wanted to reply about capacity! 10440 cells are not as evolved as the 18650 cells for the Wh capacity! For now they have approximately the same as commonly available cells. Now let's take the exemple of the L0D I have for which I just got a set of 10440 Li-Ion cells. True, this light can run on standard Alkaline/NiMH/Lithium batteries. On Li-Ion, it is much brighter but there is a drawback... Runtime is cut to approximately 1/3. This light has been tested by chevrofreak to run about 3h on primary, 6h on low and 1h on high with NiMH batteries. On 10440 Li-Ion, some unscientific tests done by me with no light meter gave me around 50 minutes on primary and 17 min 30 on high before getting a drop in output. Someone tested the runtime to be 2h on low with Li-Ion. You see the difference? More, I wouldn't consider the 10440 as lighter than a NiMH cell. The weigth difference between a Duracell 1000 mAh and a 10440 cell I have is only of 3 grams. You cannot really notice the difference by hand. But there is the advantage that every levels are brighter and that Li-Ion cells won't freeze in Canadian winters 😉.

The nice part of having a flashlight that accept Li-Ion as well as standard AAA cells is the versatility. You have the choice! For the L0D I can use it with Li-Ion to get a small and really bright light. Since they are rechargeable I don't really mind the reduced runtime. And if I want better runtimes but lower output, let say for a camping week-end, I'll only have to use NiMH batteries!

A picture of brightness difference between NiMH and Li-Ion can be seen there. This is my setup, for those interested:

http://picasaweb.google.com/picasawildchild/L0DNaturalNiMHLiIon

 

[paulr]

3 amps is not a big deal for a NiMH AA cell, even a high capacity one. The high current NiMH AA cells can produce 10+ amps without sweating. The obstacle to a tiny, powerful, NiMH powered 1-cell light is making a dc/dc converter that powerful in such a small space, from what I understand. But they could certainly make a small light designed for use with a li ion cell instead of a NiMH, so it stayed in regulation across the range of li voltages.

 

[mudman cj]

WildChild - It would be great if you could measure the current draw on the 10440 in your L0D. While the cell is not giving you longer runtimes, you mention that it is much brighter, and after looking at the pics I have to agree with you! I would like to know how hard that thing is being pushed on high, but that is beside the point and more difficult to measure than the cell current (although it may just be direct drive in this case). With information about the cell currents for both cell types, we can compare the cell energies in your application. Cell voltages under load would also help avoid assumptions.

I also saw some 500 mAh 10440s over at DX, but they probably don't deliver as promised in a flashlight going for max lumens. Even so, they ought to deliver more than 320 mAh. The title for highest capacity seems to go to Li-ion even for the small cells...but data will be more convincing than opinions, so I will stop going on about mine.

 

[WildChild]

Those 500 mAh UltraFire 10440 have been tested around 300 mAh. I got my 10440 from AW. As for the current measurement, I don't have any easy way of testing it. Sorry! But if I do (320 mAh / 17 minutes) * 60, and considering that the LED is probably direct drive with Li-Ion, the LED gets 1.129A on High.

WildChild - It would be great if you could measure the current draw on the 10440 in your L0D. While the cell is not giving you longer runtimes, you mention that it is much brighter, and after looking at the pics I have to agree with you! I would like to know how hard that thing is being pushed on high, but that is beside the point and more difficult to measure than the cell current. With information about the cell current, we can compare the cell energies in your application.

I also saw some 500 mAh 10440s over at DX, but they probably don't deliver as promised in a flashlight going for max lumens. Even so, they ought to deliver more than 320 mAh. The title for highest capacity seems to go to Li-ion even for the small cells...but data will be more convincing than opinions, so I will stop going on about mine.

 

[LEDcandle]

I direct drive a Cree P4 in my plastic anglehead with a 18650... batt draw when fully charged is about 900 ma, not sure how much is going to the LED, but I'm sure the resistance in the 'cheap' contacts and springs is helping.

