protected li-ion voltage bouncing back?

[greenLED]

The other night, my GL3 turned itself off. It's happened many times previously when the batteries reach their low voltage threshold and the protection circuit kicks in. No biggie - I used my backup light and got home no problem.

The weird thing was, once I got home, the GL3 worked again! I left it on for about 5 minutes, turned it on/off a few times, and it worked just fine... my experience has been that once the protection circuit kicks in, the light will simply not turn on. Period.

I'm puzzled, wondering why things were different this time. Could I have a defective circuit in one (or both) of my Pila cells?

 

[bezel]

My experience has been the opposite.

I've never experienced a rechargeable that didn't bounce back. Whether Ni-cad, MiMH, or Lithium Ion (protected or unprotected).
Never used Pilas, but I use protected MP RCR123s with an E1L. I hate the fact there is no warning that it is about to go out. But if I really need the light, I just wait a few seconds for bounce and the circuit to reset.

Waiting an hour or so will give enough bounce for a couple minutes of run.

bezel

 

[matrixshaman]

I'd hazard a guess that when it went out the other battery might have still had enough power to charge the one that kicked off first enough to bring it back. I think the normal scenario for a proper protection circuit is that it stays off until it gets an external charge. I am a bit fuzzy on these details so you could say I'm guessing but that's how I remember it working.

 

[45/70]

I'd hazard a guess that when it went out the other battery might have still had enough power to charge the one that kicked off first enough to bring it back.

I wouldn't mind being corrected if I'm wrong, but I believe cells in series reverse charge, rather than charge.

green, I would suspect that what happened is a sign that the protection circuit failed/is failing in one of the cells, particularly if you trip them often. They only work so many times, then .

Dave

 

[bezel]

But all of my protected cells used single-cell applications recover to the point that the light will come on again.
I would suggest that once the voltage of the cell recovers to a certain point, it is normal for the circuit to reset.

 

[greenLED]

This is really interesting - thanks for your comments, guys.

I've used my GL3 for a few years now, and I'd never noticed the batteries "bouncing back" before. Once the light died it stayed that way. I've even measured the voltage on them and they read 0, indicating the circuit has tripped them off.

I've only used Pila cells, though. Bezel's comment makes me wonder there might be differences between protection circuits? :shrug:

That said, I am using a couple of cells that are somewhat "old" (very few cycles, but that were unused for a long time). :thinking:

Along those lines... Dave, your comment makes sense. In any case, I'll have to let my light run until it dies again and pay attention to what the batteries do after I let them rest a bit.


If other people have more ideas and experiences, I'd appreciate hearing from y'all.

:thanks:

 

[Illum]

now that you mentioned it...:thinking:

one of my AW17670s have this issue...when used with the LF-EO-E2R it sometimes stop functioning in the middle of an discharge, if I take it out then put it back again it works...
this has happened maybe twice in at least 10 cycles of normal operation, and apparently "4V" was the bottle neck:green:

 

[kaidomain]

now that you mentioned it...:thinking:

one of my AW17670s have this issue...when used with the LF-EO-E2R it sometimes stop functioning in the middle of an discharge, if I take it out then put it back again it works...
this has happened maybe twice in at least 10 cycles of normal operation, and apparently "4V" was the bottle neck:green:

I was testing a new LED boost circuit with a protected RCR123A sample cell that I have received (4.08V @ no load). The cell is being discharged at around 1A, then up to 1.5A when the light is almost out after an hour. When the protection of Li-ion kicks in and I put it between a meter, it reads 2.4V . I left the battery on my desk for a couple hours and it reads 3.3V :duh2:.

I then slowly discharge the cell at around 150mA and it goes down to 0.8V and the protection circuit still would not kick in! I left the cell on my desk for a day and the voltage measures 2.7V.

I am checking with the manufacturer to see if the protection circuit actually protects against the actual voltage at no load, instead of voltage underload.

