Over-discharge of Li-ion – A Request for Flashlight Reviewers

Of late, I have become focused on the importance of knowing how a flashlight that runs on Li-ion batteries behaves when the voltage gets low. Unfortunately, flashlight reviews do not always do a good job describing how a low-voltage condition is handled.

Over-discharge of Li-ion
Depending on its design, a flashlight might do one or more of the following when voltage gets low on a Li-ion battery:

1. Nothing. The flashlight continues running until the battery is exhausted. On a protected battery, this should trip the protection. When the battery is unprotected, this could run the battery well below its safe discharge limit.


2. Provide a signal of some kind. Often this is a flashing red or blue LED on the body of the flashlight. On other flashlights, the main beam blinks periodically. Besides giving a signal, the flashlight continues to operate normally. It is the user's responsibility to respond to the signal, otherwise the dangers described in number 1 above apply.

3. Force a step-down to a lower level. Zebralights, for instance, do this.


Unfortunately, a flashlight user may not respond to any of the situations described in numbers 1-3 above. He might, for instance, put down his flashlight, and forget that it is running. When that happens, the only thing standing between a flashlight user and an over-discharge is the protection circuit of a protected battery. If that circuit fails, or if the flashlight is running unprotected batteries, an over-discharge is likely.

That brings us to the final possibility.

4. Turn off the flashlight completely using a low-voltage cutoff circuit. This is the only sure way to prevent an over-discharge. Once again, most Zebralights take this approach. On the SC62, for instance, cutoff occurs at 2.7 volts. Prior to this, you get forced step-downs from hi to med, and from med to low.

A Request for Flashlight Reviewers
If you are a flashlight reviewer, I ask that you routinely include a section describing in detail what happens when the Li-ion battery in a flashlight is over-discharged. The treatment should be systematic. It should appear in the same place, and in the same form, in every review.

Start with a direct statement describing whether there is a low-voltage cutoff. This simple yes-no statement should be as routine as the yes-no we usually get regarding tail-standing. If possible, provide the voltage at which cutoff occurs. From there, detail which of the other features described above are present. The several possibilities were numbered 1a, 1b, 2, 3a, and 3b. If possible, a review should discriminate between these five behaviors.

With this information, flashlight buyers will have a better understanding of what they need to do to make sure an over-discharge never happens in one of their lights.

good idea

Sounds reasonable to me.
+1

Great idea! This can be a level approach on the current reviews that we're all reading before buying a light. Additional details that could guide the suppose buyer. Very commendable idea!

I am on board with this. Should be helpful in pointing out safer lights so we're not on the news.

It would be a good addition indeed and personally I've always been satisfied with the way zebralight handles it. Never missed the dimming before it reaches a critical point at a bad moment. Also I made it a habit to change battery and recharge the used one when it drops below 3 flashes (so 2 or less flashes -> less than 50% remaining) leading to a lack of surprises the last few years.

But it is a feature I sorely miss in my 2xAA and other specialised torches, not terrible for battery safety with NiMH but I don't like an unexpected dead torch. With lithium ion it becomes even more important due to the potential for bad effects and their low internal resistance only making them fail to stay in regulation (on low/medium at least) until right before the very end where they'll either over discharge or turn off on you without early warning.

This really is a great idea, KIL; CPF has several stellar reviewers and this would take the already-excellent reviews yet another notch higher!

The problem is that very few reviewers can do this without risking damage to their batteries.
And seeing at what voltage the light does what (i.e. how serious the (over)discharge are) is nearly impossible only using batteries.

HKJ said:

The problem is that very few reviewers can do this without risking damage to their batteries.
And seeing at what voltage the light does what (i.e. how serious the (over)discharge are) is nearly impossible only using batteries.

Click to expand...

One thing all reviewers could do is report what a manufacturer claims. All too often, this information is glossed over in reviews. After that, reviewers can describe whatever behaviors they were able to test.

Knowing how a flashlight responds to over-discharge is an important part of safely working with Li-ion.

KeepingItLight said:

One thing all reviewers could do is report what a manufacturer claims. All too often, this information is glossed over in reviews. After that, reviewers can describe whatever behaviors they were able to test.

Knowing how a flashlight responds to over-discharge is an important part of safely working with Li-ion.

Click to expand...

Define over-discharged?

