Charging voltage for 3.8v Li-ion battery

[Rick Law]

Blimey!  What a lot of discussion this has created!  It's been very edumacational for me anyway.

I'm currently charging it with a constant current at 300mA, and the voltage across it is coming up nicely.  When it gets to 4.2v I'll switch to constant voltage and leave it 20 mins or so until the current has dropped.

I've ordered a couple of those TP4056 modules, hopefully they'll be here soon...
...

This is about the module and not all about the charge current.

After you get the module, look very carefully.  See my attached photo and look at the chip, the board looks alike but the chip is not the same.

I assume you are going to change the resistor to change the current - Don't rely on the specs alone but instead actually measure your charge and stop current and readjust the resistor as needed.  Even if you don't mind a 10% or 20% mA difference, do measure it to make sure it is not something wild.

As to the photo, I think both are not real TP's - check the photo of real TP4056 chip, there should be a logo.  Neither of these has logo.
(yike, I've been calling it 4096, I better edit those replies)

Rick

[Rick Law]

Mr. Delta,

Last reply was more about the TP4056, this one is about the current/voltage.

(I purposefully use 'I' here to emphasize this as merely my personal experience with these batteries and I am keenly aware this may not apply to you.  This is merely sharing experience.)

I use the contraption often for savaged Sony-Fukushima US18650CR battery.  Savaged from a dead laptop battery pack.

What I have learned from this is: stop voltage difference of 0.01V or even 0.26V doesn't make much difference in the total capacity.  The difference is hardly measurable between 4.00V vs 4.20V vs 4.26V stop voltage.  4.26V is what I think is the safe maximum.

All along the line from TP4056-to-Connector-to-PCM-to-battery, the voltage is dropping.  PCM is the battery Protection/Control module.  The TP4056 may be outputting an exact 4.2V, the TP's designed CV voltage, at the boards connector, it is already below that - every wire is a resistor!  At the out-board side of the connector, it drops more - every connector is a bigger resistor!  By the time it gets to the out-side of the PCM, it is even lower.  Now it hits the battery terminal and that is of course lower still.

While working with the contraption, I ponder if I should adjust the contraption's program/hardware so the battery terminal is 4.2volt - but after a moment's thought l decided not to - evenwhile knowing 4.2volt at the contraption mean the battery is likely lower than 4.2volt and can still take a bit more.

The connection drop is going to depend on the quality of the wire use, the quality of the connection...  The drop is also depending on the current flow - it drops far less when nearing full.  "Mr. A" (arduino) can of course take current into account, but that would still leave a situation where the quality of the connecting wire and the age of the gator-clip makes a difference.  The older gator-clips doesn't grip as well and thus has higher resistance and thus a higher voltage drop.

It is safer to under-charge than over-charge.  So, in the end, I didn't worry about should I take it to the limit (4.26V), and just leave the stop-volt at 4.20V.  The capacity difference is not worth the trouble.

Rick

[eas]

The cell probably has a 4.3v or a 4.35v charge termination voltage.

Samsung didn't invent a new chemistry, but there is ongoing refinement of electrode composition and electrolyte additives in pursuit of some combination of higher energy capacity, higher power delivery and lifetime. At this point, 4.35v cells can deliver higher energy capacity while still having similar lifetimes to those considered acceptable with older 4.2v-termination cells.

There is no harm charging it to 4.2v, other than leaving ~10-15% of current capacity unused (and slightly more of the power capacity) until you can give it a proper charge. Similarly, there isn't any harm in just terminating charging immediately at 4.2v or whatever the proper termination voltage is, the only downside is, again, leaving some capacity fallow.

As noted, most TP4056 boards these days seem to have knockoff chips on them. Fortunately, the knockoffs I've heard of people testing seem to perform similarly enough. Of course, who knows how consistent they are.

I wish that modules for higher termination voltages, or even better, adjustable termination voltages, were easier to come by, even at 2-4x the price of the TP4056 boards.

Oh, following up on Rick Law's suggestion of checking the actual charging current if you replace the resistor (or even if you don't), all these battery charging chips are pretty sensitive to resistance between charger and battery, so sticking a multimeter between the charger and the cell can mess with the measurement. On the other hand, the TP4056 and clones are all linear charge ICs, and don't draw all that much current on their own, so you can get a reasonable estimate of current going into the battery by measuring the supply current.

Finally, while yes, the leads protection circuits, etc all result in some voltage drop, a lot of it is related to the amount of current. So, while the charger might start ramping down from 0.5A before the battery is at 4.2v, as the current drops, the difference between the cell voltage and the voltage the charger is reacting to will drop, and by the time it is near termination current, the resistive losses are going to be pretty small.

[JohnnyBerg]

I wish that modules for higher termination voltages, or even better, adjustable termination voltages, were easier to come by, even at 2-4x the price of the TP4056 boards.

You can have a look at the MCP73831. This IC comes in 4 different versions, each with another voltage.
I could not use it in my application, because this IC is unstable when there is no battery connected. I have to detect that

Oh, following up on Rick Law's suggestion of checking the actual charging current if you replace the resistor (or even if you don't), all these battery charging chips are pretty sensitive to resistance between charger and battery, so sticking a multimeter between the charger and the cell can mess with the measurement.

100% true.

And most people don't bother to build some circuity that seperates the load from the battery when charging. Drawing current from the battery while charging also messes things up.