"Sudden Death" inevitable for LiON lights?

[brightnorm]

NOTE: I am reviving this thread in hope that some advances may have been made in preventing or signaling a warning before Sudden Death occurs in regulated Li-ON lights

Brightnorm


Many of my lithium primary lights have low battery warnings, either through automatic level step-downs or flickering, but none of my LiON lights have this feature and will abruptly switch off when battery level drops below a certain point.

Does the chemistry of rechargeable lithiums make this warning feature impossible?

Brightnorm

 

[Marduke]

Basically, yes. Either the light or the cell needs some form of protection. If you rely on the cell, the protection circuit kicks in and the light just shuts off. If the light has a warning, it has to be set higher than the standard cell protection cutoff, meaning you are getting less capacity of your cell.


This is one reason I am not fond of Li-Ion lights. They generally break my #1 rule, that a light will not suddenly cut off with absolutely no warning.

 

[VidPro]

my DD and resistered lights never have this problem, they just keep going and going and going. lower and lower , till i recharge just because its becomming lower in output.
Then
the boost lights and curcuit lights they just Blink out poof, and it always sorta surprises me , game over. sometimes after a rest i can squeak out a bit of light if i can quick enough get it set to a low level, but otherwise i am in the dark. and it always (of course) happens when i am out away from the rest of the lights.
break out the backup keylight, i dont ever carry a spare battery, cause they are as big as some of my small lights.

the dual lithium lights that will take a 18650 OR 2x123 many are already direct (sorta with ONE cell) with thier buck type drivers and they dont (usually) have that problem, dwindling down more slowly like DD, but they also arent that small.
levels are always wonderfully usefull, keep the light on Med or low most the time, use high only when needed, then i dont have to re-charge so often.

by the time the usual curcuits would have any low-bat warning between ~2.4-2.7v per cell, the protection is about to kick in anyway

 

[Mjolnir]

The L-mini has protection built into the driver itself, but it also drops to low mode when the battery gets low (around 3 volts). I believe it finally cuts off at about 2.7 volts, which is what most protection circuits cut off at. That way, you have a warning, and still a decent amount of usable light. It can also run in high after dropping down into low mode.

 

[travelinman]

I've posted this in another thread but it relates directly to this problem. The ITP C9, when on high, and using a protected RCR123 which suddenly dies when the protection kicks in has a manual "moon mode".

What you do is turn the light off (even though it is already off) by clicking the rear clicky. Then crack the head (which when on, starts the downward ramping) then turn it on. The cell has enough recover left to start the ramp and when you get to low, tighten the head again. In this low mode it will work for at least 30 min, maybe more depending on the cell you are using.

I am presently using little 3.0v Delkin cells that I can only get about 350 ma out of "just because I have lots of them and I'm too cheap to throw them away". I imagine if you have a more robust RCR123 you'd get a lot more "moon mode" time using this method.

 

[Black Rose]

The L-mini has protection built into the driver itself, but it also drops to low mode when the battery gets low (around 3 volts). I believe it finally cuts off at about 2.7 volts, which is what most protection circuits cut off at. That way, you have a warning, and still a decent amount of usable light. It can also run in high after dropping down into low mode.

It's 2.8v for the MG L-Mini II. The drivers for this light are available separately (when in stock). I modded 2 of my P60 drop-ins with this circuit as well.

I did some runtime tests with these circuits in my Solarforce L2 with AW and black Trusftire cells.
In both tests, the flashlight circuit drops to low before the protection circuit on the cells kicks in.

 

[brightnorm]

I see that there are ways to mitigate this problem, but apparently very few manufacturers have incorporated them. I can't imagine how a LEO would cope with this problem in a tense situation.

This is an important issue which doesn't appear to have been fully addressed either by designers, manufacturers, or even here on CPF.

Brightnorm

 

[Creecher]

I see that there are ways to mitigate this problem, but apparently very few manufacturers have incorporated them. I can't imagine how a LEO would cope with this problem in a tense situation.

