how should a flashlight behave when battery runs low? Ideal vs real-life behaviors.

coachclass · Apr 13, 2011

How would you like your flashlights/headlamps to behave when it runs low on battery power?

I don't have a good understanding of what's practically possible with electronics at lower battery voltage/current...so if you have more insights, feel free to write a longer response to explain.

This may also be a multi-part answer:
1) how should it behave on Max when battery can no longer sustain max.
2) how should it behave on low / moonlight when the battery can sustain low, but not max.

I guess my ideal interface would alert me if the battery is almost gone. But, if the battery is good enough for low/moonlight, but not good enough for max, should the light somehow warn me of that? Can modern battery electronics sense that the battery is too depleted for max mode without actually going into max mode? That's important because I may not use max mode very often, and using the low/moonlight modes, you loose count of how many hours you've been burning the same set of batteries. So, before stepping out the door, I'll turn on the light (on low), and if I don't get any indication that the battery is depleted, I might assume I have at least some capacity for max/med modes.

Often, you hear about how lights have such and such runtime, but there's not much focus on what happens at the end of the runtime. I'd like to hear what your flashlights do, as well as what you think they should do.

I've been testing out a Klarus ST20 lately, and found this behavior when the battery runs low and flashlight is on max - the light level goes down, and it goes into a slower strobe mode not found in any of the normal modes. Can't change modes unless I loosen the bezel for moonlight, and it seems that moonlight is the only mode available. Beacon and SOS modes were erratic. After battery gets a rest, I can cycle to all modes (moonlight, sos, beacon, low, med, high), but the medium and high were the same brightness and not very bright.

Opinions?

shane45_1911 · Apr 13, 2011

coachclass said:
1) how should it behave on Max when battery can no longer sustain max.
2) how should it behave on low / moonlight when the battery can sustain low, but not max.

1) Light should gradually decrease in output, giving you ample warning to change your batteries.
2) Light should flash or flicker to let you know that you are imminently close to "lights out" when moonlight can't be supported. Obviously, it will not be able to switch to max below a certain voltage.

Robin24k · Apr 13, 2011

Voltage-based cutoff is possible, rechargeable Li-Ion lights use it. For example, the Pelican 7060 does a series of one-, two-, and three-blinks before shutting off until recharged, whereas the Inova T4 does about 15 seconds of fast flashing that is almost too fast to be noticeable. If a voltage-based cutoff is triggered, the only way to reset would be to charge or replace batteries, so you won't have access to other modes.

With NiMH, there's no reason (or way) for a manufacturer to use voltage-based cutoff because there's no safety need, and because of differences with alkalines and NiMH. The same goes for lights that take CR123A and Li-Ion. Because of this, most lights naturally end up dimming.

coachclass · Apr 13, 2011

Specific to NiMH, when lights go dim, are they switching of the current regulated circuit and going into direct drive mode? I'm guessing if a light has only a current regulated circuit, and has no way of switching to direct drive, the light would just go OFF, because the voltage isn't high enough for the current regulation circuit?

Now I'm getting a bit into the electrical engineering of these lights, but it's interesting.

When an alkaline or NiMH battery starts to run low, is it the voltage or the current that can't be sustained? I'm assuming it's voltage (because the V drops as the cells loose power). But are the cells also current limited more so than when they were fresh? Which is the weaker link (low voltage or insufficient current supply)? I'm assuming the light needs both a power source > some volt and also a power source that can supply X mA. When the battery starts to die, does the light hit the voltage minimum first, or the current minimum?

jbrett14 · Apr 13, 2011

coachclass said:
How would you like your flashlights/headlamps to behave when it runs low on battery power?
Opinions?

I like mine to simply jump to a lower mode, if I am using it in a higher mode. But I personally hate lights, like the incans, that just start to diminish in brightness the minute they are turned on. I like a regulated output that provides a constant level of brightness, but not one that simply shuts off with no warning.

If using the light in it's lowest mode, perhaps a flash of some sort could be part of the circuitry. Or, I had a Nuwai headlamp that just had a little green\\yellow\red led indicator that let you know the condition of the batteries. And this was on a cheap $25 light.

unixguy · Apr 13, 2011

I have a Princeton APEX LED headlamp (powered by 4 AA batteries). When the amount of runtime left drops below a certain threshold, the light blinks to tell you that the end is near. (My guess is that this is not a real calculation, but rather based on dropping below a voltage threshold.)

Disclaimer: I'm a newbie and my first LED flashlights are about to be delivered to me (two Jetbeam BA20 flashlights).

cdosrun · Apr 13, 2011

coachclass said:
How would you like your flashlights/headlamps to behave when it runs low on battery power?

1) how should it behave on Max when battery can no longer sustain max.
2) how should it behave on low / moonlight when the battery can sustain low, but not max.

Opinions?

There are only 3 main types of output vs runtime curves I can think of (and ones that fall slightly in between). Here I am referring specifcally to single-celled lights that need to boost the battery voltage for the LED.

