The over-charge protection is only useful IF the charger does not decrease and ultimately stop the charging rate. If you own a analyzing charger you can see on the display that the preset current slowly decreases towards the end of the charging cycle. That’s why you have to use a real charger. I’m not confident with the outcome If you attach two magnets to the leads connected to a regular USB wallwart.
No, if the charger is working correctly it will ramp down charge current to “almost zero” when the cell is at 4.20V
It’s not exactly zero though so it is not advisable to just leave LiIon cells in the charger for days.
However … everything can and will fail, given enough time. Don’t charge LiIon cells unattended. You can minimise risk by charging on a large ceramic/glass/metal plate with nothing that could easily catch fire around and above.
1 Thank
Any good charger will stop charging at a certain threshold and the current will be zero. Even most inexpensive chargers and flashlights with internal charging won’t trickle charge.
1 Thank
Some phones effectively charge forever if left plugged in. Here’s one I measured, which goes into an infinite loop of charging to 4.35V then letting it drop to ~4.28V, then charging again, forever. (FWIW, the cell is rated at 4.35V)
After a while, it inflated the battery so much that the back of the phone popped off. Then not long after that, the battery grew thicker than the phone. And it just kept expanding. Here’s how it look shortly before it died:
I’ve measured this behavior in phones from 3 different companies, including Google. It is the reason why some versions of iOS and Android have a feature to force the battery to stop charging at 80% full… because without that, too many phones were failing while they were still young.
A torch battery like 18650 has the same problem when over-charged… except that it’s in a metal shell which doesn’t allow it to expand. So instead of just popping the back of a phone off, it can build up pressure until it does something worse.
This doesn’t happen quickly though. Pretty much any charger will try to avoid damaging the battery. But it’s still not a good idea to leave batteries in the charger for a long time, especially unattended.
More generally, if you want your batteries to last many years, try to avoid charging or discharging them all the way. They last longer when kept somewhere in the middle.
1 Thank
it might do that if it was just connected to DC over 4.20 volts - forever
but that isn’t what chargers do, what they do do, is - they limit voltage to 4.2, and also turn off when current gets below a set amount, like 5mA
another point is, usually, the wall wart is not the charger, it is just the power supply
so that will usually continue to make power available forever, but the actual charger will not be putting it into the battery forever.
[this is always true of phones, flashlight bat chargers may have those 2 parts separated, or not]
yes but you are getting half the life out of each charge… it;s almost the same either way…
that way - you gain years - but lose hours
Any decent modern charger designed for LiIon cells will do a Constant current/Constant voltage charge curve (CC/CV) and turn OFF when the voltage is at the preset level (normally 4.2 volts) and the current falls bellow a set level (maybe 5mA). Unless you are using a completely crap charger that is broken. It will not trickle charge a Lion cell.
If your charger does not do the above, get rid of it and get one that does.
But yeah, continuing to charge a liIon cell above its specifications is not good. It is bad for the cell and it could result in the cell killing itself (making it unuseable or possibly leaking). Most modern cells have internal safeguards that will prevent explosions, but still you will not like the results.
edit. I did not quote this correctly. I removed it but I’m referring to toy keepers post.
I’m assuming you had an app keeping track of that. So the phone wasn’t able to go into doze or deep doze mode. I could be wrong but I suspect you have more apps doing various things than the average person. I can understand the testing. But it basically shows why people should not be leaving things plugged in. And yes it’s a good thing that some manufacturers are allowing you to stop charging at 80 or 85 or some appear closer to 90 percent. On my phone it stops at 4.2 volts rather than 4.3. The battery indicator still shows at 100% but there’s a little emblem at the top showing that you have “battery saver” on. Some Samsungs show 85%. EV manufacturers do the same thing although some allow you to override it when you need to. All EV manufacturers and phone manufacturers explain that letting the battery sit at 100% charge is the main thing to avoid. If you are going to charge to 100%, you should only do so right before you’re going to use the battery. Apple and EV manufacturers employee charging strategies to charge right before you wake up or right before your departure time for driving.
The worst offender for this kinda thing is gaming laptops.
Theyre always plugged in for maximum performance and running at 60°C their entire lives, just constantly cycling from 4.1-4.2v or w/e at 60°C the whole time.