This is an important issue which doesn't appear to have been fully addressed either by designers, manufacturers, or even here on CPF.

Brightnorm

Agreed, a big problem with many lights. Liteflux lights offer an early warning of a couple of flashes every 2 mins, and this gives easily enough time to change the cell, or ramp the power down until it starts to flash again. And there is always the accurate voltage test available aswell.

 

[Mjolnir]

I think it would be neat if higher end lights had some sort of visual indicator on the side of the light, like a small OLED display that could be turned on to show the remaining battery.

 

[rmteo]

This one has a battery life indicator - it does not need one though.

 

[Marduke]

The problem with battery indicators is they are VERY specific for each chemistry, and can tell you little to nothing about the actual charge state of your cells.

Even the much more expensive and intricate ZTS tester has it's own issues and limitations, and you want that feature added at near zero cost as an afterthought to your favorite flashlights??

Good luck with that...

 

[Mjolnir]

I never said "zero cost;" all I said was on "higher end lights." In fact, I never envisioned it being cheap (especially if it had an OLED display), which is why it is not happening today. I'm not sure if that sarcasm was directed at me, but it really wasn't necessary.

 

[xenonk]

Some lights just aren't designed with any warning system at all, but there's nothing special about li-ions with regards to low battery detection. Most circuits use simple voltage sense to trigger the programmed warning, and different chemistries have different voltage ranges.

A lot of lights made to take li-ions are still designed with primaries in mind (which LEO/Emergency/Military are usually presumed to be using), and may have protection indicators calibrated for that voltage range. Li-Ions are nearly dead at the voltage where fresh primaries start, so they don't get low enough to set off the warning before cell protection kicks in abruptly.

Or you could use unprotected li-ions and possibly get to see the warning, but then your cells are down for the count from overdischarge.

A manufacturer could just as easily build a driver calibrated for li-ion, or even both at once by having it change behaviour based on start voltage. I know some lights designed for li-ion do at least the former, but I can't recall any off the top of my head.

 

[Marduke]

The solution would be to program the protection circuit itself to issue a warning, since it is impractical to program the light itself to issue a warning for a cell type which it cannot discern it is using.

 

[Mr. Tone]

I have recently discovered this problem with my first Lion light, the DBS. I have the 3SD and it is for 18650 only. It will be running on full tilt with no noticable difference in brightness and just completely go dark. I can then switch it on again to a lower mode but it will cut off quickly on high. At first I had no idea what was going on so I switched batteries and no more problem. After this experience several times with both batteries I figured it was the protection circuit(s). For any tactical need this would be unacceptable, but for my playing it is just a great inconvenience.

 

[Kestrel]

I'm not sure if that sarcasm was directed at me, but it really wasn't necessary.

No, Marduke directs that sarcasm at most everyone who asks questions around here.

 

[Bushman5]

the first time i experienced "sudden death" in my li-ion flashlight (Pelican 7060), i freaked out, because i thought i had broken it. I took it all apart, put it back together and charged it up and then there was :candle: < light.

My heart was racing, i honestly thought i had broken it...

 

[uk_caver]

The solution would be to program the protection circuit itself to issue a warning, since it is impractical to program the light itself to issue a warning for a cell type which it cannot discern it is using.

The problem with that is that if the cell protection gives a pre-warning, it can only really do so by one or more temporary disconnections of power (or maybe temporary insertions of some extra resistance).
Depending on the driver in the light, that could do all kinds of things - changing modes, etc. Kicking a light from a subtle low into a strobe might not be appreciated, even if the protection only gave a single warning, rather than a series of warnings over time.

 

[Marduke]

Which is more desirable?

A light just blinking out with no warning, or the light in some circumstances changing modes with no warning?

If the light didn't have a multi-click interface, it would just blink/dim accordingly.

 

[uk_caver]

It's more desirable to have a warning in the light, rather than the cells. It is at least potentially configurable there, and means a given light would behave consistently, and there's rather more opportunity for a light to do smarter things.

However, if few lights actually do incorporate such things, maybe there just isn't a great demand for them.