 

[VidPro]

different protection curcuits work different ways.
some kick off, and wont "re-trigger" ever untill a charge is applied
some kick off and as soon as the cell voltage recovers, you might get another hit on it, OR be able to run at a lower mode, slowly depleating the cell.
and some Cells themselves have high resistance, or will voltage droop by quite a bit.

and for Kai , some protection curcuits can be Tricked i have mentioned it LONG ago. i have tricked both the high and and the low end of protection curcuits, usually it requires disconnecting the battery from the protection curcuit. of course this means it can also happen from a bad connection. Disconnect the protection, overcharge the battery, then stuff it back on, and depending on the curcuit method, you can continue to overcharge the battery.
THAT
is why i always say that the battery protection is just that, for the battery, and does not suppliment the charger or the devices need to control things ALSO.
SO
run your test again, and insure that the curcuit path is solidly connected, i first discovered this phenomena when using cruddy clip leads.
and of course it must start above the low cutoff and below the high cutoff before connected.

 

[kaidomain]

different protection curcuits work different ways.
some kick off, and wont "re-trigger" ever untill a charge is applied
some kick off and as soon as the cell voltage recovers, you might get another hit on it, OR be able to run at a lower mode, slowly depleating the cell.
and some Cells themselves have high resistance, or will voltage droop by quite a bit.

and for Kai , some protection curcuits can be Tricked i have mentioned it LONG ago. i have tricked both the high and and the low end of protection curcuits, usually it requires disconnecting the battery from the protection curcuit. of course this means it can also happen from a bad connection. Disconnect the protection, overcharge the battery, then stuff it back on, and depending on the curcuit method, you can continue to overcharge the battery.
THAT
is why i always say that the battery protection is just that, for the battery, and does not suppliment the charger or the devices need to control things ALSO.
SO
run your test again, and insure that the curcuit path is solidly connected, i first discovered this phenomena when using cruddy clip leads.
and of course it must start above the low cutoff and below the high cutoff before connected.

The battery protection circuit is sealed at the bottom of the cell so I am pretty sure all connection is good. I solder all connecting wires between the LED, battery holder, and the driver, so I don't think anything is hanging loosely.

 

[VidPro]

The battery protection circuit is sealed at the bottom of the cell so I am pretty sure all connection is good. I solder all connecting wires between the LED, battery holder, and the driver, so I don't think anything is hanging loosely.

then did the battery itself "Self_discharge" ?
bring it down SLOWly to the cutoff (where it wont bounce back much) then let the battery sit self discharging itself.

you had a Slow discharge, but said it HAD bounced back, so that wouldnt apply, because you still would have to pass by the cutoff point again.

so that only leave the reaction time of the cutoff, applying the load when it was already AT the cutoff point (when loaded) and it passed the cutoff before it had a chance (curcuit reaction time).

Which is why, after reaching a cutoff, your supposed to quit using it, until its recharged, and because of any possible self discharge, not leave a discharged cell low.

on some of them its as if its like this
---|-----------------|---
low cut . normal . High cut

overcharge a battery, bypassing the protection (direct connect) and you can continue above.
Overdischarge a battery bypassing the protection (direct connected) and you can continue below.
the "little brain" inside doesnt ever see the trigger voltage things.

 

[daimleramg]

Even my AW r123 cells do the same thing so when i lend out one of my lights out to a friend with AW's in them i tell them once the light turns off by itself do not turn it on again.

 

[matrixshaman]

I wouldn't mind being corrected if I'm wrong, but I believe cells in series reverse charge, rather than charge.

green, I would suspect that what happened is a sign that the protection circuit failed/is failing in one of the cells, particularly if you trip them often. They only work so many times, then .

Dave

I think you are right - that makes sense now that I've pictured the circuit in my head. I know I've heard that a hundred times about cells getting reverse charged so I'd say I was having a momentary brain fa*t.
Next guess : the cell just recovers enough that it reinitializes the protection circuit - which just might not be working 100% right at this point.

 

[half-watt]

I've heard that a hundred times about cells getting reverse charged

addressing just the quoted text above:

happens w/cells or batteries in series when one cell/battery is discharged sooner than the others in the series (a "weak" or "old" cell perhaps, or beginning the current service/usage with different %SOC, cells/batteries of diff. capacity, etc.), the others in series may cause the weak or discharged cell to discharge to <0VDC. measuring the reverse charged cell will reveal a pole reversal (often only tens or a hundred or so millivolts negative voltage in NiMH or alkaline cells that i've measured).

this pole reversal can also happen due to self-discharge in NiMH (and other rechargeable cells/batteries) if they are stored long term.