Do you want to use Panasonic's 2.5v, or LG, Sanyo, Samsung and Sony's 2.75v?

As HKJ points out, that'll kill a poor guy's cells, so I don't know why I, as a reader of a review, would expect somebody to do that?

Now, I can see somebody testing down to 3.5v/3.6v and telling you what he/she sees in the light's output, but that's still a pretty benign level and probably doesn't tell us much?

Ultimately, playing with li-ions and knowing what they do in any given light is a user's responsibility. We should all have an idea of what our lights do when using our own cells, but it should really be on our own dime (and not a reviewer's) if we want to take things down to zero, just so we can know.

Chris

ChrisGarrett said:

Do you want to use Panasonic's 2.5v, or LG, Sanyo, Samsung and Sony's 2.75v?

As HKJ points out, that'll kill a poor guy's cells, so I don't know why I, as a reader of a review, would expect somebody to do that?

Click to expand...

Doing a single discharge to 2.5-2.75V will certainly not "kill" them. In fact, it will likely have little-to-no negative impact whatsoever. Doing such low discharges repeatedly will cause more degradation compared to terminating the discharge at higher voltages - just like, symmetrically, terminating charges at 4.2V causes more degradation than terminating at lower voltages.

KeepingItLight said:

One thing all reviewers could do is report what a manufacturer claims. All too often, this information is glossed over in reviews. After that, reviewers can describe whatever behaviors they were able to test.

Knowing how a flashlight responds to over-discharge is an important part of safely working with Li-ion.

Click to expand...

ChrisGarrett said:

As HKJ points out, that'll kill a poor guy's cells, so I don't know why I, as a reader of a review, would expect somebody to do that?

Ultimately, playing with li-ions and knowing what they do in any given light is a user's responsibility. We should all have an idea of what our lights do when using our own cells, but it should really be on our own dime (and not a reviewer's) if we want to take things down to zero, just so we can know.

Chris

Click to expand...

I guess my answer to this is in the post of mine you quoted: "One thing all reviewers could do is report what a manufacturer claims." The complete post is above.

Gauss163 said:

Doing a single discharge to 2.5-2.75V will certainly not "kill" them. In fact, it will likely have little-to-no negative impact whatsoever. Doing such low discharges repeatedly will cause more degradation compared to terminating the discharge at higher voltages - just like, symmetrically, terminating charges at 4.2V causes more degradation than terminating at lower voltages.

Click to expand...

Gauss163 certainly understands batteries better that I do. What he says about taking a battery down to 2.5-2.75 volts dovetails with what I have read elsewhere.


I hope you are just playing devil's advocate. I don't think you mean to say that reviews would be better if they said nothing on this subject. As a reviewer, if you chose not to perform testing yourself, you could easily describe what the manufacturer says will happen. For instance, a simple yes-or-no statement regarding low-voltage cutoff would be better than nothing. If the manufacturer says nothing, you could report that, and caution owners to be aware that there is probably no low-voltage cutoff.

And so on, for the other possible behaviors, such as blinky warnings and forced step-downs.

I stand by my original post. All it argues for is this: every review of a Li-ion flashlight should include a section that describes what happens when a flashlight is left on and unattended. That section should include a description of what the manufacturer says and also any independent testing that the reviewer may have done.

KeepingItLight said:

I guess my answer to this is in the post of mine you quoted: "One thing all reviewers could do is report what a manufacturer claims." The complete post is above.

I hope you are just playing devil's advocate. I don't think you mean to say that reviews would be better if they said nothing on this subject. As a reviewer, if you chose not to perform testing yourself, you could easily describe what the manufacturer says will happen. For instance, a simple yes-or-no statement regarding low-voltage cutoff would be better than nothing. If the manufacturer says nothing, you could report that, and caution owners to be aware that there is probably no low-voltage cutoff.

And so on, for the other possible behaviors, such as blinky warnings and forced step-downs.

I stand by my original post. All it argues for is this: every review of a Li-ion flashlight should include a section that describes what happens when a flashlight is left on and unattended. That section should include a description of what the manufacturer says and also any independent testing that the reviewer may have done.

Click to expand...

Well, in a perfect world, sure, we'd get everything we needed to know from the manufacturer's website and wouldn't need people to review lights for technical info.

