Protection circuit test of 3 batteries: Low Voltage condition

zebraa

Newly Enlightened
Joined
Apr 3, 2014
Messages
69
I have several Panasonic NCR18650B 3400 mAh cells that have third-party aftermarket protection circuitry added to them.

There are several claims that the protection circuitry makes, one of which is protection against dangerously low voltage conditions (over discharge). I was not able to locate information on the low-voltage cutoff feature of these batteries, so I tested it myself.


The Panasonic NCR18650B 3400 mAh must not be discharged below 2.5 volts. 2.5 volts is the Discharge Termination Voltage. Cells that are overdischarged should not be charged, and should be discarded, to avoid the risk posed by an internal short due to chemistry de-stabilization.


For this test, I could not use any Zebralights, as their lights have a 2.7 volt low voltage shutdown. Therefore I had to use lights that lent themselves better to this task (no cutoff/shutdown).


I used these three lights:
-Convoy M2. The low voltage behavior of this light was most noticable beginning at 3.0 volts, at which point the light would blink every 1 second. This is a very safe and visible warning. There was also a mode stepdown from high or medium to a lower mode. Eventually, the battery's protection cut-in, ending the test on this light. The battery, not the light, ended this test.
-Nitecore P12. This light began flashing it's "critical" LED flash pattern at 3.6 volts, however that was probably due to turbo/high mode being selected, and the resultant voltage sag. Below 3.6 volts, the light began mode step-downs, until it reached it's 1 lumen setting. Eventually, the battery's protection cut-in, ending the test on this light. The battery, not the light, ended this test.
-Sunwayman V25Cvn. This light has no modes--it is infinitely variable via magnetic control ring. The light eventually began slowly dimming--in a direct drive sort of fashion. Eventually, the battery's protection cut-in, ending the test on this light. The battery, not the light, ended this test.


Three Panasonic NCR18650B-based batteries were tested:
-Zebralight 3400 mAh.
-EagleTac 3400 mAh.
-KeepPower 3400 mAh.


I was looking to confirm the protection circuit would go open-circuit (cutting voltage to 0.00 volts) before the NCR's lower limit of 2.5 volts was hit.


The KeepPower went open-circuit first. After resetting the circuit by placing it on a charger for 1 second, the voltage read 2.90 volts. This is a safe and conservative cutoff.


The EagleTac and Zebralight batteries both went open-circuit at 2.7 volts. This cut-off level was also safe.

 

Kestrel

Flashaholic
Joined
Oct 31, 2007
Messages
7,329
Location
Willamette Valley, OR
I'm afraid I don't know much about the more modern LiIon iterations, but I know that AW has posted that his low-voltage protection circuits will not trip if the current is low enough.

It most likely doesn't apply to your testing, but it's not commonly known so I thought I'd throw it out there.
 

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
The KeepPower went open-circuit first. After resetting the circuit by placing it on a charger for 1 second, the voltage read 2.90 volts. This is a safe and conservative cutoff.

The EagleTac and Zebralight batteries both went open-circuit at 2.7 volts. This cut-off level was also safe

Checking the voltage with no load will not say much about the threshold.

I would expect the real threshold to be somewhere between 2.0 and 2.5 volt.


but I know that AW has posted that his low-voltage protection circuits will not trip if the current is low enough.

That sounds strange, the voltage limit has nothing to do with the current draw, except if there is a leak current. Do you have a link to his post?
 

Kestrel

Flashaholic
Joined
Oct 31, 2007
Messages
7,329
Location
Willamette Valley, OR
[...]That sounds strange, the voltage limit has nothing to do with the current draw, except if there is a leak current. Do you have a link to his post?
Took a bit of work to find (not bad for remembering a 4-year-old post :D), but here it is:
AW 18650 died, then came to life! Whaaa?

In general, it will take a 0.2C load to reliably trip the low voltage cut off where the voltage will rebound to a safe 3.0V when the load is removed. The design is to maximize the use of the cell's capacity. If you have to use a smaller load, you must recharge the battery ASAP to bring the voltage up to prevent permanent damage.