I've run it for 10 minutes continuous, though the output drops (as tested by DIY light box), it doesn't seem to be magic-smoke territory yet. If it was in an aluminium light, I'm sure it would perform better.

I guess I'll find out how long it lasts... Li-on direct driven Cree in a plastic host... haha... but I may be dumping in a resistor for more 'manageability'....

 

[paulr]

I think UnknownVT measured the 10440 current draw in the L0D CE at around 1.2 amps, somewhat over the 1 amp specification, and also heating the light up very fast.

 

[TORCH_BOY]

Some people like to Soup their cars, others their computers, some their bodies,
Us we like to get tho most out of our lights

 

[WildChild]

WildChild - It would be great if you could measure the current draw on the 10440 in your L0D. While the cell is not giving you longer runtimes, you mention that it is much brighter, and after looking at the pics I have to agree with you! I would like to know how hard that thing is being pushed on high, but that is beside the point and more difficult to measure than the cell current (although it may just be direct drive in this case). With information about the cell currents for both cell types, we can compare the cell energies in your application. Cell voltages under load would also help avoid assumptions.

I also saw some 500 mAh 10440s over at DX, but they probably don't deliver as promised in a flashlight going for max lumens. Even so, they ought to deliver more than 320 mAh. The title for highest capacity seems to go to Li-ion even for the small cells...but data will be more convincing than opinions, so I will stop going on about mine.

On my L0D with 10440 (I managed to insert the leads of my DMM to measure current):

Open voltage: 3.96V
Capacity: 320mAh

Primary: 0.42A
Low: 0.18A
High: 1.23A (Way over the usual 2C MAX rating of 10440 cells)

A WARNING about running a Li-Ion cell over 2C, see this post from Newbie:
http://www.candlepowerforums.com/vb/showpost.php?p=1258756&postcount=15

[EDIT] Corrected the link...

 

[mdocod]

I was just about to link to that thread.... your link doesn't seem to be working... i'll try i suppose..
http://www.candlepowerforums.com/vb/showthread.php?t=106242


guys, find out what yours li-ion cells are designed to do... then respect those limits... When you run Li-ion cells above their rated output current, you are increasing the chance of an internal short leading to thermal runaway... These UBERcompact high power setups are NOT SAFE. pushing RCR2 and 10440 size cells to an amp or more is asking for a mini-bomb with side of hydrofluoric acid.

 

[Evan]

Don't forget the circutry in the 1.5 volt lightsmay not take 4 volts. AA size LiIon worked great in a Deal Extreeme VIPPA and CHARM, so I tried it in an ELLY and fried it.

 

[zeeexsixare]

I direct drive a Cree P4 in my plastic anglehead with a 18650... batt draw when fully charged is about 900 ma, not sure how much is going to the LED, but I'm sure the resistance in the 'cheap' contacts and springs is helping.

I've run it for 10 minutes continuous, though the output drops (as tested by DIY light box), it doesn't seem to be magic-smoke territory yet. If it was in an aluminium light, I'm sure it would perform better.

I guess I'll find out how long it lasts... Li-on direct driven Cree in a plastic host... haha... but I may be dumping in a resistor for more 'manageability'....

You didn't have any kind of actual resistor at all?

 

[LEDcandle]

You didn't have any kind of actual resistor at all?

Nope, no resistor... just the resistance from the parts in the light itself, which are the switch, spring and a metal (looks like copper or brass) strip running up along the light that connects the negative.

For $7, I won't mind if the LED burns out early... at least I tried.... 😀
But so far so good!

In another mod on a plastic light (LuxV WWAS), I drilled 5 holes into the light and screwed an aluminium plate on the outside connecting to the heatsink so the heat would 'interface' with the outside. It does help quite a bit and output that used to drop in 30s now takes about 4 minutes and drops much slower overall.

I use the light continuously for 20-30 mins, but your eyes will never tell there's a drop cuz its so gradual.

The plate gets hot of course and also you kill waterproofness...