Taking Selfbuilt for example, who is the most prolific tester that I know of, he spends a lot of time testing stuff and probably a fair amount of money, just like HKJ with his cells and chargers. Sure, a lot of it is free, but leisure time is money.

Many light brochures do tell us if there's a low voltage step down, or cut-off, but obviously, not everybody does.

As far as testing down to 2.50v and/or 2.75v and whether doing that constantly, with one's own cells hurts them, or not, I'll leave that to HKJ.

I've discharged some cells below 2.50v and charged them right back up and they work, but who knows how many cycles I shaved off in the process?

Then you have the added time to run a light, let it cool, run it some more, check the voltage, reinsert it hoping that you don't go below, X, Y or Z and well, you've just added a big chunk of time.

Personally, it's just not a big factor for me and I wouldn't want the guys spending the time and effort, to trash their cells, but yeah...more info is always better.

Me? I just buy the lights that I think I'll like/want and fiddle with runtimes on my own, using my own batteries/cells and I won't be going down to 2.50v/2.75v, I can tell you that.

I guess in today's age, if you really, really need to know how a light behaves at the lower voltage ranges, you can always ask the manufacturer, which shouldn't be too hard?

Chris

ChrisGarrett said:

As far as testing down to 2.50v and/or 2.75v and whether doing that constantly, with one's own cells hurts them, or not, I'll leave that to HKJ.

Click to expand...

It is your prerogative to believe whatever you want. But be aware that the internet is flooded with unfounded "old-wives-tales" about Li-ion technology, so be careful what you accept. If you think that there is some scientific evidence to support your original claim then let's hear it. I am not aware of even a shred of evidence supporting it.

Gauss163 said:

It is your prerogative to believe whatever you want. But be aware that the internet is flooded with unfounded "old-wives-tales" about Li-ion technology, so be careful what you accept. If you think that there is some scientific evidence to support your original claim then let's hear it. I am not aware of even a shred of evidence supporting it.

Click to expand...

Because regularly taking a li-ion cell down to 2.5v/2.75v is always a good thing, right?

Gauss, I really don't care what you think. If you feel that's it's fine to regularly/routinely take your li-ion cells down to those minimums than have at it, as it's no bother to me.

Everybody else can do the same thing and I won't care one whit.

My comment first asked what the OP's definition of 'over-discharged' was and then I just threw out the minimum voltages use in testing by the Big 5: Sony, Sanyo, Panasonic, LG and Samsung. They don't go below those numbers and so I won't. Simple, ain't it?

If you have info to the contrary, that reaching, or exceeding those limits is not harmful to their cells, than post it up and I'll be glad to read it. You're really good at asking others to provide evidence, but you don't provide much yourself--unless it's a 2009 dated NASA study using one particular li-co cell.

Chris

ChrisGarrett said:

Because regularly taking a li-ion cell down to 2.5v/2.75v is always a good thing, right?

Click to expand...

Correction: the earlier discussion was about a single test, not "regularly taking....".

ChrisGarrett said:

I you have info to the contrary, that reaching, or exceeding those limits is not harmful to their cells, than post it up and I'll be glad to read it. You're really good at asking others to provide evidence, but you don't provide much yourself--unless it's a 2009 dated NASA study using one particular li-co cell.

Click to expand...

It was you who made the initial claim above that such testing will "kill a poor guy's cells". Again, where is the scientific evidence supporting your claim? Why do you believe that claim? Is it just wild guess? Probably so. Please stop posting wild guesses as if they were facts.

Gauss163 said:

Correction: the earlier discussion was about a single test, not "regularly taking....".

It was you who made the initial claim above that such testing will "kill a poor guy's cells". Again, where is the scientific evidence supporting your claim? Why do you believe that? Is it a just wild guess? Probably so. Please stop posting wild guesses.

Click to expand...

The OP was asking guys that do reviews of lights, to test for low voltage behavior. There are the prolific testers like JohnnyMac over on BLF and Selfbuilt here, who review a lot of lights, which people consult, so it wouldn't be some guy named Gauss testing a single light once, down to those levels.

The people doing the competent reviews, are doing lots of them.

Secondly, let me replace "kill" with 'damage,' as that was my mistake and since you can't see past embellishment, or hyperbole, I don't want you to keep bringing it up ad infinitum.