It's always possible that I misinterpreted what was posted - I try to keep in mind that I'm not a battery expert. :eek:

Does that add up then? :thinking:
 

archimedes

Flashaholic
Joined
Nov 12, 2010
Messages
15,780
Location
CONUS, top left
Took a bit of work to find (not bad for remembering a 4-year-old post :D), but here it is:
AW 18650 died, then came to life! Whaaa?

It's always possible that I misinterpreted what was posted - I try to keep in mind that I'm not a battery expert. :eek:

Does that add up then? :thinking:

If I am reading that thread correctly (it is confusing indeed), that OP later stated that he made a "typo" ... meaning he typed "2.20V" but meant 3.20V ??? :thinking:

If that was the case, it is not clear if the protection circuit actually did fail, or instead simply appropriately shut the cell off at ~3.17V ? :thinking:
 
Last edited:

Kestrel

Flashaholic
Joined
Oct 31, 2007
Messages
7,329
Location
Willamette Valley, OR
If I am reading that thread correctly (it is confusing indeed), that OP later stated that he made a "typo" ... meaning he typed "2.20V" but meant 3.20V ??? :thinking:

If that was the case, it is not clear if the protection circuit actually did fail, or instead simply appropriately shut the cell off at ~3.17V ? :thinking:
I know what you're saying, I'm just thinking about AW's specific post by itself - I would at least trust him to not mess up on the numbers. :)

The confusing part is the second half of the sentence:
(and how it may qualify the very first part)
In general, it will take a 0.2C load to reliably trip the low voltage cut off where the voltage will rebound to a safe 3.0V when the load is removed. The design is to maximize the use of the cell's capacity. If you have to use a smaller load, you must recharge the battery ASAP to bring the voltage up to prevent permanent damage.
There are two ways to read the very first part of what was written, it seems to me - if I've misunderstood it, I think I can see how.
 
Last edited:

archimedes

Flashaholic
Joined
Nov 12, 2010
Messages
15,780
Location
CONUS, top left
Agreed, and I have often wondered about this "low current" overdischarge risk .... I do wish that more of the CPF battery experts would drop in here to clarify :eek:
 
Last edited:

SilverFox

Flashaholic
Joined
Jan 19, 2003
Messages
12,449
Location
Bellingham WA
This topic comes up from time to time. If you look at standard Li-Ion discharge curves the voltage rapidly drops off at the end. I believe the protection circuit trips at a low voltage value and at a reasonable discharge rate it protects against over discharge.

However, in the case of parasitic drain the drop off may not be as steep and over discharge may be possible.

I was going to run some tests on this but my CBA doesn't go down to micro amp discharge levels. I believe a "typical" parasitic drain would be in the 100 - 120 micro amp range and with a fully charged cell that would be a long test. An alternate would be to discharge a cell down to the point where the protection circuit trips, reset the circuit, then apply the 100 micro amp load and see how the cell and protection circuit responds.

I have one light (First Light Tomahawk MC) that has a parasitic drain but it isn't bored out to take an 18650 cell. I use primary cells in it and when I store it I simply turn one of the batteries around to break the circuit.

HKJ, if your equipment is capable of a 100 micro amp load and you have a protected cell that you want to destroy you may want to take a look at this. It would be interesting to see how the slope of voltage drop at the end of the discharge changes under very light loads.

Tom
 

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
Took a bit of work to find (not bad for remembering a 4-year-old post :D), but here it is:
AW 18650 died, then came to life! Whaaa?

From AW:
In general, it will take a 0.2C load to reliably trip the low voltage cut off where the voltage will rebound to a safe 3.0V when the load is removed. The design is to maximize the use of the cell's capacity. If you have to use a smaller load, you must recharge the battery ASAP to bring the voltage up to prevent permanent damage.

That makes sense and it is not because the protection does not trip, but because a low discharge current will discharge the cell more thoroughly.
This is very obvious from this test I did:
AW%2018650%202600mAh%20(Black)%20-%201A-Hours.png

There the cell recovers nearly a volt when I stop the discharge (At 25 hours).