Last gaming laptop battery pack with 18650s that I disassembled had 2 leaking and the other two capacity tested to 200mah (originally 3350mah Sanyo’s) with an internal resistance of like 600mohm lol
My personal experience using that feature on my Samsung, that stops it at 85%, is that after a few months the phone gets confused and “85%” starts to creep higher and higher. Then you turn off that feature and charging a 4200mah battery from 85%-100% only requires 100mah somehow
I wouldn’t really say I lose hours. It’s more like… I need to recharge every 6 months instead of every 8 months. It’s long enough that I don’t really notice it. And in exchange, the batteries have a virtually unlimited life span. I’m not sure how long, but the cells I got 12 years ago are still working great. OTOH, the cells I’ve used for runtime testing are really showing their old age, even though they’re less than half as old.
While doing phone OS development, I was measuring power usage for each OS build, mostly using automated tools I created but also some manual tests. The measurements came from external power measurement devices connected to the network, and the phones themselves were generally running stock unmodified OSes which were freshly installed just minutes before. The phones had test probes connected to their power circuits, to get direct measurements. There were no phone apps involved, and it did not have to be awake (or even powered on) to do the measurements.
As an example, here’s a test showing the power use while a phone is asleep in airplane mode. It’s the lowest standby power available. The red shaded part of the graph is when the phone had USB connected. After a relay unplugs USB, the phone spends a short time with its screen on before it goes to sleep. The test then waits a couple minutes to make sure the phone is truly asleep before it starts “counting” the data. The green shaded section is the active measurement period. Then at the end it connects USB again and starts setting up the next test. Oh, and it runs the test several times to make sure the behavior is consistent. In this graph, it ran 3 times.
This is how that test should look:
Here is a similar test, but in this test a bug was found. Something was making the phone switch between two different power levels every 30 seconds while it was asleep. Additionally, it took 8 tries to get 3 successful test runs, and the failures for the other 5 attempts were logged in detail for later analysis. It is not supposed to look like this:
With a bunch of phones running these tests 24/7, it generated an awful lot of data. So rather than look at each individual test result, or each graph, another program gathered the results and graphed the data over time, like one data point per test type per phone model per build… with an error margin to indicate variability between individual tests. And when a test result fell outside the expected range, it would generate alerts to let people know a bug (or a fix) had been detected.
For example, here’s a case where a bug was introduced in build #321, then fixed in 322, and then another bug was introduced in 324, which wasn’t fixed until build 332. The blue shaded area is where results are expected, and anything above or below that generates a warning. It also takes note when the error margin suddenly increases, like it did in builds 295 to 298, or build 335.
Anyway, I suppose this was a long-winded way of saying no, I didn’t just install an app on my phone.
it’s probably the resting-vs-using, rather than the 80%-vs-100%, making the difference between your groups…
supposedly [you probably know] in 5 years li-ion will lose about 20% of its capacity just due to slow reactions. (they lose about 1% of charge ( not capacity ) per month, so theoretically 1 full charge would last about 8 years… ) so under that theory, charge-and-forget might last even longer.
The charger should step down before anything bad happens but don’t leave it on for days, the charger will still output a bit of power.
All of the 5 chargers that I use regularly fully shut off the charge. I have measured them. But they also will drain the cell marginally if left in. Some chargers have a feature that will resume charging if a cell voltage drops to a given user defined voltage. I always turn that “feature” off.
Bottom line is. for the best case, you should not leave cells in the charger for any longer than it takes to charge the cell fully. Chare them and remove them ASAP. But a few hours left in the charger should not harm the cells (In a properly designed and properly functioning charger)
But again, if you have a charger that is trickle charging LiIon cells. It is defective. Get rid of it.
How about if the charger is unplugged? I’ve often wondered if I was making a mistake when I pulled the plug on fully charged batteries but didn’t immediately remove them from the charger. Will there be some drain occurring?
I have not measured this. But I would bet that there is some drain.
Best to always remove cells from the charger as soon as they are fully charged.