1. Unregulated output, such as a Fenix E0 - The output declines from the moment it is turned on, much the same as a filament light. The curve will basically follow the discharge curve of the cell and the light presents an effective constant resistance load to the cell.

2. Brick-wall regulated, not sure of an example off hand - Here the light maintains the output at the required setting until the cell can no longer supply enough power for the circuit at all, and the light turns off. This sort of regulation puts a heavy load on the cell as the current rises enormously towards the end of the cell's discharge. There is little no warning to the cell's demise and the light simply switches off when the voltage drops too low for the switching regulator.

1.5 - This is not a main class, but falls somewhere between the above two scenarios and is rather more common from what I know. There are limits to the switching components in a torch based on cost, size, leakage current, heat etc. and at some point during the cell's discharge, the voltage has fallen sufficiently that the current has increased to the limit of the components. At that point, some torch regulators move to 'moon mode' and simply act as unregulated boost circuits. I think the difference between this and a 'type 2' is generally the low level cut-off voltage for the regulator logic.

3. This is the third type of regulation and is of the type in the HDS and other computerised type lights. The cell's voltage is monitored under load and the light switches proactively to a lower level as the voltage declines to preserve what is left of the dying cell.

I have seen a light for advertised on the forum that allows switching between something approaching the above 3 descriptions and think that an interesting proposition.

You asked about reporting a low cell at low outputs, I can't see this being especially practical as the cell voltage tends to 'recover' under low or no load, making it more difficult to establish charge level for many cell types. I think the best way here would be to flick up to maximum for a second and see if the torch reports a low cell level.

My preferred type of control is type 3, as the torch lets you know when the cell is low and give the choice of replacing it or turning to a low level. With type 1, one generally doesn't notice the steady decline until the level is very low, and with type 2 you are left in sudden darkness.

The actual graph would depend on the type of cell in use, and I am sure someone like Tom would be better placed to explain the above, but it might be a starting point.

Andrew

shao.fu.tzer · Apr 13, 2011

Ideal - A pleasant computerized female voice warns you that your voltage has reached a level that has made high mode unsustainable, before switching to the next lowest mode. Also, a battery indicator like on a cellphone, with bars. But it wouldn't be on all the time, rather a feature you can access, unless your light gets low, then maybe a dimly blinking red bar would display.

Real world - A few blinks and a drop in output.

Outdoorsman5 · Apr 13, 2011

I have regular experience with this because I run with a light on my head before the sun comes up (and I always carry a backup light.) I usually use a zebralight sc30, and run it on a rcr123 li-ion rechargeable. With the light on max - if the battery gets too low then the light just cuts off. I can then turn it back on at medium or low (if it's too hot though I then have to wait a few seconds for it to cool off...usually not a problem.) I timed it once after running out of juice on high. I turned it back on at 4 lumens and got another 3 hours out of it. I actually turned it off & didn't wait for it to die...was afraid of damaging my li-ion battery, and figured 3 hours was long enough.

I would like it better if the light didn't just turn off and just switched itself to a lower level. If I were in the lowest mode then I would like an occasional blink to warn me, but I would not want this blink on higher modes...just switch to the lower mode then blink if running out of juice on the lower level .

Kestrel · Apr 13, 2011

Interesting thread, particulary because there isn't a right answer IMO. There are all sorts of different light behaviors, and I can see both strengths and weaknesses in most of them.

Some of my light configs:

SureFire T1A w/ CR123 primary - strobes when it cannot sustain the selected output (usually @ maximum)
SureFire L1 w/ CR123 primary - drops to the low level
Moddoo 3-mode regulated XPG-Triple, 2x LiIon - Cutoff w/ no warning (LiIon protection circuits)
Moddoo 3-mode regulated XPG-Triple, 1x LiIon - drops to 'medium' for a few minutes, then shutoff.
Moddoo XP-G DD, 1x LiIon - gradual decrease throughout the run, then PTC cutoff
Malkoff M31W on 2xAA NiMH's - flat regulation down to ~0.8Vin, then rapidly decreasing output (only ~5 minutes of usable output or so after falling out of regulation)

Although I don't have one, I actually think that the Malkoff M61 w/ 1x18650 would be about perfect. From what I understand of those, they exhibit flat regulation for ~2/3 of the total 18650 runtime, gradually decreasing output for the last ~1/3, then protection-circuit-shutoff vis. the 18650. Plenty of flat regulation, enough declining output as a warning, then final cutoff (cell protection). A very good balance IMO.

coachclass · Apr 14, 2011

cdosrun said:
You asked about reporting a low cell at low outputs, I can't see this being especially practical as the cell voltage tends to 'recover' under low or no load, making it more difficult to establish charge level for many cell types. I think the best way here would be to flick up to maximum for a second and see if the torch reports a low cell level.
Andrew

That's kind of what I thought. Batteries (at least non-smart / dumb batteries) don't report how much chemical capacity they have left, and the voltage of a cell isn't very reliable, and changes under different loads, right?