AW%2018650%202600mAh%20(Black)%20-%203A-Hours.png

At a higher current, it is more than a volt.

How much the cell recovers will depend on the discharge current, at a very low current, there will be nearly no voltage recovery and the discharge will be much harder on the cell.


HKJ, if your equipment is capable of a 100 micro amp load and you have a protected cell that you want to destroy you may want to take a look at this. It would be interesting to see how the slope of voltage drop at the end of the discharge changes under very light loads.

I can easily do it, the problem is that types of test takes a long time.
 

kosPap

Flashlight Enthusiast
Joined
Mar 1, 2007
Messages
2,909
Location
Naoussa Greece
quote_icon.png
Originally Posted by AWIn general, it will take a 0.2C load to reliably trip the low voltage cut off where the voltage will rebound to a safe 3.0V when the load is removed. The design is to maximize the use of the cell's capacity. If you have to use a smaller load, you must recharge the battery ASAP to bring the voltage up to prevent permanent damage.

i can comfirm that from practice....

happened to me when I used an AW16340 with a 5mm white LED (remember that fad of converting P60s to 5mm LEDs)
Draw is 20mA and the protection did not trip.
It is not uncommon to measure such low Amp draw in the low mode of current flashlights, well below the 0.2C for a 16340.
So beware
 

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
HKJ, if your equipment is capable of a 100 micro amp load and you have a protected cell that you want to destroy you may want to take a look at this. It would be interesting to see how the slope of voltage drop at the end of the discharge changes under very light loads.

Would definitely be an interesting test ....

I have started on this test, but it is going to take some time (i.e. weeks).

Here is one of the preliminary curves:
Discharge%20Xtar%2018650%202600mAh%20%28Black%29%200.01A.png

The yellow line is the protection trip point.
A short time later I reset the protection with a current pulse.

But I would also like to test a battery with automatic reset of the protection, it might be different at low currents. Do anybody know some battery with automatic reset?
 

SilverFox

Flashaholic
Joined
Jan 19, 2003
Messages
12,449
Location
Bellingham WA
Hello HJK,

Better rebound than I expected, but still pretty low ending voltage.

I was under the impression that protection circuits required a minimum voltage to reset. I am not up on these circuits so I am not sure this is the way it is or not. I do know that early circuits required a minimum current as well as a voltage to reset and many chargers were unable to do this if the circuit experienced a hard shut off. The original Pila charger charged at something like 1000 mA or 1500 mA, I don't remember off hand, but it had no problems resetting the circuits used in the Pila cells. The revised Pila IBC charger dropped the charge rate to 650 mA and came with a reset button. Sometimes it took several presses of the reset button to reset the Pila cell protection circuit. In years past resetting the protection circuit was an issue that came up often. Now it doesn't seem to come up very much so I assume there has been an evolution in the circuit design.

Tom
 

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
Better rebound than I expected, but still pretty low ending voltage.

The rebounding is much large, if the battery has more time. I have increased that time significantly in my current tests.


I was under the impression that protection circuits required a minimum voltage to reset.

I had the impression that some circuits would reset automatic when the load was removed, but I have never tested it.

I do not know if I will get down to 100uA, in the first test I went directly from a 1mA discharge to a 100uA discharge. After two days with 100uA discharge, the battery voltage was still increasing and I stopped that test. Now I have added a 300uA discharge also and hope it will work a bit faster.
 

HKJ

Flashaholic
Joined
Mar 26, 2008
Messages
9,715
Location
Copenhagen, Denmark
Awesome! I am curious to see the results!

Me too, but it takes a long time.
Until now I have used a week on my test script and I do not now if I am satisfied yet. I will probably be a couple of days before I know that.

At the current time I have the above Xtar battery and a Keeppower 3400 battery that I tests with (They do definitely not use the same IC).
Depending on how long the test takes, I will probably run one or two batteries more (I am open for suggestions).
 
Top