1 Thank
At my work we have Ipads in conference rooms in special cases that have wireless charging ability. when not in use they are sitting on wireless charger cradles, every 3-4 years they pop screens, the battery inflates like a pillow. the case also is always hot, while internal i pad charger circuit shuts off, the heat does its thing, but not once they exploded, only swells to the point that screen opens up like a book,
For sure, but it will be very minimal. Maybe comparable to the drain when it’s in a flashlight that’s totally turned off. I’d still remove them as soon as you can though
The answer to the original question is: No.
Let’s consider the process of charging a cell with a benchtop power supply, which can be especially useful for charging small cells like rechargeable coin cells.
The process of charging a cell with a benchtop power supply is:
Set the voltage limit to the desired charge termination voltage, such as 4.2V.
Set the current limit to the maximum current you wish to charge at, such as 1/3C, which would be 1000mA for a 3000mAh cell.
Terminate the charge in accordance with the cell’s datasheet.
The power supply will provide constant current and constant voltage, CCCV, automatically as the cell charges.
It may at first provide a constant current (CC) at the current limit you’ve set, and the voltage will be variable, eventually climbing up to the set voltage limit.
Once the set voltage limit is reached, the voltage will be constant (CV) and the current draw will slowly drop as the voltage of the cell gets closer to the supply voltage.
A proper charge cycle of a liion cell is supposed to be terminated not when the current drops to zero, but when it reaches a certain low value, such as C/20, which for a 3000mAh cell is: 3000 / 20 = 150mA. Refer to the liion cell’s datasheet for this specification.
The Samsung 50S datasheet for example says: “Standard charge CCCV, 2.5A, 4.20 V, 250mA cut-off.” The 50S is a 5000mAh cell, so 2.5A is a C/2 charge rate, and 250mA cut-off is C/20. If you don’t have a datasheet, but know the capacity of the cell you’re charging, you could use a typical current value such as C/2 or C/3, and a typical termination current such as C/20. If you leave the cell charging beyond this point and the current draw goes even lower, it’s not a problem.
At least one cell specifies a cutoff current of .01C in the datasheet (Pytes/DLG INR18650-320) which for that cell is a cutoff of just 35mA.
Eventually the current will drop to nearly zero but this takes a long time.
Some sources [1, 2] claim that leaving a lithium ion cell charging can damage the cell. This has been discussed on some forums here and here. After looking into this, I believe that while long duration “float charging” at 4.2V is in fact bad for the life of the cell, it is of little concern within the short time frames of typical charging sessions. Here’s an article that talks about long duration float charging of liion cells: “In float charging, current entry beyond full charge is prevented by choosing a charge voltage less than or equal to the voltage naturally produced by the cell at full charge – which depends on the cell’s exact electrochemistry and temperature.”
In certain industrial applications, liion cells are actually float charged. This should be done at a voltage less than or equal to the “resting voltage” of the cell after a normal charge cycle per datasheet specifications. In other words, if you performed a charge to 4.2V and the cutoff as per the datasheet and then let the cell rest for a day or whatever, and after which you found the voltage resting at 4.18V, this would be the voltage you would not want to “float charge” beyond. I’m not advocating float charging, just making the point that if you were performing a liion charge cycle with a power supply and did not want to monitor the termination current at all, setting the voltage to a low enough value, for example 4.15V or lower, would be a sensible approach.
Typical smart chargers for lithium cells will automatically stop the charge cycle at a certain current, like 100mA or something like that. Look at HKJ’s charger tests for examples of this. Some chargers will restart the charge cycle if the voltage of the cell drops below a certain value, such as 4.0V, if you were to leave the cell in the charger. Some chargers exhibit a small parasitic drain on the cell which would drain the cell over an extended period of time. It is not advisable to leave a cell in a charger indefinitely.
What could make a liion cell self destruct is charging beyond the specified maximum charge voltage in the cell’s datasheet. A typical liion cell with have a charge termination voltage specified as: 4.20 ± 0.05 V, meaning the maximum allowable charge voltage would be 4.25V.
4.2V ± 0.05 V allows for some inaccuracy. Chargers should not be providing a voltage above 4.25V as an absolute maximum.
Also problematic is reviving a dead cell, such as one with a voltage below 2.0V. Refer to HKJ’s article about this.
2 Thanks
Yes, many chargers will drain the batteries when unplugged. Sometimes even several mA.
3 Thanks