------------------
as far as OP's observation, i've seen this same thing happen and a related phenomenon with primary cells. what i think is happening in both cases is that as the voltage in the cell/battery gets low and it's being used and gettin' hotter, the cell's/battery's internal resistance increases to the point that too much of the little remaining voltage is being dropped due to internal resistance, leaving less to power the load.

after switching off the light, the cell(s)/batter[y | ies] cool(s) and the internal resistance decreases. after sufficient cooling, the supply voltage "seen" (i.e., what can be measured with a DMM for instance) at the external poles is again sufficient to adequately power the device until it heats up again.

this is why caver's (at least in the old days, don't know 'bout nowadays) would alternate periodically (sometimes every 20-30 minutes i've been told by an "olde" caver) between their two primary ([sic], yes, two primaries, not their emergency backup/secondary light) light sources to allow the batteries to cool, thus prolonging battery life.

of course Li-ion's with a protection circuit have another factor that can come into play, but this has been discussed in prev. Posts in this Thread.

-----------------------------------------

i hope that i have followed this discussion. if i've failed to do so, please excuse this intrusion. many thanks, half-watt

 

[greenLED]

H-W, I fully agree with your observations of how voltage sag works, etc. This issue, however, is that this happened with protected rechargeable cells. In good theory (or at least, AFAI understood it) once the circuit "trips" the cells are as good as dead, or at least until you pluck'em in a charger to reset the circuit.

I'm still It's sounding like:
- "protection" circuits work differently depending on brand?
- circuits may degrade over time to the point they don't work as they should?
- even after the protection circuit quicks in, there's still the chance that once of the cells, if left inside the light (and the light is kept on?) will reverse-charge?

Am I following things right? :thinking:

------------------
as far as OP's observation, i've seen this same thing happen and a related phenomenon with primary cells. what i think is happening in both cases is that as the voltage in the cell/battery gets low and it's being used and gettin' hotter, the cell's/battery's internal resistance increases to the point that too much of the little remaining voltage is being dropped due to internal resistance, leaving less to power the load.

after switching off the light, the cell(s)/batter[y | ies] cool(s) and the internal resistance decreases. after sufficient cooling, the supply voltage "seen" (i.e., what can be measured with a DMM for instance) at the external poles is again sufficient to adequately power the device until it heats up again.

this is why caver's (at least in the old days, don't know 'bout nowadays) would alternate periodically (sometimes every 20-30 minutes i've been told by an "olde" caver) between their two primary ([sic], yes, two primaries, not their emergency backup/secondary light) light sources to allow the batteries to cool, thus prolonging battery life.

of course Li-ion's with a protection circuit have another factor that can come into play, but this has been discussed in prev. Posts in this Thread.

-----------------------------------------

i hope that i have followed this discussion. if i've failed to do so, please excuse this intrusion. many thanks, half-watt

 

[half-watt]

- "protection" circuits work differently depending on brand?
- circuits may degrade over time to the point they don't work as they should?
- even after the protection circuit quicks in, there's still the chance that once of the cells, if left inside the light (and the light is kept on?) will reverse-charge?

all good questions. sorry, no good answers.

the first might be the case. one would need to be familiar with EVERY Li-ion protection ckt on the consumer market to answer it. i sure can't. maybe SilverFox or DM51 can???

the second might be a possibility. don't know if the ckt's designed FAIL-SAFE, i.e. if that specific failure mode is not possible based upon the design.

that the third is the usual modus operandi is my understanding also. however, not sure what would prevent a ckt from being designed that would permit SOME LIMITED recovery and re-operation.

please, anyone, correct me if you feel that i'm wrong, but once the cell cools and "regains", so to speak, voltage, i don't see why it couldn't be operational again for a short period of time until more charge is drained or the cell once again heats, dropping voltage internally, and tripping the under-V prot. ckt. that said, Li-ions are supposed to have pretty low int. resistance, comparatively speaking, and so this proposed mode of operation may not prove very practical. to make this feasible during very high current draws (near or at maximal discharge rates for the cell), there would have to be sufficient voltage drop due to increased R-int caused by cell heating to make this work. don't know if this can happen to any great degree in Li-ion's (supposedly not much based upon some reading here on CPF and to a lesser degree elsewhere). again, maybe SilverFox or DM51 can speak knowledgeably on this area also???

i'm all ears, er... eyes...waiting for some authoritative answers to these questions.