I'm still waiting for your info showing that it's perfectly fine to hit, or exceed those thresholds.

Thanks/bye.

Chris

KeepingItLight said:

I stand by my original post. All it argues for is this: every review of a Li-ion flashlight should include a section that describes what happens when a flashlight is left on and unattended. That section should include a description of what the manufacturer says and also any independent testing that the reviewer may have done.

Click to expand...

Sorry, guys. I was not trying to be controversial.

The reason I feel so strongly about this is not because the experts on CPF can't take care of themselves. It's the rest of us. It's the vast majority of flashlight buyers who do not understand the potential dangers. Highlighting the issue in every review of a Li-ion flashlight is a good way to impress the issue on regular folks.

It would only take a paragraph to describe what I have outlined above. Does a flashlight have a low-voltage cutoff? Does it signal by blinking or forcing step-downs that its batteries are getting low? Does it do nothing? If a given flashlight reviewer, be it Chris, HKJ, or anyone else, does not want to perform his own tests, he can at least report what the manufacturer has to say on the subject.

My own practice would be the one that Chris recommends and HKJ follows. I would not test my batteries below the voltages that a manufacturer uses in its own tests. If that means that the low-voltage protection built into a flashlight never trips, so be it.

I am no expert, but I am pretty sure Gauss163 is 100% correct when he says that taking a battery down to the low voltages shown in its datasheet is not a problem at all. For many Panasonic batteries, that mean 2.5 volts. The low-voltage protection in most flashlights will trip before that. Perhaps reviewers should buy a couple of these batteries to use for reviews.

My understanding is that these voltage thresholds are like the red line on a tachometer. Just because you go a little over does not mean your engine will explode. I am sure the same is true of batteries. From what I have read, the bigger concern is not to store a battery at a low voltage level. If you recharge immediately, the battery should be fine.

That said, I am also pretty confident that what Chris and Gauss163 both say about voltage extremes reducing battery lifetime is also true. If you want to get the longest life from a battery, stay away from high and low voltages. Do not charge all the way up, and do not run all the way down. Chris seems to be saying that he expects a reviewer to charge all the way up for a test, but not necessarily to run all the way down.

For what its worth, selfbuilt and mhanlen are two CPF reviewers who are willing to run a battery all the way down. Both frequently run a flashlight until there is no light output at all.

We have all sorts of reviewers here. We have all sorts of reviews. Some reviewers pay for their flashlights. Some do not. Some perform submersion and drop tests. Some do not. Some create runtime charts. Others do not. I am thankful for whatever testing and reviews our many good members do share. THANK YOU!

At the end of the day, I agree with Chris that we should not demand that a reviewer spend unlimited amounts of cash for batteries and flashlights. Let him perform the tests he is comfortable doing. Let him decide how much to spend.

In this thread, I am asking that flashlight reviewers systematically include information about how Li-ion flashlights behave when their batteries run low. That's all. Details of battery voltage should not sway your opinion, one way or the other. I worry that we are losing track of the main point.

The questions are simple. The answers are short. If a flashlight has a low-voltage cutoff, a review should say so. If not, the review should explicitly say that. If a flashlight blinks when the power runs low, or forces step-downs to lower levels, a review should describe that. If a flashlight does nothing at all, a review should explain that.

I say this information should be included in every review of Li-ion flashlight. It can be gathered from manufacturer's statements, testing, or a combination of the two. It should be presented in one place, for example, in a section entitled What Happens When the Batteries Run Low?.

What do you think?

@KIL To be sure, the above arguments against your proposed test are nonsense.

(1) First, generally it doesn't requires a battery to perform low voltage cutoff tests. One can instead do a voltage sweep using a variable-voltage DC power source (if you don't have one then you can purchase a buck regulator for a few dollars on eBay).

(2) Second the author of the argument has been unable to supply even a single shred of scientific evidence supporting his claim that using 18650s for such tests will "kill" them. He has failed to even explain why he believes this claim. Why does he believe that the cell being momentarily at the minimum spec voltage is so much worse than it being at the highest spec voltage? Why doesn't he believe that charging the cells to the maximum spec voltage (e.g. 4.2V) will also "kill" them? He has given no logic or scientific evidence in support of his claim. So, as it stands, the claim is simply a guess (like much that is written about Li-ion tech on the web). But that's a moot point anyhow, given (1).