So it would seem that if you run your flashlight on low most of the time, you might be surprised to find yourself lacking the high mode when you need it. This is true even on lights that have a dedicated auxiliary red indicator for low power; the indicator only trips when the flashlight mode creates a low enough voltage; ie on high mode.

Even the realistic ideal methods for low battery indication require the user to go into high mode at some point to "test" their batteries. How long do you have to remain in high mode to do such a test? A quick check might not be long enough because the battery might just be able to do high mode for a few seconds, but in actuality is pretty low.

Maybe this problem can't be solved through flashlight circuitry; maybe this problem requires smart batteries.

andrewnewman · Apr 14, 2011

Kestrel said:
Although I don't have one, I actually think that the Malkoff M61 w/ 1x18650 would be about perfect. From what I understand of those, they exhibit flat regulation for ~2/3 of the total 18650 runtime, gradually decreasing output for the last ~1/3, then protection-circuit-shutoff vis. the 18650. Plenty of flat regulation, enough declining output as a warning, then final cutoff (cell protection). A very good balance IMO.

Yes. I do own a Malkoff MD2 with M61 running an 18650. It behaves exactly as you describe and it is pretty much perfect. I often run it with an unprotected 18650 (I never lend this light out!) as it is almost impossible to over-discharge the battery unless you are trying to. I figure in a real emergency I might be willing to sacrifice a LiIon for a few additional minutes of light.

DaveyJones · Apr 14, 2011

shao.fu.tzer said:
Ideal - A pleasant computerized female voice warns you that your voltage has reached a level that has made high mode unsustainable, before switching to the next lowest mode

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This.
and with a wifi or bluetooth connection so u can download different voices and messeges

Click to expand...

MatNeh · Apr 22, 2011

Real-life: Zebralight H31w running on high for 30 minutes with an AW protected RCR123: light turned off and killed the battery (measured 3.05V)
Ideal: Would have turned off earlier and NOT KILLED MY BATTERY!

beamis · Apr 22, 2011

shao.fu.tzer said:
Ideal - A pleasant computerized female voice warns you that your voltage has reached a level that has made high mode unsustainable

Perhaps that would be ideal if the aim is to drive someone to commit themselves to psychiatric care.

srfreddy · Apr 22, 2011

Well, I think that they should run flat until almost dead, then go into moonlight.

JWRitchie76 · Apr 22, 2011

This is an interesting subject. Recently my Ra Clicky started to indicate that the battery was getting low (AW RCR123) so just for the fun of it I pulled the battery out to test the voltage and it measured 3.4 volts? So under load it would actually measure less? If this isn't the case I kind of find the HDS indicator unreliable because all of my HDS lights seem to behave differently from battery to battery as far as the low battery indicator is concerned. I've also been wondering how my Haiku will act when the battery starts to go in it? Maybe someone can chime in and let me know what to look for. Good thread!

radioactive_man · Apr 22, 2011

MatNeh said:
Real-life: Zebralight H31w running on high for 30 minutes with an AW protected RCR123: light turned off and killed the battery (measured 3.05V)
Ideal: Would have turned off earlier and NOT KILLED MY BATTERY!

How can a light kill a protected battery? I thought the protection circuit would prevent this.

And what is the minimum safe voltage for a Li-Ion cell?

wingspar · Apr 22, 2011

My experience with my Fenix lights is the brightest setting will not work when the batteries start to reach the end of their life, and scrolling thru the lower settings becomes difficult. First time it happened, I had no idea what was going on, and thought I was going to have to return the light for service. The next day, it happened to another of my Fenix lights. (I bought them both at the same time, so batteries pooped out at the same time) Replacing the batteries fixed the problem, so now I know. The lights remain usable when the batteries start to reach the end of their life, not like cheap hardware store lights that start getting dimmer as soon as you turn them on.

I've got used to that. I would not care for flashing lights to tell me it's time to change batteries. That would tell me it's time to change flashlights.

shao.fu.tzer · Apr 23, 2011

DaveyJones said:
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This.
and with a wifi or bluetooth connection so u can download different voices and messeges

Nice one... Even better...

beamis said:
Perhaps that would be ideal if the aim is to drive someone to commit themselves to psychiatric care.

Everything will talk to you in the future, even your toaster. Dude, don't you watch Sci-Fi movies?

how should a flashlight behave when battery runs low? Ideal vs real-life behaviors.

Newly Enlightened

Enlightened

Flashlight Enthusiast

Newly Enlightened

Enlightened

Newly Enlightened

Enlightened

Flashlight Enthusiast

Flashlight Enthusiast

Flashaholic

Newly Enlightened

Enlightened

Newly Enlightened

Newly Enlightened

Newly Enlightened

Enlightened

Flashlight Enthusiast

Enlightened

Newly Enlightened

Flashlight Enthusiast

Similar threads