----
EDIT:

GL, as far as your last concern. i think that it stands to reason that if the prot. ckt FAILED as you envision that it might, that the cell in question (if it's in series with other cells) might become over-discharged - that said, i've never seen a prot. ckt schematic, so, i don't know how it might fail or what can possibly happen for each potential failure mode of the prot. ckt. if it's a non-protected cell, it can be over dis-charged if the light doesn't have "smarts" built in to prevent it [once ran a unprot. RCR123A down to, as measured on my Fluke DMM, no-load ~0.8V *HOT*, right out of the device - the light was supposed to have "smarts" that prevented such a thing from occurring. of course, the voltage increased quite a bit after the cell cooled off - my "old-timers" is actin' up so i don't remember what the no-load recovery voltage was. chucked the cell rather than try to recharge and reuse it.].

 

[45/70]

Interesting post by half-watt! :thumbsup:

green, I think in the case of your GL3, when it's turned off, nothing is going to happen as far as activity between the cells. It's an open circuit.

Different protection circuits between different brands? No doubt, this is part of the problem understanding this situation.

Protection circuits degrading over time? I think this is a pretty well known fact. That is the primary reason to avoid tripping the protection circuit. It's not really there to tell you when you need to charge your cell, it's there to protect you from "catastrophic failure".

I'll be the first one to say I'm not an "expert", but there's my 3 cents.

Dave

Edit: half-watt, you snuck another post in while I was typing. :thumbsup:

 

[greenLED]

...but once the cell cools and "regains", so to speak, voltage, i don't see why it couldn't be operational again for a short period of time until more charge is drained or the cell once again heats, dropping voltage internally, and tripping the under-V prot. ckt.

Thanks for expanding on this, Gents. You know, HW, I can't think of a reason why what you state wouldn't be possible.

Protection circuits degrading over time? I think this is a pretty well known fact. That is the primary reason to avoid tripping the protection circuit. It's not really there to tell you when you need to charge your cell, it's there to protect you from "catastrophic failure".

I charge my batts periodically so I don't hit the low voltage cutoff. Hoewever, this particular set of batteries were unused for a long time and I can't discard circuit degradation as a source of what I experienced. Now, why would the protection circuit in unused cells degrade faster than that in cells I use frequently? Could it have mroe to do with degrading battery chemistry than circuit performance?

 

[45/70]

Now, why would the protection circuit in unused cells degrade faster than that in cells I use frequently? Could it have mroe to do with degrading battery chemistry than circuit performance?

Ya, well, good point. I don't have an answer for that. There has been discussion about this in several threads, although none recently. There was one where SilverFox, and I think NewBie, were discussing some other aspect of Li-Ions, and this came up. Can't find it now, but I think the main issues involved were heat and number of times the circuit was tripped. So, as for your cells that haven't been used, I don't know. Electronics, of course, do break down over time, but I don't think in the time frame we're talking about.

Dave

 

[matrixshaman]

Actually I would say electronic circuits do NOT normally break down over time IF they are properly designed and IF they are protected from too much heat. I've got a transistor radio that is at least 45 years old and it still works fine. Components in a circuit normally only fail from heat or excess voltage and or current. If battery protection circuits are different - possibly because of exposure to heat and limited heat sinking than that's understandable. I just wanted to clarify that since it sounded like what you were saying is all electronic circuits have a limited life. That may have been true with the very old radio tube circuits where tubes like light bulbs had limited life but solid state electronics if designed well can far outlast most of their owners.
But back to the question at hand. Perhaps you have nothing more than a loose connection internally on the protection circuit? There are a number of possibilities but I'd assume you may just have a one-off deal here with one or two batteries since they don't normally behave this way. Your first assumption may have been as close as it's going to get - a defective protection circuit.